Previous Page
THE MIRAGE OF THE H2 ECONOMY 477
pockets and resulting perturbations. J. Chem. Phys. 117:
3753–3761.
8. Bowron, D.T., Soper, A.K., and Finney, J.L. (2001). Temper-
ature dependence of the structure of a 0.06 mole frac-
tion tertiary butanol–water solution. J. Chem. Phys. 114:
6203–6219.
9. Hummer, G. et al. (1996). An information theory model
of hydrophobic interactions. Proc. Natl. Acad. Sci. 93:
8951–8955.
10. Botti, A. et al. (2004). The microscopic structure of concen-
trated NaOH solutions. J. Chem. Phys. 120: 10154–10162.
11. Bowron, D.T. and Finney, J.L. (2003). Structure of a
salt–amphiphile–water solution and the mechanism of salt-
ing out. J. Chem. Phys. 118: 8357–8372.
12. Sacco, A., De Cillis, F.M., and Holz, M.J. (1998). Chem. Soc.
Faraday Trans. 94: 2089.
THE MIRAGE OF THE H2 ECONOMY
REUEL SHINNAR
The City College of the CUNY
New York
In a recent paper (1), I analyzed the new ‘‘hydrogen
economy’’ and explained why, for most uses, it makes
no sense to convert fuels, electricity, or any other form of
energy to hydrogen, only to use the hydrogen as a fuel for
fuel cells. As H2 is not a resource but an inefficient energy
carrier, it would appear that the H2 economy is just a
mirage. If alternative energy sources are ever developed,
an economy based on electricity would be cheaper (by a
factor of three) and much easier to adopt than H2 . Oddly,
I had covered the same topic twenty years ago in a paper
that analyzed the hydrogen economy of the 1970s (2).
Then, the investment in H2 research was about 10 billion
wasted dollars before it was phased out. In this column, I
will try to explain why the scientific establishment engages
in such obviously futile behavior and why this mirage
attracts so many followers.
In the mid 1970s, I became aware of a strange
phenomenon. Many leading scientific journals were
suddenly giving enormous coverage to hydrogen, the
fuel of the future; a new term, hydrogen economy,
was born. International meetings attracted thousands of
participants and the United States, Germany, and Japan
funded the necessary development with vast resources.
The central idea, put forward by the International
Atomic Energy Research Center in Vienna, used a high-
temperature nuclear reactor as the heat source. In this
reactor, pressurized helium was used as the coolant. To
generate electricity, the hot helium was first expanded
through a turbine and then used for steam generation
before being fed back to the reactor.
The ‘‘thermochemical cycle,’’ the name applied to this
new process, does not include a turbine, and the heat from
cooling the hot helium is used to drive a set of closed-cycle
chemical reactions; the net result is that water is split into
hydrogen and oxygen.
Following is a hypothetical example of such a reaction:
MeO + Heat −−−→ Me + 1/2O2
Me + H2 O −−−→ H2 + MeO + Heat
where Me is any metal. No such reaction was ever
discovered, and all cycles proposed were quite complex.
One does not need to be an experienced cost estimator to
realize that such a cycle would be very expensive and have
a low thermal efficiency.
The H2 was supposed to be used to replace natural
gas in existing pipelines. Several obvious questions were
not asked:
1. A pipeline for methane cannot be used to transport
H2 . To avoid leakage, one needs totally different
valves and connectors. Furthermore, as the volume
of H2 is triple that of methane, one needs larger
pipes and different compressors. Of course, the same
factors would apply to the distribution of the gases
to customers. None of the valves or controls would
be tight enough, and all burners would have to
be changed.
2. As nuclear reactors are available, why not use
electricity instead of methane for most applications?
Electricity is far more efficient than H2 or methane
(by a factor of two) for most uses. Let me illustrate.
Today, in the United States and the world, a large
amount of electricity is generated from natural gas.
This electricity, on a BTU basis, costs by a factor
of about 2.2 more than the natural gas feed. The
fact that we do it obviously means that electricity
is more valuable than natural gas. At the same
price, it would have driven out natural gas. H2 as
a fuel has an even lower value than natural gas. If
we now convert electricity to H2 , we again lose at
least one-third of the energy. Thus, H2 made from
electricity has a value about four times lower than
the electricity. This simple calculation illustrates
the stupidity of the idea to convert nuclear or
solar energy to hydrogen. At that time, no one was
foolhardy enough to suggest using H2 generated by
a nuclear reactor as fuel for a fuel cell. Furthermore,
an electricity distribution system is already in place
and can grow with demand.
3. Why not make H2 by electrolysis, a well-known
process? Nobody asked why—in contradiction to our
cumulative technical experience—a set of chemical
reactions should generate a large increase in T
more cheaply or efficiently than electrolysis. The
main item in the cost of electricity is the nuclear
reactor, and if a thermochemical cycle is less
efficient, it is inherently more expensive.
4. The cost of switching to hydrogen is astronomical.
To produce the equivalent of one million BTU of
hydrogen per day by electrolysis would require a
25-kilowatt nuclear reactor at an investment cost
of $150,000, which means that a standard nuclear
reactor of one gigawatt costing $6 billion will be
able to supply 40,000 million BTU/day of H2 the
478 THE MIRAGE OF THE H2 ECONOMY
equivalent of 6700 barrels of oils. Building the
capacity for supplying a new pipeline that could
carry 12% of the current U.S. capacity of natural gas
would have to be built from the start at a cost of
close to $1 trillion. For thermochemical cycles, the
cost would be significantly higher. All of this must be
invested before a pipeline can be started. The same
energy needs could be supplied by electrical energy
at only one-quarter the cost because electrical energy
has a higher efficiency for almost all purposes.
The investment for electricity can be spread over
many years because a grid is already in existence.
Furthermore, just the energy from the nuclear
reactions would cost $50 per million BTU H2 .
5. The most dangerous of all fuels, pressurized H2
ignites with an invisible flame. Its explosive limits
are very wide, and its minimum ignition energy is
ten times smaller than that of any other gaseous
fuel. This problem occurs not only when H2 is being
distributed or used, but even more so when it is
generated in a thermochemical cycle. When H2 is
produced by electrolysis, one can locate the plant
far away from the nuclear reactor generating the
electricity, which, however, is not true for helium,
which must be pumped in such large quantities that
it cannot be sent over long distances at reasonable
cost. Society has accepted the risks of explosion
in conventional fuels but tries to limit exposure.
Thus, even a small propane storage tank may not
be transported through a tunnel, but what level of
risk from a potential explosion at a nuclear power
plant complex is acceptable? None of these critical
questions were asked then—nor are they being
asked today.
These questions have been on my mind for a long
time. In the ‘‘age of technology,’’ how could the scientific
community have been misled by such obvious nonsense?
And why had so many engineering firms and large
companies come up with totally unrealistic cost estimates,
at least by a magnitude too low, and thermal efficiency
estimates that are unrealistically high?
The intellectual climate of the time can be illustrated by
a small story. Science carried a six-page article, featured
on the cover, about a new cycle developed by General
Electric that promised to produce H2 at $2.50/million
BTU H2 (at that time it would have cost $7.00/million
BTU to produce H2 from oil). For fun, I flow-sheeted the
process and published the mass and heat balance in a
letter to Science (3) suggesting that if GE can reduce costs
to $2.50, they should forget about hydrogen and change
the economy. By comparison, us poor chemical engineers
could not achieve such results at a cost of $250/million
BTU H2 . Three months later, I was invited to GE’s central
laboratory where the Director of Chemical Research
offered me a consulting job. The head of the lab was a
chemist who had read about the need for a thermochemical
cycle, and he actually developed the first one that worked.
His engineers had given him this cost estimate. When my
letter was published, he asked Corporate Engineering
to make their own estimate and they confirmed my
numbers. This story illustrates a key aspect of the
hydrogen economy. Many excellent scientists lack the
skills needed to deal with economics and costs. Later, I
published another paper demonstrating, from elementary
design principles, that any thermochemical process must,
by necessity, cost much more than electrolysis (2).
It is true that many scientists could have erred because
they were not knowledgeable about process economics.
However, I also came to realize that other forces were
driving the program. Although they are inherently safer,
high-temperature nuclear reactors of the type required by
thermochemical cycles were deemed less competitive and
more costly than boiling water reactors, which created an
opportunity for their developers to seek more government
support. A large nuclear research community was hungry
for grants and contracts.
Interestingly and unfortunately, the episode of the
1970s has been ignored or forgotten. However, before
we place large national resources behind an effort to
again deal with such major problems as renewable energy,
global warming, and energy independence, we have to
understand what went wrong the first time a so-called
‘‘hydrogen economy’’ was attempted.
The paper I just published is about the reincarnation
of the hydrogen economy. This time, all the proposed
processes are feasible, but the program itself seems to
make even less sense. As before, the main emphasis is
on the replacement of conventional fuels with a national
distribution system of H2 at a tremendous loss in efficiency
in the production and transmission steps. As before, the
same intractable safety and cost problems still persist.
Similar driving forces remain at work. Large research
budgets are projected to find a use for a technology in
which DOE has invested huge sums of money to little or
no avail. This time, it is fuel cells powered by hydrogen,
a technology that is good for space applications, where
H2 is used as a rocket fuel, and therefore available in
small quantities, and other specialty applications, but
useless for applications that couple H2 generation with
the generation of electricity. Such systems have half the
efficiency of combined cycle turbines, cost over five times
more (1,4), and create at least double the greenhouse
emission, probably much more.
Today, there even is talk about switching cars from
gasoline to H2 , which makes the transition problem even
more difficult as the new supplying system cannot start
locally, and this would also require the use of pressurized
H2 at 6000 psi, a pressure very seldom used in industry.
Nobody mentions that in industry an H2 tank with
the capacity proposed for a family car requires storage
in a special room with a blowout wall. Any technician
approaching such a pressurized tank to check the valves
swings a wooden broom to prevent being fried by an
invisible flame which may come from a leaking valve,
as above 1500 psi, an H2 leak is self-igniting (1). For a
gas station, a minimum-sized storage tank would have the
explosive power of two thousand tons of TNT or two of the
largest bombs used by the Air Force. A terrorist could just
open a valve and detonate a small bomb after 20 seconds.
The most obvious fallacy of the new H2 economy,
however, is that it solves all of our problems (5,6). For

those of us who accept the first and second laws of
thermodynamics, it is obvious that the use of H2 would
increase the use of other energy resources. Arguments
are made that, once we have a hydrogen distribution
system, we could capture the CO2 produced and sequester
it. However, we have no idea how to sequester such vast
amounts of CO2 safely. A successful policy can only be
formulated after all the elements are solved and a good
deal is known about costs. At present, the United States
releases 100 million tons of sequestered CO2 back into the
atmosphere each year.
In a system in which the basic science is known,
costs are reduced by innovative ideas that develop during
the implementation stage of a project rather than by
more research. The cost of liquid natural gas plants and
combined cycle power plants decreased in the last ten
years by a factor of two even though the basic technology
did not change because many were built to compete in the
market. So, research, to be effective, needs to be coupled
with a clear plan for implementation.
In order to promote the use of fuel cells or other
new technologies, which are by one to two orders of
magnitude more expensive than available technologies,
the research community has invented the myth of
‘‘learning curves.’’ In the last 40 years, computers became,
through miniaturization, cheaper by several magnitudes,
but this option does not apply to energy systems. In the last
20 years, the cost of solar cells was reduced by less than a
factor of two despite a tremendous research budget. Better
engineering and mass production can reduce the cost of
most new technologies by a factor of two to three. However,
this would require the building of large plants based on
competitive bidding, not a new research breakthrough.
Large research budgets over the past 40 years have not
made the cost of fuel cells more competitive.
The alternative sources that really reduce CO2
emissions include solar and nuclear energy and biomass.
The use of H2 just makes every one of these options several
times more expensive. Solar and nuclear energy produce
electricity, and, for most purposes, electricity is far more
efficient to use than H2 . To provide electricity to the motor
in a H2 car requires three to four times the solar or nuclear
capacity required for providing the electricity directly via
a battery. Although we don’t have good electric cars yet,
Toyota has just come out with a hybrid car with a plug-in
battery large enough for a 40-mile drive. Such a car could
reduce total gasoline consumption by 80%, but electric cars
would only reduce greenhouse emissions if the electricity
were generated by solar energy. As for biomass, which
is a valuable but limited future resource, conversion by
fermentation to alcohol followed by conversion to gasoline
or diesel is far cheaper and thermally more efficient than
conversion to hydrogen.
In the early 1990s, an Israeli company, Luz (7), built
solar electric power plants using a high-temperature heat
transfer fluid, which included sufficient storage to supply
Los Angeles with 300 megawatts of electricity on a reliable
basis. The Luz plant had another major advantage; the
design allowed for conventional fuel to be used as a backup
because, even in the desert, 10% of the days may be
rainy. To be profitable, all Luz needed was a 50% tax
THE MIRAGE OF THE H2 ECONOMY 479
break on the investment. When Luz became large, the tax
breaks were taken away. If Luz had been allowed to build
a few gigawatts, the price would have come down and
other companies would have come in; we might have had
affordable solar energy today.
We have today the resources and the technology to
do something useful about global warming if we really
had the resolve. For example, for $50 billion a year
over twenty years, we could replace 40% of coal power
plants with solar power plants using proven technology;
we just have to engineer them better and provide them
with sufficient storage capacity. We are the only western
nation who has enough sun and enough area for solar
power. Alternatively, we could eliminate about 80% of
our gasoline used for cars. We could easily afford to do
both, but we have to realize that no research can make
solar, nuclear, or other alternative energy competitive
with cheap oil or gas.
Our society has solved far more difficult large-scale
technical problems successfully, e.g., space travel and the
most advanced defense weapons in the world. Applying
our capabilities to global warming or to resource depletion
to preserve a livable world to our grandchildren would
require not only conviction but a strong political will to
do so.
To give a recent example, thirty years ago, we faced the
problem of increasing pollution from our coal power plants.
We had the technology (scrubbers) to reduce the pollution
substantially. It would have cost about $20–30 billion, but
politics and the influence of senators from coal-producing
states prevented the adoption of such a solution. Instead,
we spent $20 billion on ‘‘clean coal’’ research programs
that failed to produce results that could help solve this
problem. However, the ‘‘clean coal’’ effort did provide large
research budgets along with the illusion that we were
accomplishing something. Thirty years later, the same
plants are producing more electricity and consuming more
coal—still poisoning the fish we eat with mercury, causing
smog that blocks us from seeing the sky, destroying our
forests, and, most importantly, impairing the health of all
of us more than any other form of pollution.
The private car is another example of the obstacles
that a serious national policy faces. In the 1970s, in
response to the Arab oil embargo, oil price increases,
and an alleged likelihood that the oil resource base
would become depleted within the next several decades,
Presidents Nixon, Ford, and Carter proposed new energy
policy initiatives that included the goal of improving auto
fuel economy. These initiatives resulted in the Corporate
Average Fuel Economy standards (CAFÉ), which required
that the fleet fuel economy of new light-duty vehicles
(5500 lbs or less) reach 27.5 mpg by a certain, date a
30% reduction in gasoline demand. This standard limited
the ability of the companies to produce large cars and
stationwagons. The same provision set lower standards in
mileage and emissions for light-duty trucks weighing less
than 5500 lbs.
Nine years ago, the car industry realized that one could
produce large cars (SUVs) that fit all the definitions of
a light truck, such that one should be able to fold all
seats to the floor to provide a major space occupying
480 HYDROGEN ION
the car and several others. It was a great success. As
light trucks convertible to luxury passenger cars do not
count as passenger cars for the CAFE laws and are not
considered in the fleet average, this allowed car companies
again to produce gas guzzlers. Eight years later, 30% of
the passenger cars sold in the United States are SUVs, and
all the efficiency gains of 30 years have been wiped out.
As we face the potential of global warming and
resource depletion that might endanger the future of
our grandchildren, we escape our responsibilities by
pretending that all problems can be solved better and at
lower cost through the mirage of more research. Thus, the
H2 mirage allows car companies to continue to produce
gas-guzzling, highly polluting SUVs while touting the
glorious future, an ‘‘H2 ’’ car that will miraculously be
energy efficient and pollution-free. And, as in the 70s,
thousands in the scientific and political communities are
promoting this mirage to the public.
Thus, mirages like the ‘‘hydrogen economy’’ play an
important psychological function in our society, and many
in the technical community play along. The role engineers
and scientists should be playing today is to educate the
public and the politicians about the real options. We should
be reminded of an old cartoon in the Pogo series that bears
the caption ‘‘We have met the enemy and he is us.’’
BIBLIOGRAPHY
1. Shinnar, R. (2003). The hydrogen economy, fuel cells, and
electric cars. Technology in Society 25(4): 453–576.
2. Shinnar, R., Shapira, D., and Zakai, S. (1981). A thermochem-
ical and hybrid cycles for hydrogen production. A differential
economic comparison with electrolytes. IEC Process Design &
Development 20: 581.
3. Shinnar, R. (1975). Thermochemical hydrogen generation heat
requirements and costs. Science 188: 1036.
4. Shinnar, R. (2003). Fuel-cell faux pas. Chemical & Engineering
News 81(14): 4.
5. Rifkin, J. (2003). The hydrogen economy. Environmental
Magazine. January.
6. Bush, G.W. (2003). State of the Union Address. January 28.
7. Shinnar, R. (1993). The rise and fall of luz. ChemTech 23:
50–53.
READING LIST
Energy Research of the Department of Energy. (2001). Was It
Worth It? National Research Council.
Effectiveness and Impact of Corporate Average Fuel Economy
(CAFE) Standards. (2002). National Research Council.
HYDROGEN ION
GHOLAM A. KAZEMI
Shahrood University of
Technology
Shahrood, Iran
INTRODUCTION
This article includes an introduction to the ion and its
properties followed by an overview of the characteristics
of the hydrogen atom. The latter has been extracted
mostly from Sienko and Plane (1). This article should be
studied in conjunction with the pH and Hydronium Ion
articles.
An ion in general terms can be any grouping of one or
more atoms that carries a charge (1). The characteristics
of an ion include:
1. the sign (type) of charge: positive (cation) or
negative (anion);
2. the number of charges or valence (z): monovalent
divalent, etc;
3. the ionic radius (r): the distance from the center of
the nucleus to the outer shell of the atomic particle;
4. the ionic potential: z/r, the ratio of the charge of the
ion to the ionic radius;
5. the electronegativity: the power of the ion to
attract electrons;
6. the molar volume: the volume of one mole
(cm3 /mol); and
7. the ionic conductance: the ability to carry electri-
cal current.
When an ion is surrounded by many other ions/molecules
such as in a solution, its other properties such as
ionic strength and ion activity coefficient come to light.
Ottonello (2) is a major reference where many properties
of various ions are tabulated.
Hydrogen, the lightest of all gases, is colorless, odorless,
and tasteless. Together with helium, it occupies the first
period of the periodic chart of elements and is the most
abundant element in the solar system. After helium, it has
the second lowest boiling temperature of all elements (2).
Because of its low boiling point, liquid hydrogen is used
as a cryogenic fluid (to produce low temperatures), and
because of its high heat of combustion (120 kJ/g), it
is a valuable rocket fuel. Molecular hydrogen, H2 , can
combine directly with most elements. Mixtures of H2 and
O2 can be explosive. Much of the industrial consumption
of hydrogen is in hydrogenation reactions, the addition of
H2 to other molecules. Water and power are required to
produce hydrogen. In its compounds, hydrogen is found in
three oxidation states, +1, −1, and 0. Hydrogen is such
an important element that it has its own association. The
National Hydrogen Association (WWW. Hydrogenus.org),
for example, was formed in 1989 (in the U.S.) to act as a
focal point in hydrogen related matters. Anyone interested
in water related matters should know the important
characteristics of hydrogen because of its place in the
water molecule.
DEFINITION OF A HYDROGEN ION
A hydrogen ion, H+ , is a monovalent cation that occurs
when a hydrogen atom loses its only electron:
H −−−→ H+ + e−
Therefore, the hydrogen ion has only one proton and
no electron. This is why ‘‘proton’’ is another name for

hydrogen ion. Some references state that there is no such
ion as a hydrogen ion and what occurs, is actually a
hydronium ion, H3 O+ , rather than a hydrogen ion (3).
The hydride ion, H− , occurs when a hydride such as NaH
dissociates:
NaH −−−→ Na+ + H−
In this reaction, the hydride ion radius is 1.46 Å. Molecular
hydrogen is formed from the reaction of a hydride ion and
a hydrogen ion (4):
H+ + H− −−−→ H2
H3 + is an ion that is artificially generated by the
interaction of H+ and H2 . For instance, generation of
H3 + is the most critical specification of the ion source in a
mass spectrometer (5).
CHARACTERISTICS OF THE HYDROGEN ION
Some major characteristics of the hydrogen ion are
presented in Table 1.
In chemistry and geochemistry, it is assumed that
the standard heat capacity, C0p ; standard partial molal
Gibbs free energy from the elements, G0f ; standard partial
molal enthalpy of formation from the elements, Hf0 ;
p
standard partial molal entropy, S0 ; and standard partial
0
molal volume, Vp , of the hydrogen ion equal zero at all
temperatures and pressures (2,7).
MEASUREMENT OF HYDROGEN ION CONCENTRATION
More than 100 years ago, Arrhenius suggested that acid
properties are attributable to the presence of an easily
detachable hydrogen ion, H+ (1). Now, everyone knows
that the concentration of hydrogen ion in any solution is
a measure of the solution’s acidity or pH (pH = − log H+ ).
Thus, the measurement of H+ concentration is better
described in the pH article of the encyclopedia and is
only briefly dealt with here.
The H+ ion is the easiest ion concentration to measure.
This is done using a hydrogen electrode, which is formed
by bubbling H2 over a platinum electrode surface (7). A
variety of these electrodes are commercially available;
Metrohm pH meter Model 744 and Orion pH meter
Table 1. Notable Characteristics of the Hydrogen Ion, H+a
Type Cation
Valence 1
Radiusb —
Ionic potential (IP)c —
Electronegativity 2.2
Molar volume 0 (This is based on convention)
Ionic conductance 349.8 cm2 × (equiv )−1
Charge to mass ratio (q/m) 9.6 × 107 C/kg
a
References 4, 6, 7.
b
It is difficult to define because the electron of the hydrogen atom has gone,
and only the proton remains. So, there is no outer shell left to measure its
distance to the proton.
c
It cannot be calculated because there is no specified ionic radius.
HYDROGEN ION 481
Model 420A are examples. Measurement of pH cannot
be more accurate than 0.02 pH units because the standard
solutions or buffers which have been calibrated by the U.S.
Bureau of Standards (NBS) have an absolute accuracy in
the range of ±0.01 to 0.02 pH units (7). In field conditions,
where disturbing factors are ample, the accuracy is
usually ±0.05 pH units. Colorimetry is one older method
of measuring H+ concentration.
Some Examples
Pure neutral water at 25 ◦ C has a pH of 7, a H+
concentration of 10−7 mg/l. A 1-molal solution of HCl
will have an H+ concentration of 1 and a 0.01-molal
solution will have an H+ concentration of 10−2 mg/l. The
concentration of hydrogen ion in a 0.01-molal solution
of a weak such as H2 CO3 is only 10−4.2 mg/l, and it is
10−10.8 mg/l in an aqueous solution of pH 3.2.
Many authors prefer to use the term ‘‘hydrogen ion
activity’’ rather than hydrogen ion concentration. The
reason that the ‘‘pH’’ concept has been developed is the
very low concentration of hydrogen ion and the difficulty in
comprehending such low, sometime ambiguous, numbers.
In solutions, the concentration of H+ never reaches zero;
there is no solution that is 100% basic or 100% acidic.
PRODUCTION OF HYDROGEN ION IN WATER
Hydrogen ion is produced in natural water from the three
following main sources as well as from humic and fluvic
acids, volcanic gases, acid rain, and short-chain organic
acids present in some oil field brines (8).
1. Hydrolysis
H2 O −−−→ H+ + OH−
As stated above, however, it might be more appropriate to
write the previous equation in the form:
2H2 O −−−→ H3 O+ + OH−
2. Dissociation
H2 CO3 −−−→ H+ + HCO3 −
3. Oxidation
2FeS2 + 7.5O2 + 7H2 O −−−→ 2Fe(OH)3
+ 8H+ + 4SO4 2−
IMPORTANCE OF THE HYDROGEN ION
Some of the important functions and behaviors of the
hydrogen ion are described as following. Considering the
huge and universal importance of H+ , these should be
regarded only as examples.
1. The radius of the hydrogen ion is very close to the
size of the neutron. This similarity is the principle
used to design and develop compensated neutron
482 THE HYDRONIUM ION
logs (CNL) and neutron probes. The former is used
as an important tool for calculating the porosity of
petroleum reservoirs and aquifers (9,10), and the
latter is of extensive use in agricultural practices for
measuring soil moisture (11).
2. The concentration of hydrogen ion, pH, is a
master variable of the groundwater system because
the hydrogen ion participates in many important
chemical reactions such as (12):
a: mineral dissolution—precipitation
CaCO3 + H+ = Ca2+ + HCO3 −
b: aqueous complexation
Fe3+ + H2 O = Fe (OH)2+ + H+
c: adsorption –desorption
≡ FeOH + Cu+ = ≡ FeOCu + H+
3. The hydrogen ion plays a major role in denitrification
through which various forms of nitrogen in the soil
revert to N2 (6):
4NO3 − + 4H+ −−−→ 2N2 + 5O2 + 2H2 O
It also plays a major role in redox reactions and
acid–base reactions.
4. In sanitary engineering and water treatment, pH
is an important factor in chemical coagulation,
disinfection, water softening, corrosion control, and
biological processes (13).
5. The most remarkable function of hydrogen ion is its
control of the acidity or alkalinity of solutions.
BIBLIOGRAPHY
1. Sienko, M.J. and Plane, R.A. (1979). Chemistry: Principles
and Applications. McGraw-Hill, New York.
2. Ottonello, G. (1997). Principles of Geochemistry. Columbia
University Press, New York.
3. Fetter, C.W. (1988). Applied Hydrogeology. 2nd Edn. CBS
Publishers and Distributors, Delhi.
4. Lagowski, J.J. (1973). Modern Inorganic Chemistry. Marcel
Dekker, New York.
5. Attendorn, H.G. and Bowen, R.N.C. (1997). Radioactive and
Stable Isotope Geology. Chapman & Hall, London.
6. Evangelou, V.P. (1998). Environmental Soil and Water
Chemistry: Principles and Applications. John Wiley & Sons,
New York.
7. Langmuir, D. (1997). Aqueous Environmental Geochemistry.
Prentice-Hall, Upper Saddle River, NJ.
8. Hounslow, A.W. (1995). Water Quality Data: Analysis and
Interpretation. Lewis, New York.
9. Hearst, J.R., Nelson, P.H., and Paillet, F.L. (Eds.). (2001).
Well Logging for Physical Properties: A Handbook for
Geophysicists, Geologists, and Engineers, 2nd Edn. John
Wiley & Sons, New York.
10. Todd, D.K. (1980). Groundwater Hydrology, 2nd Edn. John
Wiley & Sons, New York.
11. Wilson, E.M. (1984). Engineering Hydrology, 3rd Edn.
Macmillan, London.
12. Deutsch, W.J. (1997). Groundwater Geochemistry: Funda-
mentals and Applications to Contamination. Lewis, Boca
Raton, FL.
13. Tebbutt, T.H.Y. (1995). Principles of Water Quality Control.
Reprint of 4th Edn. Butterworth Heinemann, Oxford.
THE HYDRONIUM ION
CARL W. DAVID
University of Connecticut
Storrs, Connecticut
Naked protons cannot exist in aqueous solution. Even
when one writes
HCl(g) → H+ (aq) + Cl− (aq) (1)
one knows that the H+ (aq) is a representation of a
significantly more complicated moiety (or, to be more
precise, set of moieties). The simplest representation of
H+ (aq) is the hydronium ion (sometimes also referred to
as the oxonium ion or the hydroxonium ion),
H3 O+ (aq), but this is just the first in a series the next
few members of which are H5 O2 + , H7 O3 + , H9 O4 + , and
in summary (H2 O)n H3 O+ , the hydrated hydronium ions
(Fig. 1).
The hydronium ion is isoelectronic with ammonia (NH3 )
and therefore, ought to be pyramidal with the oxygen at
the apex of the pyramid and the three protons disposed
below, with symmetry C3v . Like ammonia, it should invert,
with the O[or N] passing through the plane of the three
protons from one side to the other. In the planar transition
state, when all four nuclei are coplanar, the symmetry
should be D3h . Also, like ammonia, there should be a
barrier to this umbrella inversion; appropriate tunneling
has been observed spectroscopically (ν2 ). It is thought
that the O–H distance is about 0.96 Å, and the H–O–H
angle about 115◦ . The barrier to inversion, a saddle point
on the energy hypersurface, is thought to be between 1
and 5 kcal/mol (Fig. 2).
The simplest occurrence of the hydronium ion can be
seen in the gas phase water dimer, (H2 O)2 . The proton
transfer reaction inside the dimer would be
H2 O–H2 O(g) → H3 O+ (g) : OH− (g) (2)
forming an ion pair (Fig. 3).
+
H
O
H H
Figure 1. The hydronium cation.
 THE HYDRONIUM ION 483

H
H +
O H
H
O H − +
HO O
O H
H
H H
H H
Energy

O
H O

55.35 cm −1 H H
Figure 4. With three molecules of water, one can form a hydrated
+ hydronium cation and a hydroxide anion.
H + H

O H H O
Clearly, in liquid water, the hydroxide and hydronium
H H ions are solvated but not statically enough to allow
definition of a moiety.
Inversion coordinate
In aqueous media, for the analog of the proton transfer
Figure 2. Potential energy surface for the hydronium ion’s reaction, one could write
inversion coordinate.
2H2 O(l) → H3 O+ (aq) + OH− (aq) (3)

the prototype of all acid–base chemistry in an aque-

H
ous medium.
O H In aqueous acidic media, the hydronium ion is rarely
H
O H − written. Instead, the older notation is employed. For
O +
O instance, for acetic acid, CH3 COOH, one would write
H H
H H CH3 COOH(aq) ⇔ H+ (aq) + CH3 COO− (aq) (4)
Figure 3. Dimer water forming an ion pair through pro-
ton transfer. partial dissociation into a proton and a acetate anion,
rather than the more accurate

H2 O(l) + CH3 COOH(aq) ⇔ H3 O+ (aq) + CH3 COO− (aq)

In the dimer, one molecule of water is a proton donor to
the other (which is a proton acceptor). The former water
is a Bronsted acid, the latter a Bronsted base. On the line
of the hydrogen bond, HO–Hdonor ::::OH2 , the movement of
the donor proton from one oxygen to another is equivalent
to the proton transfer (above). The O–O distance in the
dimer is about 3 Å, and the OH distances are about 1 Å,
so the shift of 1 Å in a proton’s position (between two
oxygens) creates an OH− :::H3 O+ ion pair.
Hydrated versions of this ionization process illustrate
that the actual process in liquid water is not cleanly
definable. Figure 4 shows the results of three water
molecules involved in ionization, whereas Fig. 5 shows
the results with four water molecules.
(5)
The conductance properties of water are attributable in
part to the hydronium cation and its peculiar properties.
The Grotthuss mechanism for proton transfer in ice
employs hydronium ions hydrogen bound to adjacent water
molecules, in which, like the dimer, an excess proton
migrates (or tunnels) from one oxygen to an adjacent one,
moving only a tiny distance, whereupon the charge has
shifted enormously (one O–O distance).
A similar mechanism for OH− (aq) charge migration
most likely also obtains. Here a water donor adjacent
to the hydroxide anion shifts its donating proton to
the hydroxide, thereby itself becoming a hydroxide anion
(Fig. 6).

O H
H
O H −
O +
H H O H
H H H
O H
O O
H O
H H Figure 5. With four waters, hydrated hydroxide
H H
and hydrated hydronium ions are possible.
484 INFILTRATION AND SOIL MOISTURE PROCESSES

H
Ground
O Exfiltration
H O surface
H H
Infiltration
H
O
H
O
H
H
H O Redistribution
Plant
H uptake
Interflow

Water Capillary rise

table
Recharge
H −
H Ground water
H
O
O H

H O Figure 1. Unsaturated zone definition and active processes (1).

H
O
+
H
H CONTROLS ON INFILTRATION AND SOIL MOISTURE
H O

To characterize soil moisture and infiltration, the

H physical controls on these processes must be considered.
Figure 6. The Grotthuss mechanism for positive and negative
charge creation and migration in aqueous solution.
INFILTRATION AND SOIL MOISTURE
PROCESSES
PAUL R. HOUSER
(from The Handbook of Weather,
Climate, and Water:
Atmospheric Chemistry,
Hydrology, and Societal
Impacts, Wiley 2003)
Infiltration is the process of water entry from surface
sources such as rainfall, snowmelt, or irrigation into
the soil. The infiltration process is a component in
the overall unsaturated redistribution process (Fig. 1)
(1) that results in soil moisture availability for use
by vegetation transpiration, exfiltration (or evaporation)
processes, chemical transport, and groundwater recharge.
Soil moisture, in turn, controls the partitioning of
subsequent precipitation into infiltration and runoff, and
the partitioning of available energy between sensible and
latent heat flux.
Because of the importance of soil moisture on multiple
processes, its definition can be elusive (2); however, it is
most often described as moisture in the unsaturated sur-
face layers (first 1 to 2 m) of soil that can interact with
the atmosphere through evapotranspiration and precipi-
tation (3).
The primary soil controls will be considered in this
chapter; however, other factors such as soil chemistry,
thickness, soil layering or horizons, and preferential flow
paths, as well as vegetation cover, tillage, roughness,
topography, temperature, and rainfall intensity also exert
important controls (4).
A soil’s particle size distribution has a large impact
on its hydraulic properties. Soil particles less than 2 mm
in diameter are divided into three texture groups (sand,
silt, and clay) that help to classify broad soil types and
soil water responses (Fig. 2) (5). The type of clay and the
coarse material over 2 mm in diameter can also have a
significant impact on soil water properties. An overview of
methods for determining particle size properties is given
by Gee and Bauder (6).
Bulk density, ρb (M/L3 ) is the ratio of the weight of
dry solids to the bulk volume of the soil, and porosity, φ
(M3 /M3 ), is the total volume occupied by pores per unit
volume of soil:
Va + Vw ρb
φ= =1− (1)
Vs ρm
where Vs (L3 ) is the total volume of soil, Va (L3 ) is the
volume of air, Vw (L3 ) is the volume of water, and ρm
(ML−3 ) is the particle density (normally about 2.65 g/cm3 ).
The volumetric water content, or soil moisture, θ (L3
L−3 ) is the ratio of water volume to soil volume:
Vw Ww ρb
θ= = (2)
Vs Wd ρw
where Ww (M) is the weight of water, Wd (M) is the
weight of dry soil, and ρw (M/L3 ) is the density of water.
 INFILTRATION AND SOIL MOISTURE PROCESSES 485

100 0.60
10
90
20 Porosity, f
80
0.50
30
lay

Pe
70
tc

Clay

rc
igh

40

en
60
we

tb
0.40

yw
50
by

Proportion by volume
50

eig
nt

Silty
Field
rce

Sandy clay
60

ht
clay
capacity, qfc
Pe

40

sill
Silty Clay
Clay loam loam 70 0.30
30
Sandy clay loam
80
20
Loam Silt loam
Sandy Loam 90 0.20 Wilting
10 Loamy Silt point, qpwp
Sand Sand 0
10
100

90

80

70

60

50

40

30

20

10

0.10
Percent by weight sand Clay
Figure 2. Soil textural triangle describing the relationship Loam Heavy
between texture and particle size distribution (5). clay loam
0 Fine Clay loam
Sand sandy loam
Light clay loam
sandy loam Silt loam
“Absorbed” Water
Fine sand
Figure 4. Water holding properties of various soils (8).
“Capillary”
water
Suction

Clayey soil

Particles

Figure 3. Capillarity and adsorption combine to produce suction Sandy soil

(7).

Soil moisture can vary in both time and space, with a
theoretical range from 0 to ϕ, but for natural soils the
range is significantly reduced due to isolated pore space
and tightly held or ‘‘adsorbed’’ water (Fig. 3) (7). If a soil is
saturated, then allowed to drain until the remaining water
held by surface tension is in equilibrium with gravitational
forces, it is at field capacity, θf . Vegetation can remove
water from the soil until the permanent wilting point, θw ,
is reached. Therefore, the available water content for plant
use, θa = θf − θw . Typical ranges of porosity, field capacity,
and wilting point for different soils are given in Fig. 4 (8).
In unsaturated soils, water is held in the soil against
gravity by surface tension (Fig. 3). This tension, suction, or
Water content
Figure 5. Effect of texture on water retention characteristics (9).
matric potential, ψ (L), increases as the radii of curvature
of the meniscus or water content decreases (Fig. 5) (9).
Matric potential is expressed in reference to atmospheric
pressure, so for saturated soil ψ = 0 and for unsaturated
soil ψ < 0.
The hydraulic conductivity, K (L/T), is a measure of the
ability of the soil to transmit water that varies nonlinearly
over a large range depending on both soil properties and
water content (Fig. 6) (10). Many laboratory and field
hydraulic conductivity measurement methods exist for
486 INFILTRATION AND SOIL MOISTURE PROCESSES
10−3
10−4
Sand
Hydraulic conductivity, K, ms−1
10−5
10−6
10−7
10−8
10−9 0 0.1 0.2 0.3
Water content, q,
Figure 6. Effect of texture and soil moisture on hydraulic
conductivity (10).
use with various soils; see Bouwer and Jackson (11) or
Green et al. (12) for details.
Soil water content can significantly impact infiltra-
tion by (1) increasing the hydraulic conductivity, which
increases infiltration, and (2) reducing the surface tension
that draws moisture into the soil, which reduces infiltra-
tion. The net effect of these impacts depends on the water
content itself, the water input rate, and duration and the
distribution of hydraulic conductivity.
The water retention characteristic describes a soil’s
ability to store and release water and is defined by
the relationship between soil moisture and the matric
potential (Fig. 5). This is a power function relationship
that has been described by Brooks and Corey (13)
and Van Genuchten, (14) among others. The water
tension characteristic is usually measured in air pressure
chambers where the water content of a soil sample can be
monitored over a wide pressure range (15).
The water retention relationship may actually change
between drying and wetting due to the entrapment of air
in soil pores (Fig. 7) (16). For practical applications, this
effect, called hysteresis, is usually neglected (17).
PRINCIPLES OF SOIL WATER MOVEMENT
Through experiments on saturated water flow through
sand beds, Darcy (18) found that the rate of flow, Q
(L3 /T), through a cross-sectional area A (L2 ), is directly
proportional to head loss (e.g., water elevation difference),
H (L), and inversely to the flow path length, l (L):
H
Q = KA
l
Loam
0.4 0.5 Figure 7. Changes in water retention characteristics between
sorption and desorption (16).
m3 m−3
Combining Darcy’s law with the law of conservation of
mass results in a description of unsaturated flow called
Richards equation (19):


∂θ ∂ K ∂θ ∂K
= − (4)
∂t ∂z C ∂z ∂z
where C = −∂θ/∂ψ is the water content change in a unit
soil volume per unit matric potential, ψ change. The
Richards equation is the basis for most simulations of
infiltration and redistribution of water in unsaturated
soil. Using some approximations, analytical solutions of
the Richards equation are available (20,21) that show good
agreement with observations (22). The Richards equation
is based on saturated flow theory, and does not account
for all of the processes active in natural systems, so it may
not always perform well (23).
INFILTRATION ESTIMATION
Some basic principles that govern the movement of water
into the soil can be used to predict infiltration. The
infiltration capacity, f (L), is the maximum rate that a
soil in a given condition can absorb water and generally
decreases as soil moisture increases. If the rainfall rate
is less than the infiltration capacity, then infiltration
proceeds at the capacity rate. However, if the rainfall rate
exceeds the infiltration capacity, then infiltration proceeds
at the capacity rate, and the excess rainfall ponds on the
surface or runs off. As the time from the onset of rainfall
increases, infiltration rates decrease due to soil moisture
increases, raindrop impact, and the clogging of soil pores,
until a steady-state infiltration rate is reached (Fig. 8) (24).
Existing infiltration models use empirical, approximate, or
physical approaches to predict infiltration.(25)
Empirical. Empirical infiltration models generally uti-
(3) lize a mathematical function whose shape as a function of
 INFILTRATION AND SOIL MOISTURE PROCESSES 487

5 where S is a parameter called sorptivity, t is time

from ponding, and A is a constant that depends on soil
Rate of rainfall, infiltration, and runoff (cm/hr)

properties. In this model, the infiltration rate approaches a

Rainfall
4 constant equal to the hydraulic conductivity at the surface
Runoff water content, and the wetting front advances without
changing its shape and approaches a constant velocity.
3 Physical. Recent advances in numerical methods and
computing has facilitated the practical application of the
Richards equation to realistic flow problems. Such pack-
2 ages can simulate water infiltration and redistribution
using the Richards equation and including precipitation,
runoff, drainage, evaporation, and transpiration processes
1 (30).
Infiltration
INFILTRATION MEASUREMENT

0 Infiltration rates can be measured at a point using

20
Minutes from onset of rainfall
Figure 8. Idealized relationship between rainfall, infiltration,
and runoff rates (24).
time, t, matches observations and then attempts a physical
explanation of the process.
Kostiakov (26) proposed the simple infiltration rate,
f (L/T) model:
f = αγ tα−1 (5)
where α and γ are constants that have no particular
meaning and must be evaluated by fitting the model to
experimental data.
Horton’s (27) infiltration model has been widely used
in hydrologic simulation. It relates infiltration capacity to
initial infiltration rate, and f0 , the constant infiltration
rate at large times, fc :
where β is a soil parameter describing the rate of decrease
of infiltration.
Approximate. Analysis approximations to the Richards
equation are possible if several simplifying assumptions
are made. Most approximate infiltration models treat the
soil as a semi-infinite medium, with the soil saturating
above a wetting front.
Green and Ampt (28) assumed in a soil with constant
hydraulic properties, the matric potential at the moving
wetting front is constant, leading to a discontinuous
change in soil moisture at the wetting front:
 
(ϕ − θi )Sf
f =K 1+ (7)
F
where Sf (L) is the effective suction at the wetting front, θi
is the initial water content, and F (L) is the accumulated
infiltration.
Phillip (29) proposed that the first two terms in a series
of powers of t1/2 could be used to approximate infiltration:
40 60 80 100 120 140
a variety of methods described here, each appropriate
for certain conditions. However, because of the large
temporal and spatial variability of infiltration processes,
catchment average infiltration rates may be desired, which
can be obtained through the water balance analysis of
rainfall–runoff observations (31).
Ring Infiltrometer. This simple method is most appro-
priate for flood irrigation or pond seepage infiltration. A
cylindrical metal ring is sealed at the surface and flooded.
Intake measurements are recorded until steady-state con-
ditions are reached (32). If the effects of lateral flow
are significant, then a double-ring infiltrometer can be
used. Due to ponding conditions within the ring, observed
infiltration rates are often higher than under natural
conditions (33).
Sprinkler Infiltrometer. This method is appropriate for
quantifying infiltration from rainfall. Artificial rainfall
simulators are used to deliver a specified rainfall rate
f = fc + (f0 − fc )e−βt (6) to a well-defined plot. Runoff from the plot is measured,
allowing computation of the infiltration rate (34,35).
Tension Infiltrometer. The tension or disk infiltrometer
employs a soil contact plate and a water column that
is used to control the matric potential of the infiltrating
water. By varying the tension, the effect of different size
macropores can be determined (36,37).
Furrow Infiltrometer. This method is useful if informa-
tion on infiltration of flowing water in irrigation furrows
is desired. Either the water added to a small section of
blocked off furrow to maintain a constant depth or the
inflow—outflow of a furrow segment can be monitored
to determine the infiltration characteristics of the system
(38).
SOIL MOISTURE MEASUREMENT
Soil water content can be determined directly using
gravimetric techniques or indirectly by inferring it from a
property of the soil (39,40).
Gravimetric. The oven-drying soil moisture measure-
f = 12 St1/2 + A (8) ment technique is the standard for calibration of all
488 INFILTRATION AND SOIL MOISTURE PROCESSES
other methods but is time consuming and destructive.
The method involves obtaining a wet soil sample weight,
drying the sample at 105 ◦ C for 24 h, then obtaining the
dry sample weight [see Eq. (2)].
Neutron Thermalization. High-energy neutrons are
emitted by a radioactive source into the soil and are
preferentially slowed by hydrogen atoms. The number
of slow neutrons returning to the detector are a measure
of soil moisture.
Gamma Attenuation. The attenuation in soil of gamma
rays emitted from caesium-137 is directly related to
soil density. If the soil’s bulk density is assumed to be
constant, then changes in attenuation reflect changes in
soil moisture (41).
Time-Domain Reflectometry (TDR). TDR measures the
soil’s dielectric constant, which is directly related to soil
moisture, by measuring the transmit time of a voltage
pulse applied to a soil probe.
Tensiometric Techniques. This method measures the
capillary or moisture potential through a liquid-filled
porous cup connected to a vacuum gage. Conversion to
soil moisture requires knowledge of the water retention
characteristic.
Resistance. The electrical resistance or conductivity of
a porous block (nylon, fiberglass, or gypsum) imbedded
in the soil depends primarily on the water content of
the block. However, because of salinity and temperature
sensitivity, measurements of these sensors are of limited
accuracy (42).
Heat Dissipation. Changes in the thermal conductivity
of a porous block imbedded in the soil depend primarily
on the water content of the block. The dissipation of a
heat pulse applied to the block can be monitored using
thermistors, then the soil water content can be determined
from calibration information.
Remote Sensing. Soil moisture can be remotely sensed
with just about any frequency where there is little
atmospheric absorption (43). But, it is generally accepted
that long wavelength, passive microwave sensors have
the best chance of obtaining soil moisture measurements
that contain little error introduced by vegetation and
roughness and offer great potential to remotely sense soil
moisture content with depth due to differential microwave
absorption with varying dielectric constant (44).
SPATIAL AND TEMPORAL VARIABILITY
Natural soils exhibit considerable spatial heterogeneity
in both the horizontal and vertical directions, and at
all distance scales from the pore to the continent to
a degree that it is difficult to capture this variability
in routine measurements [45,46]. This large variations
in soil properties, infiltration, and soil moisture over
relatively small areas makes it difficult to transfer
the understanding of processes developed at a point to
catchment scales. Many hydrological models assume that
a single spatially representative average soil property can
be used to characterize catchment (or even larger) scale
processes. It is clear from the nonlinear character of soil
water processes that catchment average infiltration cannot
be computed based on catchment average soil properties. It
is also clear that the physical meaning of a soil property,
say porosity, is relative to the volume over which it is
averaged [47]. However, there is a need to understand
and reduce this complexity for the purposes of prediction
and management. Several approaches, including dividing
the catchment into hydrologically similar subareas [48],
various statistical approaches [49], and scaling and
similarity theory [50,51] have made headway toward an
understanding of infiltration and soil moisture spatial
variability, but are not being widely used in practical
applications.
One of the most important recent findings in this regard
is the scale invariance of soil water movement(i.e., the
Richards equation) remain uniform across spatial scales
[52]. This new understanding of the underlying symmetry
of the Richards equation may help to facilitate a workable
scale invariant analytical soil water dynamical model.
Finally, there is acontinuing need for the observation of
soil properties, soil moisture, and infiltration processes at
multiple scales to facilitate understanding and prediction
of these complex and socially significant processes. It is
likely that remote sensing of soil moisture and other land
surface factors will be instrumental in this respect.
BIBLIOGRAPHY
1. Dingman, S.L. (1994). Physical Hydrology. Prentice-Hall,
Englewood Cliffs, NJ, p. 211.
2. Lawford, R.G. (1992). An Overview of soil moisture and its
role in the climate system. In: F.J. Eley, R. Granger and
L. Martin (Eds.). Soil Moisture; Modeling and Monitoring
for Regional Planning, National Hydrology research Centre
Symposium No.9 Poceedings, pp. 1–12.
3. Schmugge, T., Jackson, T.J., and McKim, H.L. (1979). Survey
of in-situ and remote sensing methods for soil moisture
determination. In: Satellite Hydrology. American Water
Resources Association.
4. Rawls, W.J., Ahuja, L.R., Brakensiek, D.L., and Shirmoham-
madi, A. (1993). Infiltration and soil water movement. In:
Handbook of Hydrology. D.R. Maidment (Ed.). McGraw-Hill,
New York, pp. 5.1–5.51.
5. Hillel, D. (1982). Introduction to Soil Physics. Academic, New
York, p. 29.
6. Gee, G.W., and Bauder, J.W. (1986). Particle size analysis. In:
Methods of Soil Analysis, Part I—Physical and Mineralogical
Methods, 2nd Edn. A. Klute (Ed.). American Society of
Agronomy Monograph 9, pp. 383–411.
7. Hillel, D. (1980). Fundamentals of Soil Physics. Academic,
New York, p. 143.
8. Dunne, T. and Leopold, L. (1978). Water in Environmental
Planning. W.H. Freeman, New York, p. 175.
9. Hillel, D. (1980). Fundamentals of Soil Physics. Academic,
New York, p. 150.
10. Marshall, T.J., and Holmes, J.W. (1988). Soil Physics, 2nd
Edn. Cambridge University Press, p. 87.
11. Bouwer, H. and Jackson, R.D. (1974). Determining soil prop-
erties. In: Drainage for Agriculture. J. van Schilfgaard (Ed.).
American Society of Agronomy, Madison, WI, pp. 611–672.

12. Green, R.E., Ahuja, L.R., and Chong, S.K. (1986). Hydraulic
conductivity, diffusivity, and sorptivity of unsaturated
soils—field methods. In: Methods of Soil Analysis, Part
I—Physical and Mineralogical Methods, 2nd Edn. A. Klute
(Ed.). American Society of Agronomy Monograph 9,
pp. 771–798.
13. Brooks, R.H. and Corey, A.T. (1964). Hydraulic Properties of
Porous Media, Hydrology Paper 3, Colorado State University,
Fort Collins, CO.
14. Van Genuchten, M.Th. (1980). A closed-form equation for
predicting the hydraulic conductivity of unsaturated soils.
Soil Sci. Soc. Am. J. 44: 892–898.
15. Cassel, D.K. and Klute, A. (1986). Water potential: Tensiom-
etry. In: Methods of Soil Analysis, Part I—Physical and
Mineralogical Methods, 2nd Edn. A. Klute (Ed.). American
Society of Agronomy Monograph 9, pp. 563–596.
16. Hillel, D. (1980). Fundamentals of Soil Physics. Academic,
New York, p. 153.
17. Rawls, W.J., Ahuja, L.R., Brakensiek, D.L., and
Shirmohammadi, A. (1993). Infiltration and Soil Water
Movement. In: Handbook of Hydrology. D.R. Maidment (Ed.).
McGraw-Hill, New York, pp. 5.5–5.6.
18. Darcy, H. (1856). Les fontaines publiques de la ville de Dijon.
Dalmont, Paris.
19. Richards, L.A. (1931). Capillary conduction of liquids in
porous mediums. Physics 1: 318–333.
20. Kuhnel, V., Dooge, V.C.I., Sander, G.C., and O’Kane, J.P.J.
(1990). Duration of atmosphere-controlled and soil-controlled
phases of infiltration for constant rainfall at a soil surface.
Ann. Geophys. 8: 11–20.
21. Sposito, G. Recent advances associated with soil water in the
unsaturated zone. In: Reviews of Geophysics, Supplement,
U.S.National Report to International Union of Geodesy and
Geophysics 1991–1994, pp. 1059–1065.
22. Whisler, F.D. and Bouwer, H. (1970). A comparison of
methods for calculating vertical drainage and infiltration in
soils. J. Hydrol. 10: 1–19.
23. Nielsen, D.R., Biggar, J.W., and Davidson, J.M. (1962). Exper-
imental consideration of diffusion analysis in unsaturated
flow problems. Soil Soc. Am. Proc. 26: 107–111.
24. Dunne, T. and Leopold, L. (1978). Water in Environmental
Planning, W.H. Freeman, New York, p. 169.
25. Rawls, W.J., Ahuja, L.R., Brakensiek, D.L., and
Shirmohammadi, A. (1993). Infiltration and soil water
movement. In: Handbook of Hydrology. D.R. Maidment (Ed.).
McGraw-Hill, New York, pp. 5.21–5.23.
26. Kostiakov, A.N. (1932). On the dynamics of the coefficient of
water-percolation in soils and on the necessity for studying
it from a dynamic point of view for purposes of amelioration.
Trans. Sixth Comm. Intern. Soil. Sci. Soc. Russian, part A,
17–21.
27. Horton, R.E. (1940). An approach toward a physical inter-
pretation of infiltration-capacity. Soil Sci. Soc. Am. J. 5:
399–417.
28. Green, W.H., and Ampt, G.A. (1911). Studies on soil physics:
1. Flow of air and water through soils. J. Agric. Sci. 4: 1–24.
29. Phillip, J.R. (1957). The theory of infiltration: 1. The
infiltration equation and its solution. Soil Sci. 83: 345–357.
30. Ross, O.J. (1990). Efficient numerical methods for infiltration
using Richards equation. Water Resour. Res. 26: 279–290.
31. Soil Conservation Service, Hydrology. (1972). In: SCS
National Engineering Handbook. U.S. Department of Agri-
culture, Washington, DC, Sec. 4.
INFILTRATION AND SOIL MOISTURE PROCESSES 489
32. Bouwer, H. (1986). Intake rate: Cylinder infiltrometer. In:
Methods of Soil Analysis, Part I—Physical and Mineralogical
Methods, 2nd Edn. A. Klute (Ed.). American Society of
Agronomy Monograph 9, pp. 825–844.
33. Lukens, R.P. (Ed.). (1981). Annual Book of ASTM Standards,
Part 19: Soil and Rock, Building Stones, pp. 509–514.
34. Agassi, M., Shainberg, I., and Morin, J. (1988). Effects on
seal properties of changes on drop energy and water salinity
during a continuous rainstorm. Aust. J. Soil Res. 26: 1–10.
35. Peterson, A. and Bubenzer, G. (1986). Intake rate sprinkler
infiltrometer. In: Methods of Soil Analysis, Part I—Physical
and Mineralogical Methods, 2nd Edn. A. Klute (Ed.).
American Society of Agronomy Monograph 9, pp. 845–870.
36. Ankeny, M.D., Kaspar, T.C., and Horton, R. (1988). Design
for an automated tension infiltrometer. Soil Sci. Soc. Am. J.
52: 893–896.
37. Perrouix, K.M. and White, I. (1988). Designs of disc
permeameters. Soil Sci. Soc. Am. J. 52: 1205–1215.
38. Kincaid, D.C. (1986). Intake rate: Border and furrow. In:
Methods of Soil Analysis, Part I—Physical and mineralogical
Methods, 2nd Edn. A. Klute (Ed.). American Society of
Agronomy Monograph 9, pp. 871–887.
39. Gardner, W.H. (1986). Water content. In: Methods of Soil
Analysis, Part I—Physical and Mineralogical Methods,
2nd Edn. A. Klute (Ed.). American Society of Agronomy
Monograph 9, pp. 493–544.
40. Schmugge, T.J., Jackson, T.J., and McKim, H.L. (1980).
Survey of methods for soil moisture determination. Water
Resour. Res. 16(6): 961–979.
41. Jury, W.A., Gardner, W.R., and Gardner, W.H. (1991). Soil
Physics, 5th Edn. Wiley, pp. 45–47.
42. Hillel, D. (1982). Introduction to Soil Physics. Academic, New
York, 1982, pp. 57–89.
43. Schmugge, T.J. (1985). Remote sensing of soil moisture. In:
Hydrological Forecasting. John Wiley & Sons, New York.
44. Ulaby, F.T., Bradley, G.A., and Dobson, M.C. (1979). Potential
application of satellite radar to monitor soil moisture. In:
Satellite Hydrology, American Water Resources Association,
pp. 363–370.
45. Jury, W.A., Gardner, W.R., and Gardner, W.H. (1991). Soil
Physics, 5th Edn. John Wiley & Sons, New York, pp. 268–
293.
46. Dingman, S.L. (1994). Physical Hydrology. Prentice-Hall,
Englewood Cliffs, NJ, pp. 247–250.
47. Bear, J. (1972). Dynamics of Fluids in Porous Media. Elsevier,
New York.
48. Springer, E.P. and Cundy, T.W. (1987). Field-scale evaluation
of infiltration parameters from soil texture for hydrologic
analysis. Water Resour. Res. 23: 325–334.
49. Berndtsson, R., and Larson, M. (1987). Spatial variability
of infiltration in a semi-arid environment. J. Hydrol. 90:
117–133.
50. Sharma, M.L., Gander, G.A., and Hunt, G.C. (1980). Spatial
variability of infiltration in a watershed. J. Hydrol. 45:
101–122.
51. Wood, E.F., Sivaplan, M., and Beven, K. (1990). Similarity
and scale in catchment storm response. Rev. Geophys. 28:
1–18.
52. Sposito, G. (1995). Recent advances associated with soil
water in the unsaturated zone. In: Reviews of Geophysics,
Supplement, U.S. National Report to International Union of
Geodesy and Geophysics 1991–1994, pp. 1059–1065.
490 ION EXCHANGE AND INORGANIC ADSORPTION

ION EXCHANGE AND INORGANIC Table 1. Distinction Between Physical Adsorption and
Chemisorption (1)
ADSORPTION
Physical Adsorption Chemical Adsorption
K. KADIRVELU
Low heat of adsorption (<2 High heat of adsorption (> or 3
JYOTSNA GOEL
or 3 times latent heat of times latent heat of
Centre for Fire, Explosives, and evaporation.) evaporation.)
Environmental Safety Nonspecific. Highly specific.
Timarpur, India Monolayer or multilayer. Monolayer only.
No dissociation of adsorbed May involve dissociation. Possible
species. Only significant at over a wide range of
The presence of inorganic ions in water is of major relatively low temperature.
concern for water quality because they can interfere with temperatures.
legitimate use of water for various applications. Inorganic Rapid, nonactivated, Activated, may be slow and
ions, particularly heavy metals, are seriously questionable reversible. irreversible.
because of their toxicity and threat to human life and the No electron transfers, Electron transfer leading to bond
although polarization of formation between sorbate and
environment.
sorbate may occur. surface.
Ion exchange and adsorption share so many common
features in regard to removal of inorganic ions from water
that they can be grouped together as unified processes
for inorganic ion treatment. These processes involve the sulfonic acid groups; less commonly, groups are carboxylic,
increase in concentration of a particular component at the phosphonic, and phosphinic, and so some resins involve
surface or interface between two phases. quaternary ammonium groups (strongly basic) and amino
The objective of this article is to enumerate application groups (weakly basic).
of adsorption and ion exchange processes for inorganic Most ion exchangers in large-scale use are based on
ion removal from water. This article presents a review synthetic resins—either preformed and then chemically
of potential adsorbents and ion exchangers for inorganic reacted as styrene or formed from active monomers.
ion removal. It introduces important topics regarding Natural zeolites were the first ion exchangers, and natural
inorganic ion removal by ion exchange and adsorption and synthetic zeolites are in use today.
processes. Factors effecting metal removal are also Ion exchange may be thought of as a reversible reaction
presented and discussed in sections of the article. Research and of chemically equivalent quantities. A common
results from batch studies are presented to illustrate basic example of ion exchange is the familiar water-softening
concepts. Accordingly, emphasis is placed on providing reaction
a review of adsorbents and ion exchangers available for Ca++ + 2NaR ⇔ CaR2 + 2Na+ (1)
inorganic ion removal.
where R represents a stationary univalent anionic site in
ADSORPTION AND ION EXCHANGE: IN GENERAL the electrolyte network of the exchanger phase.

Adsorption involves, in general, the accumulation (or
depletion) of solute molecules at an interface (including
gas-liquid interfaces, as in foam fractionation, and liquid-
liquid interfaces, as in detergency). Adsorbent surfaces
are often physically and/or chemically heterogeneous, and
bonding energies may very widely from one site to another.
In discussing the fundamentals of adsorption, it is useful
to distinguish between physical adsorption, involving only
relatively weak intermolecular forces, and chemisorption,
which involves, essentially, the formation of a chemical
bond between the sorbate molecule and the surface of
the adsorbent. The general features, which distinguish
physical adsorption and chemisorption, are presented in
Table 1.
Ion exchange is an effective means of removing
heavy metals from wastewaters. It is a reversible
chemical reaction, where the removal of heavy metals
is accomplished by the exchange of ions on the resin for
those in wastewater. In ion exchange, ions of positive
charges (cations) and negative charges (anions) from
the fluid (from aqueous solution) replace dissimilar
ions of the same charge in solids. The ion exchanger
contains permanently found functional groups of opposite
charge-type. Cation exchangers generally contain bound
FACTORS AFFECTING INORGANIC IONS REMOVAL BY
ADSORPTION
Factors affecting inorganic ion removal by adsorbent
include (a) nature of the adsorbent (b) solution pH,
(c) presence of foreign ions, and (d) temperature.
Nature of the Adsorbent
Surface Area and Pore Structure. Surface area with
active functional groups is one of the principle charac-
teristics affecting the adsorption capacity of an adsorbent.
The adsorption capacity of a solid adsorbent is generally
proportional to the specific surface area of the mass adsor-
bent; that is, the adsorption of certain solutes increases
with an increasing surface area. Table 2 shows the physi-
cal properties of a few adsorbents that have been already
employed for inorganic removal.
However, the specific surface area alone is frequently
inadequate to explain the adsorption capacity of porous
solids, such as activated carbons, for different solutes. In
view of this, pore size distribution in activated carbon
indicates that microspores contribute a major portion of
the specific surface area.
 ION EXCHANGE AND INORGANIC ADSORPTION 491

Table 2. Properties of Adsorbents Employed for Inorganic Ion Removal (2–6)

Calgon Coir-Pith AC from AC Carbon
Property CAL Carbon Arudo Donax Cloth Aerogel

Surface area (m2 /g) 1100–1300 575 1142 1689 700

Porosity (%) 65.0 93.11 76.0 66.5 36.2
Micropore Volume (g/ml) 0.12 0.0913 0.47 0.964 0.10
Bulk density (g/ml) 0.56 0.12 0.5 0.13 0.644
Moisture content (%) 2.0 5.88 9.3 2.0 18.0
Ash content (%) 7.0 7.87 21.56 13.0 3.5

Depending on their preparation, active carbons also
contain larger pores, known as macro and mesopores in
the classification proposed by Dubinin (7) and now adopted
by the International Union of Pure and Applied Chemistry
(IUPAC) (8). The definition of the different types of pores
is based on their width w, which represents the distance
between the walls of a slit-shaped pore or the radius
of a cylindrical pore. One distinguishes micropores, for
which w does not exceed 2 nm; macropores, for which w
exceeds about 50 nm; and mesopores of intermediate size
(2.0–50 nm).
A linear relationship has also been demonstrated
between the specific surface area and sorbate molecular
size (9).
Chemistry of Adsorbent Surface. The nature and the
presence of specific functional groups on the surface
of the adsorbent impart certain characteristics that
affect the adsorption process. It is now known that
surface functional groups are formed during the carbon
(adsorbent) activation process.
Attempts have been made to identify and estimate the
surface oxygen chemical structures (functional groups)
using several physical and chemical techniques, which
include neutralization of acid and base (known as Boehm’s
method); desorption of the oxide layer; potentiometer,
thermometric, and radiometric titrations; direct analysis
of oxide layers by specific chemical reaction; Infrared
(IR) spectroscopy; and X-ray photoelectron spectroscopy
(XPS). The surface morphology and metal ion distribution
adsorbent samples can be visualized via scanning electron
microscopy (SEM). Figure 1 shows SEM micrograph for
carbon aerogel equilibrated with Pb (II). SEM of activated
carbon surfaces before and after the adsorbent was
equilibrated with the metal ion solution clearly shows
the presence of Pb (II) ions.
As a result of these investigations, various researchers
have postulated the existence of functional groups,
such as carboxyic, phenolic, lactonic, aldehydes, ketones,
quinones, hydroquinone, anhydrides, and ethereal struc-
tures. Also, analysis of physicochemical properties of
adsorbent ions provides a deep insight into adsorption
mechanism involved during the adsorption process.
Solution pH
Extensive research investigations have been reported for
adsorption of inorganic ions at different solution pH
values (9–11). The dependence of adsorbent’s adsorption
capacity for inorganic ion removal from aqueous solution
pH has been attributed to changes in surface chemical
characteristics of adsorbent and species distribution
of the inorganic ions varying with solution pH. In
general, the cationic species removal will increase with
increasing solution pH (Fig. 2), whereas in the case
of inorganic ions present as anionic species, removal
generally increases with decreasing pH, which is mainly
because of electrostatic repulsion, which can be understood
by removal of Cr (VI) by various adsorbents at lower
pH range.
Specifically, pH affects the status of the outer hydration
sheaths of the metal ion, metal speciation of inorganic ion
in aqueous medium, complexation and solubility, and the
electrochemical behavior of the carbon surface.

(a) (b)

Figure 1. (a) SEM of carbon aerogel; (b) SEM of carbon aerogel loaded with lead.
492 ION EXCHANGE AND INORGANIC ADSORPTION

98

96

% Removal of Pb (II)
94

92

90

91.8 xxx xxx

88 xxx xxx
10 0.1
8
6 0.08
Figure 2. Plot for percent removal of Pb (II) versus 4 Adsorbent dosage (g)
pH 2 0.06
pH and adsorbent dosage using carbon aerogel as
0 0.04
adsorbent.

Figure 3 depicts possible inorganic ion removal mecha-
nisms for carbon aerogel in relationship to pH, at which
solution precipitation begins for Pb (II), Hg (II), and Ni
(II). Figure 3 shows clearly that at any pH, the removal
of metal ions is greater by adsorption in the presence of
carbon aerogel than by precipitation in the absence of
carbon. A rapid increase in percentage removal of metal
ions is observed after pH 5.4 and 7.5 for Pb (II) and Ni
(II), respectively, because of precipitation in the absence
of carbon aerogel.
The adsorption of metal ions depends on both the nature
of the adsorbent surface and the species distribution of the
metal ions in the aqueous solution, which is, in turn,
governed by solution pH. Removal noticed in a pH range
of 3–5 for Pb (II) and Hg (II), because of their partial
hydrolysis resulting in the formation of hydrolyzed metal
ion species [PbOH]+ , [HgOH]+ Pb (OH)2 , and Hg (OH)2 .
In the case of Ni (II), the maximum adsorption range
was shifted to 4–7, which may be because of formation of
[NiOH]+ and Ni (OH)2 at these pH (12). Low solubilities
of hydrolysis metal ion species may be another reason
for maximum adsorption. At pH values below pHprec ,
120
100
80
% Removal
adsorption was hypothesized to be the primary removal
mechanism. Also, it has been reported that metal removal
by activated carbon was found to be inversely proportional
to the solubility of the metal (13).
Namasivayam and Kadirvelu (14) reported an ion
exchange phenomenon for Hg (II) removal by the formation
of the following surface complexes when Hg (II) is present
in the solution with the functional groups present on GAC
(ACG-50).
2(CX OH+ ) + Hg2+ −−−→ (CX O)2 Hg2+ + 2H+ (2)
According to the surface complexation theory, Hg
(II) removal increases as pH function increases can
be explained on the basis of a decrease in adsorption
competition between proton and metal ion species for
the adsorbent surface sites by the decrease in positive
surface charge, which results in a lesser columbic repulsion
forward of the Hg (II) ion.
Presence of Foreign Ions
The presence of complexing ligands and competing
adsorbates can alter metal removal from that observed
in the mono-component system. The degree of competition
is dependent on the type and concentration of competing
ions, number of surface sites, and affinity of the surface for
adsorbate and the ionic property of adsorbate. Presence
of other cocations can influence the adsorption capacity of

Pb(II) Adsorption adsorbent with respect to one adsorbate. This phenomenon

60 Hg(II) Adsorption is illustrated in Fig. 4, where the adsorption equilibrium
Ni(II) Adsorption curve for Hg (II) removal by GAC is considerably effected
40 Pb(II) Precipitation by the presence of other metal ions like Cu (II) and
Hg(II) Precipitation Pb (II) (15,16). Similar results have been reported for
20 Ni(II) Precipitation the effect of EDTA on lead removal using GAC column
0 studies (17).
0 1 2 3 4 5 6 7 8 9 10 11 12
Initial pH
Temperature Effect
Figure 3. Effect of solution pH on the removal of lead, mercury, As the process of adsorption is spontaneous, it is
and nickel using carbon aerogel. accompanied by a decrease in the system’s free energy. A
 ION EXCHANGE AND INORGANIC ADSORPTION 493

20 Hg(II) Table 3. Thermodynamic Parameters for the Mercury

Hg(II) + Cu(II) Adsorption on Carbon Aerogel
Hg(II) + Pb(II) Temperature G0 H 0 S0
15
Hg(II) + Pb(II) + Cu(II) (K) % Removal (J mol−1 ) (Jmol−1 ) (Jmol−1 K −1 )
qe (mg g−1)

293 60 −987.714 +25708.82 +91.11444

10
303 68 −1898.86 +18990.88 +68.94302
313 73 −2588.29 +32921.47 +113.4497
5 323 80 −3722.79 +46023.63 +154.0137
333 87 −5262.92 +28133.9 +100.2908
343 90 −6265.83 — —
0
0 10 20 30 40 50 60 70
Ce (mg l−1) Raji et al. (20) have also presented alike thermodynamics
Figure 4. Equilibrium curve for Hg (II) ion in binary and tertiary for the uptake of copper and mercury, respectively.
system with Cu (II) and Pb (II) ions using GAC as adsorbent.
ADSORPTION MECHANISM—ION EXCHANGE PROCESS

decrease in entropy always exists because of loss of degrees With some of the adsorbents, like coir pith carbon,
of freedom of the solute in passing from the dissolved state adsorption mechanisms have been explained based on
to the adsorbate state. It follows from the relationship the ion exchange model (12). A pure carbon surface is
considered to be nonpolar, but in actual practice, some
G = H − TS (3) carbon–oxygen complexes Cx O, COx and Cx O2 are usually
present (21–23). The surface oxygen complexes hydrolyze
that the adsorption process must always be exothermic; water molecules as shown below:
that is H must have negative value regardless of the
nature of interaction forces. An increase in temperature, Cx O + 2H2 O −−−→ Cx OH2 2+ + 2OH− (8)
therefore, will result in a reduction of the equilibrium
adsorption capacity, whereas lower temperature will favor COx + xH2 O −−−→ C(OH)x + xOH− (9)
an increased capacity. 2 2+ −
Cx O + H2 O −−−→ Cx O + 2OH (10)
As reported by J. Goel et al. (6), according to the
adsorption isotherm, the amount of Hg (II) by carbon As the carbon is prepared with H2 SO4 , (NH4 )2 S2 O8 , and
aerogel increases with increasing temperature. Values of NaHCO3 , groups such as Cx ONa+ , Cx ONa2 2+ , Cx SO3 Na,
the thermodynamic parameters, such as G0 , H 0 , and and Cx ONH4 are also present. Na+ in the above groups
S0 , describing mercury uptake by carbon aerogel were also exchange with H+ in the medium as follows:
calculated using the following thermodynamic equations:
Cx ONa+ + H+ −−−→ Cx OH+ + Na+ (11)
G = −RT. ln K
0
(4)
Cx ONa+ + 2 H+ −−−→ Cx OH2 +2 + Na+ (12)
H 0 = [R × T1 × T2 /(T2 − T1 )] × Ln (K2 /K1 ) (5) 2+ + +2 +
Cx ONa2 + H −−−→ Cx OH2 + 2 Na (13)
S0 = (H0 − G0 )/T (6) + +
CxSO3 Na + H −−−→ Cx SO3 H + Na (14)
where R is the gas constant and K, K1 , and K2 are
Excess of Na+ was introduced into carbon where it was
the equilibrium constant at temperature T, T1 , and T2 ,
washed with NaHCO3 to neutralize any free sulfuric acid
respectively. The equilibrium constants was calculated
(H2 SO4 ) during the preparation of carbon. When metal
from:
ion is present in solution, its adsorption will free some
K = Ceq,s /Ceq,1 (7)
H+ and the pH increase will be lower than in the blank.
At the same time, Na+ will also be released according to
where Ceq,s and Ceq,1 were the equilibrium concentration
reactions as given below:
of Hg (II) in solution and on the adsorbent, respectively.
Values found for G0 , H 0 , and S0 , as presented in 2Cx OH+ + M2+ −−−→ (Cx O)2 M2+ + 2H+ (15)
Table 3, are indicative of the spontaneous nature of the
2+ 2+ 2 +
uptake process. CX OH2 +M −−−→ CX OM + 2H (16)
The positive H 0 value confirms the endothermic + 2+ 2+ +
2Cx ONa + M −−−→ (Cx O)2 M + 2Na (17)
nature of the sorption process. As diffusion is an
2+ 2+ 2 +
endothermic process, it would be expected that increased CX ONa2 +M −−−→ CX OM + 2Na (18)
solution temperature would result in increased uptake of + 2+ + +
2Cx SO3 H + M −−−→ (Cx SO3 )2 M + 2H (19)
Hg (II) ions from aqueous solution.
2+ +
Anoop Krishnan et al. (18) has reported that the 2Cx SO3 Na + M −−−→ (Cx SO3 )M + 2Na (20)
endothermic nature of the adsorption process, where
uptake of mercury on the bagasse piths activated carbon, is which indicates that apart from exchange of H+ ions on the
favored by increase in temperature. Panayotova (19) and adsorbent with M2+ ions, significant amount of Na+ ions,
494 ION EXCHANGE AND INORGANIC ADSORPTION

which were present in the adsorbent, were also exchanged
with M2+ ions.
ION EXCHANGERS FOR INORGANIC ION REMOVAL
By ion exchange, undesirable ions are replaced by
others, which do not contribute to contamination of the
environment. The method is technologically simple and
enables efficient removal of even traces of impurities
from solutions. Examples of selective removal of heavy
metal ions by ion exchange are presented in Table 4. They
include removal of Pb (II), Hg (II), Cd (II) Ni (II), V
(IV, V), Cr (III, VI), Cu (II), and Zn (II) from water and
industrial wastewaters by various means of modern type
of ion exchangers.
By ion exchange, either all ions can be removed
from a solution or substances are separated. Therefore,
selective removal of ionic contamination and complete
deionization can be distinguished. The choice between
selective removal and complete deionization depends
mainly on the composition of solution and on the extent of
decontamination required.
Hg (II) ions sorption on the ion exchanger Imac TMR
(Styrenedivinylbenzene Copolymer) can be described by
means of the following reactions:
2+ +
2R–SH + Hg −−−→ R–S–Hg–S–R + 2H
+ +
RSH + HgCl −−−→ R–S–HgCl + H
where R is the styrenedivinylbenzene copolymer. After
Hg (II) ions sorption, the ion exchanger Imac TMR
can be regenerated by means of the concentrated
hydrochloric acids.
Metal ions and other cationic impurities are removed
on the cation exchanger, whereas chromates are removed
on the anion exchangers:
nRSO3 H + M+ −−−→ nRMn+ + nH+ (23)
2ROM + CrO4 2− −−−→ R2 CrO4 + 2OH− (24)
R2 CrO4 + CrO4 2− + H+ −−−→ R2 Cr2 O7 + OH− (25)
The NaOH solutions are applied for anion exchanger
regeneration. Further developments of conventional ion
exchangers, like polystyrenesulphonate cation exchangers
and strongly basic anion exchangers, were aimed at
large exchange capacity and better chemical, thermal,
and mechanical stability. Kinetic capability of ion
exchangers could be improved by means of macroporous
ion exchangers developed in the 1970s. The range of
application of the ion exchange method was remarkably
extended, first by invention and application of new organic
and inorganic ion exchangers.
ADSORBENTS FOR INORGANIC ION REMOVAL
(21)
Large numbers of adsorbents have been developed so
(22) far for the removal of inorganic ions, which primarily

Table 4. Selective Removal of Heavy Metal Ions by Ion Exchangers

Metal Ion Ion Exchangers Composition and Functional Groups

Pb (II) Leawtit TP 207 (24) Chelating ion exchanger with

functional iminodiacetate group
Ethanotricarboxylate, Gel type Copolymer Of Vinyl benzene
1,1-ddicarboxylate-2- Chloride, Styrene and Divinyl
etanophosphonate (25) Benzene
Polyacrylamidoglycol Resin (26) Amide, OH and carboxylic groups
Clinoptilolite (27) Aluminosilicate structure
Hg (II) Imac TMR Styrenedivinylbenzene Copolymer
Varian BTAM (Nike), Varian BTKM, Containing nitrogen and sulphur
Varian BSM, Ionac SR-3 (28–31) functional group (22),
Thiocarbamate group (23),
Mercapathile group (24), Thiourea
functional group (25)
Dowex A-1 (32) Iminodiacetate group
Cd (II) Dowex 50 W-X4 Polystyrenesulphonic cation exchanger
Diphonix (33) Sulphonic groups and diphosphonic
groups
KB-2 and KB-2-12P (34) Polyacrylate carboxylic cation
exchangers
Phillipsite and Chabazite (35) Natural zeolite from volcanic rocks of
Central Italy
Ni (II) and V Wofatit CA-20 (36) Macro porous carboxylic cation
(IV, V) exchanger
Chelex −100 (37) Iminodiacaetae ion exchanger with
Na/H/Ni in exchange system
Sodium titnate, Na4 Ti9 O20 (38) Layered material with exchangeable
Na ions in the titanium oxide layer.
Cr (III, VI) Lewatit MP-500A (39) Basic anion exchanger
Amberlite IR-120 (40) Polystyrenesulphone cation exchanger
 ION EXCHANGE AND INORGANIC ADSORPTION 495

Table 5. Summary Table of Maximum Reported Adsorption Capacities (mg/g)

Hg Pb
Sorbent Source Cd (II) Cr (III) Cr (VI) (II) (II)

Bark (41–44) 32 19.45 400 182

CPEI cotton (45) 1000
Chitin Cr, Hg, Pb (46) 558 92 100 796
Chitosan (47, 48) 16.50 57 1123 58
Clay (49, 50) 28 116
Dead biomass (51) 43.0
Leaf mould (46–52) 150 1587
Lignin Cd, Cr, Hg and Pb (46) 87 17 632 1587
Modified wool (53) 46.65
Moss (54, 55) 5.058 76 43.9 16.2 135
Peat (54, 55)
Sawdust (56) 16.05 230
Seaweed (57, 58) 215 344
Xanthate (59, 60) 33.27 19.67 1.149 18
Coir pith carbon (3, 61) 94.0 154.0 264.0
Fly ash (62) 198 .073 444.0
Red mud (63) 67.0 7.1 165.0
Peanut hull carbon (64) 89.0 220.0 210.0
Furnace Sludge (65) 7.4 68.0

include activated carbon prepared from a variety of
materials of biological origin, industrial wastes, dead
biomass, organic wastes, etc. Besides, several low-cost
adsorbents such as chitin, chitosin, clay, fly ash, peat,
moss, zeolites, xanthates, seaweed/algae/alginate, lignin,
and bark/tannin-rich material have also been explored
for heavy metal containing wastewaters (37–56). Table 5
presents some of the highest adsorption capacity reported
for various low-cost adsorbents.
Although a significant number of low-cost adsorbent
materials from sources, the commercial activated carbon
(CAC) is still being used intensively to date. Based
on its size and shape, activated carbon is classified as
follows: Powder (PAC), granular (GAC), Fibre (ACF),
and cloth (ACC). As a result of the different sources
of raw materials, the extent of chemical activation,
and the physicochemical characteristics, each type of
activated carbon has its specific application with respect
to physicochemical properties, as well as depending on
the inherent advantage and disadvantages in wastewater
treatment.
Acknowledgment
Authors are grateful to Sh. A. K. Kapoor, Director, Dr.
Chitra Rajagopal, Scientist ‘‘F’’, CFEES, DRDO, for granting
permission to write this article and their constant encouragement
and support.
BIBLIOGRAPHY
1. Ruthven, D.M. (1984). John Wiley & Sons, New York.
2. Netzer, A. and Hughes, D.E. (1984). Water Res. 18: 927.
3. Kadirvelu, K. and Namasivayam, C. (2000). Environ. Tech-
nol. 21: 1091–1097.
4. Basso, M. C., Cerella, E.G., and Cukierman, A.L. (2002). Ind.
Eng. Chem. Res. 41: 180–189.
5. Kadirvelu, K., Faur-Brasquet, C., and Le Cloirec, P. (2000).
Langmuir 16: 8404–8409.
6. Goel, J., Kadirvelu, K., Rajagopal, C., and Garg, V. K. (2004).
Carbon. Elsevier Press, Amsterdam, In press.
7. Dubinin, M. M. (1983). Carbon 19: 321.
8. IUPAC. (1972). Manual of symbols and terminology,
Appendix 2, Pt.1, colloids and surface Chemistry. Pure and
Appl. Chem. 3: 578.
9. Calgon Corporation. Basic Concept of Adsorption on Activated
Carbon. Calgon, Pittsburgh, PA.
10. Kadirvelu, K., Thamaraiselvi, K., and Namasivayam, C.
(2001). Sep. and Purifi. Technol. 24: 497–505.
11. Ricou, P., Lecuyer, I., and Le Cloirec, P. (1998). Environ.
Technol. 19: 1005–1062.
12. Kadirvelu, K. (1998). Characterization of coir pith carbon
and its utilization in the treatment of metal bearing
wastewater. Ph.D. Thesis, Bharthiar University, Coimbatore,
India, pp. 59–60.
13. Sigworth, E.A. and Smith, S.B. (1972). J. AWWA 64:
386–391.
14. Namasivayam, C. and Kadirvelu, K. (1999). Carbon 37:
243–256.
15. Goel, J., Kadirvelu, K., and Rajagopal, C. (2004). Ads. Sci.
Tech. 22: 257–274.
16. Goel, J., Kadirvelu, K., and Rajagopal, C. (2004). Environ.
Technol.
17. Reed, B.E. and Arunachalam, S. (1995). ASCE J. Env. Eng.
121(9): 653–663.
18. Panayotova, M.I. (2001). Waste Mgmt. 21: 671–676.
19. Anoop Krishnan, K. and Anirudhan, T.S. (2002). J. Haz.
Mater. B92: 161–183.
20. Raji, C. and Anirudhan, T.S. (1996). Removal of Hg (II) from
aqueous solution by sorption on polymerized sawdust. Ind. J.
Chem. Tech. 3: 49–54.
21. Huang, C.P. and Ostovic, F.B. (1978). J. Env. Eng. Div. ASCI.
104: 863–878.
22. Sharma, D.C. and Forster, C.F. (1994). Bioresource Technol.
47: 257–264.
23. Yenki, M. K.N. and Natarajan, G. S. (1991). Separation Sci.
Technol. 26: 661–674.
496 IRON
24. Bayer, A.G. (2000). Leverkusen.
25. Trochimczuk, W.A. (1999). Europ. Polym. J. 35: 1457–1464.
26. Rivas, B.L., Maturana, H.A., Villegas, S., and Pereira, E.
(1998). Polym. Bull. 40: 721–728.
27. Petruzzelli, D., Pagano, M., and Tiravanti, G. (1995). Ind.
Eng. Chem. Res. 34: 2612–2617.
28. Bolto, B.A., Pawlowski, L.E., and Spoon, F.N. Ltd. (1987).
London.
29. Warshawsky, A. and Erlich-Rogozinski, S. (1977). Microm. J.
22: 362–368.
30. Dorfne, K. (1991). Walter de Gruyter, Berlin.
31. Inoue, K. et al. (1993). Bunseki Kagaku. 42: 725–731.
32. Rengan, K. (1997). J. Radioanal. Nucl. Chem. 219: 211–215.
33. Hubicki, Z., Jakowicz, A., and Lodyga, A. Application of
the ions from waters and sewages. In: Adsorption and
Environmental Protection, Studies in Surface Science and
Catalysis. A. Dabrowski (Ed.). Vol. 120. Elsevier, New York.
34. Kauspedience, D., Snukiskis, J., and Gefeniene, A. (1998). J.
Radioanal. Nucl. Chem. 229: 129–132.
35. Colella, C., De Gennaro, M., Langella, A., and Pansini, M.
(1993). Cadmium removal from wastewater using chabazide
and phillipsite. In: Natural Zeolites 93. D.W. Ming and
F.A. Mupton (Eds.). Brockport, NY, p. 377.
36. Halle, K., Fischwasser, K., and Fenz, B. (1982). Technol.
Umweltschutz. 25: 1201032.
37. Leinonen, H. and Lehto, J. (2000). Rect. Funct. Polym. 43:
1–6.
38. Lehto, J. et al. (1996). J. Radio Nucl. Chem. 208: 435–443.
39. Untea, I., Tudorache, E., and Florea, C. (2000). Revista de
Chimie. 51: 123–126.
40. Drela, I. (1996). Rudy Met. Niezelaz. 41: 321–323.
41. Randell, J. M., Bermann, R.L., Garrett, V., and Waiss, A.C.
(1974). J. Foorest Prod. J. 24(9): 80–84.
42. Teles de Vasconcelos, L.A. and Gonzalez Beca, C. G. (1994).
Eur. Water Poll. Control 4(1): 41–51.
43. Masri, M. S. and Friedman, M. (1974). J. Appl. Polymer Sci.
18: 2367–2377.
44. Randall, J. M., Hautala, E., and Waiss, A.C., Jr. (1974).
Removal and recycling of heavy metals ions from mining
and industrial waste water streams with agricultural
byproducts. Proceedings of the Fourth Mineral Waste
Utilization Symposium, Chicago, IL, pp. 329–334.
45. Roberts, E.J. and Rowland, S.P. (1973). Env. Sc. Technol.
7(6): 552–555.
46. Masri, M. S., Reuter, F.W., and Friedman, M. (1974). J. Appl.
Poly. Sc. 18: 675–681.
47. Prades, E.G. et al. (1994). J. Chem. Tech. Biotechnol. 59:
289–295.
48. Cadena, F., Rizvi, R., and Peters, R.W. (1990). Feasibility
studies for the removal of heavy metals from solution
using tailored bentonite. In hazardous and industrial wastes.
Proceedings of the 22nd Mid-Atlantic Industrial Waste
Conference. Drexel University, Philadelphia, PA, pp. 77–94.
49. Matis, K.A. and Zoublious, A.I. (1994). Biotech. Bioeng. 44:
354–360.
50. Niu, H., Xu, X.S., and Wang, J. H. (1993). Biotech. Bioeng. 42:
785–787.
51. Sharma, D.C. and Forster, C.F. (1994). Biores. Technol. 49:
31–40.
52. Srivastava, S. K., Singh, A.K., and Sharma, A. (1994). Envi-
ron. Technol. 15: 353–361.
53. Low, K. S., Lee, C. K., Cheong, F.Y., and Lim, G. S. (1993). J.
Environ. Sci. Health A 28(8): 1689–1704.
54. Tong, P., Baba, Y., Adachi, Y., and Kawaza, K. (1991). Chem
Lett. 9: 1529–1532.
55. Kertman, S. V., Kertman, G. M., and Chibrikova, Z.H. (1993).
J. Appl. Chem. USSR 66(2): 465–466.
56. Dikshit, V.P. (1989). Nat. Acad. Sci. Lett. 12(12): 419–421.
57. Holan, Z.R., Volesky, B.N., and Prasetyo, I. (1993). Biotech.
Bioeng. 41(8): 819–825.
58. Leush, A., Holan, Z.R., and Volesky, B. (1995). J. Chem. Tech.
Biotechnol. 62: 279–288.
59. Bricka, R.M. and Hill, D.O. (1989). Metal immobilization
by solidification of hydroxide and xanthate sludges. In:
Environmental Aspects of Stabilisation and Solidifications
of Hazardous and Radioactive Wastes. P.L. Cote and
T.M. Gilliam (Eds.). ASTM, Philadelphia, PA, pp. 257–272.
60. Flynn, C. M., Jr., Carnahan, T.G., and Lindastrom, R.E.
(1980). Adsorption of heavy metal ions by xanthated sawdust.
Report of Investigation #8427. United States. Bureau of
Mines.
61. Kadirvelu, K. and Namasivayam, C. (2003). Advances in
Environmental Research.
62. Apak, R., Tutem, E., Hugul, M., and Hizal, J. (1998). Water
Res. 32: 430–440.
63. Suemitsu, R., Venishi, R., Akashi, I., and Nakana, M. (1986).
J. Appl. Poly. Sci. 31: 75–83.
64. Periasamy, K. (1993). Adsorptive removal of heavy metals
from synthetic indisustrial wastewater using peanut hull
carbon. Ph.D. Thesis, Bharthiar University, Coimbatore,
India.
65. Lopez-Delgado, C.P. and Girgis, B.S. (1998). Water Res. 32:
1169–1177.
IRON
WILLIAM R. ROY
Illinois State Geological Survey
Champaign, Illinois
CHEMISTRY OF IRON
The average concentration of iron in the earth’s crust
is about 5%, and it occurs in a wide variety of oxide,
carbonate, silicate, and sulfide compounds either as
metallic ores or in rock-forming minerals. Iron also occurs
naturally in water from the dissolution of iron-containing
minerals. Iron in water can occur in a dissolved form,
a colloidal form that is combined with organic matter,
and as suspended, iron-containing particles. Dissolved
iron occurs in water in two oxidation states: ferrous
iron, Fe (II), and ferric iron, Fe (III). In solution, iron
can occur as a cation (Fe2+ ) and in combination with
other ions to form complexes and neutral ion pairs. Iron
does not generally occur as the trivalent cation (Fe3+ )
in significant concentrations in water. Trivalent iron
generally exists in combination with other constituents.
Depending on the occurrence of anions, such as sulfate,
carbonate, and phosphate, and the pH and reduction-
oxidation potential (Eh ), the major iron species found
in natural waters include a sulfate complex (FeSO4 0 ),

16
Fe(OH)3
14
12
Fe2+
Iron, mg/L
10
8 drinking water and alter the taste of food cooked in
6 the water. The taste threshold for ferrous iron is about
Fe(OH)2+
4
2 contaminant level for public water supplies is 0.3 mg/L,
0 and 1.0 mg/L for freshwater aquatic life (4). However,
−300 −200 −100 0 100 200 300 400 500 600 700
Eh, mV
Figure 1. The distribution of 10 mg/L of iron in water at 15 ◦ C
and pH 7.0 as a function of oxidation-reduction potential (Eh).
a carbonate form [Fe(HCO3 )2 ], phosphate complexes
(FeHPO4 0 , FeH2 PO4 + ), and various hydroxide species
such as FeOH+ , Fe(OH)2 + , Fe(OH)3 0 , Fe(OH)4 − , and
FeOH2+ . The relative proportion of ferrous iron and ferric
iron depends on the Eh of the water; ferrous iron is the
dominant form in oxygen-poor water (reducing conditions).
In oxygen-rich water (oxidizing conditions), ferric forms
of iron dominate. For example, in a solution containing
10 mg/L of iron [Fig. 1], about 98% of the iron would occur
as Fe2+ in a relatively reducing system. At an Eh greater
(i.e., more oxidizing) than about 200 mV, ferrous iron is
converted to the ferric forms Fe(OH)3 0 and Fe(OH)2 + . At
an Eh of 700 mV, about 96% of the iron occurs in ferric
forms. When a groundwater sample is collected, it may
initially appear clear. However, if the sample is allowed to
react with air and oxidize, the ferrous iron in the initially
clear, reduced sample will be converted to ferric forms:
4Fe2+ + O2 + 8OH− + 2H2 O 
 4Fe(OH)3 (1)
As a consequence, the sample may develop a reddish-
brown color, and iron-containing particles may settle at
the bottom of the container. The rate at which ferrous iron
is oxidized depends on temperature, pH, and the amount
of dissolved oxygen and iron.
Microorganisms can influence the oxidation state and
distribution of iron in water. Iron can be used as an
energy source via oxidation by microbial activity or
reduced because iron can act as an electron acceptor. For
example, Crenothrix and Leptothrix are diverse groups
of iron-precipitating bacteria that are widely distributed
in freshwater and soil. After ferrous iron is oxidized to
the sparingly soluble ferric form, it may precipitate as
ferric hydroxide. The bacterial cell sheaths can become
coated with the precipitate, and the resulting gelatinous
mass can clog water pipes and drains. There are bacteria
and fungi that can reduce ferric iron. However, much
less is known about iron-reducing microorganisms (1).
The Geobacter are a group of bacteria that can reduce
ferrous iron.
HEALTH AND REGULATORY ISSUES
Iron is not toxic in human consumption. Iron is an
essential element for the formation of hemoglobin. The
IRON 497
recommended dietary allowance for iron in healthy
adults is 10 mg per day for men and 15 mg per day
for premenopausal women (2). Dissolved iron in potable
water is more of a nuisance than a potential heath
issue. It can impart an unpleasant metallic taste to
0.1 mg/L. Concentrations exceeding 0.3 mg/L of iron can
stain plumbing and clothing (3). The secondary maximum
bioassays by Roy et al. (5) suggested that concentrations
less than 1.0 mg/L may be toxic to freshwater aquatic life.
LEVELS OF IRON IN WATER
The amount of dissolved iron in water depends on pH
and Eh, but concentrations of less than 0.01 to 10 mg/L
are typical for natural waters (6). Well-aerated surface
water usually contains less than 0.05 mg iron/L. Iron
concentrations in aquifers in Illinois have ranged from
<0.05–21 mg/L (7). Iron in acid-mine drainage from coal
mines has ranged from 10–5,625 mg/L (5,6).
SOLUBILITY OF IRON IN WATER
The amount of dissolved iron in water can be limited
by the solubility of iron-containing solid phases. Schwab
and Lindsay (8) concluded that when pH + pE (pE is
related to the redox potential Eh by pE = (EhF)/(2.3RT)
where R = gas constant, T = absolute temperature, and
F = Faraday constant) is less than 8.5, the concentration
of ferrous iron in groundwater is limited by the solubility
of siderite, FeCO3 :
Fe2+ + CO3 2− 
 FeCO3 (2)
When pE + pH is greater than 8.5, the solubility of ferrosic
hydroxide, Fe3 (OH)8 , limits the amount of ferrous iron:
+
2Fe3+ + Fe2+ + 8H2 O 
 Fe3 (OH)8 + 8H (3)
When the concentration of iron is greater than the
solubility product of either siderite or ferrosic hydroxide,
the iron precipitates, from solution. Ferrosic hydroxide
is thermodynamically unstable, and other minerals may
form, causing ferrosic hydroxide to dissolve. For example,
amorphous ferric hydroxide, [Fe(OH)3,a ] and goethite (α-
FeOOH) are less soluble than ferrosic hydroxide and given
time, may control iron solubility. Moreover, amorphous
ferric hydroxide can dehydrate (9) to form the more stable
mineral hematite (Fe2 O3 ):
 Fe2 O3 + 3H2 O
2 Fe(OH)3,a  (4)
The prevalent form of iron in surface water in the
iron–carbon dioxide–water system can be illustrated by
constructing a pH–pE stability diagram [Fig. 2]. This
type of diagram illustrates which forms of iron are
most thermodynamically stable at various values of pH
498 IRON

+20

Fe3+ Up
per
Fe(OH)2+ lim
it o
fw
ate
r st
abi
lity

+10

amorphous
Fe2+ Fe(OH)3
pE

Fe(OH) −4
Low
er l FeCO3
imi
t of
wa
ter st
−10 abi
lity
Fe(OH)2

Figure 2. pE–pH diagram for the iron–

carbon dioxide–water system at 25 ◦ C 0 3 6 9 12 15
and 1 atm (11). pH

and pE rather than at specific concentrations. Ferrous
iron is the stable form in equilibrium with relatively
acidic and reducing waters, whereas ferric iron, as both
Fe3+ and FeOH2+ , is in equilibrium with acidic and
oxidizing waters. In relatively less oxidizing and alkaline
waters, the most thermodynamically stable forms may
be amorphous iron hydroxide, siderite (FeCO3 ), or solid
ferrous hydroxide.
The concentration of iron in water may also be influ-
enced by sorption–desorption interactions with mineral
surfaces. Iron oxides and hydroxides have been studied
as sorbents for other metals, but ionic iron is sub-
ject to the same electrostatic interactions that attract
cations to clays and organic matter. Ferric iron can be
sorbed by silica, clay minerals, and iron and manganese
oxides (10). The sorption of ferrous iron has not been
studied in detail.
METHODS FOR DETERMINING IRON
The amount of iron dissolved in water can be determined
quantitatively by atomic absorption spectroscopy (AAS),
atomic emission spectroscopy, and by colorimetric methods
(12). In the AAS method, a water sample is aspirated into
an air-acetylene flame and atomized. A monochromatic
beam of light at a wavelength corresponding to iron is
passed through the flame. The amount of light energy
sorbed is proportional to the amount of iron present.
When analyzed at a wavelength of 248.3 nm, the optimal
concentration range for iron is about 0.3 to 10 mg/L.
Dissolved iron can be determined by argon-supported
inductively coupled plasma (ICP) spectroscopy. In ICP,
the liquid sample is aspirated into a high-temperature
(6000–8000 K) flame, and atomic emission is measured
to determine the concentration of the iron present. The
concentration of iron is proportional to the intensity of
the emissions. Iron has many such emission lines in its
spectrum, but the wavelength of 259.94 nm is commonly
used for concentrations ranging from about 0.007 to
100 mg/L.
Dissolved iron also can be determined by the phenan-
throline method. In this method, iron in solution is reduced
to the ferrous form, and then allowed to react with 1, 10-
phenanthroline to form a reddish-orange product. The
absorbance of the complex is measured in the range of
460–520 nm. Both AAS and ICP methods determine total

iron, but the amount of ferric and ferrous iron can be
determined separately by the phenanthroline method (13).
METHODS FOR REMOVING IRON FROM WATER
The presence of large concentrations of dissolved iron is
a common water-quality problem, especially for ground-
water (1). Methods for removing iron are summarized in
Lehr et al. (3) and Vigneswaran and Visvanathan (2). The
oxidation of soluble, ferrous iron into relatively insoluble
ferric iron, followed by filtration of the suspended mass
is common to many of the methods. Depending on its
concentration in the untreated water, the oxidation of
ferric iron can be promoted by aeration, catalytic oxida-
tion with manganese oxides, and chlorination. Household
water softeners are designed to remove calcium and mag-
nesium from hard water using an ion exchange resin, but
they can also remove dissolved iron. Lime softening of
municipal water supplies can also promote the precipita-
tion of ferric iron. The presence of iron-fixing bacteria may
necessitate the use of relatively large amounts of bleach
(‘‘shot chlorination’’) to reduce the amount of iron in the
final product.
BIBLIOGRAPHY
1. Chapelle, F.H. (2001). Ground-Water Microbiology and Geo-
chemistry. John Wiley & Sons, New York.
2. Vigneswaran, S. and Visvanathan, C. (1995). Water Treat-
ment Processes. Simple Options. CRC Press, Boca Raton,
FL.
3. Lehr, J.H., Gass, T.E., Pettyjohn, W.A., and DeMarre, J.
(1980). Domestic Water Treatment. McGraw-Hill, New York,
pp. 172–178.
4. U.S. EPA. (1999). National Recommended Water Quality
Criteria–Correction. EPA 822-Z-99-001.
5. Roy, W.R., Skelly, T.M., Krapac, I.G., Griffin, R.A., Dickerson,
D.R., Schuller, R.M., Suloway, J.J., and Shimp, N.F. (1984).
Geochemical Properties and Toxicological Effects on Aquatic
Life of Coal Wastes. Environmental Geology 107, Illinois State
Geological Survey, Champaign, IL, p. 124.
6. Patterson, J.W. (1975). Wastewater Treatment Technology.
Ann Arbor Science, Ann Arbor, MI.
7. Warner, K.L. (2000). Analysis of Nutrients, Selected Inor-
ganic Constituents, and Trace Elements in Water from Illinois
Community-Supply Wells, 1984–91. Water-Resources Inves-
tigations Report 99-4152, U.S. Geological Survey.
8. Schwab, A.P. and Lindsay, W.L. (1983). Effect of redox and
pH on the solubility and availability of iron. Soil Sci. Soc. Am.
J. 47:201–205.
9. Garrels, R.M. and Christ, C.L. (1965). Solutions, Minerals,
and Equilibria. Freeman Cooper, San Francisco, CA.
10. Anderson, M.A. and Rubin, A.J. (1981). Adsorption of Inor-
ganics at Solid–Liquid Interfaces. Ann Arbor Science, Ann
Arbor, MI.
11. Stumm, W. and Morgan, J.J. (1981). Aquatic Chemistry. John
Wiley & Sons, New York.
12. American Public Health Association. (1992). Standard
Methods for the Examination of Water and Wastewater, 18th
Edn. Washington, DC, Section 3500. Iron, 3-65–3-68.
13. Stucki, J.W. (1981). The quantitative assay of minerals
for Fe2+ and Fe3+ using 1, 10-phenanthroline: II. A
photochemical method. Soil Sci. Soc. Am. J. 45: 638–641.
ISOTOPES 499
ISOTOPES
ALDO CONTI
Frascati (RM), Italy
Isotopes are atoms with the same atomic number, Z, but
different atomic weight, A. Different isotopes, therefore,
have the same chemical but different physical properties,
such as melting and boiling point. The name isotope means
‘‘at the same place’’ and comes from the periodic table,
where different isotopes of the same element are at the
same place. The scientific way of labeling isotopes reflects
this. They have the symbol of the corresponding element,
but followed by a hyphen and the number of nucleons
(neutrons and protons) of the isotope. So, C-12 is the
normal carbon (six protons and six neutrons) and C-14 is
a heavy unstable isotope (six protons and eight neutrons).
Isotopes of the same element share the same chemical
properties because they depend only on the number of
protons, Z. But different isotopes can have very different
weights, which is particularly evident in hydrogen. Normal
hydrogen has only one proton and no neutrons. Deuterium,
or H-2, has one neutron and its weight is nearly twice that
of hydrogen. Then, tritium, H-3, has two neutrons and
is three times as heavy, which is the reason why, when
deuterium combines with oxygen, the resulting molecule is
called heavy water. Due to the extra weight of deuterium,
heavy water boils at 101.42 ◦ C (214.56 ◦ F) and freezes at
3.81 ◦ C (38.86 ◦ F), at standard pressure. The difference in
weight affects the speed of chemical reactions involving
water. For this reason, heavy water is not really safe to
drink. Experiments performed with mice have shown that
large amounts of heavy water in the body can inhibit
mitosis, which affects rapidly dividing tissues, such as the
stomach lining.
Isotopes do not differ only in physical properties. The
extra neutrons make some of the nuclei highly unstable,
which is the reason that some isotopes are radioactive.
A typical example is uranium. Its isotopes, such as U-
235, decay radioactively following a long chain, which
leads to lead. Some of this decay involves a neutron
becoming a proton by emitting an electron. In this case, the
nucleus becomes a new element, with the atomic number
increased by one. The energy emitted during the decay
can be harvested in nuclear plants or used destructively
in atomic bombs.
The instability of their nuclei makes isotopes useful for
many applications. A famous one is carbon dating. Carbon
comes mainly in two isotopes, the normal C-12 and the
heavy, unstable, C-14. When an organism grows, it uses
them in the same proportion in which they are found in the
environment. After the death of the organism, the ratio
changes continuously, as the heavy C-14 decays. C-14 has
a half life of 5730 years, which means that after this period
half of it has gone. By this method, it is possible to date
any biological remains or artifact up to about 50,000 years
old. For older objects, the amount of C-14 becomes too
small to be measured reliably. Isotopes are often used as
labels. By replacing a stable atom in a molecule with an
unstable radioactive isotope, it becomes possible to follow
500 ISOTOPE FRACTIONATION
the chemical reactions the molecule undergoes. Our ability
to detect radioactive isotopes makes them very useful in
the fields of medicine and biology, where they are used
as tracers.
ISOTOPE FRACTIONATION
GHOLAM A. KAZEMI
Shahrood University of
Technology
Shahrood, Iran
INTRODUCTION
The following paragraph extracted from the major
textbook by Bowen (1) describes fractionation as it is dealt
with in general chemistry. The subsequent paragraph
defines fractionation in general hydrochemistry. Then,
isotope fractionation is discussed.
In general chemistry, fractionation is the separation of
a mixture, usually of chemically related components, into
fractions that possess different properties. A mixture of
liquids having different boiling points, may be separated
by collecting fractions that boil at different temperatures
in a fractionating column. This is a long, vertical column
containing plates or bubble caps attached to a still. As
a consequence of internal reflux, a gradual separation
occurs between high and low boiling point fractions of the
liquid mixture.
In hydrochemistry, the fractionation factor is a ratio
which shows the depletion or addition of ions in rainwater
compared to seawater (2). In other words, the fractionation
factor is the ratio of two ions in rainwater to the ratio of
the same ions in seawater. For example, for Cl− relative
to Na+ , the fractionation factor FNa is (2):
(Cl/Na)rainwater
FNa =
(Cl/Na)seawater
In this equation, the units of concentration can be mg/l,
mmol/l, meq/l, etc.
Example: What is the fractionation factor of Na
for a rainwater sample whose chloride and sodium
concentrations are 1 and 0.5 mg/l, respectively?
Cl/Narainwater sample = 1/0.5 = 2
Cl/Naseawater = 1.81
(2)
FNa = = 1.11
(1.81)
A fractionation factor greater than unity indicates
that the rainwater sample is depleted (in comparison to
seawater) in the ion in question. If it is less than unity, it
shows the enrichment (in comparison to seawater) of the
sample in the ion. Appelo and Postma (2) also propose the
fractionation factor as a quantitative approach to find the
activity of a single species that forms part of a given
total concentration, for example, for HCO3 , mHCO3 − =
α × CO2 , where α is the fractionation factor. Some
researchers use ‘‘enrichment factor (Ef)’’ in an equation
slightly different from the fractionation factor to deduce
source/s of solutes in atmospheric precipitation (3,4). If
the Ef is greater than unity, the ion in question (in the
sample) is derived partly or wholly from a source other
than oceanic origin. If it is less than unity, the ion could
have been derived wholly from oceanic sources:
(Ion/Cl)sample
Ef = ||
(Ion/Cl)seawater
In this example, the enrichment factor equals 0.91.
(0.5)
Ef = = 0.91
(0.55)
DEFINITION
Isotope fractionation is defined by Domenico and
Shwartz (5) as a change in an isotopic ratio (for example,
18
O/16 O) because of a chemical reaction (for example, water
changes to water vapor). It is also defined by Coplen (6) as
the fractionation or partitioning of isotopes by physical or
chemical processes which is proportional to the differences
in their masses. Coplen also classifies isotopic fractiona-
tion processes into two categories, chemical and physical
isotopic fractionation processes: physical processes such
as ultrafiltration or gaseous diffusion of ions or molecules.
Chemical processes are those that lead to redistribution
of the isotopes of an element among phases or chemical
species and are divided into two broad categories equilib-
rium isotopic reactions and kinetic isotopic reactions (6).
Gilreath (7) in a simple yet valuable text describes the
fundamentals of isotope fractionation in detail and refers
to it as ‘‘isotope separation.’’ Anyone intending to deal
with isotope fractionation is advised to consult this text
first. The point of primary importance is that fraction-
ation of isotopes is important in the geologic environ-
ment only for elements of low atomic number. Bowen (1)
describes three main processes which cause fraction-
ation: (1) isotope exchange reactions; (2) unidirectional
reactions; and (3) physical processes such as evapora-
tion, condensation, crystallization, melting, adsorption,
desorption, and diffusion. Isotopes may also be fraction-
ated in the laboratory by a variety of means (URL Ref.
http://www.c14dating.com/frac.html). One method that
may result in considerable fractionation during isotopic
analysis is the recently introduced technique of secondary-
ion mass spectrometry (SIMS) (8).
The following are the factors that control the rate of
isotope fractionation:
1. Fractionation is minimized when reactions are
fast (9).
2. The higher the mass difference between various
isotopes of the element, the more effective the
fractionation process. Deuterium, for example, has
the largest mass difference from common hydrogen
compared to isotopes of other elements. Hence, it is
subject to greater fractionation.

3. Isotope fractionation decreases as temperature
increases—at finite temperatures, all isotopic
species are well mixed (6).
4. Isotopic exchange between minerals and water in
deep, basinal flow systems or in geothermal systems
induced by high temperature helps the fractionation
process to proceed (5).
5. In evaporation from a pond, fractionation is more
efficient if the vapor produced is constantly removed,
for example, by wind (10).
Fractionation Factor
The fractionation factor is the ratio of isotopes (for
example, 18 O/16 O) in a liquid phase to the same ratio
in the vapor phase;
(18 O/16 O)liquid
α= (11)
(18 O/16 O)vapor
or more clearly/generally defined by Haldorsen et al. (11)
‘‘as the increase or decrease of a given isotope in a
substance that undergoes a phase change or simply in
terms of delta notation:
1000 + δ 18 Oliquid
α=
1000 + δ 18 Ovapor
Example: In the condensation of water vapor to liquid,
typical values of δ 18 O for the liquid and vapor are, −5% and
−14%, respectively. Therefore, the fractionation factor is,
1000 + (−5) 995
α= = = 1.0092
1000 + (−14) 986
Each fractionating process results in a specific frac-
tionation factor. For example, the oxygen fractionation
factor in freezing differs from that from condensation. The
former and the latter are 1.0111 and 1.003, respectively.
This means that freezing is more effective than conden-
sation for isotope fractionation; freezing leads to stronger
fractionation than condensation.
Fractionation factors approach unity as temperature
increases; when α = 1, the isotope ratios become equal
in both phases being considered, and hence no further
isotopic fractionation can occur (1). Higher fractionation
factors mean more effective fractionation processes.
Isotopic Fractionation Curve
A curve defines the equilibrium isotope fractionation
between coexisting phases as a function of temperature (1).
It usually takes the form of an exponentially decreasing
curve; there is a negative correlation between isotope
fractionation and temperature. The shape of this curve is
different for different isotopes.
Fractionation of Various Environmental Isotopes
As can be inferred from the information provided so
far, isotope fractionation is an extensive topic in chem-
istry/geochemistry and has been dealt with in detail in
many references such as Bowen (1), Gat (12), Gilreath (7),
ISOTOPE FRACTIONATION 501
and Ottonello (13); the last is a thorough mathematical
treatment of the topic. Here, in the Encyclopedia of Water,
the discussion of isotope fractionation is limited to the iso-
topes of hydrogen and oxygen, specifically, the atoms that
form the water molecule, as well as other isotopes that
are of frequent use in water related/hydrologic studies.
The latter group includes isotopes of the elements C, N,
Cl, S, Sr, and less used U, Th, and Si. These are, often,
referred to as environmental isotopes (6). Some workers
such as Haldorsen et al. (11) classify the environmental
isotopes into two groups: hydrologic isotopes (O, D, T)
and biochemical isotopes (C, N, S). Although this classi-
fication covers only a limited number of environmental
isotopes, it is sound for fractionation. For the first group,
the main fractionating processes are physical or chemi-
cal reactions, biological/bacteriological/microbial processes
are dominant in the fractionation of the second group.
Before discussing all environmental isotopes one by one,
it should be pointed out that in some cases, considerable
isotope fractionation has led to the usefulness of those
isotopes in hydrologic studies, for example, H and O
isotopes (enrichment and depletion processes are the
fundamentals for studying lake–groundwater interaction,
recharge temperature, and many other applications). In
contrast, in some other cases, lack of isotope fractionation
has been helpful. Strontium isotopes, for example, are
used to study the origin of solutes in the streams because
they are not subjected to substantial fractionation.
H and O
Hydrogen and oxygen atoms have three isotopes each
(see also Heavy Water and Deuterium articles): 1 H, 2 H,
3
H, and 16 O, 17 O, and 18 O. Water molecules containing
the heavy isotopes are less likely to evaporate and more
likely to condense than those containing lighter isotopes.
Hence, when a body of water evaporates, the vapor
phase is depleted and the residue, the liquid phase, is
enriched in heavy isotopes, 2 H, 3 H, 17 O, and 18 O. During
melting/freezing of water, heavy isotopes are slightly
more inclined to freeze and less inclined to melt than
lighter isotopes, making the melt water depleted in 18 O
and deuterium compared to the snowpack. Ultrafiltration
(passage of water through shale micropores) enriches
the residue (not passed through water) in O and H
isotopes (6), but transpiration is not a fractionating
process. The principles of hydrogen isotope fractionation
are the scientific foundation for producing heavy water,
which is used in the nuclear industry.
The fractionation factor for deuterium is greater
than that for oxygen-18 because of a larger difference
between deuterium and hydrogen masses compared to the
difference between 16 O and 18 O masses. The fractionation
factors of these two elements at various temperatures are
calculated via the following equations (1):
For oxygen,
1000 ln (α 18
O) = 2.644 − 3.026 × 103 /T + 1.534 × 106 /T 2
For hydrogen,
ln (α D) = 0.0771 + 13.436 × 103 /T 2
502 ISOTOPE FRACTIONATION
(note that the temperature is in K).
Example:
The fractionation factor for 18 O at 293.15 K (20 ◦ C) is
ln (α 18
O) = (2.644 − 3.026 × 103 /293.15 + 1.534
× 106 /293.152 )/1000 = 1.01000
The fractionation factor for 18 O during calcite precipitation
from water is 1.0349, which means that calcite is 35%
more enriched in 18 O relative to the water from which it
precipitates. However, not all precipitated calcite is richer
in 18 O compared to ‘‘parent water.’’ For instance, if calcite
precipitation is due to freezing, the 18 O content of calcite
is lower (11).
C the fractionation and enrichment of 34 S depend on the
Oxidation–reduction processes in which the heavier
isotope is concentrated in the more oxidized forms are
important in carbon fractionation. The most influential
factor in carbon isotope fractionation is selective uptake
of lighter CO2 by plants. During photosynthesis, plants
discriminate against the heavier isotopes of carbon; the
isotope 13 C in organic matter (soil) is depleted by 1.8%
compared to its natural ratios in the atmosphere. It is
believed that all organisms discriminate against 14 C about
twice as much as against 13 C and the ratio between
the stable 12 C and 13 C atoms can be used to correct
for the initial depletion of 14 C. The amount of isotope
fractionation depends on the photosynthetic pathway used
by the plant. Plants are divided into three categories
in this regard (11). At high temperatures, equilibrium
fractionation of isotopes may occur between CH4 and CO2 ,
and this is sensitive to changes in pH, oxygen fugacity, and
ionic strength as well as to changes in temperature (13).
As an example, the fractionation factor for 13 C in CaCO3
in equilibrium with CO2 is 1.01074.
Cl
The author has not come across a remark in the
literature about the occurrence of fractionation in chlorine
isotopes. This may have happened because chloride is
a highly mobile, soluble, and geochemically conservative
(is not removed from solution by mineral interaction or
secondary mineral formation) element and has a relatively
high mass.
N
Fractionation of nitrogen isotopes occurs mostly through
biological processes and some isotope exchange reactions.
Kinetic isotope fractionation often dominates nitrogen
reactions and because of this, during denitrification,
δ 15 N values of residual nitrate increase exponentially
as concentrations decrease (6). Bowen (1) reports that
conversion of ammonia to nitrate (nitrification) by
nitrifying organisms produces isotope fractionation up to
20% and even greater values (up to 30%) are associated
with denitrification.
S
Sulfur in near-surface environments frequently changes
between the oxidized and the reduced forms, so it acquires
a wide variety of 34 S/32 S ratios. Isotope fractionation
occurs in many processes of the sulfur cycle. The principal
reaction leading to the fractionation of sulfur isotopes
in nature is the reduction of sulfate ion by anaerobic
bacteria (13). Bacterial reactions are important because
they are kinetic, fast, favor the lighter isotope, and result
in depleting 34 S in natural biogenic H2 S and hence enrich
the remaining sulfate (11). Brucher et al. (14) studied
isotope fractionation during bacterial sulfate reduction
and found that in addition to the variations in the rates
of bacterial sulfate reduction, genetic and physiological
differences between various genera of sulfate-reduction
bacteria strongly influence isotope fractionation.
In the precipitation of sulfide minerals from solution,
relative bond strengths and the temperature of solu-
tion (13). Sulfide minerals in newly deposited sediment
can be enriched in 32 S by about 50% compared with
associated marine sulfate (1). The equilibrium isotope frac-
tionation factor between SO4 2− and H2 S is 1.074 and is
1.020 between water and dissolved sulfate at 88 ◦ C (6).
During dissolution of evaporites, the largest reservoir of
sulfate, no sulfur fractionation takes place.
Sr
Because of its relatively high mass, no detectable isotope
fractionation accompanies precipitation of Sr within
minerals (6).
ARTIFICIAL ISOTOPE FRACTIONATION FOR
‘‘SEPARATION OF ISOTOPES’’
The most interesting and applied part of the isotope
fractionation topic is artificial separation (‘‘fractionation’’)
of isotopes, a subject that is scientifically simple, but, in
engineering terms very complicated. Historically, the first
successful attempt to separate isotopes was that by Aston
(the inventor of the mass spectrograph and the Nobel prize
winner in 1922) who separated neon into two fractions of
different atomic weights by repeated diffusion. The topic
gained importance with the discovery, in 1939, that 235 U
is a fissionable material and in succeeding years that
uranium-235 is the key material for producing nuclear
weapons and power (7).
VARIOUS TECHNIQUES FOR SEPARATING ISOTOPES
ARTIFICIALLY
As stated before, separating isotopes is an intensive
complicated engineering task, and only a few countries
have the capacity. Four principles that constitute the
foundation of isotope separation should be considered
when devising a separation process (7):
1. At a given temperature, gaseous molecules, regard-
less of mass, have the same average kinetic energy.

2. The gravitational or centrifugal forces acting on
a particle of matter are directly proportional to
its mass.
3. The electric or magnetic forces affecting ionized
atoms or molecules follow Coulomb’s laws for
these forces.
4. Among the lighter elements, isotopic species may
vary slightly in their chemical reactivities.
Considering these four principles, a variety of techniques
have been developed for isotopic enrichment (7). Each of
these techniques has advantages and disadvantages and
is suitable for specific types of conditions and isotope(s).
1. The gaseous-diffusion method is used for separating
and enriching uranium isotopes. The production
plant which uses this technique is huge; the huge
diffusion plant at Oak Ridge, USA, in operation
from 1945, represents one of the greatest technical
achievements in the history of humankind.
2. The electromagnetic method is suitable only where
cheap electricity is available.
3. The thermal diffusion method is good for laboratory-
scale production.
4. The centrifuge method was not developed into a
large-scale production plant.
5. The electrolytic method is used only for separating
hydrogen and deuterium.
6. The chemical exchange method is used mostly for
separating C and N.
7. The distillation method is only for enrich-
ing deuterium.
A sideline and an interesting coincidence is that while
this article was being written, the Board of Governors
of International Atomic Energy Agency (IAEA) postponed
its decision on Iran’s (the home country of the author)
uranium enrichment and nuclear program activities (CNN
Online edition, Nov. 22, 2003). The uranium enrichment
and nuclear activities program of Iran has been in the
news headlines for quite a long time.
BIBLIOGRAPHY
1. Bowen, R. (1988). Isotopes in the Earth Sciences. Elsevier,
London.
2. Appelo, C.A.J. and Postma, D. (1999). Geochemistry, Ground-
water and Pollution. Balkema, Rotterdam.
3. Kazemi, G.A. and Mehdizadeh, H. (2001). Solute content
and chemical composition of atmospheric precipitation of
Shahrood Region, Proc. of 5th Iranian Geol. Convention,
Tehran, September 2001, pp. 401–405 [in Persian].
4. Panettiere, P., Cortecci, G., Dinelli, E, Bencini, A., and
Guidi, M. (2000). Chemistry and sulfur isotopic composi-
tion of precipitation at Bologna, Italy. Appl. Geochem. 15:
1455–1467.
5. Domenico, A.P. and Shwartz, F.W. (1991). Physical and
Chemical Hydrogeology. John Wiley & Sons, New York.
6. Coplen, T.B. (1993). Uses of environmental isotopes New
York. In: Regional Ground-Water Quality. W.M. Alley (Ed.).
ITP, pp. 255–293.
MARIOTTE BOTTLE—USE IN HYDROLOGY 503
7. Gilreath, E.S. (1985). Fundamental Concepts of Inorganic
Chemistry. 19th printing, International Student Edition.
McGraw-Hill, Singapore.
8. MacRae, N.D. (1995). Secondary-ion mass spectrometry and
geology. Can. Mineralogist 33: 219–236.
9. Langmuir, D. (1997). Aqueous Environmental Geochemistry.
Prentice-Hall, Upper Saddle River, NJ.
10. Mazor, E. (1991). Applied Chemical and Isotopic Groundwa-
ter Hydrology. Open University Press, Milton Keynes.
11. Haldorsen, S., Riise, G., Swanson, B., and Sletten, R.S. (1997)
Environmental isotopes as tracers in catchments. In: Geo-
chemical Processes, Weathering and Groundwater Recharge in
Catchments. O.M. Saether and P.D. Caritat (Eds.). Balkema,
Rotterdam, pp. 185–210.
12. Gat, J.R. (1981). Isotopic fractionation. In: Stable Isotope
Hydrology, Deuterium and Oxygen-18 in the Water Cycle.
J.R. Gat and R. Gonfiantini (Eds.). IAEA, technical report
series, No. 210, pp. 21–34.
13. Krauskopf, K.B. (1979). Introduction to Geochemistry, 2nd
Edn. McGraw-Hill, New York.
14. Bruchert, V., Knoblauch, C., and Jorgensen, B.B. (1999). Sul-
fur isotope fractionation during bacterial sulfate reduction.
In: Geochemistry of Earth’s Surface. H. Armannsson (Ed.).
Balkema, Rotterdam, pp. 303–305.
MARIOTTE BOTTLE—USE IN HYDROLOGY
JOSEPH HOLDEN
University of Leeds
Leeds, United Kingdom
A Mariotte bottle is a device that allows constant flow
of a fluid from a container, even when the fluid level in
the container changes. It is named after Edme Mariotte
(1620–1684), a French physicist, who, independently from
Boyle, discovered the law that states that the volume of
a gas varies inversely with its pressure. Mariotte bottles
have a wide range of applications in engineering and
science (such as fuel supply) and are used in a variety
of hydrologic applications ranging from river discharge
gauging to infiltration measurement.
Figure 1 shows the general design of a Mariotte bottle.
It consists of a bottle that has an escape pipe or tap to allow
fluid to exit and a vent tube to allow air to enter the bottle.
The bottle is sealed so that the only location where air can
enter is at that bottom of the vent tube, b. The fluid (such
as water) exits through a tap (or any sort of escape tube) at
c, at a rate that is controlled by the size of the escape tube
and the head of water. The head of water, h will always be
equal to the distance between the outlet tap and the bottom
of the vent tube. Therefore, regardless of the water level
inside the bottle, there will be a constant head of water.
If the vent tube is kept at a constant level, then as
a result of the reduction of the fluid in the bottle, the
air above it expands and its pressure drops. When the
pressure of the air has dropped so that it plus the pressure
of the water column from c to b (the vent tube opening) is
less than the atmospheric pressure of the air outside the
Mariotte bottle (and therefore also inside the vent tube),
then air enters the Mariotte bottle at b and rises through
the fluid. Therefore, the pressure at c will always be equal
504 MARS EXPLORATION ROVER MISSION
Vent tube
Air
c
Fluid
h Figure 2. Mariotte bottle used to keep a constant water level
a then it can be used to determine how much water has been
Outlet pipe
Figure 1. A Mariotte bottle.
to atmospheric which is also equal to the pressure at b.
The excess pressure, under which the fluid is ejected, is
the pressure of the fluid between the levels b and a. It
remains constant until the fluid drops to the level of b, and
the discharge from the escape pipe will remain constant
until then. Thus, Mariotte bottles can be used wherever
there is an application for a constant flow of water from a
supply reservoir.
If the vent tube is raised to a greater height, the head
will increase, and therefore, the rate of water flowing out
of the bottle will increase. It is therefore possible to control
the rate of water release from the bottle by controlling the
head of water, and this is, in turn, controlled by the height
of the base of the vent tube above the outlet at a. If the
vent tube is moved down to the level of the escape opening,
a, the flow stops because then the height of the column
becomes zero. The discharge also ceases if the vent tube is
sealed preventing air from entering the bottle.
There is a wide range of common hydrologic applica-
tions for Mariotte bottles. They are used frequently for
adding a constant rate of solution (such as a dye or tracer)
to watercourses. Most commonly this is for discharge gaug-
ing where, if a constant rate of solution (such as sodium
chloride solution) is added to river, then the discharge of
the river can be calculated based on the equation,
Q = q(Ci − Cd )/(Cd − Cb )
where Q is the river discharge, q is the rate of injection of
tracer solution, Ci is the concentration of the tracer, Cb is
the background concentration in the river, and Cd is the
downstream sampled concentration of the tracer.
It is often useful to use a Mariotte bottle to supply a
constant rate of water and also, if the bottle is graduated,
Air
Water
d
e
Water
above the ground surface.
supplied during any given period. Frequently, Mariotte
bottles are used in instruments designed to measure soil
infiltration rates such as rain simulators or other types of
infiltrometers where a constant rate of water supply to the
soil surface might be required.
In constant head infiltrometers (see entry for
INFILTROMETERS), the Mariotte principle is used to ensure
that there is a constant depth of ponded water above the
soil surface. This is done by placing the vent tube for the
Mariotte bottle at the same end of the bottle as the water
exit pipe as in Fig. 2. Whenever the water level on the
ground drops below the height of the base of the vent tube,
d, air is allowed to enter the water supply bottle, and water
can be released from e until the water level on the ground
has risen back to the base of the vent tube. In this way, a
constant height or head of water can be maintained on a
soil surface. This principle can also be used for irrigation
to maintain a constant surface water level or a constant
groundwater level.
MARS EXPLORATION ROVER MISSION
GUY WEBSTER
Jet Propulsion Laboratory
Pasadena, California
DONALD SAVAGE
NASA Headquarters
Washington
OPPORTUNITY ROVER FINDS STRONG EVIDENCE
MERIDIANI PLANUM WAS WET
Scientists have concluded the part of Mars that NASA’s
Opportunity rover is exploring was soaking wet in
the past.
This article is a US Government work and, as such, is in the
public domain in the United States of America.

This image, taken by Opportunity’s microscopic imager, shows a portion of the rock outcrop at
Meridiani Planum, Mars, dubbed ‘‘Guadalupe.’’ View Opportunity images from this press release.
Evidence the rover found in a rock outcrop led scientists
to the conclusion. Clues from the rocks’ composition,
such as the presence of sulfates, and the rocks’ physical
appearance, such as niches where crystals grew, helped
make the case for a watery history.
‘‘Liquid water once flowed through these rocks. It
changed their texture, and it changed their chemistry,’’
said Dr. Steve Squyres of Cornell University, Ithaca,
N.Y., principal investigator for the science instruments
on Opportunity and its twin, Spirit. ‘‘We’ve been able to
read the tell-tale clues the water left behind, giving us
confidence in that conclusion.’’
Dr. James Garvin, lead scientist for Mars and lunar
exploration at NASA Headquarters, Washington, said,
‘‘NASA launched the Mars Exploration Rover mission
specifically to check whether at least one part of Mars ever
had a persistently wet environment that could possibly
have been hospitable to life. Today we have strong evidence
for an exciting answer: Yes.’’
Opportunity has more work ahead. It will try to deter-
mine whether, besides being exposed to water after they
formed, the rocks may have originally been laid down by
minerals precipitating out of solution at the bottom of a
salty lake or sea.
The first views Opportunity sent of its landing
site in Mars’ Meridiani Planum region five weeks
ago delighted researchers at NASA’s Jet Propulsion
MARS EXPLORATION ROVER MISSION 505
Laboratory, Pasadena, Calif., because of the good fortune
to have the spacecraft arrive next to an exposed slice of
bedrock on the inner slope of a small crater.
The robotic field geologist has spent most of the past
three weeks surveying the whole outcrop, and then turning
back for close-up inspection of selected portions. The rover
found a very high concentration of sulfur in the outcrop
with its alpha particle X-ray spectrometer, which identifies
chemical elements in a sample.
‘‘The chemical form of this sulfur appears to be
in magnesium, iron or other sulfate salts,’’ said Dr.
Benton Clark of Lockheed Martin Space Systems, Denver.
‘‘Elements that can form chloride or even bromide salts
have also been detected.’’
At the same location, the rover’s Mössbauer spectrom-
eter, which identifies iron-bearing minerals, detected a
hydrated iron sulfate mineral called jarosite. Germany
provided both the alpha particle X-ray spectrometer and
the Mössbauer spectrometer. Opportunity’s miniature
thermal emission spectrometer has also provided evidence
for sulfates.
On Earth, rocks with as much salt as this Mars rock
either have formed in water or, after formation, have been
highly altered by long exposures to water. Jarosite may
point to the rock’s wet history having been in an acidic
lake or an acidic hot springs environment.
506 REMOVAL OF ORGANIC MICROPOLLUTANTS AND METAL IONS FROM AQUEOUS SOLUTIONS BY ACTIVATED CARBONS
The water evidence from the rocks’ physical appearance
comes in at least three categories, said Dr. John
Grotzinger, sedimentary geologist from the Massachusetts
Institute of Technology, Cambridge: indentations called
‘‘vugs,’’ spherules and crossbedding.
Pictures from the rover’s panoramic camera and
microscopic imager reveal the target rock, dubbed ‘‘El
Capitan,’’ is thoroughly pocked with indentations about a
centimeter (0.4 inch) long and one-fourth or less that wide,
with apparently random orientations. This distinctive
texture is familiar to geologists as the sites where crystals
of salt minerals form within rocks that sit in briny water.
When the crystals later disappear, either by erosion or
by dissolving in less-salty water, the voids left behind are
called vugs, and in this case they conform to the geometry
of possible former evaporite minerals.
Round particles the size of BBs are embedded in the
outcrop. From shape alone, these spherules might be
formed from volcanic eruptions, from lofting of molten
droplets by a meteor impact, or from accumulation of
minerals coming out of solution inside a porous, water-
soaked rock. Opportunity’s observations that the spherules
are not concentrated at particular layers in the outcrop
weigh against a volcanic or impact origin, but do not
completely rule out those origins.
Layers in the rock that lie at an angle to the main
layers, a pattern called crossbedding, can result from
the action of wind or water. Preliminary views by
Opportunity hint the crossbedding bears hallmarks of
water action, such as the small scale of the crossbedding
and possible concave patterns formed by sinuous crestlines
of underwater ridges.
The images obtained to date are not adequate for
a definitive answer. So scientists plan to maneuver
Opportunity closer to the features for a better look. ‘‘We
have tantalizing clues, and we’re planning to evaluate this
possibility in the near future,’’ Grotzinger said.
JPL, a division of the California Institute of Technology
in Pasadena, manages the Mars Exploration Rover project
for NASA’s Office of Space Science, Washington.
For information about NASA and the Mars mission on
the Internet, visit http://www.nasa.gov/.
Images and additional information about the project
are also available at http://marsrovers.jpl.nasa.gov/ and
http://marsrovers.jpl.nasa.gov/relocate.html?relocate=
http://athena.cornell.edu/.
http://marsrovers.jpl.nasa.gov and http://athena.cornell
.edu.
REMOVAL OF ORGANIC MICROPOLLUTANTS
AND METAL IONS FROM AQUEOUS
SOLUTIONS BY ACTIVATED CARBONS
BASU SAHA
Loughborough University
Loughborough, United Kingdom
INTRODUCTION
World population growth has triggered agricultural and
industrial growth for the sustenance of the day-to-day
demands of the population. However, the increase in
agricultural and industrial growth has also led to a
corresponding growth in pollution related to the two
activities. The use of pesticides to boost food production
has caused increased contamination of drinking water,
thus putting pressure on governments to invest in the
provision of clean water. The same can also be said
about the increase in industrial activities, which has
seen, among many of the pollutants, the increase of
heavy metals in the water bodies. Since the installation of
water pollution control schemes are not profit-generating
activities, there is a lot of ongoing research into finding
efficient but inexpensive ways of separating pollutants
from drinking water.
The ban on the use of atrazine for nonagricultural
purposes in the United Kingdom in 1993 has resulted in
the introduction of alternative herbicides. Among them is
benazolin, which can be a potential groundwater pollutant
owing to its high water solubility when compared with
that of atrazine. A report compiled by the Environment
Agency (1) revealed that, for the first time in England
and Wales, benazolin has become one of the 10 most
frequently occurring herbicides in surface water. The
potential harmful effects of benazolin as a herbicide
necessitates that it must be removed from water bodies
if present at concentrations higher than that required
by environmental regulations. However, conventional
water treatment processes, such as filtration, disinfection,
coagulation, and sedimentation, have been reported to be
more effective in removing less water-soluble and easily
degradable pesticides (2). Ozone technology is a potentially
attractive technique in wastewater treatment. However,
in some cases, the ozone degradation products can be
equally harmful. Lambert et al. (3) have investigated
degradation efficiency of benazolin and other herbicides
using ozone and a combination of ozone and hydrogen
peroxide. They realized that for benazolin, doses of
1.0 mg O3 /l, 2.0 mg O3 /l, and 3.1 mg O3 /l, degradation
efficiencies of 55%, 68%, and 88% could be achieved,
respectively. The combination of ozone and hydrogen
peroxide further increased the degradation efficiency,
but they recommended complementary techniques for
other herbicides. Adsorption on activated carbon is
one of the well-established and effective techniques for
the removal of herbicides from water. The adsorptive
efficiency of activated carbon is because of its well-
developed internal pore structure, surface area, and
surface reactivity. Typically, activated carbon consists
of interwoven micropores, mesopores, and macropores,
and these are responsible for the high surface area of
the carbons. The existence of surface functional groups
in an activated carbon matrix means that they can be
manipulated by thermal or chemical treatments to produce
adsorbents that are tailored for particular functions.
As wastewater streams consist of a mixture of
pollutants that can generally be classified as organic and
inorganic pollutants, tailoring an adsorbent to remove
both classes of pollutants would give a significant process
advantage. Among the inorganic pollutants, heavy metals
form a major part in the classification. Heavy metals
constitute a class of environmental pollutants that are
 REMOVAL OF ORGANIC MICROPOLLUTANTS AND METAL IONS FROM AQUEOUS SOLUTIONS BY ACTIVATED CARBONS
of particular concern in the treatment of industrial
wastewater because of their potential danger to humans
and the environment. Therefore, they must be treated if
present at concentrations higher than those specified by
the environmental regulations.
Activated carbons have been proven to be an effective
adsorbent for the removal of organic compounds from
water (4,5). Research has also shown that they have
a potential to remove heavy metals, especially in its
modified state. For example, a number of authors have
oxidized activated carbons to enhance sorption of heavy
metals (6–9). The oxidation processes have mainly been
by use of nitric acid, ozone, and air. The three methods
have always produced the same results, that of producing
weakly acidic functionality on the surface of the carbon.
Saha et al. (10) compared the metal sorption capacity of a
conventional carbon and its oxidized form. The adsorption
of herbicides and trace metal ions onto activated carbon
is discussed in this chapter. In particular, we discuss the
effect of modification of activated carbon on the uptake of
copper and benazolin. The effect of post-treatment of the
oxidized carbons on the uptake of the target pollutants is
also highlighted. The adsorbents were characterized in a
bid to understand their chemical and physical properties,
which gives a better understanding of the nature of the
association of the target pollutants and the adsorbents.
BACKGROUND
Herbicides have undoubtedly contributed to the growth
of agricultural productivity. For example, it has been
reported that a $4 billion U.S. investment in herbicide
control saved $16 billion U.S. in crops every year (11).
However, the benefit realized from the use of herbicides
does not factor in the negative effects they have on
human health.
As a result of their negative impact on human
health and the environment, stringent legislation has
been imposed on the purity of drinking water. The
European Union (E.U.) agreed to have the drinking water
Directive Admissible Concentration for single herbicide
or pesticide at a maximum of 0.1 µg/L. In Great Britain,
the National Rivers Authority (12) showed that eulan and
permethrin (mainly in effluents from the textile industry)
had exceeded their Environmental Quality Standard
(EQS) limits. Also, diazinon, a sheep dip insecticide,
had also exceeded its EQS limits. One major concern
was that the most commonly used, herbicides atrazine,
diuron, bentazone, isoproturon, and mecoprop exceeded
the standard 0.1 µg/L in surface waters, which has been
and always will be a major concern simply because most
groundwater sources used for drinking water have no
treatment facilities designed to remove herbicides. Also,
the most common water treatment processes, such as
sedimentation, disinfection, coagulation, and filtration,
can effectively remove only selected herbicides, especially
those that have low solubility in water. On the other
hand, advanced water treatment processes (mainly using
adsorption onto activated carbon) have proved to be the
most efficient and reliable method for the removal of
aqueous dissolved organic herbicides. Hence, a lot of
507
studies have gone into trying to improve the removal
efficiency of the existing adsorbents. In improving the
adsorption capacity of adsorbents, attention has been
turned on to the number of variables that affect the
adsorption capacity of the activated carbons.
As pollution in wastewater is because of a cocktail
of pollutants, the use of activated carbon could be a
viable way of simultaneously adsorbing pollutants of a
different nature, which has a further advantage of cutting
down on the number of wastewater processing units.
Hence, the objective of this research is to investigate
the effect of simultaneous adsorption of a heavy metal
(copper) and a selected herbicide (benazolin) on as-
received and surface-modified activated carbons. The
adsorbents chosen for this work are as-received F400
supplied by Chemviron (USA) and various surface-
modified samples of F400 that were prepared in the
author’s laboratory. Benazolin is highly soluble in
water (solubility 600 mg/l) when compared with atrazine
(solubility 33 mg/l). Properties, structural formula, and
three-dimensional representation of benazolin are given in
Table 1 and Fig. 1, respectively. Copper has been chosen as
a representative of heavy metals. In order to understand
the sorption mechanism of the target pollutants, physical
and chemical characterization were performed for all
the adsorbents. The other objective of the study is to
investigate the mechanism of adsorption of the chosen
pollutants. If the mechanism is fully understood, it is
possible to produce tailor-made engineered adsorbents for
various purposes in water treatment.
EXPERIMENTAL
Adsorbent Materials
A coal-based commercial granular activated carbon, F400
was used for surface modification with nitric acid. The
Table 1. Properties and Structural Formula of Benazolin
Solubility
Chemical in Water
Herbicide Formula Structural Formula (mg/L) pKa
Benazolin C9 H6 ClNO3 S S 600 3.04
O
N
CI CH2CO.OH
Figure 1. 3-D molecular representation of benazolin.
508 REMOVAL OF ORGANIC MICROPOLLUTANTS AND METAL IONS FROM AQUEOUS SOLUTIONS BY ACTIVATED CARBONS
carbon sample was obtained from Chemviron (USA).
Before the oxidation process, the carbon was washed
in distilled water and dried at 383◦ K for 24 hours. The
dried sample was then subjected to different modes of
modification using part of a method described by Saha
et al. (7) and Chingombe et al. (13).
Sample Preparation
A known amount of the washed and dried sample was
introduced into the reaction vessel that contained nitric
acid and distilled water at a ratio of 1:1 (v/v). The reaction
mixture was heated to 363◦ K, and the reaction was
allowed to continue for 6 hours with continued stirring.
The reaction solution was removed and a fresh reaction
mixture of nitric acid and water was introduced. The
reaction was allowed to continue for an additional 3 hours
under the same reaction conditions. The oxidized carbon
was then washed with distilled water until no further
change in pH could be detected.
The resultant water-washed samples (after oxidation)
were divided into three portions. The first portion was
denoted AC1, and this is a sample that was washed with
water after the oxidation reaction. AC2 was a sample
that was washed with 0.1 M sodium hydroxide to remove
humic acids that were formed during the oxidation process.
The washing was continued until no further coloration
from the humic acids could be detected. Reconditioning
of the sample to the hydrogen form was performed using
0.1 M hydrochloric acid, and further washing with distilled
water was done until the pH of the supernatant solution
stabilized to a pH of about 4.0. The conversion was deemed
complete when the inlet and the outlet pH of the liquid
flowing in the column were the same. The third sample,
denoted AC3, was a result of heating the third portion
of the oxidized sample at a temperature of 580◦ K under
vacuum of 2 mmHg for 12 hours to remove humic acids.
ANALYSIS OF METAL ION AND HERBICIDE
Copper
The copper ion content was analyzed using a Var-
ian SpectraAA-200 atomic absorption spectrophotometer
(AAS). The AAS was set on the flame mode using air-
acetylene mixture.
HPLC Analysis Method
As preconcentration of benazolin before analyzing on the
HPLC was required, a trace analysis method had to
be developed for the quantification of the herbicide. A
Hewlette Packard 100 series HPLC consisting of a diode
array detector, column thermostat, auto sampler, and a
binary pump was used for analysis. Jones Chromatograph
supplied a Genesis C18 silica column (4 µm, 150 × 3.0 mm)
combined with a 1 cm Genesis guard column. The detector,
flow rate, and injection volume were set at 222 nm,
0.88 ml/min, and 100 µl, respectively, and this gave the
retention time of benazolin at 2.01 minutes. The column
temperature was also set at 311◦ K. The eluent consisted
of a solution of 40% acetonitrile and 60% buffer (10 mmol
KH2 PO4 at pH 3). The calibration curve for benazolin
Table 2. Procedures for Solid Phase Extraction (SPE)
Conditioning 500 mg, s-triazine, 6 ml SPE columns were
conditioned by passing 10 ml of
HPLC-grade methanol followed by 10 ml of
2% MeOH in water under gravity.
Extraction 500 ml sample solution were passed through
the column at a flow rate of 1 ml/min.
Elution 6 ml of acetonitrile was allowed to soak in the
columns for 20 minutes before being passed
through at approximately 0.5 ml/min and
the eluent collected in 7 ml vials. The
columns were then blown dry with
nitrogen and the solvent was evaporated to
dryness using a stream of nitrogen.
Reconstitution 500 µl of 40% acetonitrile, 60% 10 mmol
KH2 PO4 buffer at pH 3.0 was added to the
vials to redissolve the analytes. The vials
were shaken vigorously and their contents
transferred to HPLC auto-sample vials.
showed that correlations were linear over the range
(0–8 ng/µl).
Solid phase extraction (SPE) technique was employed
for the preconcentration of benazolin before analysis. The
complete procedure for SPE is given in Table 2.
CHARACTERIZATION OF ADSORBENTS
All samples were characterized to investigate the effect
of the treatment and the chemical as well as the physi-
cal characteristics of these materials. The characterization
included scanning electron micrographs (SEM), Brunauer-
Emmett-Teller (BET) and Langmuir surface area mea-
surements, pore size distribution, Fourier Transform Infra
Red spectroscopy (FTIR) analysis, sodium capacity mea-
surement, pH titration, zeta potential measurements,
and X-ray photoelectron spectroscopy (XPS) analysis and
elemental analysis. The detailed characterization proce-
dure and the results have been reported by Chingombe
et al. (13). The results indicated that a significant change
occurred in the structure of the precursor by oxidation
and subsequent treatments. Characterization of the adsor-
bents revealed that modification of the carbon by oxidation
resulted in the introduction of weakly acidic functional
groups, and the presence of such groups was confirmed
by FTIR, pH titration, zeta potential measurements, and
sodium capacity results. Sodium capacity was enhanced
by a factor of up to 8 for the oxidized carbons compared
with as-received samples.
Batch Adsorption Experiments
The sorption of copper on the F400 carbon range is
shown in Fig. 2. The results clearly show that AC1 and
AC2 have the same adsorptive capacity for copper when
compared with AC3 and unoxidized F400. However, among
the adsorbents employed, as-received F400 showed the
least adsorptive capacity. The enhanced capacity observed
on AC1, AC2, and AC3 could be as a result of the
oxidation process, which generated carboxylic functional
groups as evidenced in the FTIR spectra and the high
sodium capacity of the above sorbents. These groups
 REMOVAL OF ORGANIC MICROPOLLUTANTS AND METAL IONS FROM AQUEOUS SOLUTIONS BY ACTIVATED CARBONS
1.4
Copper uptake, mmol/g
1.2
1
F400
0.8 AC3
0.6 AC1
AC2
0.4
0.2
0 adsorption capability of the adsorbents is the electrostatic
0 0.05 0.1 0.15 0.2
Equilibrium concentration, mmol/l
Figure 2. Equilibrium sorption results of copper on F400 and
the oxidized carbons.
would be dissociated in aqueous solution, and they are
responsible for copper sorption. The oxidized and heat-
treated sample AC3 showed reduced capacity compared
with AC1 and AC2, which could be because of the
decomposition of the carboxylic acid groups during the
heat treatment. Toledo-Bautista et al. (14) made similar
observations when they adsorbed Cr(III) on a range of
conventional and modified carbons.
The formation of metal surface complexes on the
oxidized carbon involving cooperative action seems quite
likely. Approximate calculation of oxygenated functional
group density per unit area for F400(ox) yields a value
of 0.02 functional groups per square angstrom. Given
that the oxidized carbons evaluated in the present study
possess a large proportion of pores in the region of 10–20
angstrom and the diameter of the hydrated metal ions is
approximately 8 angstrom (15), it is reasonable to assume
a cooperative binding mechanism (see Fig. 3).
Figure 4 shows the adsorption of benazolin on all the
adsorbents under study. The results clearly show that
benazolin adsorption on F400 is slightly higher than AC2
and reasonably higher than AC1 and AC3. However, AC1
and AC3 seem to have similar sorptive performance,
whereas AC2 showed a higher performance than the
other oxidized carbons, which can be expected because
there was a significant loss in micropores for samples
AC1 and AC3 as obtained by the pore size distribution
O
C
O O
O C
O O
C OH
OH C H
O H
O O O O
C
Cu2+ (H2O)6
O C O Cu O
C O
O C
O
C Figure 3. Postulated complexation re-
O action between copper (II) and oxidized
509
analysis (13). It is generally accepted that sorption of an
organic molecule takes place mainly in the micropore
region of the adsorbents, because in this region, the contact
points between the adsorbent and the molecules is higher
than in mesopores and macropores. Also, it is reported that
an overlap of potential forces exists in the micropores,
which results in increased adsorption forces. Therefore,
naturally, the capacity of AC1 and AC3 is expected to
be low. Another factor in the difference between the
interaction between the adsorbent and the adsorbate.
The zeta potential results show that the isoelectric
point (IEP) of the oxidized carbons are significantly lower
(1.5) than that of the conventional carbon (6.0) because
of the introduction of weakly carboxylic acid groups with
dissociation between pH 2 and 6. As the experiments are
performed at pH 5, it would imply that the surface of the
oxidized carbons is predominantly negative because of the
dissociation of the carboxylic groups. Also, the pKa value
of benazolin (3.04) is lower than the pH of the solution
experiment, and this condition would render mutual
repulsion between the adsorbent and the adsorbate.
Adsorption of benazolin is more favorable on F400 because
of its high IEP. Streat and Horner (4), when comparing
adsorption of some herbicides (including benazolin) on
F400 and hypercross-linked polymers, realized that F400
outperformed the polymeric sorbents. They postulated
25
Equilibrium concentration
on adsorbent (µmol/g)
20
15 F400
AC1
AC2
10 AC3
5
0
0 0.02 0.04 0.06 0.08 0.1 0.12 0.14
Equilibrium solution concentration (µmol/l)
Figure 4. Adsorption of benazolin on F400 and the oxi-
dized carbons.
C
C
C
O O
H
H
carbon surface.
510 REMOVAL OF ORGANIC MICROPOLLUTANTS AND METAL IONS FROM AQUEOUS SOLUTIONS BY ACTIVATED CARBONS

that adsorption of benazolin on F400 was through the 25

Equilibrium concentration
on adsorbent (µmol/g)
dispersion interaction of the π electrons in the respective
20
aromatic systems by a donor-acceptor mechanism. They
further noted that the chloro (–Cl) group on benazolin is 15
an electron-withdrawing group, therefore implying that a
reduction of electron density occurs in the π ring system 10
Benaz + 10 ppm Cu
of the herbicide. Hence, benazolin would act as an electron 5 Benaz + 5 ppm Cu
acceptor in the donor-acceptor mechanism involving F400. Benaz
This observation is consistent with investigations carried 0
out by Radovic et al. (16) when they considered the 0 0.05 0.1 0.15 0.2
adsorption of aromatics possessing electron-withdrawing Equilibrium solution concentration (µmol/l)
and electron-donating functional groups on conventional Figure 6. Simultaneous adsorption of copper and benazolin
and chemically modified activated carbons. Chemical on AC1.
modification involved samples that were subjected to
nitration and others that were oxidized. Heat-treated
1.2

Copper uptake, mmol/g

carbons were also considered. They used nitrobenzene and
aniline as the target adsorbates. Nitrobenzene contains 1
nitro (–NO2 ) group, which is known to be electron 0.8 AC3
withdrawing, whereas aniline possesses an amino (–NH2 ) AC2
0.6 AC1
group, which is electron donating. As nitration and
0.4 F400
oxidation reduces the π -electron density in the graphene
layers of the carbon surface, which would in turn favor 0.2
adsorption of aromatics with electron donor functional 0
groups, it was realized that maximum adsorption uptake 0 20 40 60 80
of aniline (with electron donor group) were attained on Initial benazolin concentration, ppb
oxidized carbon surfaces. On the other hand, maximum Figure 7. Effect of the initial concentration of benazolin on
adsorption uptake for nitrobenzene were found on heat- copper uptake.
treated carbon surfaces (which possess higher π -electron
density than the oxidized carbons). Hence, from the
uptake is distinctly affected in the presence of copper ions.
view point of Radovic et al. (16), it would be reasonable
Also, Fig. 7 shows the effect of the initial concentration of
to postulate that benazolin adsorption is favored by
benazolin on the uptake of copper. The results clearly show
adsorbents with high π -electron density. However, in
that as the initial concentration of benazolin increases, a
the oxidized carbons employed in this work, such a
corresponding decrease occurs in the uptake of copper up
mechanism would be highly unlikely because of the
to the point when the initial concentration of benazolin
introduction of oxygen groups in the form of carboxylic
is at approximately 20 ppb, that is when the uptake
acids. The carboxylic acid functional groups are electron
of copper normalizes. The immediate conclusion that
withdrawing, and this condition reduces the π -electron
could be drawn from the two sets of results is that a
density in the ring systems of the carbon. Therefore,
competition exists for the same adsorption sites by copper
the dispersion interaction between the herbicide and the
and benazolin. As mentioned previously, the adsorption
oxidized carbon is reduced. Hence, unoxidized F400 would
of benazolin is favored through electron donor-acceptor
be expected to have a higher benazolin capacity than the
type of a mechanism, which involves π electron in the
oxidized carbons.
ring systems of the carbon. Biniak et al. (6) showed that
Figures 5 and 6 show the adsorption isotherms of
copper ions had the capacity to interact with the π
benazolin in the presence of copper at 5 ppm and
electron in the ring systems of the carbon. In their
10 ppm for unoxidized F400 and AC1 and AC3 samples,
work, they employed XPS and FTIR to study the state
respectively. For both cases, it is apparent that benazolin
of the copper adsorbed on different carbon surfaces. They
investigated three types of modified activated carbons.
25 One of the modified carbons was obtained by annealing
a conventional carbon in a vacuum at 1000◦ K. This
Equilibrium concentration
on adsorbent (µmol/g)

20 carbon had a low oxygen content of 2.72% compared

with oxidized and ammonia-treated carbons, which had
15
10.10% and 6.20%, respectively. The annealed carbon was
10 virtually devoid of acidic functional groups. XPS and FTIR
results indicated that the dominant mechanisms of copper
5 Benaz adsorption on heat-treated active carbons could be dipole-
Benaz + 10 ppm Cu
Benaz + 5 ppm Cu dipole (π -d) interactions between graphene layers and
0 metal ionic species as well as spontaneous electrochemical
0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 reduction of copper ions. Hence, it is expected that
Equilibrium solution concentration (µmol/l) benazolin and copper would compete for the same sites
Figure 5. Simultaneous adsorption of copper and benazolin derived from the π electrons in the graphene layers of
on F400. the modified adsorbents. However, conclusive evidence of

such competition would only be proved by employing an
annealed carbon adsorbent.
CONCLUSIONS
Oxidation of the conventional carbon and post-treatment
of the resultant carbons had an influence on their
uptake for copper and benazolin. Oxidation resulted in
the introduction of weakly acidic carboxylic functionality,
which markedly enhanced the uptake of copper ions. The
isoelectric point (IEP) was also reduced as a result of
oxidation. Oxidation of the carbon resulted in altered
pore size distribution. Reduction in microporosity was
observed, and it also explained the low uptake of
benazolin by oxidized sample compared with as-received
F400. However, post-treatment by washing with sodium
hydroxide seemed to open up the pores for AC2 sample.
Simultaneous adsorption of benazolin and copper showed
that direct competition existed for the active sites, and
possibly the two adsorbates targeted the π electron cloud
on the graphene layers of the adsorbent. Copper uptake by
the conventional and modified samples decreased in the
presence of benazolin and vice versa. The decrease in metal
uptake was dependent on the concentrations of metals in
multicomponent aqueous solutions. The effect of copper
on benazolin uptake was more pronounced than the effect
of benazolin on copper uptake from water. Further work
is necessary to fully understand the precise mechanisms
of adsorption of complex aromatic molecules onto carbon,
and this forms the basis of our continued work. Research is
in progress in the author’s laboratory to provide conclusive
evidence on the interaction of these adsorbates with the
modified carbon surface.
The environment is challenged by other micropol-
lutants, namely polychlorinated hydrocarbons, organic
oxygen compounds, surfactants, pesticides, aliphatic
intermediates developing from the chemical indus-
try, and, of course, endocrine disrupting compounds
(EDCs). Further study of activated carbon continues to
explore for potential solutions to these important prob-
lems.
Acknowledgments
I would like to express my sincere gratitude to Professor Michael
Streat for his support, continuous encouragement, and advice
for conducting the research in the area of adsorption and ion
exchange and development of tailored sorbent materials for
environmental remediation. I am indebted to Dr. Hadi Tai and
Purazen Chingombe for their meticulous experimental work.
Funding from European Community, EPSRC, Severn Trent
Water Ltd., and Commonwealth Scholarship Commission is
gratefully acknowledged.
BIBLIOGRAPHY
1. Environment Agency. (2002). The Annual Report of the
Environment Agency pesticide monitoring programme. UK.
2. Kouras, A., Zouboulis, A., Samara, C., and Kouimtzis, T.
(1998). Environ. Pollut. 103: 193–202.
3. Lambert, S.D., Graham, N.J.D., and Croll, B.T. (1996). Ozone
Sci. Engineer. 18: 251–269.
4. Streat, M. and Horner, D.J. (2000). Trans. IChemE. 78(B2):
363–382.
MOLECULAR NETWORK DYNAMICS 511
5. Streat, M., Malik, D.J., and Saha, B. (2004). In: Ion Exchange
and Solvent Extraction. Vol. 16. A.K. SenGupta and
Y. Marcus (Eds.). Marcel Dekker, New York, pp. 1–84.
6. Biniak, S., Pakula, M., Szymanski, G.S., and Swiatkowski, A.
(1999). Langmuir 15: 6117–6122.
7. Saha, B., Tai, M.H., and Streat, M. (2001). Trans. IChemE.
79(Part B): 211–217.
8. Jia, Y.F., Xiao, B., and Thomas, K.M. (2002). Langmuir 18:
470–478.
9. Saha, B., Tai, M.H., and Streat, M. (2003). Chem. Eng. Res.
Des. 81: 1343–1353.
10. Saha, B., Tai, M.H., and Streat, M. (2001). Trans. IChemE.
79(Part B): 345–351.
11. Department of the Environment. (1996). Pesticides in Water.
Department of the Environment Report, The Working Party
on the Incidence of Pesticides in Water.
12. National Rivers Authority. (1998). Pesticides in the aquatic
environment. Water Quality Series 26: 12.
13. Chingombe, P., Saha, B., and Wakeman, R.J. (2004). 16th
International Congress of Chemical and Process Engineering,
Prague, Czech Republic.
14. Toledo-Bautista, I., Utrilla-Rivera, J., Garcia-Ferro, M.A.,
and Castillo-Moreno, C. (1994). Carbon 32(1): 93–100.
15. Nightingale, E.R.J. (1959). Phys. Chem. 63: 1381–1386.
16. Radovic, L.R. et al. (1997). Carbon. 35(9): 1339–1348.
MOLECULAR NETWORK DYNAMICS
D.L. MARRIN
Hanalei, Hawaii
Perhaps the twentieth century’s most startling discovery
about liquid water is that it functions as a vast
interconnected network, rather than as a random
collection of independent molecules. Individual water
molecules serve as network elements, which constitute
the building blocks for the fundamental water network.
Elements in the water network are connected to one
another via magnetic-type linkages known as hydrogen
bonds. In solid water (ice), each water molecule hydrogen
bonds with all four of its nearest neighbors in forming
a perfect tetrahedron, which is a three-sided pyramid
comprising the most basic molecular geometry of water
(see Fig. 1). As ice melts into liquid water, some
of the hydrogen bonds between water molecules are
broken (causing the tetrahedron to distort), whereas the
remaining bonds transition to an ultra-dynamic state
whereby they are switched as rapidly as a trillion times per
second. The dynamics of this unique molecular network are
believed to underlie many of water’s anomalous properties.
NETWORK BEHAVIORS
This frantic switching of hydrogen bonds, which consti-
tutes the underlying dynamism of water’s molecular net-
work, permits water to flow and to behave as a liquid even
though it retains much of the molecular geometry of a solid.
Physical chemists have long modeled liquid water as a
space-filling network of individual water molecules, within
512 MOLECULAR NETWORK DYNAMICS
Hydrogen atom
H
Covalent bond
Oxygen atom
H O
H biomolecules (5). Water releases its hydrogen bonds with
Lone pair of electrons
H
Figure 1. This diagram represents the simplest tetrahedral unit
of water’s H-bonded network (1). Two of the four hydrogen atoms
comprising the tetrahedron are covalently bonded to the central
oxygen atom (forming an individual H2 O molecule), whereas
the remaining two hydrogen atoms (donated by each of two
neighboring water molecules) are hydrogen bonded to the oxygen’s
lone pairs of electrons. Every water molecule in the network can
act as both a hydrogen donor and a hydrogen acceptor in bonding
with its nearest neighbors.
which all potential hydrogen bonds were described as
either unbroken (ON) or broken (OFF) (2). More recently,
scientists have characterized the shuffling (i.e., breaking
and forming) of hydrogen bonds comprising liquid water’s
network as a nonlinear process and, consequently, have
been unable to decipher the rules that govern the chang-
ing of molecular interconnections. This inability does not
imply that the switching rules are haphazard—only that
science’s observational and mathematical techniques are
currently overwhelmed by water’s dynamism. Even scien-
tists who study much simpler computer-generated systems
have found it difficult to identify switching rules solely
by analyzing network patterns and behaviors. Instead,
researchers arbitrarily set the switching rules for these
artificial systems and then observe corresponding network
behaviors.
Whereas network hierarchies within liquid water have
only recently been investigated, structural hierarchies are
well described. Water exhibits structural self-organization
on at least three distinct levels: molecules, clusters
of molecules (e.g., hydration shells), and clusters of
clusters (e.g., micelles). Self-organization implies that
water molecules or molecular clusters arrange themselves
into increasingly complex states. It is this self-organization
that reportedly permits a network of simple interconnected
elements, along with their switching rules, to display
coherent behaviors (3). Although the primary water
network is held together by short-range hydrogen bonds
between neighboring molecules, it may either behave
as an extensive system with global coherence or as
individual smaller regions with local order (4). These
two behaviors have been classically linked to network
patterns displaying chaotic or disordered (i.e., bulk liquid
water) and more ordered (i.e., water clusters) properties,
respectively.
BIOLOGICAL STRUCTURE AND FUNCTION
Molecular biologists originally considered water to be
merely an inert solvent within which biomolecules (e.g.,
proteins and nucleic acids) structured and activated
themselves. More recently, water has been perceived
as a key structural and functional component of many
biomolecules, whose interaction with the surrounding
environment is often mediated through hydration shells,
which are immense water networks surrounding the
biomolecular components so that protein or DNA assembly
may begin. Conformational changes in the biomolecule
then restructure both the internal (bound or integral) and
external (hydrating) water that, in turn, facilitates more
changes in the biomolecule. This iterative process appears
to repeat until assembly and folding mechanisms yield a
bioactive crystal.
In addition to the structural importance of water in
biomolecules, the insertion and removal of water also has
profound effects on their functioning. One of the most
remarkable functions that proteins perform is that of a
biochemical catalyst or enzyme, which permits chemical
reactions to proceed at rates that are millions of times
faster than would occur under ambient conditions. Water
appears to play several roles in the functioning of enzymes.
First, water molecules located at the interface of the
protein-substrate complex often provide the free energy
that actually binds the enzyme to its substrate. Second,
water may be excluded from the active site until the
moment of catalysis; at which time, it is allowed to enter
the active site and reconfigure hydrogen bonds (permitting
the reaction to proceed). Finally, the enzymes actually
structure the water network or networks around them,
thus enhancing the rate of catalysis (6).
In addition to its role as a structural component of
biomolecule, water hydrates DNA as it would any other
solute in an aqueous solution. It appears that water is
most highly ordered adjacent to the DNA molecule and
is less ordered with each successive hydration shell. This
highly ordered water is sometimes referred to as vicinal
or biological water (6). As was the case for proteins, each
subunit of the DNA seems to have its own water network
that interacts with other networks and, through successive
hydration shells, with bulk water. If the geometry of
the DNA molecule is contorted, the hydration shells and
associated geometric clusters seem to be likewise affected.
Similarly, the DNA molecule may contort in response
to changes initiated by the water that hydrates it. The
reasons for twisting and folding DNA molecules may
include stuffing it into chromosomes or unraveling it
during transcription and replication.
COMPLEX DYNAMICS
Scientists working with the pentagonal water cluster,
which serves as a basic building block for many large
hydration shells, observed that water’s vibrational states
are caused by a combination of bending and stretching
motions of its bonds. Molecular vibrations are expressed
in the units of cycles per second, or hertz. Although
 MOLECULAR NETWORK DYNAMICS 513

water’s covalent bonds vibrate at about 1014 hertz (one- H

hundredth of a picosecond per cycle), its hydrogen bonds Total H H O H
vibrate within the lower frequency range of 1012 to 1013 shell H H
O O H
hertz (one-tenth to one picosecond per cycle). The time O H H O H O H O
H
frames for one cycle of these bond vibrations are shorter H H H
H O H O H
than those measured for the exchange hydrogen bonds in H H O
H O O H H H
H
H
water’s bulk liquid network (more than one picosecond). H H
H O H O O O
One picosecond equals a time period of one-trillionth of H O H
H
a second (10−12 seconds). In addition to the dynamics Primary
O
H H
O H
shell O H O
associated with vibrating and exchanging bonds, water’s H H
H O H O H
molecular network performs several concurrent activities O O
H O
that have been clocked at both faster and slower rates. H H H H H
H
As an example, ultrafast vibrational energy is trans- O H H O H O
H
H
O O
ferred in water on timescales less than one-tenth of a H H H H
picosecond (7). This ultrafast transfer could play a role H H O H O
H
in liquid water’s transporting vibrational energy among O H H O
Solute H H
certain solutes (e.g., proteins, DNA) and, perhaps, in trans-
ferring information within water’s local network. Recent Figure 2. This diagram illustrates the primary and total
spectroscopy studies have suggested that liquid water’s hydration shells surrounding an ionic, or electrically charged,
network may also function on timescales of less than a solute (1). In this case, a positively charged solute (e.g., a sodium
femtosecond (one-thousandth of a picosecond), which is ion in seawater) attracts the negatively charged oxygen atoms of
at least 100 times faster than the ultrafast vibrational surrounding water molecules. A solute’s effect on the orientation
energy. Moreover, some of the studies indicate that most of water molecules in the liquid network is greatest within the
molecules in the liquid network form only two (rather primary shell and diminishes as a function of distance from the
than three or four) hydrogen bonds with their neigh- hydrated ion. Theoretically, the total hydration shell includes all
water molecules in the network that are affected by the solute.
bors, perhaps creating long chains of molecules within a
disordered cluster network (8). In contrast to these sub-
picosecond dynamics, the reorientation of water molecules
and (3) varying orientations of the water molecules that
in a spatially restricted environment (e.g., the interface of
comprise ion hydration shells (9). These factors are
a micelle) occurs on the much slower timescale of hundreds
believed to profoundly affect the structure and dynamics
to thousands of picoseconds. Although further research is
of the water network comprising the world’s oceans, which
required to substantiate the recent observations, models
contain 97% of the planetary water. Perhaps scientists
describing both the structure and dynamics of water’s
will better understand the oceans when the secrets of
molecular network are rapidly evolving.
seawater’s molecular network dynamics are more fully
revealed.
SEAWATER
BIBLIOGRAPHY
Thus far, the discussion of water’s network has dealt
with either pure water or dilute aqueous solutions in
1. Marrin, D.L. (2002). West. In: Universal Water. Inner Ocean
which solutes comprise only a small fraction (e.g., less
Publ., Maui, HI, pp. 100–114.
than 0.1%) of the total solution. By contrast, the salts
2. Rice, S. (1975). Conjectures on the structure of amorphous
(ions) in seawater account for about 3.5% of the total
solid and liquid water. Topics Curr. Chem. 60: 109–200.
solution. Unlike the structure of pure water’s bulk phases
3. Varela, F., Thompson, E., and Rosch, E. (1991). The Embodied
and clusters, the molecular structure of seawater is
Mind. MIT Press, Cambridge, UK, pp. 87–90.
poorly described as a result of difficulties associated
4. Blumberg, R.L., Stanley, H.E., Geiger, A., and Mausbach, P.
with employing the investigative techniques that routinely
(1984). Connectivity of hydrogen bonds in liquid water. J.
probe the structure of pure water networks. Moreover, the Chem. Phys. 80: 5230–524.
high concentration of ions in seawater is hypothesized to
5. Westhof, E. (1993). Structural water bridges in nucleic acids.
significantly affect the water network. As the percentage In: Water and Biological Macromolecules. E. Westhof (Ed.).
of water molecules comprising hydration shells increases, CRC Press, Boca Raton, FL, Chapter 7.
the network structure of bulk water is compromised (see 6. Clegg, J. (1979). Metabolism and the intracellular environ-
Fig. 2). A salting-out effect is observed at very high ment: the vicinal-water network model. In: Cell-Associated
ion concentrations, when the interaction between water Water. W. Drost-Hansen and J. Clegg (Eds.). Academic Press,
molecules and ions is sufficiently weakened to permit the New York, pp. 363–413.
formation of salt crystals. 7. Nitzan, A. (1999). Ultrafast relaxation in water. Nature 402:
The most pertinent changes in the molecular water 472–475.
network that have been observed or hypothesized in 8. Zubavicus, Y. and Grunze, M. (2004). New insights into the
transitioning from dilute aqueous solutions to seawater structure of water with ultrafast probes. Science 304: 974–976.
include (1) the aforementioned ion–ion interactions, 9. Robinson, G., Wilse, S.B.Z., Singh, S., and Evans, M. (1996).
(2) overlapping hydration shells in which a single water Water in Biology, Chemistry and Physics. World Scientific,
molecule simultaneously hydrates more than one ion, Singapore, pp. 245–259.
514 IN SITU CHEMICAL MONITORING
IN SITU CHEMICAL MONITORING
MELISSA A. SINGER PRESSMAN
JOSEPH H. ALDSTADT, III
Genetic Technologies, Inc.
Testing Institute
Waukesha, Wisconsin
INTRODUCTION
For routine monitoring of environmental systems, in situ
measurements offer several distinct advantages over
laboratory analysis (1). In situ analysis allows much
higher rates of sample throughput. As a result of the rapid
dynamics of most environmental systems, it is important
to collect frequent samples to prevent undersampling of
the phenomena under study, which is essential to allow
for a more accurate characterization of the true dynamics
of the system and, therefore, a better understanding of
the biogeochemical processes involved (2). Additionally,
in situ analysis eliminates sample contamination that
inevitably develops during the transport and storage of
samples to the laboratory. Furthermore, current research
efforts in several laboratories have demonstrated that
relatively inexpensive monitors can be fielded that can
self-calibrate, can be remotely controlled, and can provide
reliable measurements in realistic environments (3). In
particularly, it is the inability to provide high data quality
at trace levels in complex matrices that has prevented
in situ measurements from gaining more favor.
In situ monitors have been classified broadly as either
‘‘chemical sensors’’ or ‘‘chemical analyzers.’’ Chemical
sensors depend on diffusion to bring the analyte to the
transducer surface. This type of sensor is mechanically
simple, sensitive, and practical from the standpoint
of low power requirements and small footprint, but
often suffers from poor selectivity and the inability
to self-calibrate. More importantly, sensors are based
on reversible chemical reactions, i.e., the interaction of
the analyte with the transducer surface, which makes
it difficult for these sensors to detect environmentally
relevant levels; that is, to detect trace levels of a chemical
species, the equilibrium binding (formation) constants
must be very large. For the reaction of analyte A with
ligand B to form a complex AB, the rate equations are (3):
aA + bB → ABrate = −(d/dt) [A] = −(d/dt) [B]
= k1 [A]a [B]b
xAB → A + Brate = −(d/dt) [AB] = k−1 [AB]x
The equilibrium constant for the dissociation of the
complex is equal to the ratio of the rate constants for
the dissociation of the analyte-ligand complex and the
formation of the complex, as shown in the following
equation (3):
KD = k−1 /k1
If the complex dissociates slowly, the time required to
return the signal to a steady baseline (‘‘clear down’’
time) is introduced. One way to avoid this problem
is to continuously renew the reagent at the sensor
surface. This strategy depends on molecular diffusion to
transport the analyte to and from the sensor surface;
this type of sensor is called a ‘‘renewable reagent
sensor’’ (4–6).
Chemical analyzers, on the other hand, depend on
active transport of the analyte into and away from
the monitor. Although analyzers lack the mechanical
simplicity of sensors, the addition of valves and pumps
in the design can more importantly provide a means by
which to accommodate self-calibration of the monitor.
The addition of these mechanical capabilities also
allows for multistep, complex chemistries to be used
as a way to increase sensitivity and selectivity of the
monitor.
Continuous flow analysis (CFA) is a more general
term than so-called ‘‘chemical analyzers,’’ referring to
any measurement process in which the concentration
of an analyte is determined based on a steady flow of
sample through the instrument (7). Nonchromatographic
CFA-based measurement systems are widely used for
industrial process control and environmental monitoring.
As the instrument is often placed in a hostile environment
in these types of applications, the relative simplicity
of the CFA approach is attractive (8). Initial attempts
to use CFA-based instruments were direct adaptations
of benchtop (i.e., laboratory) CFA systems that were
ruggedized (e.g., water-proofed) and taken to the sampling
site (9). This approach was burdensome, expensive, and far
from robust.
An example of an early effort of this type is
illustrated by the work of Elliott et al. in using a
segmented continuous flow analyzer, often called an
‘‘autoanalyzer,’’ for shipboard chemical analyses (10).
These researchers monitored nitrate concentrations in the
Everglades using a commercially available autoanalyzer
that they modified for the nitrate derivatization chemistry.
In their design, air bubbles were created in-line to
segment the liquid sample stream as a means to prevent
uncontrolled dispersion
More recently, in situ CFA monitors have been
developed that are not merely ruggedized bench-top
instruments. One of the more interesting instruments,
and the model which inspired this research, was described
in work published in 1994 by Jannasch et al. at the
Moss Landing Marine Laboratory in Monterey, CA (11).
This group of chemical oceanographers was interested
in accurately measuring nitrate ion in seawater with
high sampling frequency. To accomplish this goal, they
designed a CFA monitor ‘‘from the ground up’’ and
incorporated several novel features. In doing so, they
created a new paradigm in the science of chemical
measurements.
In the system developed by Jannasch et al., a well-
characterized, irreversible colorimetric method was used
to measure nitrate. The dissolved nitrate was determined
by reduction to nitrite in a neutrally buffered (imidazole)
solution by an in-line copperized cadmium surface. The
nitrite then merged with an incoming reagent stream
containing sulfanilamide (p-aminobenzenesulfonamide or

PABSA) in acid to undergo reaction to form a diazo-
nium ion. This species was then reacted with N-(1-
naphthyl)ethylenediamine dihydrochloride (NED) to form
an azo dye of high molar absorptivity that was detected by
molecular absorbance spectroscopy.
The novel in situ monitor that the Moss Landing
group developed used nonmechanical osmotic pumps to
both deliver reagents to the system and to propel the
continuous flow sample stream through the microscale
manifold (∼1 mm diameter conduits) (11). For the two
reagent osmotic pumps (i.e., for PABSA and NED), the flow
was driven by the osmotic pressure differential developed
across the semipermeable membrane that separated a
saturated brine solution from a solution with deionized
water (12). The pressure created from the diffusion of
deionized water across the semipermeable membranes
resulted in compression of inner bags containing the
reagents and their pulse-free delivery to the manifold (at
∼4 µL per hr). For the sample pump, conventional osmotic
pumps were modified by removing the inner bags and
mounting the rigid permeable membranes in an enclosed
container. Four of these modified osmotic pumps were
then coupled to draw sample through the manifold at
12 µL per hr.
Microscale flow manifolds can be fabricated by ‘‘cold-
forging’’ PVC: A wire bent in the desired flow microconduit
configuration was pressed onto the surface of a flat-
milled sheet of PVC. Following removal of the wire,
∼0.5-mm diameter holes were drilled into the plate for
connection of components. The optical system in the Moss
Landing nitrate analyzer used a simple light-emitting
diode (LED) source with photodiode detector, with data
collected by a low-power programmable datalogger. The
monitor exhibited a detection limit of 0.11 µM and a
linear response to 20 µM nitrate. Measurements were
taken every 2 min and the response time was ∼30 min.
Quality control samples were measured automatically
every 48 hours. Jannasch et al. demonstrated the in situ
monitor’s stability by operating it continuously for
month-long periods in large tanks at the Monterey
Bay Aquarium.
The Moss Landing group’s research incorporated
several novel features—mainly the microscale flow design
using nonmechanical osmotic pumps for reagent delivery
and sample intake—into an in situ monitor capable of
frequent and reliable measurements in environmentally
complex samples. In addition, through the periodic
addition of nitrate standards, the monitor automatically
self-calibrated by correcting for baseline drift. Another
advantage to this system was its flexibility; using different
irreversible chemistries, the in situ monitor could be
readily adapted to measure other analytes for optical
detection and quantitation. A drawback of the system
was that it detected relatively high levels of analyte
(i.e., micromolar levels), albeit levels that are relevant to
studying this important nutrient. Concerns over particles
in the water eventually clogging the microconduit channels
are less serious, because small particles (approx. <50 µm)
cannot be drawn into the manifold at the low flow
rates used.
IN SITU CHEMICAL MONITORING 515
In 1999, Byrne and coworkers at the University of
South Florida developed a monitor based on long path-
length absorbance spectroscopy (LPAS), which could
be used for the determination of nitrate and several
other biologically important chemicals in seawater at
much lower levels. These workers used the same well-
characterized chemistry as Jannasch et al., but they
benefited from a 1-m liquid core waveguide as the
absorbance flow cell. Another difference with the Byrne
group’s approach was in the use of mechanical (peristaltic)
pumps instead of osmotic pumps—resulting in a five-
fold increase in response time. This monitor was field
tested recently in the Gulf of Mexico, and detection
limits for nitrate were observed to be approximately
2 nM (13,14).
The next key development was reported recently by
Worsfold and coworkers at the University of Plymouth.
A portable in situ monitor for the determination of trace
phosphate in surface waters was developed (15,16). This
work demonstrated a monitor that was built with a
forged plate similar in design to the work done by
Jannasch et al. The phosphate monitor had a throughput
of ∼225 determination per hr and a detection limit of
0.15 µM with 2.0% RSD (n = 9) reported. This monitor
was evaluated by mapping the phosphate concentration in
Port Phillip Bay in Australia for a brief (2 hr) period. The
agreement between determinations obtained using the
in situ monitor and data obtained using manual sampling
aboard the ship followed by laboratory analysis was within
±10%.
INSTRUMENT DEVELOPMENT FOR IN SITU MONITORING
OF CR(VI)
For the in situ measurement of relatively low levels
of important biological nutrients such as nitrate or
phosphate, the work of the Jannasch, Byrne, and Worsfold
groups (and others) was pioneering. However, clearly
the next challenge in developing in situ monitoring
instruments was the ability to measure relevant levels
of trace metal ions in surface waters by the in situ
approach. Hexavalent Cr was chosen as the target
analyte because it is regulated at nanomolar levels (low
parts per million) in surface waters. In addition, the
metropolitan-industrial areas of southern Lake Michigan,
specifically the Milwaukee Harbor area, contain industrial
sources that generate a concentration of total Cr in
surface water that ranges from 0.1 to greater than
100 µg/L (17). These reasons, coupled with the well-
characterized spectrometric chemistry (18–21), made Cr
a logical choice as a representative analyte to study for
the development of an in situ monitor that measures trace
metals in the environment.
A benchtop method for determining hexavalent
chromium in high particulate-containing surface waters
by sequential injection (SI) was developed (22). For utility
in particulate-rich waters, the relative performance of two
membrane-based methods for sampling was compared.
The first membrane approach was based on a commer-
cial design known as the ‘‘supported capillary membrane
516 IN SITU CHEMICAL MONITORING
sampler (SCMS)’’ that uses tubular membranes; the sec-
ond approach was based on a conventional parallel-plate
dialyzer (PPD) design that uses planar membranes. The
membranes were evaluated using the colorimetric method
for the determination of hexavalent chromium by com-
plexation with 1,5-diphenylcarbazide (DPC). Thin-walled
(∼200 µm) microporous (pore size ∼0.2 µm) polypropy-
lene membranes were equilibrated with DPC during each
sampling period. Formation of the DPC-Cr(VI) complex
allowed for efficient membrane transport; without the
membrane, Cr(VI) transport decreased by ∼90%. Factors
optimized included reagent concentration, sampling time,
flow rate, and spectrophotometric conditions. Optimal con-
ditions for the reagents were 2.00 mM DPC and 0.100 M
nitric acid, and sampling times of 600 s and 900 s for
the planar and tubular designs, respectively. The planar
(PPD) design increased the sensitivity relative to the tubu-
lar (SCMS) design by ∼225%. The PPD-SI method was
applied to the determination of dissolved Cr(VI) in high
particulate-containing surface water samples. Figures-of-
merit included a detection limit of <20 µg/L, precision of
1.1% relative standard deviation at 100 µg/L (n = 4), and
selectivity for dissolved Cr(VI) in several surface water
samples with high levels of particulate matter and poten-
tial interferences.
The optimized method conditions from the SI studies
was then applied to the design, fabrication, and testing of
an in situ remote monitor for the measurement of sub-part
per million levels in surface waters (23,24). The monitor
was based on a continuous flow analysis (CFA) design
in which a narrow conduit (1.0 mm) manifold was hot-
forged between Plexiglas plates. A miniature low-power
peristaltic pump was used to draw sample (1.5 mL/hr)
into the manifold where chromium (VI) in the sample
was derivatized to its diphenylcarbazide complex. The
absorbance of the complex was then measured in a
light-emitting diode/photodiode detection cell. Solenoid-
actuated osmotic pumps were used to introduce quality
control standards at user-determined periodicities to
achieve acceptable data quality. Figures-of-merit included
a detection limit of 0.25 mg/L (7 min cycle time), precision
of 2.5% relative standard deviation at 1 mg/L (n = 3),
and selectivity for dissolved Cr(VI) in several surface
water samples with moderate levels of particulate matter.
Continuous testing of the monitor over a three-week
period using Milwaukee Harbor water fortified with Cr(VI)
yielded results that were in close agreement (±5%) with
an ICP-MS reference method.
BIBLIOGRAPHY
1. Hanrahan, G., Ussher, S., Gledhill, M., Achterberg, E.P.,
and Worsfold, P.J. (2002). High temporal and spatial
resolution environmental monitoring using flow injection with
spectroscopic detection. TrAC, Trends Analytic Chem. 21: 233.
2. Rawls, R. (1998). Wanted: instruments to view ocean’s
dynamics. Chem. Engineer. News. 31.
3. Johnson, K.S. and Jannasch, H.W. (1994). Analytical chem-
istry under the sea surface: monitoring ocean chemistry
in situ. Naval Research Rev. 3: 4.
4. DeGrandpre, M.D., Baehr, M.M., and Hammar, T.R. (1999).
Calibration-free optical chemical sensors. Analytical Chem.
71: 1152.
5. Inman, S.M., Stromvall, E.J., and Lieberman, S.H. (1989).
Pressurized membrane indicator system for fluorogenic-based
fiber-optic chemical sensors. Analytica. Chimica. Acta. 217:
249.
6. Degrandpre, M.D. (1993). Measurement of seawater pCO2
using a renewable-reagent fiber optic sensor with colorimetric
detection. Analytical Chem. 65: 331.
7. Ruzicka, J. and Hansen, E.H. (1988). Flow Injection Analysis,
2nd Edn. John Wiley & Sons, New York.
8. Johnson, K.S., Coale, K.H., and Jannasch, H.W. (1992).
Analytical chemistry in oceanography. Anal. Chem. 64:
1065A.
9. Valcarcel, M., Cardenas, S., and Gallego, M. (2002). Contin-
uous flow systems for rapid sample screening. Trac-Trends
Analytic Chem. 21: 251.
10. Elliott, C.L., Snyder, G.H., and Cisar, J.L. (1989). A modified
Autoanalyzer-II method for the determination of NO3 − in
water using a hollow-Cd reduction coil. Communications in
Soil Science and Plant Analysis 20: 1873.
11. Jannasch, H.W., Johnson, K.S., and Sakamoto, C.M. (1994).
Submersible, osmotically pumped analyzers for continuous
determination of nitrate in-situ. Analytic Chem. 66: 3352.
12. Pump, A.O. (2000). Technical Note. Alzet, Inc.
13. Kaltenbacher, E.A., Byrne, R.H., and Steimle, E.T. (2001).
Design and applications of a chemical sensor compatible
with autonomous ocean-sampling networks. IEEE J. Oceanic
Engineer. 26: 667.
14. Bryne, R.K. (1999). Autonomous in situ analysis of the upper
ocean. Sea Technol. 71.
15. Hanrahan, G., Gledhill, M., Fletcher, P.J., and Worsfold, P.J.
(2001). High temporal resolution field monitoring of phos-
phate in the River Frome using flow injection with diode
array detection. Analytica. Chimica. Acta. 440: 55.
16. Lyddy-Meaney, A.J., Ellis, P.S., Worsfold, P.J., Butler,
E.C.V., and McKelvie, I.D. (2002). A compact flow injection
analysis system for surface mapping of phosphate in marine
waters. Talanta. 58: 1043.
17. Environmental Defense Fund. (2004). Available: http://
www.scorecard.org/.
18. Miller-Ihli, N.J. (1992). Chromium. Technol. Instrum. Anal.
Chem. 12: 373.
19. Lynch, T.K. and Wilson, J. (1984). Speciation of metals
in solution by flow injection analysis. Part 1: Sequential
spectrophotometric and atomic absorption detectors. Analyst.
109: 839.
20. Losi, M.E., Amrhein, C., and Frankenberger, W.T. (1994).
Rev. Environ. Contamin. Toxicol. 136: 91.
21. Kotas, J. and Stasicka, Z. (2000). Chromium occurrence in the
environment and methods of its speciation. Environ. Pollut.
107: 263.
22. Pressman, M.A. and Aldstadt, J.H. (2003). A comparison
of diffusion samplers for the determination of hexava-
lent chromium by sequential injection spectrophotometry.
Microchem. J. 74: 47–57.
23. Pressman, M.A. (2003). Development of Novel Analytical
Techniques and Methods for the Determination of Trace
Chromium in the Environment, Ph.D. dissertation, University
of Wisconsin-Milwaukee.
24. Pressman, M.A. and Aldstadt, J.H. (2005). Development and
Testing of an in situ Chemical Monitor for Quantitation of
Hexavalent Chromium in Natural Waters, in preparation.
 NITROGEN 517

NITROGEN to be mineralized in the soil by a chain of microbial

processes (Fig. 1). The release of ammonium and ammonia
KENNETH F. STEELE (total ammonia) from organic matter is referred to as
KRISTOFOR R. BRYE mineralization (3) or ammonification. Total ammonia is
University of Arkansas readily oxidized to nitrate. The conversion of nitrogen gas
Fayetteville, Arkansas to fixed form and from fixed form to nitrogen gas includes
four main conversion reactions that form the essential
parts of the nitrogen cycle (Fig. 1):
INTRODUCTION
1. fixation
Nitrogen (N) is ubiquitous in the environment and in 2. ammonification
the correct form is an essential nutrient for plants and 3. nitrification
animals. About 26.5% of Earth’s nitrogen is in rocks and
essentially all of the remainder (73%) is in the atmosphere 4. denitrification
as an inert monotomic gas (N2 ) that comprises 78% of
the atmosphere (1). Only a small amount of nitrogen is Fixation
present in the hydrosphere and biosphere, but nitrogen For N2 gas to be used by plants, it must first be fixed
is a crucial nutrient there. For nitrogen to be used by with hydrogen to form ammonium ions. Fixation is a
organisms, it must be in a usable form; it must be combined microbial process, and even the few types of plants that
or ‘‘fixed’’ with oxygen or hydrogen. The most common can use N2 directly are ultimately dependent on microbes
fixed form of nitrogen in ground water is the anionic also. For example, legumes can use N2 because of the
oxide—nitrate (NO3 − ), and the second most common form symbiotic relationship between the plant and Rhizobium
is the cationic form–ammonium (NH4 + ). Approximately bacteria that live in nodules in the plant roots. Other
85% of the nitrogen in stream water is in the form of means of fixation include conversion of atmospheric N2
organic N and most of the rest (NO3 − and NH4 + ) is to nitrate by lightning and industrial processes, such as
derived from the decomposition of organic matter (2). the Haber–Bosch process for producing ammonia (NH3
N2 is converted to organic nitrogen predominantly by gas) (4).
microorganisms and certain plants. About 96% of the
nitrogen in soil is bound in organic matter, and nitrate Ammonification
is a product of organic matter degradation (3). Nitrogen
occurs in five oxidation states, −3, 0, +1, +3 and +5, that Ammonification (also referred to as mineralization), the
are primarily the result of bacteria-induced oxidation. The process of converting organic nitrogen (e.g., urea) to
nitrogen species are listed in Table 1. ammonium (Eq. 1), generally occurs during decomposition
of animal and plant matter (4). Dissolution of ammonium
carbonate releases ammonium ions:
NITROGEN CYCLE
H2 NCONH2 +2H2 O −−−→ (NH4 )2 CO3 (1)
Fixation by plants can occur; leguminous plants convert Urea Ammonium carbonate
N2 to organic nitrogen compounds, but most plants cannot
use N2 directly. Most plants require atmospheric nitrogen As indicated by Eqs. 2 and 3, the relationship between
converted to ammonium or nitrate ions (4). Organic NH3 and NH4 + is pH dependent. Under alkaline
nitrogen is not directly available to plant roots and has conditions (pH > 9), formation of NH3 is favored due to
neutralization of NH4 + by hydroxide (OH− ) (Eq. 2). In
acid conditions (Eq. 3), ammonia gas is transformed into
Table 1. Common Forms of Nitrogen in the Environment
and Typical Concentrations of Nitrogen
Typical Nitrogen gas
Oxidation Concentration in N fixation N2 (g)
Species State Typical Source Water as N (e.g., NH3 formed)
Denitrification—
N2 0 Atmosphere 14 mg/L Immobilization NO3−
NO3 − +5 Oxidation of <0.6 mg/L background (conversion to protein)
organic matter/ to 250 mg/L for
atmosphere contaminated Nitrate
Organic N
sources NO3−
NO2 − +3 Intermediate form <0.1 mg/L
in nitrification,
denitrification, Ammonification
Nitrification—
(mineralization)
and reduction of NO2−
NO3 −
Ammonium
NH3 −3 Fixation of N2 <0.01 mg/L
NH4+
NH4 + −3 Ionization of NH3 <0.01 mg/L
N2 O +1 Reduction of NO2 − Negligible
Figure 1. Simplified nitrogen cycle.
518 NITROGEN
ammonium ions. For example, at 15 ◦ C, ammonia gas
makes up 0.027% of total ammonia (NH3 + NH4 + ) at a pH
of 6.0, but makes up 2.7% of total ammonia at a pH of
8.0. Although 2.7% is a small amount, it is a hundredfold
increase of ammonia (5).
NH3 + H2 O = NH4 + + OH−
NH3 + H+ = NH4 +
Other factors that affect the concentration of NH3 in water
are temperature and ionic strength. The concentration of
NH3 increases as temperature increases and decreases as
ionic strength increases. At 15 ◦ C and pH 8.0, the percent
NH3 in solution is 2.7%, whereas, at 30 ◦ C and pH 8.0,
the concentration increases to 7.5% (5). The effect of ionic
strength is minor for fresh to slightly saline waters.
Nitrification
Nitrification is the biological oxidation of ammonium ions
to nitrite followed by conversion of nitrite to nitrate.
Biological oxidation is carried out by Nitrosomonas and
Nitrobacter bacteria (6,7) as shown in Eqs. 4 and 5. In
addition to these two bacteria, (8) indicate that several
different autotrophic organisms mediate nitrification:
Nitrosomonas
2NH4 + + 3O2 −−−−−−−→ 2NO2 − (nitrite) + 4H+
bacteria
+ 2H2 O
Nitrobacter
2NO2 − + O2 −−−−−−→ 2NO3 − (nitrate)
bacteria
Because oxidation of nitrite to nitrate is a rapid
reaction and nitrite is metastable, nitrite is not abundant
(Table 1) (9). Note that if both oxygen and nitrate are
present in soil, bacteria preferentially use oxygen because
it yields more energy (1). Therefore, nitrification typically
occurs in oxic conditions, whereas denitrification occurs in
anoxic conditions.
Denitrification
Denitrification is the biological reduction of nitrate
to nitrogen gas by heterotrophic bacteria, such as
Pseudomonas denitrificans and Thiobacillus denitrifi-
cans (8). These bacteria require organic carbon for energy
(Equations 6 and 7).
Heterotrophic bacteria
NO3 − + organic carbon −−−−−−−−−−−−→ NO2 −
Heterotrophic bacteria
NO2 − + organic carbon −−−−−−−−−−−−→ N2 + CO2
+ H2 O
Because oxidation of organic carbon by oxygen yields more
energy, generally the oxygen concentration must be very
low (anoxic conditions) for nitrate to be used (4,6).
Analytical Terminology
Analytically, organic nitrogen (proteins, peptides, nucleic
acid, urea, and other compounds) and ammonia are
determined together and are referred to as ‘‘Kjeldjahl
nitrogen,’’ a term that refers to the analytical method
used. By expressing all nitrogen species as N, the
(2)
concentrations of the various species can be added and
(3) subtracted. Organic N (−3 charge) is determined by
analyzing ammonia by other methods and subtracting
this value from the Kjeldhal nitrogen concentration (10).
Inorganic N is the sum of nitrate, nitrite, and total
ammonia (ammonia plus ammonium). Some analytical
methods determine the sum of nitrate and nitrite, but
the results are sometimes incorrectly reported as nitrate
because nitrite concentrations are considered very low.
Total N is the sum of organic N and inorganic N.
AQUATIC NITROGEN
Despite the large proportion of nitrogen in the atmosphere,
its abundance in water is low-measured as a few mg/L
N, mostly in the inert form of nitrogen. Typically, the
most environmentally important form of nitrogen is
nitrate; however, dissolved organic nitrogen makes up the
majority of total dissolved nitrogen in watersheds in New
England (11). Note that for primary producers ammonium
may be more available to organisms than nitrate (12). The
various common species of nitrogen in the environment
(4) and the estimated concentration of each are presented
in Table 1. Mueller and others (13) have demonstrated
that background concentrations of nitrate as N are less
(5)
that 0.6 mg/L.
Nitrogen species are subject to oxidation/reduction
(redox) and reactions with hydrogen ions (pH). In Fig. 2,
the NO3 − /NO2 − pe–pH boundary is dotted to indicate that
NO2 − is metastable. Nitrate, at a pe above the NO3 − /NO2 −
boundary, is reduced at the NO3 − /N2 boundary to form
nitrogen. Thus, there is little NO3 − available to form
NO2 − . Because nitrite is a metastable species, it occurs
only in low concentrations in natural waters where
denitrification is occurring (9). Nitrate is present in the
N2 field, but under natural conditions, it is present in
concentrations less than those of N2 . Ammonium ions are
the dominant nitrogen species at low pe and pH. Ammonia
is the dominant species above pH 9 (Fig. 2) and thus is not
present in significant concentrations in natural waters;
most natural waters have a pH range of 5–9. Dissolved
organic N can also be relatively mobile.
(6) Mobility
Nitrate is very soluble in water and because of its negative
charge, is not adsorbed by humus or clay. Nitrate not
taken up by plants can be transported from the upper soil
zone to lower soil zones and eventually to groundwater.
(7)
Ammonium ions are not nearly as mobile as nitrate ions, in
part, because their positive charge allows their transport
to be retarded by adsorption on humus and clay. In some
situations, ammonium ions can be transported to deep soil
horizons (3).

20
16
NO3−
12
NO3−
O2
8 NO2−
H2O
pe
N2(g)
0 H2O
H2
NH4+
−4
−8
−12 NH30
0 2 4 6 8 10 12
pH
Figure 2. Stability fields for common nitrogen species in natural
waters. The diagram is based on the following conditions:
temperature of 25 ◦ C, partial pressures of H2 and O2 are each
1 atmosphere, partial pressure of N2 is 0.77 atmosphere, and
total aqueous N concentration is 0.001 moles per liter.
Toxicity
Ammonia is a pungent, colorless gas that is very soluble
in water and is toxic to fish at concentrations greater than
0.2 mg/L. Dissolved ammonia reacts with water to form
ammonium ions, which are not nearly as toxic (1/50th)
as the gas (3). The ammonia–ammonium equilibrium is
controlled to a large extent by temperature and pH; only
highly alkaline waters favor the formation of ammonia.
The pH range of most natural waters is such that
ammonium ions are dominant compared to ammonia
gas (Fig. 2). The U.S. Environmental Protection Agency
bases its ammonia criteria for water on a complex set of
equations that include the effect of temperature and pH on
ammonia concentration. The equations were developed to
account for acute and chronic toxicity to particular types of
fish. In addition, the criteria are based on specific periods
of time and the frequency of exceeding these criteria (14).
Typically nitrate is toxic to warm-blooded animals only
where large amounts of nitrite are formed by the reduction
of nitrate. Nitrate can be reduced in the gastrointestinal
tract to nitrite, especially in infants. Nitrite enters the
bloodstream and reacts with hemoglobin to produce
methemoglobin that reduces the number of hemoglobin
cells available for oxygen exchange. These reactions can
NITROGEN 519
be life threatening for children less than 3 years old.
Babies who consume water with nitrate concentrations
as low as 1 mg/L as N risk methemoglobinia or ‘‘blue
baby disease’’ because of the impairment of oxygen
transport. There are also indications that nitrate (as
related to nitroso compounds) may be associated with
gastric cancer and perhaps liver and kidney cancer. The
relative consumption of nitrate from water compared to
consumption of dietary nitrate is a factor that concerns
researchers attempting to determine the risk of cancer
from water (15,16). The maximum concentration limit
(MCL) of 10 mg/L N for nitrate in drinking water is based
on minimizing methemoglobinia and is considered safe
concerning cancer.
CONTAMINATION
The causes of pollution of water are divided into point
sources and nonpoint sources. Nitrogen contamination of
water is generally related to nitrate from sewage and septic
systems (point sources) and from agricultural sources.
The agricultural sources are commercial and animal
manure fertilizers (nonpoint sources) (17,18) and from
confined animal operations (point sources). Agriculture is
generally considered the main source of nitrogen pollution
of water (19). During rainstorms, streams often exhibit
a dilution of nitrate concentration that indicates that
14 most of the nitrate is transported to these streams by
groundwater. Excessive nutrients in runoff can result in
eutrophication of surface waters, for example, the Neuse
River in North Carolina, U.S.A. (20). Best management
practices have been developed to minimize the negative
impacts of fertilizers (17).
ISOTOPES
There are two stable isotopes of nitrogen, 14 N and 15 N.
Although the whole earth abundance of 14 N is 99.63‰ of
total nitrogen, due to fractionation, the 15 N:14 N ratio is
0.003677 (21) in the atmosphere. Differences in the ratio
of the two isotopes in the various nitrogen species are a
result of the faster chemical reaction of the lighter isotope
than the heavier isotope. Isotopic ratios of nitrogen are
expressed as
δ 15 N = [(Rsample /Rair ) − 1] × 1000
where R is the abundance ratio of 15 N to 14 N. The
international standard for δ 15 N is air, which by definition
has a δ 15 N of 0‰. Values greater than 0‰ indicate
enrichment in 15 N, and values less than 0‰ indicate
depletion of 15 N. Sources of nitrate have been determined
using δ 15 N. The premise is that agricultural activity
and sewage (including septic tank effluent) may have
distinct δ 15 N values. A definitive determination of source
is sometimes thwarted by the possibility of overlapping
δ 15 N values between possible sources and changes in
δ 15 N due to volatilization of NH3 from the soil and
denitrification of nitrate in water. The origin of the nitrate
in water can be found by using δ 18 O to recognize different
520 OSMOSIS-DIFFUSION OF SOLVENT OR CAUSED BY DIFFUSION OF SOLUTES?
sources of nitrate and the influence of denitrification on
δ 15 N (21,22). A number of studies have had success using
N and often O isotopes to determine the source of nitrate
contamination (23,24).
BIBLIOGRAPHY
1. Hitchon, B., Perkins, E.H., and Gunter, W.D. (1999).
Introduction to Ground Water Geochemistry. Geoscience,
Alberta, Canada, pp. 195–200.
2. Berner, E.K. and Berner, R.A. (1987). The Global Water Cycle.
Prentice-Hall, Inc., Englewood Cliffs, NJ, p. 397.
3. Canter, L.W. (1996). Nitrates in Groundwater. CRC Lewis,
Boca Raton, FL.
4. Jeurgens-Gschwind, S. (1989). Ground water nitrates in
other developed countries (Europe)–relationships to land
use patterns. In: Nitrogen Management and Ground Water
Protection. R.F. Follett (Ed.). Elsevier, New York, pp. 75–138.
5. U.S. EPA. (1976). Quality Criteria for Water. U.S. Environ-
mental Protection Agency, Washington, DC, pp. 10–13.
6. Manahan, S. (1999). Environmental Chemistry. CRC Press,
Boca Raton, FL.
7. Stumm, W. and Morgan, J.J. (1996). Aquatic Chemistry, 3rd
Edn. John Wiley & Sons, New York, pp. 927–934.
8. Kendall, C. and Aravena, R. (1999). Nitrate isotopes in
groundwater systems. In: Environmental Tracers in Subsur-
face Hydrology. P.G. Cook and A.L. Herczeg (Eds.). Kluwer
Academic, Boston, pp. 261–298.
9. Kehew, A.E. (2001). Applied Chemical Hydrogeology. Prentice-
Hall, Upper Saddle River, NJ.
10. Eaton, A.D., Lenore, C.S., and Greenberg, A.E. (Eds.).
(1995). Standard Methods for the Examination of Water
and Wastewater. American Public Health Association,
Washington, DC, pp. 4-75–4-96.
11. Campbell, J.L. et al. (2000). Dissolved organic nitrogen
budgets for upland, forested ecosystems in New England.
Biogeochemistry 49: 123–142.
12. Peterson, B.J. et al. (2001). Control of nitrogen export from
watersheds by headwater streams. Science 292: 86–90.
13. Mueller, D.K., Hamilton, P.A., Helsel, D.R., Hitt, K.J., and
Ruddy, B.C. (1995). Nutrients in ground water and surface
water of the United States—an analysis of data through
1992. U.S. Geological Survey Water-Resources Investigations
Report 95-4031, p. 74.
14. U.S. EPA. (1999). 1999 Update of Ambient Water Quality
Criteria for Ammonia. U.S. Environmental Protection Agency.
National Service Center for Environmental Publications,
Cincinnati, OH.
15. Bogardi, I. and Kuzelka, R.D. (Eds.). (1991). Nitrate Con-
tamination: Exposure, Consequence and Control. NATO ASI
Series. Springer-Verlag, New York.
16. Follett, R.F. and Walker, D.J. (1989). Ground water quality
concerns about nitrogen. In: Nitrogen Management and
Ground Water Protection. R.F. Follett (Ed.). Elsevier, New
York, pp. 1–22.
17. Steele, K.F. (Ed.). (1995). Animal Waste and the Land–Water
Interface. CRC Press, Lewis, Boca Raton, FL.
18. Steele, K.F. and McCalister, W.K. (1991). Potential nitrate
pollution of ground water in limestone terrrain by poultry lit-
ter, Ozark Region, USA. In: Nitrate Contamination: Exposure,
Consequence and Control. I. Bogardi and R.D. Kuzelka (Eds.).
NATO ASI Series, Springer-Verlag, New York, pp. 209–218.
19. Spalding, R. and Exner, M. (1993). Occurrence of nitrate in
groundwater—A review. J. Environ. Qual. 22: 392–402.
20. Paerl, H.W., Pinckney, J.L., Fear, J.M., and Peierls, B.L.
(1998). Ecosystem responses to internal and watershed
organic matter loading: Consequences for hypoxia in the
eutrophying Neuse River estuary, North Carolina, USA. Mar.
Ecol. Prog. Ser. 166: 17–25.
21. Clark, I.D. and Fritz, P. (1997). Environmental Isotopes
in Hydrogeology. Lewis, CRC Press, Boca Raton, FL,
pp. 150–152.
22. Ostrom, N.E., Peggy, H., and Ostrom. (1999). Nitrogen
isotopes. In: Encyclopedia of Geochemistry. C.P. Marshall
and R.W. Fairbridge (Eds.). Kluwer Academic, Dordrecht,
The Netherlands, pp. 431–433.
23. Karr, J.D., Showers, Hinson, W.J., and Thomas, H. (2002).
Nitrate source identification using d 15 N in a ground water
plume near an intensive swine operation. Ground Water
Monitoring and Remediation 68–75.
24. Aravena, R., Evans, M.L., and Cherry, J.A. (1996). Stable
isotopes of oxygen and nitrogen in source identification of
nitrate from septic systems. Ground Water 31: 180–186.
OSMOSIS-DIFFUSION OF SOLVENT OR
CAUSED BY DIFFUSION OF SOLUTES?
KARL ERIK ZACHARIASSEN
Norwegian University of
Science and Technology
Trondheim, Norway
Osmosis is the movement of water through a semiperme-
able membrane separating two solutions to the solution
with the higher solute concentration. The membrane is
semipermeable in the sense that it is permeable to water,
but not to the solutes. The osmotic movement of water
can be blocked by the application on the high concentra-
tion side of a hydrostatic pressure, which counteracts the
so-called osmotic pressure that moves the water through
the membrane.
The prevailing theory ascribes osmosis to a diffusion
of water molecules through the membrane down a water
concentration gradient across the membrane. The solutes
are supposed to dilute the water, and the dilution will be
strongest at the side with the highest solute concentration.
However, several investigators claim that this explana-
tion fails to explain important characteristics of osmotic
phenomena. Osmotic movement of water is considerably
faster than should be expected from a diffusive movement
of water. In addition, the water concentration of solutions
of different solutes are not correlated with the osmotic
activities of the solutions. Furthermore, the osmotic pres-
sure is related to the total concentrations of dissolved
particles, and van t’Hoff pointed out that the number of
solute particles and the solution volume are related to
the osmotic pressure in the same way as the number of
molecules, pressure, and volume of a gas, i.e., osmotic
phenomena, follow the universal gas laws. No theory has
been proposed to explain the agreement between osmotic
phenomena and gas laws on the basis of diffusion of water
down a water concentration gradient.

In 1903, George Hulett (1) published a theory that
explains the agreement between osmotic phenomena
and gas laws and that is in agreement with the other
observations on osmotic phenomena. Hulett emphasized
that the solute particles diffuse inside solutions in the
same way as gas molecules diffuse within the space in
which the gas is distributed. In the same manner, as gas
molecules create a pressure when they collide with the
gas boundary, solute particles create a pressure when
they collide with the boundary of the solution. Both
for gases and solutions this pressure will tend to cause
the volume to expand, which is easy for a gas if the
boundaries can yield to the pressure and allow expansion
of the volume. However, the water in which the solutes of
a solution are distributed is essentially incompressible,
and the expansive pressure created by the diffusing
solutes cannot lead to an expansion unless more water
can enter the solution and increase the solution volume.
The semipermeable membrane of osmotic processes allows
such an influx of water from a solution where a lower
solute concentration creates a lower expanding pressure,
whereas the solutes that create the pressure are prevented
from passing. Water will subsequently move across the
membrane until the pressures on both sides are equal.
According to the gas laws, an ideal 1 molar solution
(1 mole of particles dissolved in a volume of 1 L), which
is separated from pure water by a semipermeable, will
create a pressure of 22.1 atm, which corresponds to a water
column of impressive 224 m. It will have a melting point
of −1.86 ◦ C, and a solution with this osmotic concentration
is said to be a 1 osmolar solution.
Proportionality exists among the osmolality, the
osmotic pressure, and the melting point depression. This
proportional relationship with osmotic pressure is also
displayed by vapor pressure and boiling point elevation.
Because of the exact proportionality that exists among
these parameters, they are frequently referred to as
the colligative parameters (from the Latin colligare =
bind together).
A closer inspection of the relationship between
mole-based concentration measures and the colligative
properties reveals that it is the molal concentration that is
most closely related to the colligative parameters. For this
reason, some investigators prefer the term osmolality.
UNITS FOR OSMOTIC CONCENTRATIONS
Some confusion exists about the units for osmotic
concentration. The officially approved unit for osmolality
is Osmoles/kg water. Some investigators find that this
unit is misleading, because it invites us to consider an
Osmole as a defined amount of solute, which when diluted
in 1 kg of water, gives an osmolality of 1 Osmole/kg
water. Accordingly, when diluted in 2 kg of water, the
same osmole of solutes should give an osmolality of 0.5
Osmole/kg water, but this is never the case. An Osmole
is not a defined amount of solute, and the concept of
osmolality has a meaning only for solutions. For this
reason, some investigators use the unit Osm, where
the letter m emphasizes the close relationship between
PARTITIONING AND BIOAVAILABILITY 521
osmotic and molal concentrations, which have m as the
official unit.
To add to the confusion, the official unit for osmolality
has led some authors to introduce Osmoles/L solution
as a unit for osmolarity. However, the melting point is
claimed to be −1.86 ◦ C both for a 1 osmolal and for a 1
osmolar solution. Hence, the distinction between units
for osmolality and osmolarity is merely semantic and
of no practical consequence. Osmolality and osmolarity
are different terms for the same thing, and furthermore,
they are just quasi-measures for concentration. Solutions
of different solutes with the same molar or molal
concentrations all have different osmolalities, and for
a given type of solute, no exact proportionality exists
between molal concentrations on one side and colligative
parameters on the other, which is at least in part because
the amount of hydration water associated with each solute
particle differs substantially from substance to substance,
and for dissolved salts, the degree of dissociation varies
from substance to substance and with their concentration.
Although osmolality is not a precise chemical concentra-
tion measure, it is a useful tool for comparing total solute
concentrations of fluids in physiology, and it is a useful
measure to control the chemical solute concentrations of
body fluids and solutions.
BIBLIOGRAPHY
1. Hulett, G. (1903). Beziehung zwischen negativem Druck und
osmootischen Druck. Z. Phys. Chem. 42: 353–368.
Hammel, H.T. and Scholander, P.F. (1976). Osmosis and
Tensile Solvent. Springer Verlag, Berlin.
PARTITIONING AND BIOAVAILABILITY
JIM PHILP
FIONA STAINSBY
SANDRA DUNBAR
Napier University
Edinburgh, Scotland, United
Kingdom
INTRODUCTION
As soon as a chemical is spilled it becomes a pollutant,
but the fate of that pollutant depends to a large degree
on its chemistry, which determines how it will partition to
various ‘‘compartments,’’ or phases, of the environment.
This idea is enshrined in risk-based strategies for
assessing contaminated land and water. In many cases
a contaminant spill on land quickly becomes a water
pollution issue, be it surface or groundwater. As many
countries are heavily reliant on groundwater as a source
of potable water, investigating the fate of pollutants is
an essential prerequisite to any remedial action plan for
contaminant cleanup.
Partitioning is governed by some fundamental prop-
erties of chemicals, but partitioning should be viewed in
relation to the toxicity of the chemical in question and its
522 PARTITIONING AND BIOAVAILABILITY
degradability in the various environmental compartments.
In this article, the important properties of chemicals that
relate to partitioning are examined, and the theory is
reinforced by example. The examples chosen are ben-
zene, phenol, and some chlorinated phenols. All of these
compounds are important industrial chemicals.
Bioavailability is a simple concept, but practically it
is a complex issue that remains to be resolved. The
bioavailability concept is also enshrined in the risk
assessment approach to contaminated land and water. If a
pollutant is present in soil or water in high concentration,
but is not available to the biota, then surely it presents
minimal risk. Regardless of the apparent simplicity of
the concept, there is great urgency to find methods for
quantifying bioavailability. Some approaches to solving
this quandary are discussed.
PROPERTIES THAT DETERMINE PARTITIONING
The compartments of the environment relevant to
partitioning are air, water, soil, and sediment and, through
bioaccumulation/magnification through food chains, biota.
Partitioning into the biota is beyond the scope of this
article, and it will deal principally with partitioning to
nonliving compartments. Figure 1 represents a common
scenario in environmental remediation, contamination
from old gas stations.
Vapor Pressure
This is the partial pressure of a chemical in a gas phase
that is in equilibrium with the pure liquid or solid phase
of the chemical. Vapor pressure is greatly influenced by
temperature, and figures are usually quoted at a constant
temperature (usually 20 ◦ C or 25 ◦ C). If the volume of the
containing vessel is also known, the vapor pressure can be
used to calculate the concentration of the chemical in the
headspace of the vessel from the ideal gas law:
n P
= (1)
V RT
where P = vapor pressure (atm)
V = volume
Vapours
Contaminated soil
Figure 1. Partitioning of fuel as a con-
taminant from a gas station that has
Fuel on groundwater surface
a leaking underground storage tank.
Because fuels are multicomponent liquids,
different components partition to different
compartments of the environment.
n = number of moles of the chemical
R = gas constant
T = absolute temperature
Another way to think of vapor pressure is that it is
the pressure that a vapor exerts on the atmosphere:
thus it will be clear that it is the driving force for
volatilization. Vapor pressures of organic compounds
are often compared to the vapor pressure of mercury
(atmospheric pressure is 760 mmHg). Thus the vapor
pressure of benzene is 87.2 mmHg at 25 ◦ C. Volatilization
potential can be misleading in the absence of information
on the water solubility of the compound, however. A
compound that has a low vapor pressure may still have
a high tendency to ‘‘escape’’ if the water solubility is
low. For example in Table 1, 2,4,6-trichlorophenol has
a much lower vapor pressure than 3-chlorophenol, and
yet its tendency to escape to air is higher. This can be
explained when the water solubilities are compared: the
much lower value for 2,4,6-trichlorophenol results in it
having a Henry’s constant about a magnitude order higher
than 3-chlorophenol, even though its vapor pressure is
much lower.
Henry’s Law Constant
To study partition phenomena, then, Henry’s law constant
is more valuable than vapor pressure. Henry’s law
constant is the ratio of the equilibrium concentration
of a compound in air and its equilibrium concentration
in water. It can therefore be regarded as the partition
coefficient between air and water, and is the best indicator
of the tendency of a chemical to volatilize from water.
Henry’s law constant can be written as:
P
H= (2)
S
where H = Henry’s law constant (atm-m3 /mol)
P = vapor pressure (atm)
S = water solubility (mol/m3 )
Water Solubility
This is of central importance to understanding partition
phenomena in the environment. It is the maximum, or
Direct contact Public area
soil/vegetable
intake
River
Vapours Vapours
Operational
spills Fuel tank
Contaminated soil
Underground
leaks
Fuel dissolved in groundwater
 PARTITIONING AND BIOAVAILABILITY 523

Table 1. Important Partition Variables for a Range of Compounds. The Figures were Computed from U.S. EPA Software
(EPI Suite v.3.11 and ECOSAR)
Vapor Level III
Pressure, Water Fugacity Toxicity Fish Biodeg.
mmHg, Henry’s Solubility, Log Log Mass Amount, 96-h LC50 , Probability,
25 ◦ C Constant mg/L Kow Koc % mg/L MITI Non-Linear

Benzene 87.2 5.55 × 10−3 1790 2.13 2.22 Air: 37.6 59.2 .73
Water: 48.4
Soil: 13.8
Sediment: 0.20
Phenol 0.323 3.33 × 10−7 77, 900 1.46 2.43 Air: 1.33 27.7 .71
Water: 47.3
Soil: 51.3
Sediment: 0.098
3-chloro-phenol 0.119 3.45 × 10−7 26, 000 2.50 2.64 Air: 1.35 15 0.39
Water: 39.4
Soil: 59.1
Sediment: 0.21
2,4,6-trichloro-phenol 0.00536 2.6 × 10−6 800 3.69 3.07 Air: 2.55 3.7 0.043
Water: 22.9
Soil: 72.4
Sediment: 2.17
Pentachloro-phenol 1.08 × 10−5 2.45 × 10−8 14 5.12 3.53 Air: 0.0665 0.8 .0031
Water: 5.59
Soil: 70.9
Sediment: 23.4

saturation, concentration of a compound that dissolves in
water. It also is significantly affected by temperature. As
the water solubility of compounds is defined by a host of
factors, comparisons should always be made at a constant
temperature, usually 25 ◦ C. In natural water systems, the
concentration of pollutant chemicals is usually rather low,
making this especially important.
Polarity refers to the extent to which charge is unevenly
distributed within a molecule. Water itself is highly polar.
Therefore charged or highly polar organic and inorganic
molecules, hydrophilic molecules, are readily soluble in
water. Conversely, nonpolar, hydrophobic molecules find
solution in water a less energetically favorable state, and
they have low water solubility. In the examples in Table 1,
benzene has moderate water solubility, but the addition
of the polar hydroxyl group makes phenol much more
water soluble. The solubility of any solute in any solvent
is mediated by forces of attraction; the main ones are:
• van der Waals forces;
• hydrogen bonding;
• dipole–dipole interactions.
The size and shape of a molecule is also critical in
determining its overall water solubility. Large molecules
tend to have lower water solubility because they have
a higher molar volume. The addition of halogens to
an organic molecule might be expected to increase the
water solubility of the new molecule as halogens are
electronegative (electron-withdrawing) species and would
thus tend to increase the polarity of the molecule.
However, in the increasingly chlorinated series of phenols
in Table 1, this is not the case: increasing chlorination
very markedly decreases the solubility of chlorophenols
because the halogens increase the molecular volume of
organic compounds (1).
Other chemical functional groups affect water solu-
bility in rather specific ways. The presence of polar
functional groups tends to increase the water solubility of
molecules compared to unsubstituted hydrocarbons: alco-
hols, amines, ethers, ketones, and organic acids are good
examples. Even with these examples, though, a general
rule of increasing molecular size and decreasing water sol-
ubility still exists. The short-chain fatty acid acetic acid is
much more water soluble than a long-chain fatty acid. The
additions to the molecule are not more polar functional
groups but simply more nonpolar C–H2 groups that make
the molecule larger but less polar.
The situation is further complicated by more specific
chemical reactions. The chlorophenols in Table 1 illustrate
this point. Chlorophenols behave as weak acids when
dissolved in water due to the loss of a proton from
the hydroxyl group to leave a phenoxide ion. The more
chlorines that are added to the ring, the more electrons
are withdrawn from the electron-rich phenoxide ion. This
lowers the phenoxide negative charge and therefore lowers
its ability to hold a proton. Thus the more chlorines on
the ring, the more easily a proton is released, and the
greater the acidity. The ionized form is usually several
orders of magnitude more water soluble than the neutral
un-ionized form. The relative proportion of each species is
highly dependent on pH.
Log Kow , a Measure of Hydrophobicity
Polarity and water solubility do not give direct measures
of the hydrophobicity of a compound. A hydrophobic
compound might be expected to dissolve in a lipophilic
phase, and this is the basis for the octanol–water
524 PARTITIONING AND BIOAVAILABILITY
partition coefficient, Kow . This is the ratio of a chemical’s
concentration, at equilibrium, when dissolved in octanol
and in water, when both phases are saturated with
each other.
This test was devised by the pharmaceutical industry
to study partitioning of drugs as it roughly equates to the
partitioning between water and body fat. It has proven
very useful in predicting how chemicals partition in the
environment. Large molecules of low polarity have a high
Kow and are more likely to partition to solids in the
environment. Small, polar molecules dissolve in water
rather than octanol, have a low Kow , are much less likely
to partition to solids in the environment, and so are more
likely to remain in an aqueous phase:
concentration of pollutant in
water-saturated octanol (mg/L)
Kow = (3)
concentration of pollutant
in octanol-saturated water (mg/L)
The range of values for Kow is very wide, and tabulated
values are more often expressed as log Kow . In the
examples in Table 1, it can be seen that, as the molecule
becomes larger through increasing chlorination, the water
solubility decreases and log Kow increases.
Log Koc , Partitioning to Soil
The most basic statement of the composition of soil is
that it contains sand, clay, organic matter, water, and air.
Although typically the organic matter content of soil is in
the range of 0.1 to 5%, this is where organic pollutants sorb,
not to the soil as a whole. The soil adsorption coefficient,
Koc , takes account of this. For a liquid organic pollutant,
mass of pollutant sorbed
to soil organic matter (mg/g)
Koc (mL/g) = (4)
mass of pollutant in
the aqueous phase (mg/mL)
The range of values for Koc is very wide, and therefore the
use of log Koc is preferred. The soil partition coefficient
is greater, the more hydrophobic the pollutant. However,
it has been measured for far fewer chemicals than Kow .
It might be expected to be a more realistic measure than
Kow , but as data are relatively rare for Koc , a series of
correlation equations has been developed that relate Koc to
Kow . The increase in log Kow seen in Table 1 as chlorination
increases is mirrored by increasing log Koc .
Fugacity
The term fugacity literally means the ‘‘tendency to flee,’’
and refers to a molecule’s driving force to escape from
the compartment in which it currently exists. With
knowledge of all of a chemical’s partition coefficients, it
is possible to predict its relative concentration in air,
water, and soil at equilibrium. The gradient of fugacity
between two compartments (e.g., soil and soil porewater)
determines the potential for the chemical to move from
one compartment to the other. Fugacity therefore has
units of pressure, but this can be related to the chemical
concentration through modeling, and thus the distribution
through various compartments predicted (Table 1).
THE RELATIONSHIP TO TOXICITY AND
BIODEGRADATION
Hydrophobicity and Toxicity
Particularly Kow and log Kow have become increasingly
used as predictors of the behavior of pollutants in the envi-
ronment. It has been correlated with toxicity. The most
common mode of action regarding the toxicity of industrial
pollutants is probably narcosis. Nonpolar narcosis results
from the perturbation of cellular membranes (2) due to
the entry of hydrophobic (lipid-soluble) molecules into the
phospholipid bilayer of the membrane. Narcosis has been
shown to be a nonspecific physiological effect which is
independent of chemical structure, and log Kow is highly
correlated with nonpolar narcosis toxicity (3).
Phenols have greater toxicity than predicted by
nonpolar narcosis (4). They act by polar narcosis, which
probably results from a strong hydrogen-bonding group
that makes the compound more polar. Nevertheless,
there is still a strong correlation between the toxicity
of these compounds and log Kow . Pentachlorophenol has
greater toxicity as a result of a further mechanism:
its acidity makes it a weak acid respiratory uncoupler
(WARU). Terada (5) noted that WARUs possess an acid-
dissociable moiety (e.g., a hydroxyl group), a bulky
hydrophobic moiety, such as an aromatic ring, and
multiple electronegative groups, such as halogens. During
weak acid uncoupling, ATP synthesis is inhibited with
no effect on the respiratory chain (6) because electron
transport and phosphorylation reactions are uncoupled
from each other.
Hydrophobicity and Biodegradability
It is believed that only molecules of hydrocarbons
that are dissolved in the aqueous phase are available
for intracellular metabolism (7). The rate at which a
particular organic compound dissolves in water is critical
to both the toxicity and biodegradability of the compound
as this governs the rate of transfer to the organism (in
biodegradability, microorganisms are important). The rate
of transfer is determined by the equilibrium and actual
concentration in the bulk phase and the aqueous phase.
This is central to the concept of bioavailability.
From the compounds in Table 1, the increasing degree
of chlorination decreases the rate of biodegradation.
Increasing chlorination also leads to a marked drop in
water solubility and a corresponding increase in hydropho-
bicity, which influence biodegradability. The toxicity of
the compound also adversely affects biodegradation. The
interaction between toxicity and biodegradation is not
well understood, but they are clearly linked and compet-
ing processes: at low concentrations, a compound may
be biodegradable, but this is arrested at higher concen-
trations by toxicity. If the concentration is very low for
biodegradation, then this may not be enough to support
microbial growth and reproduction, the concept of thresh-
old elaborated in detail by Alexander (8).

BIOAVAILABILITY
Bioavailability is a concept recently borrowed from
nutrition to explain the behavior of chemical pollutants
in the environment (9). There are several definitions in
circulation. In its simplest form (10), bioavailability is
defined as ‘‘the amount of contaminant present that can
be readily taken up by microorganisms and degraded.’’ But
this ignores bioavailability to higher organisms. Another
definition that takes account of this and alludes to the risk
assessment practice (11) is ‘‘a measure of the potential of
a chemical for entry into ecological or human receptors.’’
The bioavailability quandary mentioned previously has
been accurately articulated by the International Union of
Pure and Applied Chemistry (IUPAC).
‘‘In the IUPAC Conference of Pesticide Chemistry in London in
1998, there was a controversial discussion on the topic whether
agrochemicals in the environment are really hazardous. The
conventional wisdom is that chemicals in soil are available to
microorganisms, plant roots, and soil fauna like earthworms
and animals via dermal exposure. Then bioaccumulation
through the food web may induce indirect exposure to
higher organisms. National governments are simply reducing
the threshold levels of chemicals in the environments for
larger safety margins in their guidelines. However, the
question raised at the Congress was that chemical residues
in the environment are not always bioavailable, so that
the actual exposure of biota to chemicals is rather different
from the amount (concentration) present in the environment
(authors’ italics). In addition, the persistence and efficacy
of agrochemicals are affected by their bioavailability in
soil. There is a lack of comprehensive understanding of
the bioavailability of chemicals with different chemical
characteristics, such as non-polar organochlorines, polar
agrochemicals and etc. Thus, there is a strong need to clarify
the scientific basis for bioavailability: definition, estimation
methods and affecting factors.’’
Measuring Bioavailability
Measuring or even estimating bioavailability is problem-
atic as it is affected by many individual and interacting
conditions relating to soil and water chemistry, pollutant
chemistry and partitioning, and biological transformation
and concentration factors (12). Moreover, the ‘‘aging’’ pro-
cess brings about a time-dependent decrease in bioavail-
ability as compounds become sequestered in soil over time
by ill-understood mechanisms (13). The techniques cur-
rently being researched can be categorized broadly as
(a) Out In (b)
Bioavail.
M element initiates transcription resulting in synthe-
Nonbioavail. M o/p lux genes
M Direction of transcription
PARTITIONING AND BIOAVAILABILITY 525
direct or indirect and biological or chemical (14), although,
as pointed out by these authors, only organisms can deter-
mine whether a chemical is bioavailable, so direct chemical
methods are not possible.
Direct Biological Methods. These are techniques that
measure the actual amount of a chemical taken up by a
target organism, and this may ultimately be the most
accurate measure of bioavailability, although it is by
no means routine. For example, the determination of
chemical levels in earthworms is relatively expensive and
time-consuming. The great strength of these direct tech-
niques, however, is that they integrate all the biotic and
abiotic modifying factors of chemical bioavailability (14).
Indirect Biological Methods. It is possible to observe a
response to a chemical in an organism without actually
measuring the concentration of the chemical. A range
of responses is possible, such as lethality, enzyme
induction or inhibition, and reproductive effects. These are
regarded as indirect biological methods because the effect
may be quantifiable in the organism, but the chemical
concentration remains unknown.
A promising, novel approach is the use of genetically
modified microorganisms that can detect and quantify
specific pollutants. This type of biosensor is based on
the highly specific genetic control mechanisms used
by microorganisms to ensure that specific proteins are
expressed only when needed; for example, to detoxify
a particular toxic substance. This control is exerted by
inducible promoters, consisting of a specific DNA sequence
upstream of the genes to be controlled and a DNA-binding
protein that either activates or prevents transcription
in response to the presence or absence of the target
compound. Biosensors of this type are easily generated
by fusing such a controllable promoter to a reporter gene
that generates a detectable signal when the promoter is
activated (15). The most popular reporter is the bacterial
lux system that produces light (Fig. 2), because light can
be easily quantified and, due to its rarity in biology, there
is no interference from the background biochemistry of the
host organism.
Such biosensors can be both highly specific and
responsive to very low concentrations, and the microbial
cell membrane makes this approach truly a measure of
bioavailability. There are limitations to developing such
biosensors for detecting organic pollutants as often the
Figure 2. (a) Schematic of a genetically modified
whole bacterial cell biosensor for detecting bioavail-
able metal. The metal may be expected to be
transported into the bacterial cell via a specific
transporter protein mediating uptake and efflux. If
the metal is in a nonbioavailable form, for example,
as a precipitate, it will not be transported. The inter-
action of the metal with the specific metal regulatory
sis of the luciferase enzyme and the production
of light. (b) The naturally bioluminescent bacterium
Vibrio fischeri producing light from colonies and
liquid culture.
526 PARTITIONING AND BIOAVAILABILITY
required specificity is lacking. For heavy metals, specificity
may not be a problem. Nevertheless, careful selection of
the host strain is required for a variety of reasons (16).
Such biosensors offer the prospect of a much more routine
measure of bioavailability.
Indirect Chemical Methods. Indirect chemical methods
usually involve extracting a fraction of the chemical
(metals or organics) from a soil; extractability is defined
by the chemical itself, the nature of the extractant(s), and
the experimental conditions applied. The origins of this
approach are in sequential extraction procedures (17)
that attempt to quantify the speciation of toxicants into
those weakly bound and those strongly bound to the soil
matrix. These tests often assume that the weakly bound
toxicants are those that are more bioavailable, although
evidence for this is incomplete. A recent advance in this
technique is sequential accelerated solvent extraction (18)
that reduces sample preparation time and maintains
relatively constant extraction conditions. A technique
under development is the use of selective supercritical
fluid extraction (19).
A number of methods are based around the human
gastrointestinal tract (the physiologically based extraction
test, PBET; 20). To mimic human conditions, they
incorporate gastric juices and enzymes, mix at 37 ◦ C, and
use soil residence times similar to those found in children
after ingestion of food. Analysis of all solutions produced in
the PBET extractions is completed by analytical chemistry
techniques under matrix-matched conditions.
However sophisticated, any such technique is ulti-
mately limited by the lack of an organism. There will be
no universal chemical test for bioavailability as the phe-
nomenon is organism- and even species-dependent (21).
BIBLIOGRAPHY
1. Yalkowsky, S.H. and Valvani, S.C. (1979). Solubilities and
partitioning. 2. Relationships between aqueous solubilities,
partition coefficients and molecular surface areas of rigid
aromatic hydrocarbons. J. Chem. Eng. Data 24: 127–129.
2. Cronin, M.T.D. and Schultz, T.W. (1998). Structure-toxicity
relationships for three mechanisms of action of toxicity to
Vibrio fischeri. Ecotoxicol. Environ. Saf. 39: 65–69.
3. Cronin, M.T.D. and Dearden, J.C. (1995). Review: QSAR in
toxicology. 1. Prediction of aquatic toxicity. Quant. Struct.
Act. Relationships 14: 1–7.
4. Cronin, M.T.D. and Schultz, T.W. (1997). Validation of Vibrio
fischeri acute toxicity data: mechanism of action-based QSARs
for non-polar narcotics and polar narcotic phenols. Sci. Total
Environ. 204: 75–88.
5. Terada, H. (1990). Uncouplers of oxidative phosphorylation.
Environ. Health Perspect. 87: 213–218.
6. Schultz, T.W. and Cronin, M.T.D. (1997). Quantitative struc-
ture-activity relationships for weak acid respiratory uncou-
plers to Vibrio fischeri. Environ. Toxicol. Chem. 16: 357–360.
7. Sikkema, J., de Bont, J.A.M., and Poolman, B. (1995). Mecha-
nisms of membrane toxicity of hydrocarbons. Microbiolo. Rev.
59: 201–222.
8. Alexander, M. (1999). Threshold. In: Biodegradation and
Bioremediation. Academic Press, San Diego, pp. 105–116.
9. Madsen, E.L. (2003). Report on Bioavailability of Chemical
Wastes with Respect to the Potential for Soil Bioremediation.
U.S. Environmental Protection Agency report EPA/600/R-
03/076.
10. Maier, R. (2000). Bioavailability and its importance to
bioremediation. In: International Society for Environmental
Biotechnology: Environmental Monitoring and Biodiagnos-
tics. J.J. Valdes (Ed.). Kluwer, Boston, pp. 59–78.
11. Loehr, R.C. and Webster, M.T. (1997). Effects of treatment
on contaminant availability, mobility, and toxicity. In:
Environmentally Acceptable Endpoints in Soil. D.G. Linz
and D.V. Nakles (Eds.). American Academy of Environmental
Engineers, Annapolis, MD, pp. 137–386.
12. Hund-Rinke, K. and Kördel, W. (2003). Underlying issues in
bioaccessibility and bioavailability: experimental methods.
Ecotoxicol. Environ. Saf. 56: 52–62.
13. Chung, N. and Alexander, M. (2002). Effect of soil properties
on bioavailability and extractability of phenanthrene and
atrazine sequestered in soil. Chemosphere 48: 109–115.
14. Lanno, R., Wells, J., Conder, J., Bradham, K., and Basta, N.
(2004). The bioavailability of chemicals in soil for earthworms.
Ecotoxicol. Environ. Saf. 57: 39–47.
15. Philp, J.C., French, C., Wiles, S., Bell, J.M.L., Whiteley, A.S.,
and Bailey, M.J. (in Press). Wastewater toxicity assessment
by whole cell biosensor. In: Handbook of Environmental
Chemistry. D. Barceló (Ed.). Vol. 5. Water Pollution, Emerg-
ing Organic Pollutants in Wastewaters, Springer-Verlag,
Berlin.
16. Rensing, C. and Maier, R.M. (2003). Issues underlying use
of biosensors to measure metal bioavailability. Ecotoxicol.
Environ. Saf. 56: 140–147.
17. Tessier, A., Campbell, P.G.C., and Bisson, M. (1979). Sequen-
tial extraction procedure for the speciation of particulate trace
metals. Anal. Chem. 51: 844–851.
18. Tao, S., Guo, L.Q., Wang, X.J., Liu, W.X., Ju, T.Z., Daw-
son, R., Cao, J., Xu, F.L., and Li, B.G. (2004). Use of sequen-
tial ASE extraction to evaluate the bioavailability of DDT and
its metabolites to wheat roots in soils with various organic
carbon contents. Sci. Total Environ. 320: 1–9.
19. Björklund, N.T., Bøward, S., Pilorz, K., Mathiasson, L., and
Hawthorne, S.B. (2000). Introducing selective supercriti-
cal fluid extraction as a new tool for determining sorp-
tion/desorption behavior and bioavailability of persistent
organic pollutants in sediment. J. Biochem. Biophys. Methods
43: 295–311.
20. Ruby, M.V., Davis, A., Schoof, R., Eberle, S., and Sell-
stone, C.M. (1996). Estimation of lead and arsenic bioavail-
ability using a physiologically based extraction test. Environ.
Sci. Technol. 30: 422–430.
21. Reid, B.J., Jones, K.C., and Semple, K.T. (2000). Bioavailabil-
ity of persistent organic pollutants in soils and sediments—a
perspective on mechanisms, consequences and assessment.
Environ. Pollut. 108: 103–112.
READING LIST
Hemond, H.F. and Fechner-Levy, E.J. (2000). Chemical Fate and
Transport in the Environment. Academic Press, San Diego.
Landis, W.G. and Yu, M-H. (2003). Introduction to Environmental
Toxicology, 3rd Edn. Lewis, Boca Raton, FL.
Nirmalakhandan, N.N. and Spreece, R.E. (1988). Prediction of
aqueous solubility of organic compounds based on molecular
structure. Environ. Sci. Technol. 22: 328–338.
Watts, R.J. (1997). Hazardous Wastes: Sources, Pathways,
Receptors. John Wiley & Sons, New York.

PHYSICAL PROPERTIES
KEVIN S. MCLEARY
STRUCTURE
Water is a molecule composed of two atoms of hydrogen
bonded to an atom of oxygen. It has a volume of 2.97 ×
10−11 µ3 (29.7 Å3 ) (1). As a result of the hybridization
of the oxygen atom, the molecule is bent at an angle
of 104.45◦ . Due to the nonlinear structure and to the
difference in electronegativity between oxygen (3.44) and
hydrogen (2.20), the molecule has a dipole moment of 1.85
Debye (2).
Water molecules are held together by intermolecular
attractions. In water, the intermolecular attractions are
hydrogen bonds. The negatively charged oxygen atom
in one molecule of water is electrically attracted to the
positively charged hydrogen atom of another molecule.
Hydrogen bonding governs the structure of water up to
the boiling point (2).
SPECIFIC WEIGHT (γ )
The specific weight of a substance is its weight per unit
of volume. In U.S. Customary Units, it is often expressed
as pounds-force per cubic foot (3). In SI units, it is often
expressed in kN/m3 . Although a value of 62.4 pounds-
force per cubic foot (9.81 kN/m3 ) is commonly used for the
majority of engineering purposes, the specific weight of
water varies with temperature, from 62.42 pounds-force
per cubic foot (9.805 kN/m3 ) at 32 ◦ F (0 ◦ C) to 59.8 pounds-
force per cubic foot (9.399 kN/m3 ) at 212 ◦ F (100 ◦ C) (4).
Table 1 shows the variability in the specific weight of
water with temperature in U.S. Customary Units. Table 2
shows the variability in the specific weight of water with
temperature in SI Units.
Specific weight is also a function of gravity, and it varies
slightly with elevation. One cubic foot of water at 32 ◦ F
at sea level has a specific weight of 62.42 pounds-force
per cubic foot (9.805 kN/m3 ); the same volume of water at
10,000 feet above sea level has a specific weight of 62.36
pounds-force per cubic foot (9.796 kN/m3 ). Because the
earth is not a perfect sphere, but bulges at the equator,
specific weight is also a weak function of latitude. However,
these variations are small enough to warrant the use
of a constant value of 62.4 pounds-force per cubic foot
(9.81 kN/m3 ) at all elevations and latitudes (4).
DENSITY (ρ)
The density of a fluid is defined either as its mass per
unit of volume or as the ratio of its specific weight to the
acceleration due to gravity. Density ranks among the most
important hydraulic properties of water (1). It is a factor
in the settling velocity of particles and in the filtration
of water through a porous medium. A fluid’s Reynolds
number, the most important dimensionless number in
hydraulics, is a direct function of density. The Reynolds
PHYSICAL PROPERTIES 527
number is used in calculating friction losses, pipe sizing,
and pumping requirements
In U.S. Customary Units, density is expressed in
slugs/ft3 , and in SI units, it is expressed in kg/m3 .
Like specific weight, density varies with temperature.
The density of pure water varies from 1.940 slugs/ft3
(999.8 kg/m3 ) at 32 ◦ F (0 ◦ C) to 1.860 slugs/ft3 (958.4 kg/m3 )
at 212 ◦ F (100 ◦ C). Tables 1 and 2 show the variation
of density in U.S. Customary Units and SI units,
respectively.
SPECIFIC GRAVITY
Specific gravity is defined as the ratio of the density
of one fluid at a specified temperature to that of a
reference fluid at a reference temperature. For liquids,
the reference liquid is usually water, although varying
reference temperatures have been used, for example, 39◦
and 68 ◦ F (4◦ and 20 ◦ C). Although specification of the
reference temperature is required to ensure accuracy, it is
often not given because the density of water varies by only
0.2% over the usual temperature range (5). The specific
gravity of water is usually taken as 1.0.
ABSOLUTE (DYNAMIC) VISCOSITY (µ)
The viscosity of a fluid is expressed as a measure of the
resistance to flow. Mathematically, it is defined as the
ratio of the shear stress of a fluid to its velocity gradient or
as the proportionality constant between the shear stress
and the velocity gradient:
F
=µ
dv
(1)
A dy
Viscosity can be measured by taking two plates of area A,
which are separated by a fluid whose thickness is y. The
bottom plate is held stationary. The top plate is moved at
a constant velocity v. The force required to maintain the
constant velocity is measured and used to calculate the
viscosity µ of the fluid.
The chief cause of viscosity in fluids is molecular
cohesion. Molecular cohesion decreases as temperature
is increased, therefore, viscosity also decreases as
temperature increases. Changes in pressure do not affect
viscosity (5). Impurities also exert little influence on
viscosity (1). Tables 1 and 2 show the effect of temperature
on viscosity.
KINEMATIC VISCOSITY (ν)
There are many times in fluid mechanics and hydraulics
when the fluid’s absolute viscosity is divided by its
density. To ease calculations, this term has been defined
as kinematic viscosity because it has kinematic units of
length and time, ft2 /sec or m2 /s.
Both absolute viscosity and density are a function of
temperature, so kinematic viscosity is also a function of
temperature. Tables 1 and 2 show how kinematic viscosity
varies with temperature.
528 PHYSICAL PROPERTIES

Table 1. Physical Properties of Water in U.S. Customary Unitsa

Modulus of Dynamic Viscosity, Kinematic Viscosity, Surface Tension,c Vapor Pressure,
Specific Weight, Density, ρ, Elasticityb µ × 105 , ν × 105 , σ, pv ,
Temp., ◦ F γ , lbf /ft3 slug/ft3 E/103 , lbf /in.2 lbf · sec/ft2 ft2 /sec lbf /ft lbf /in.2

32 62.42 1.940 287 3.746 1.931 0.00518 0.09

40 62.43 1.940 296 3.229 1.664 0.00614 0.12
50 62.41 1.940 305 2.735 1.410 0.00509 0.18
60 62.37 1.938 313 2.359 1.217 0.00504 0.26
70 62.30 1.936 319 2.050 1.059 0.00498 0.36
80 62.22 1.934 324 1.799 0.930 0.00492 0.51
90 62.11 1.931 328 1.595 0.826 0.00486 0.70
100 62.00 1.927 331 1.424 0.739 0.00480 0.95
110 61.86 1.923 332 1.284 0.667 0.00473 1.27
120 61.71 1.918 332 1.168 0.609 0.00467 1.69
130 61.55 1.913 331 1.069 0.558 0.00460 2.22
140 61.38 1.908 330 0.981 0.514 0.00454 2.89
150 61.20 1.902 328 0.905 0.476 0.00447 3.72
160 61.00 1.896 326 0.838 0.442 0.00441 4.74
170 60.80 1.890 322 0.780 0.413 0.00434 5.99
180 60.58 1.883 318 0.726 0.385 0.00427 7.51
190 60.36 1.876 313 0.678 0.362 0.00420 9.34
200 60.12 1.868 308 0.637 0.341 0.00413 11.52
212 59.83 1.860 300 0.593 0.319 0.00404 14.70
a
References 3 and 7.
b
The modulus of elasticity is given at atmospheric pressure.
c
Surface tension is given for water in contact with air.

Table 2. Physical Properties of Water in SI Unitsa

Modulus of Dynamic Viscosity,
Specific Weight, Density, ρ, Elasticity,b µ × 103 , Kinematic Viscosity, Surface Tension,c Vapor Pressure,
Temp., ◦ C γ , kN/m3 kg/m3 E/106 , kN/m2 N·s/m2 ν × 106 , m2 /s σ , N/m pv , kN/m2

0 9.805 999.8 1.98 1.781 1.785 0.0765 0.61

5 9.807 1000.0 2.05 1.518 1.519 0.0749 0.87
10 9.804 999.7 2.10 1.307 1.306 0.0742 1.23
15 9.798 999.1 2.15 1.139 1.139 0.0735 1.70
20 9.789 998.2 2.17 1.002 1.003 0.0728 2.34
25 9.777 997.0 2.22 0.890 0.893 0.0720 3.17
30 9.764 995.7 2.25 0.798 0.800 0.0712 4.24
40 9.730 992.2 2.28 0.653 0.658 0.0696 7.38
50 9.689 988.0 2.29 0.547 0.553 0.0679 12.33
60 9.642 983.2 2.28 0.466 0.474 0.0662 19.92
70 9.589 977.8 2.25 0.404 0.413 0.0644 31.16
80 9.530 971.8 2.20 0.354 0.364 0.0626 47.34
90 9.466 965.3 2.14 0.315 0.326 0.0608 70.10
100 9.399 958.4 2.07 0.282 0.294 0.0589 101.33
a
References 3 and 7.
b
The modulus of elasticity is given at atmospheric pressure.
c
Surface tension is given for water in contact with air.

SURFACE TENSION (σ )
Surface tension is caused by the attractive forces between
molecules in a liquid. In a bulk liquid, each molecule is
attracted to its neighboring molecules, and the total force
on the molecule is zero. However, a molecule at the surface
experiences an attraction to other molecules deeper in the
liquid, but there are no molecules outside to balance these
attractions. The molecule at the surface experiences a net
attractive force toward the bulk liquid.
Surface tension is the amount of work necessary to
increase the surface area of the liquid by one unit. It
has units of force per unit area (ft-lbf /ft2 , or lbf /ft) and
decreases as the temperature of the liquid increases. It is
also a function of the gas contacting the liquid (5). Table 1
gives the variation of surface tension of water in contact
with air as a function of temperature in U.S. Customary
Units, and Table 2 shows the variation in SI units.
VAPOR PRESSURE (PV )
Water is a collection of a large number of molecules in a
liquid phase that move about at random. The molecules
 ENVIRONMENTAL PHOTOCHEMISTRY IN SURFACE WATERS
collide with each other and rebound in different directions.
Each molecule moves with a certain velocity (energy).
Molecules on the surface of a liquid moving away from
the bulk liquid have no other molecule to collide with.
If such a molecule has sufficient energy to escape the
intermolecular attractions, it may escape the liquid into
the vapor space above the liquid. It evaporates. Evaporated
water molecules also move about at random. Some of the
water vapor condenses into the liquid water phase.
If the water were placed in a closed container at
a constant temperature and the air evacuated from
the vapor space above the liquid, one could measure
the pressure exerted by the water vapor. The pressure
would increase until equilibrium is attained—when the
rate of molecules evaporating and the rate of molecules
condensing are equal. The pressure of the water vapor then
is constant. This is the vapor pressure of the water for the
given temperature. Table 1 gives the vapor pressure of
water in psi. Table 2 gives the vapor pressure in kN/m2 .
MODULUS OF ELASTICITY (E )
Although water is usually considered incompressible, it is
slightly compressible. The modulus of elasticity is defined
as the increase in pressure that results in a specific
percentage change in volume:
p
E= (2)
V/V
where p is the increase in pressure, V is the resulting
change in volume, and V is the original volume of water.
The modulus of elasticity has units of pressure and is a
function of temperature. Tables 1 and 2 show the variation
of the modulus of elasticity with temperature.
SPEED OF SOUND (C )
Sound is the propagation of a pressure wave through a
material and is dependent on the material’s modulus of
elasticity and density. The speed of sound, c, in a liquid is
given by

Egc
c= (3)
ρ terrestrial surface area, accounting for over 70% of the
The speed of sound in water is important in calculating the
pressure increase due to water hammer. When the velocity
in a pipe is decreased suddenly, the pipe experiences an
increase in pressure. The increase in pressure, p, caused
by a decrease in velocity of v, is
ρcv
p = (4)
gc
HEAT CAPACITY (CP )
Heat capacity is defined as the amount of heat needed
to raise the temperature of a given amount of substance
by 1◦ . It is usually expressed in units of heat per unit of
substance per degree. The term ‘‘heat capacity’’ is often
529
used for this quantity of heat on a molar basis. If a mass
basis is used, the term ‘‘specific heat’’ is frequently used.
The heat capacity of water is the basis of the unit of heat
in both the English and SI systems. Therefore, the heat
capacity (specific heat) of water is 1 Btu/pound-mass/ ◦ F in
the English system and 1 cal/g/ ◦ C in the SI system.
Heat capacities of liquids are not functions of
temperature, except at high reduced temperatures.
(A reduced temperature is a thermodynamic quantity
relating the temperature of a material to its critical
temperature. The critical temperature for water is
647.1 K.) (6). At reduced temperatures exceeding 0.7–0.8,
liquid heat capacities are large and are strong functions of
temperature.
BIBLIOGRAPHY
1. Fair, G.M., Geyer, J.C., and Okun, D.A. Water and Wastewater
Engineering. John Wiley & Sons, New York.
2. Atkins, P.W. Physical Chemistry, 2nd Edn. W.H. Freeman,
San Francisco.
3. Metcalf & Eddy, Inc. Wastewater Engineering: Treatment,
Disposal, and Reuse, 3rd Edn. McGraw-Hill, New York.
4. F.S. Merritt (Ed.). Standard Handbook for Civil Engineers, 3rd
Edn. McGraw-Hill, New York.
5. Lindeburg. Civil Engineering Reference Manual.
6. Smith, J.M. and van Ness, H.C. Introduction to Chemical
Engineering Thermodynamics, 3rd Edn. McGraw-Hill, New
York.
7. Culp/Wesner/Culp. Handbook of Public Water Systems. Van
Nostrand-Reinhold, New York.
ENVIRONMENTAL PHOTOCHEMISTRY IN
SURFACE WATERS
MARK A. YOUNG
University of Iowa
Iowa City, Iowa
INTRODUCTION
Surface waters make up a large fraction of the available
total area with most of that consisting of ocean waters. A
correspondingly large fraction of the incoming solar flux
will impinge on natural waters, penetrating to a depth
determined by the scattering and absorption properties
of the water system. While most of the short wavelength
solar radiation is attenuated in the atmosphere, mainly
by absorbing species such as O2 and O3 , the surface flux
of energetic photons in the near-UV and visible remains
appreciable. While water itself is not photoactive, the
irradiated volume of typical surface waters contains a
plethora of potential chromophores. The chromophoric
material consists of dissolved species and suspended
particulate matter that can have strong electronic
absorptions that significantly overlap with the available
solar flux. Absorption of solar photons can initiate a rich
variety of photochemical processes with sufficient kinetic
530 ENVIRONMENTAL PHOTOCHEMISTRY IN SURFACE WATERS

rates to manifest a substantial impact on the chemistry

Dissolved and colloidal chromophores
taking place in the aqueous environment. Photochemically x
driven chemistry, in turn, can have enormous implications
for global geochemical cycles, climate, and biogenic DOM, NO3−/NO2−, > Fe(III)-L,
processes, making photochemistry a critical consideration
in the study of environmental processes.
+ hv

H2O, O2, R
BACKGROUND

Natural surface waters can be either saltwater (marine, or

ocean, waters) or freshwater (continental waters). Within
each of these reservoirs, widely varying environments are
found: remote and coastal areas in the oceans, and rivers,
lakes, streams, bogs/marshes on the continents, as well
as interfacial environments, such as estuaries. In addition Reactive intermediates
to differences in salinity and pH, there will be variations
e−(aq), O2−, HO2−, RO2−, 1O2+
in the types and concentrations of chromophoric species
residing in each reservoir. Potential chromophores include OH−, H2O2, > Fe(II), DOM+ −, R−
dissolved organic material (DOM), inorganic NOx − ions,
metal complexes, anthropogenic pollutants, and colloidal
H2O, O2, R
matter consisting of crustal minerals and particulate
organic material (POM). Organic chromophores may also
adsorb on, or complex with, the surface of mineral
colloids and participate in charge transfer processes.
Likewise, metals can become incorporated into organic
material. The distinction between dissolved and colloidal
material is not sharp and often depends on the analytical
procedure used to assay concentrations. The relative
importance of the various chromophores will depend on Final products
the details of the aqueous environment. Chromophores
can absorb solar photons and undergo photochemical Figure 1. Schematic representation of photochemical processes
transformation while yielding a variety of primary and in surface waters. Chromophores consist of DOM, nitrate/nitrite
secondary reactive species, as depicted schematically in ions, dissolved and colloidal metal complexes [represented by
Fig. 1. The photogenerated intermediates can, in turn, Fe(III)], and anthropogenic pollutants (such as substituted
react with other species present in the water at rates that aromatics). Reactive intermediates consist of solvated electrons,
various radical species, both cationic and anionic, peroxides, and
are often diffusion controlled to form final products. In
reduced metal centers.
some instances, the chromophore may be regenerated,
resulting in a photocatalytic cycle that can be highly
efficient (Fig. 1). Photophysical processes, luminescence (fluorescence and
Excitation of a chromophore can initiate a number phosphorescence), internal conversion, intersystem cross-
of photophysical and photochemical processes that are ing—Eq. 2—and collisional quenching—Eq. 3—will re-
outlined in the following equations: turn the chromophore to the electronic ground state and
convert the photon energy to heat. The excited state may
J also ionize, Eq. 4, or react, Eq. 5, converting the photon
S + hν −−−→S∗ excitation; (1) energy into chemical potential. The reaction could be uni-
kL ,kI molecular, such as dissociation, or bimolecular in nature,
S∗ −−−→S (+hν) luminescence, internal
involving other components of the water system. Products
conversion, and intersystem
might include the reactive intermediates in Fig. 1, which
crossing; (2)
can then participate in secondary chemistry. Energy trans-
kQ fer, or sensitization, Eq. 6, could also induce a concurrent
S∗ + Q−−−→S + Q quenching; (3)
change in the electronic spin state, yielding a reactive
∗ + −
S −−−→ S + e photoionization; (4) triplet excited state.
kR
The photochemical sequence is initiated by absorption
−−−→ of a photon, Eq. 1, with a photolytic rate constant, J. From
S∗ P reaction; (5) a simple consideration of Beer’s Law, J is seen to depend
[R]
kET
on the incident solar flux, I, and the (molar, decadic)
S∗ + R−−−→S + R∗ energy transfer; (6) extinction coefficient of the chromophore, ε. The relative
kCT
efficiency, or yield, of a reaction can be specified by the
S∗ + R−−−→S+/− + R−/+ charge transfer. (7) quantum yield, φ, which is the number of product species

formed divided by the number of photons absorbed. The
absorption coefficient and quantum yield are generally
wavelength dependent. The flux depends on wavelength
as well as the solar zenith angle and other factors, such as
latitude, longitude, and meteorological conditions, such as
cloud cover. The rate constant for a photoinduced process
can then be written as an integral over the wavelength
region of interest:
 
J = 2.303 ε(λ)φ(λ)I(λ)dλ = 2.303 ε(λi )φ(λi )I(λi )λi
i While highly diverse in nature, an average chromophore
(8)
The flux in this case is termed the spectral actinic flux and
is a spherically averaged irradiance that takes into account
radiation incident on the chromophore over 4π steradians.
For practical purposes, the integral is often evaluated
as a discrete sum over small wavelength intervals, as
expressed in the last term of Eq. 8, where ε, φ, and
I represent average quantities over the interval λi
centered at λi .
The solar actinic flux at noon (zenith angle = 0◦ ) for a
northern latitude (47.4◦ N) is shown in Fig. 2. The flux is
expressed in terms of the photon flux per unit wavelength
interval with units of einstein · m−2 · s−1 · nm−1 , where an
einstein corresponds to 1 mole of photons (Avogadros’s
number, NA = 6.022 × 1023 mol−1 ). The UV-B radiation
(280–320 nm) is strongly attenuated by the ozone layer,
effectively limiting the useful actinic flux to about 300 nm.
Shorter wavelength radiation does reach the surface but
at 293 nm, the flux has decreased by more than two orders
of magnitude relative to 300 nm. UV-A (320–400 nm) is
not strongly absorbed by ozone and the flux increases up
to the maximum in the visible at approximately 575 nm.
A photon has a molar energy of
hc
E = NA J · einstein−1
λ particular reaction to calculate the reaction half-life as
A 300-nm photon has an energy of 399 kJ · einstein−1 ,
which is on the order of typical bond enthalpies (e.g.,
1.0
0.8
(e)
0.6
Absorbance
0.4 (d)
(c)
0.2 (b)
(a)
0.0
200 250 300 350 400
Wavelength, nm
ENVIRONMENTAL PHOTOCHEMISTRY IN SURFACE WATERS 531
C–C, 348 kJ; O–O, 146 kJ; C–OH, 322 kJ; benzylic C–H,
368 kJ; phenolic O–H, 362 kJ; C–Cl, 342 kJ). Over the
wavelength range of photochemical interest, 300–500 nm,
Fig. 2 indicates that the noon-time flux to surface
waters is about 4 einstein · m−2 · h−1 . Typical penetration
depths range from 1 to 20 m and so the molar rate
of photon absorption is 0.2–4.2 × 10−3 einstein·L−1 ·h−1 .
Dissolved organic carbon (DOC) is the major absorbing
species in most natural waters (see Fig. 2) and has
an approximate average concentration of 5 mg C·L−1 .
molecular weight of 120 g C·mol−1 (10 carbon atoms)
can be assumed and so the absorber concentration is
approximately 0.04 mM. The estimated photon absorption
rate for a well mixed volume would correspond to
an average excitation rate of 5–100 h−1 , meaning that
a near-surface chromophore is excited approximately
once each minute (2,3). Furthermore, it is estimated
that the accumulated annual radiation dose is about
1300 times larger (2). Given the time scales for transport
and chemistry in surface waters, it is clear that
photochemically driven reactions will be an important
component of chemical cycles in the hydrosphere (Fig. 2).
REACTIVE INTERMEDIATES
A variety of reactive intermediates have been identified
from the irradiation of natural waters at relevant solar
wavelengths (see Fig. 1). Selected properties for some of
the important species are listed in Table 1, along with
estimated steady-state concentrations in typical surface
waters (generally corresponding to a noon-time solar
flux). Under conditions of constant solar illumination,
pseudo-first-order kinetics result and the steady-state
concentration for an intermediate, [I]SS , can be used in
(9) conjunction with the second-order rate constant, k, for a
ln 2
τ= (10)
k[I]SS
10
8
Einstein × 106 m−2⋅s−1⋅nm−1
4
Figure 2. Left ordinate: Absorption spectra of
2 surface water samples with varying concentra-
tions and sources of DOC in mg C·L−1 : (a) = 3.0,
(b) = 7.8, (c) = 13.4, (d) = 13.4, (e) = 15.4. Path
length is 1.0 cm. Data are adapted from Ref-
0
erence 7. Right ordinate: Solar actinic flux for
450 500
noon-time sun (zenith angle = 0◦ ) for a northern,
midlatitude. Data is from Reference 1.
532 ENVIRONMENTAL PHOTOCHEMISTRY IN SURFACE WATERS

and, thus, estimate the importance of a given reaction
pathway. For example, half-lives for reactions with phenol,
a representative organic, aromatic compound, have been
included in Table 1.
Primary transient species are produced rapidly, on
the order of <1 ns. These transients can then react
with other components in the aquatic environment to
produce secondary products. Besides H2 O, one of the
most important interactions involves dissolved oxygen,
which, in well aerated surface waters, will be present in
concentrations determined by Henry’s Law,
[O2 ] = KH PO2
where KH = 1.3 × 10−3 M−1 · atm−1 is the Henry’s Law
constant (8). The mixing ratio of O2 in the atmosphere
is 21%, yielding a concentration of [O2 ] = 0.27 mM at
298 K. Primary radicals will react with dissolved O2
at diffusion controlled rates (diffusion controlled rate
constant, kD ≈ 7 × 109 M−1 · s−1 at 298 K) corresponding
to a time scale of ≈ 0.4 µs, which will limit the lifetime of
many photogenerated transients (the ‘‘oxygen wall’’) (9).
reaction with e− (aq) has been shown to degrade organic
pollutants (2,10), it is unlikely to play an important role
in photochemical processes in most surface waters.
O2 −
Superoxide, O2 − , can be formed from reaction with e− (aq),
as above, in charge transfer reactions with sufficiently
reducing species, and via electron trapping at the
surface of semiconductor colloids. Photolysis of DOM is
a major source of superoxide but photolysis of trace metal
complexes may also contribute to O2 − formation (11).
(11) Superoxide is a mild reductant/oxidant (Table 1) and
can participate in both one-electron oxidation and
reduction reactions. The steady-state concentration is
uncertain but from limited formation rate measurements,
the concentration is [O2 − ]SS ≈ 10−9 –10−7 M (9,10). The
relatively high concentration suggests that superoxide
may be important in photochemical degradation of various
species (7). However, O2 − is unstable in water with respect
to disproportionation to form hydroperoxyl radical and
hydrogen peroxide (3,12)

e− (aq) K
O2 − + H+ −−
−−−−HO2 log K = −4.8 (13)
Electrons generated from photoionization of DOM are
HO2 /O2 − + H+ −−−→ H2 O2 + O2 (14)
rapidly solvated in aqueous environment to produce,
e− (aq). The solvated electron reacts with dissolved oxygen
to form the superoxide, RO2 •

k
e− (aq) + O2 −−−→O2 −
where k ≈ 2 × 1010 M−1 · s−1 [which is larger than the
diffusion limit due to a tunneling mechanism for e− (aq)
transport] (4,5). The time scale for Eq. 12 is on the order
of 200 ns. The short lifetime coupled with measured
typical production rates leads to a very small steady-state
concentration of [e− (aq)]SS ≈ 10−17 –10−15 M (2,9). While
Peroxyl radicals can be formed from the photolysis of DOM
(12) and metal complexes. Formation rates for RO2 species are
thought to linearly scale with the DOM concentration in
typical waters (2), but the reactivity of the radical is highly
dependent on the type of humic substance involved, which
is itself not well characterized. These uncertainties make
it difficult to estimate a steady-state concentration for
peroxyl radicals. HO2 is a mild oxidant/reductant (Table 1)
but DOM-derived peroxyl radicals exhibit significant

Table 1. Selected Properties of Some Important Reactive Intermediates in

Aqueous Photochemistrya
Radical

Property −
e (aq) O2 −
HO2 OH 1
O2 ∗

Gf 0 (kJ·mol−1 ) 276b 13b

Hf 0 (kJ·mol−1 ) 277b 80c 138c −7b 94.3d
E0 (V)e −2.9f −0.33g 0.79h 1.9i 0.65j
[I]SS (M) 10−17 –10−15 10−9 –10−7 10−19 –10−17 10−14 –10−13
k (M−1 ·s−1 )k 3.0(7) 5.8(2) 8.5(3) 1.0(10) 0.2(7)
τ (days × 103 ) 0.3–30 0.0001–0.01 0.08–8 0.04–0.4
a
The rate constant data is for reaction with C6 H5 OH (phenol) except for the HO2 radical, which is for
reaction with 1,4-C6 H4 (OH)2 .
b
Data from Reference 4.
c
Data from Reference 5.
d
Gas phase value.
e
Data from Reference 6.
f
H2 O + e− → e− (aq)
g
O2 + e− → O2 −
h
HO2 + e− → HO2 −
i
OH + e− → OH−
j1
O 2 ∗ + e− → O 2 −
k
Number in parentheses indicates the exponent value. Data from Reference 7.
 ENVIRONMENTAL PHOTOCHEMISTRY IN SURFACE WATERS
reactivity with alkylphenols (13) and so may be an
important reaction channel for some pollutants.
1
O2 ∗
Singlet oxygen, 1 O2 ∗ , is an electronically excited state
of oxygen with 94.3 kJ of excess energy. Triplet excited
3
DOM∗ , generated by intersystem crossing from the
initially excited singlet state, can sensitize the formation
∗
of 1 O2 through energy transfer, Eq. 6. Most of the singlet
oxygen will be rapidly quenched back to the triplet ground
∗
state by H2 O (τ ≈ 4 µs) and only a fraction of the 1 O2 will
be available for subsequent reaction (2). The production of
1 ∗
O2 generally scales with the amount of DOC and steady-
state concentrations of 0.3–3.0 × 10−14 M per mg·L−1 of
DOC have been measured for typical surface waters (14).
Singlet oxygen is a reactive electrophile, particularly
toward cyclic 1,3 dienes, phenolate anions, and some
polynuclear aromatics (2,14,15) and can play an important
role in aqueous photochemistry.
OH
The hydroxyl radical is a powerful oxidant (Table 1)
that exhibits high reactivity toward a range of organic
species. It is mainly formed from photodissociation of
NO3 − /NO2 − but also via photolysis of metal-hydroxides,
decomposition of H2 O2 , and DOM photochemistry. The
hydroxyl radical is rapidly scavenged by DOM and
the steady-state concentration of OH in surface waters
is generally small, [OH]SS ≈ 10−19 –10−17 M (9,10). The
OH radical readily reacts with many organic species,
via abstraction and addition to double bonds, but the
low steady-state concentration suggests that hydroxyl
will not be the dominant photooxidant for most surface
waters. For example, a ‘‘typical’’ OH rate constant (4)
is 5 × 109 M−1 · s−1 , corresponding to a reaction half-life
of τ ≈ 160 d at the higher OH concentrations. However,
for environments with a larger concentrations of NOx −
relative to DOC, hydroxyl radical reactions can be
important pathways due to near diffusion-controlled OH
rate constants (2).
H2 O2
Hydrogen peroxide is a relatively stable, nonradical
species produced from photochemical processes in natural
surface waters. It is not a primary product but,
rather, is generated in secondary reactions such as
the disproportionation of superoxide, Eqs. 13 and 14.
Hydrogen peroxide absorbs only weakly in the UV
(ε = 1 M−1 · cm−1 at 310 nm) (16), and decomposition to
yield hydroxyl radicals may be dominated by biological
mediation and redox cycles involving trace metals
H2 O2 −−−→ 2OH (15)
The slow decomposition of H2 O2 leads to relatively high
steady-state concentrations in surface waters, [H2 O2 ]SS =
10−8 –10−7 M (2,9).
533
CHROMOPHORES
NO3 − /NO2 −
Inorganic nitrate and nitrite ions are perhaps the most
well defined and easily understood photoactive component
of natural waters. Nitrate concentration in surface
waters is on the order of 1 mg·L−1 although it may be
substantially higher in waters subject to an influx of
agricultural drainage due to the heavy use of nitrogen
fertilizers. The absorption band maxima for nitrate and
nitrite are at 302 nm (ε = 3.1 M−1 ·cm−1 ) and 352 nm (ε =
9.6 M−1 ·cm−1 ), respectively (17,18). Ultraviolet excitation
of nitrite leads to photodissociation (18):
NO2 − + hν −−−→ NO + O
ž −
(16)
The ž O− species, the conjugate base of hydroxyl radical, is
rapidly protonated at pH < 12 [pKa (ž OH) = 11.9] (4):
O− + H+ −−−→ OH
ž ž
(17)
The quantum yield for ž OH production from nitrite
varies from 0.015 to 0.08 over the wavelength range
298–371 nm (18,19). Photolysis of the nitrate ion yields
either nitrite ion or nitrogen dioxide, which can then
reform nitrate and nitrite:
NO3 − + hν −−−→ NO2 − + O (18)
−
NO3 − + hν −−−→ NO2 + O
ž ž
(19)
− − +
2 NO2 + H2 O −−−→ NO2 + NO3 + 2H
ž
(20)
For wavelengths ≥290 nm, the quantum yield for
formation of ž OH from NO3 − , via Eq. 19, is in the
range of 9.2–17 × 10−3 (17,18). Hydroxyl is a highly
reactive radical and a strong oxidizing agent (see
above). Various laboratory studies have shown that
trace organic molecules, such as biologically refractory
pollutants, are degraded by irradiation in the presence of
nitrate/nitrite (17–19).
DOM
Dissolved organic matter (DOM) is a major and ubiquitous
component of natural surface waters, having numerous
sources such as soils, sediments, and aqueous biological
activity (20). DOM concentrations are often expressed in
terms of the dissolved organic carbon (DOC) fraction,
which comprises about 60% of the total DOM. Continental
waters have typical levels of 1–10 mg C·L−1 DOC,
although some waters, bogs, marshes, and shallow
streams, for instance, can have concentrations exceeding
50 mg C·L−1 . Marine DOC is about 1 mg C·L−1 and
represents an enormous carbon reservoir, approximately
equal in size to the CO2 reservoir in the atmosphere (2).
DOM is nominally defined as the organic content of
water that passes through a filter of some pore size,
0.1–0.5 µm, and so will have a colloidal component.
The DOC in surface waters can be divided into two
fractions; a lower molecular weight nonhumic fraction and
higher molecular weight humic component. The nonhumic
534 ENVIRONMENTAL PHOTOCHEMISTRY IN SURFACE WATERS
fraction is about 10–60% of the DOC and is made
up of small compounds of mainly biogenic origin, such
as peptides, carbohydrates, fats, and chlorophyll, which
are subject to rapid biological degradation. In addition,
anthropogenically derived species, such as biocides
(often halogenated compounds) and polycyclic aromatic
hydrocarbons (PAHs, from combustion processes) are
present. The higher molecular weight fraction (>300 Da),
comprising 40–90% of the DOC, is termed humic matter
(HM). Humic matter is a complex material composed
of condensed biopolymers and is much more biologically
refractory. HM is mildly acidic, has a significant aromatic
content, and is highly functionalized with carboxyl, phenol,
hydroxyl, carbonyl, ether, and ester functional groups
present. HM can be further divided based on solubility
properties into humic acids (HAs), which precipitate under
acidic conditions, and fulvic acids (FAs), which are soluble
at all pH values. FAs are more highly substituted and
have a higher oxygen content than HAs and also have a
lower average molecular weight.
DOM is a highly colored material with an absorption
spectrum that overlaps significantly with the solar actinic
flux (see Fig. 2). The UV absorption of surface waters
has been found to be highly correlated with the DOC
content (10,21,22). Studies have shown that photoexcited
DOM can yield many of the radical species outlined
in Fig. 1 and Table 1 through primary and secondary
processes (9). The radical production rate varies with the
character of the DOM and the concentration, as well as
the location and type of surface water environment. One
study found that coastal marine waters yielded radical
production rates of 0.18–1.8 µM·d−1 in the summer due
to irradiation of humic material in the surface layer (to
a depth of 1.5 m) (23). The active spectral region was
determined to be 300–385 nm, suggesting a quantum
yield for radical production of 1–2 × 10−3 . Measurement
of light absorption rates in highly colored, organic-rich
waters that exhibit significant absorption at longer, visible
wavelengths suggest the radical production rate will be
orders of magnitude higher (9). These radicals dominate
the photochemistry of most natural water samples,
particularly in marine environments.
The DOM will be altered by photolysis, yielding
a cationic radical product, for instance, but it can
also be attacked by the photogenerated radicals. DOM
decomposition can lead to photobleaching, allowing solar
radiation to penetrate deeper into the water and impacting
microbial photosynthesis. DOM is generally biologically
refractory, resulting in long carbon reservoir residence
times, estimated at 6000 years for deep ocean DOM (24).
Photodecomposition of DOM may be an important removal
pathway with significant implications for the global carbon
cycle (9,15,24,25). The lower molecular weight degradation
products are biologically labile and can provide a major
fraction of the biological carbon uptake in certain surface
waters (21,25).
Fe(III/II)
Iron is an abundant crustal metal and is one of
the most important trace metals in natural water
systems. The Fe-(hydrous)oxides (e.g., α-FeOOH, goethite;
α-Fe2 O3 , hematite; γ -FeOOH, lepidocrocite; Fe(OH)3 ,
ferrihydrite) may be present as dissolved species but are
predominantly in colloidal form. Iron has two common
oxidation states, Fe(III) and Fe(II), and participates
in redox reactions that are an important component
of aqueous chemical cycles. Fe(III) can form strong
complexes with naturally occurring organic ligands, such
as polycarboxylate anions (e.g., oxalate, malonate, citrate),
and humic and fulvic acids (11,26). Photoexcitation of a
Fe(III) complex promotes a ligand to metal charge transfer
process yielding a free, oxidized ligand and a reduced,
Fe(II), center. Subsequent reactions generate reactive
intermediates (11,12), as schematically depicted below for
the oxalato (C2 O4 2− ) ligand:
>Fe(III)C2 O4 − + hν −−−→ >Fe(II) + C2 O4 − (21)
−
C2 O4 −−−→ CO2 + CO2 −
(22)
− −
CO2 + O2 −−−→ CO2 + O2 (23)
>Fe(II) + O2 −−−→ >Fe(III) + O2 − (24)
>Fe(II) + H2 O2 −−−→ >Fe(III) + OH (25)
where ‘‘>Fe(III/II)’’ represents either an iron colloid
surface site or a dissolved species. The superoxide will form
hydrogen peroxide and hydroperoxyl radicals, as discussed
above. In colloidal material, the surface Fe(II) is generally
oxidized back to Fe(III) by the ligand or O2 . However, due
to weakened binding with neighbor atoms, the reduced
species may also be released to the solution (3,26). The
result is photodissolution of the mineral colloid and an
increase in soluble iron species, facilitating redox reactions
and promoting microbial activity.
BIBLIOGRAPHY
1. Finlayson-Pitts, B.J. and Pitts, J.N., Jr. (1986). Atmospheric
Chemistry: Fundamentals and Experimental Techniques.
Wiley-Interscience, Hoboken, NJ.
2. Hoigne, J., Faust, B.C., Haag, W.R., Scully, J.F.E., and
Zepp, R.G. (1989). Aquatic humic substances as sources
and sinks of photochemically produced transient reactants.
In: Aquatic Humic Substances. Vol. 219, I.H. Suffet and
P. MacCarthy (Eds.). American Chemical Society, Washing-
ton, DC, pp. 363–381.
3. Stumm, W. and Morgan, J.J. (1996), 3rd Edn. John Wiley &
Sons, New York.
4. Buxton, G.V., Greenstock, C.L., Helman, W.P., and Ross, A.B.
(1988). Critical review of rate constants for reactions of
hydrated electrons, hydrogen atoms and hydroxyl radicals
(OH/O− ) in aqueous solution. J. Phys. Chem. Ref. Data 17:
513–886.
5. Spinks, J.W.T. and Woods, R.J. (1990). An Introduction to
Radiation Chemistry, 3rd Edn. John Wiley & Sons, Hoboken,
NJ.
6. Wardman, P. (1989). Reduction potentials of one-electron
couples involving free radicals in aqeous solution. J. Phys.
Chem. Ref. Data 18: 1637–1755.
7. NDRL Radiation Chemistry Data Center. Available:
http://www.rcdc.nd.edu/.
8. Seinfeld, J.H. and Pandis, S.N. (1998). Atmospheric Chem-
istry and Physics: From Air Pollution to Climate Change.
John Wiley & Sons, Hoboken, NJ.

9. Mopper, K. and Kieber, D.J. (2000). Marine photochemistry
and its impact on carbon cycling. In: The Effects of UV
Radiation in the Marine Environment. Vol. 10. S.D. Mora,
S. Demers, and M. Vernet (Eds.). Cambridge University
Press, Cambridge, UK, p. 101–129.
10. Cooper, W.J., Zika, R.G., Petasne, R.G., and Fischer, A.M.
(1989). Sunlight-induced photochemistry of humic substances
in natural waters: major reactive species. In: Aquatic Humic
Substances. Vol. 219. I.H. Suffet and P. MacCarthy (Eds.).
American Chemical Society, Washington, DC, pp. 333–362.
11. Faust, B.C. (1994). A review of the photochemical redox reac-
tions of iron(III) species in atmospheric, oceanic, and surface
waters: influences on geochemical cycles and oxidant for-
mation. In: Aquatic and Surface Photochemistry. G.R. Helz,
R.G. Zepp, and D.G. Crosby (Eds.). Lewis Publishers, Boca
Raton, FL, pp. 3–37.
12. Zuo, Y. and Hoigne, J. (1992). Formation of hydrogen peroxide
and depletion of oxalic acid in atmospheric water by photolysis
of iron(III)-oxalato compounds. Environ. Sci. Technol. 26:
1014–1022.
13. Faust, B.C. and Hoigne, J. (1987). Sensitized photooxidation
of phenols by fulvic acid and in natural waters. Environ. Sci.
Technol. 21: 957–964.
14. Haag, W.R. and Hoigne, J. (1986). Singlet oxygen in surface
waters. 3. Photochemical formation and steady-state concen-
tration in various types of waters. Environ. Sci. Technol. 20:
341–348.
15. Aguer, J-P., Richard, C., and Andreux, F. (1997). Comparison
of the photoinductive properties of commercial, synthetic and
soil-extracted humic substances. J. Photochem. Photobiol. A
103: 163–168.
16. Bolton, J.R. and Cater, S.R. (1994). Homogeneous pho-
todegradation of pollutants in contaminated water: an intro-
duction. In: Aquatic and Surface Photochemistry. G.R. Helz,
R.G. Zepp, and D.G. Crosby (Eds.). Lewis Publishers, Boca
Raton, FL, pp. 467–490.
17. Zepp, R.G., Hoigne, J., and Bader, H. (1987). Nitrate-induced
photooxidation of trace organic chemicals in water. Environ.
Sci. Technol. 21: 443–450.
18. Alif, A. and Boule, P. (1991). Photochemistry and the
environment part XIV. Phototransformation of nitrophenols
induced by excitation of nitrite and nitrate ions. J. Photochem.
Photobiol. A 59: 357–367.
19. Marbury, S.A. and Crosby, D.G. (1994). The relationship
of hydroxyl reactivity to pesticide persistence. In: Aquatic
and Surface Photochemistry. G.R. Helz, R.G. Zepp, and
D.G. Crosby (Eds.). Lewis Publishers, Boca Raton, FL,
pp. 149–161.
20. Beckett, R. (1990). The surface chemistry of humic substances
in aquatic systems. In: Surface and Colloid Chemistry in
Natural Waters and Water Treatment. R. Beckett (Ed.).
Plenum Press, New York, pp. 3–20.
21. Kieber, R.J., Zhou, X., and Mopper, K. (1990). Formation of
carbonyl compounds from UV-induced photodegradation of
humic substances in natural waters: fate of riverine carbon
in the sea. Limnol. Oceanogr. 35: 1503–1515.
22. Miller, W.L. (1994). Recent advances in the photochemistry
of natural dissolved organic matter. In: Aquatic and Sur-
face Photochemistry. G.R. Helz, R.G. Zepp, and D.G. Crosby
(Eds.). Lewis Publishers, Boca Raton, FL, pp. 111–127.
23. Zafiriou, O.C. and Dister, B. (1991). Photochemical free
radical production rates: Gulf of Maine and Woods Hole-
Miami transect. J. Geophys. Res. 96: 4939–4945.
ISOTOPE EXCHANGE IN GAS-WATER REACTIONS 535
24. Mopper, K. et al. (1991). Photochemical degradation of
dissolved organic carbon and its impact on the oceanic carbon
cycle. Nature 353: 60–62.
25. Kieber, D.J., McDaniel, J., and Mopper, K. (1989). Photo-
chemical source of biological substrates in sea water: impli-
cations for carbon cycling. Nature 341: 637–639.
26. Sulzberger, B., Laubscher, H., and Karametaxas, G. (1994).
Photoredox reactions at the surface of iron(III) (hydr)oxides.
In: Aquatic and Surface Photochemistry. R. Helz, R.G. Zepp,
and D.G. Crosby (Eds.). Lewis Publishers, Boca Raton, FL,
pp. 53–73.
ISOTOPE EXCHANGE IN GAS-WATER
REACTIONS
TOM WYDRZYNSKI
WARWICK HILLIER
The Australian National
University
Canberra, Australia
INTRODUCTION
Molecular oxygen began to accumulate in Earth’s
atmosphere about 2.2–2.7 billion years ago, and it is
widely accepted that it originated from the photosynthetic
oxidation of water (1). The accumulation of O2 in the
atmosphere marked the most important event in the
evolution of life on Earth as the atmosphere changed from
a highly reducing to an oxidizing condition. All preexisting
life forms had to either find an anaerobic ecological niche or
adapt to the ever-increasing O2 levels of the atmosphere.
Today on Earth, the photosynthetic O2 generators are
widely distributed on land and in the oceans in the form of
plants, algae, and cyanobacteria. The current atmospheric
pool of O2 contains about 4 × 1019 mols and takes about
3.7 million years to turnover. The unique enzyme complex
that produces the O2 by photosynthetic organisms is called
photosystem II (PSII), which acts as a water/plastoquinone
oxido-reductase (2). PSII contains a special photochemical
reaction center, which at excitation can generate strong
oxidants that can be coupled to the splitting of two
water molecules into O2 , four protons, and four electrons.
The protein motif within PSII that catalyzes the water
oxidation reaction contains a cluster of four Mn ions, a Ca
ion, and a redox-active tyrosine and is often referred to as
the O2 evolving complex (OEC).
The oxidation of water by PSII is driven photochemi-
cally by the capture of light quanta (3,4). The chlorophyll
binding proteins associated with peripheral regions of
the PSII complex serve as light-harvesting antenna and
direct spectrally diverse quanta to the special reaction
center chlorophyll cluster called P680. At absorption of a
light quantum, P680 generates an excited state with its
pheophytin partner (P680-Pheo)∗ , which in turn is in equi-
librium with the exciton in the antenna and the radical
pair P680+ -Pheo− . The time course for this reaction has
been under some debate, but it is in the order of 3–20 ps.
The primary charge separation is rapidly followed by a
charge stabilization event, in which the electron is rapidly
536 ISOTOPE EXCHANGE IN GAS-WATER REACTIONS
transferred to a tightly bound plastoquinone molecule, QA ,
in 300–500 ps, which results in the formation of P680+ -
Pheo QA − . The electron hole in P680+ is subsequently filled
with an electron derived from a redox-active tyrosine, YZ ,
located in the D1 protein subunit within some 40–380 ns,
which results in the formation of YZ ox P680-Pheo QA − .
The electron at QA − is then transferred to a second bound
plastoquinone molecule, QB , in 100–200 µs, whereas YZ ox
is reduced by the Mn4 Ca cluster with S-state-dependent
kinetics (see below) in the range of 30–1300 µs.
One key finding for understanding the water oxidation
reaction is the observation that for dark-adapted, PSII-
containing samples illuminated by a series of brief,
saturating light flashes, O2 is released with a damped
periodicity of four. The peak O2 yields occur on the third,
seventh, eleventh, and so on flashes and eventually damp
out after the third or fourth cycle (5, see Figs. 2B and 2C).
The phenomenological model to explain this observation
was developed by Kok et al. (6,7) and entails a cyclic
reaction sequence in which the OEC traverses through
five intermediate states called the S states. Beginning in
S0 and progressing to S4 , each S state is advanced by a
single quantum event at P680. When it reaches S4 , O2 is
released within 1–2 ms, S0 is regenerated, and the cycle
begins anew (see Fig. 1).
The damping in the O2 flash yield oscillations (which
is caused by the desynchronization of the forward S
state transitions within a given population of O2 evolving
centers) is explained by a certain probability for a ‘‘miss’’
(α) to occur during a flash in which an electron hole fails
to accumulate in the OEC and, depending on the flash
duration, by a certain probability for a ‘‘double hit’’ (β) to
occur in which a double excitation event takes place. The
maximum O2 yield on the third flash is explained by the
condition that the cycle initially starts in the S1 state after
complete dark-adaptation. The higher S states (S2 and
S3 ) exhibit lifetimes of tens of seconds to several minutes,
depending on the temperature, and deactivate to S1 if not
advanced by another incoming quantum, whereas the S0
state slowly relaxes in the dark to the S1 state (8). The S4
state being metastable reacts immediately to produce O2
and the S0 state.
DURING WHICH STEP IN THE S STATE CYCLE DOES THE
SUBSTRATE WATER BIND?
In the original Kok et al. hypothesis, it was implied that
the two substrate water molecules entered the reaction
sequence during the last step, just before O2 release.
Although this notion has been considered throughout the
literature, the most recent models for the OEC generally
predict that the substrate water binds to the catalytic
site at the beginning of the S state cycle and that a
hν hν hν hν
S0 S1 S2 S3 S4
O2
Figure 1. OEC Cycle.
concerted four-electron event occurs during the S3 to S4
to S0 transition to create the O–O bond (9,10). However,
definitive experimental evidence to show at which step in
the S state cycle the substrate water binds and the O–O
bond forms has until recently remained elusive.
The binding of the substrate water to the catalytic site
can be probed by oxygen isotope exchange measurements
between the solvent water and the photogenerated O2
using mass spectrometry. Basically, the measurement
involves the transfer of a photosynthetic sample into
labeled water of known oxygen isotopic composition (i.e.,
18
O-enriched water) and then determining the extent of
isotopic incorporation into the photogenerated O2 .
In the first oxygen isotope exchange measure-
ments (11–13), the fastest detectable rate of exchange
was limited to ∼30 s, mainly because of the long stabi-
lization time of the ‘‘open’’ chamber system that intro-
duced isotopically labeled water to the sample. However,
in the more recent measurements from our group, the
stabilization time was greatly improved, by a factor of
nearly 5000, through the introduction of a ‘‘closed’’ cham-
ber system (14,15). The results from this experimental
approach revealed for the first time that one substrate
water molecule is bound to the catalytic site through-
out the S-state cycle, whereas the second substrate water
molecule is bound in at least the S2 and S3 states, and
that the O–O bond indeed forms during the last step of
the reaction sequence. This article documents the exper-
imental approach and the arguments that come to these
conclusions.
RAPID 18 O EXCHANGE MEASUREMENTS DURING
PHOTOSYNTHETIC O2 EVOLUTION
To probe the substrate water bound at the OEC, excess
18
O isotope (added as 18 O-labeled water) is introduced into
the sample to set up a new isotopic distribution, because
natural abundance 18 O in the solvent water is low at
about 0.2%. The effect of the new isotopic distribution
on the photogenerated O2 can then be measured at
m/e = 32, m/e = 34, and m/e = 36 with an in-line mass
spectrometer (e.g., the VG Instruments IsoPrime stable
isotope spectrometer Manchester UK) for the 16 O16 O,
16 18
O O, and 18 O18 O di-oxygen products, respectively. From
these measurements, the extent of 18 O incorporation
during the O2 evolving process can be determined.
Figure 2(a) shows the O2 flash yields at these three
masses for spinach thylakoid membrane samples after
the injection of an aliquot of H2 18 O (e.g., 98.5% enriched
water from ISOTECH, Miamisberg, Ohio) that gives a
final 18 O isotopic enrichment in the sample of ε ∼ 12%.
In this case, after complete equilibration, the absolute
yields (as determined by the extrapolated lines shown in
the figure for the third flash) at the three masses follow
closely the theoretical isotope equilibrium distribution:
32:34:36 = (1 − ε)2 : 2ε(1 − ε) : ε2 = 100%, which indicates
that no major isotopic discrimination occurs during the
O2 evolving process, as shown earlier from steady-state
measurements (16).
Figure 2(b) shows a plot of the normalized O2 yields
at m/e = 32 as a function of flash number. The oscillatory

(a)
Mass spectrometer oxygen measurements
m/e = 32
(1940) (b)
Normalised O2 yield
m/e = 34
(507)
(c)
m/e = 36
Normalised O2 yield
(31)
Light flash
pattern in the mass spectrometric measurements follows
very closely the flash pattern in Fig. 2(c), which was
obtained by the conventional amperometric method using
a so-called Joliot-type oxygen electrode. Both types of
measurements give rise to similar Kok et al. parameters,
which indicates that the two techniques are measuring
the same phenomenon.
To get information on the binding affinities of the
substrate water in the OEC, the rate of isotope exchange
between 18 O-labeled water and the photogenerated O2
needs to be determined. To achieve this, our group
designed a ‘‘closed’’ chamber system in which the sample
volume is considerably reduced (160 µl) and a stirrer
is incorporated to facilitate rapid mixing. By employing
a solenoid-actuated Hamilton syringe, 18 O-labeled water
can be rapidly injected and equilibrated with the sample
within 6–8 ms. Thus, by measuring the extent of isotope
incorporation as a function of the time after injection of
the 18 O-labeled water, the rate of isotope exchange can be
determined.
To evaluate the S-state dependence on the rate of
isotope exchange, the ‘‘closed’’ chamber is optically coupled
to a bank of three xenon flash lamps (e.g., FX-193 lamp,
4 µF @ 1 kV capacitor from EG & G, Salem, MA) via
a 3-to-1 fiber optic to variously illuminate the sample
with brief (FWHH ∼8 µs), saturating light flashes. The
samples are first given a series of activating flashes to
preset the sample into a particular S state (see below).
Then, after the rapid addition and equilibration of the
18
O-labeled water, a subsequent turnover flash sequence
(spaced 5–10 ms apart, depending on S state, see below)
ISOTOPE EXCHANGE IN GAS-WATER REACTIONS 537
m/e = 32 Flash pattern
4
2
1
0
2 4 6 8 10 12 14
Flash number
Joliot electrode flash pattern
3
0
2 4 6 8 10 12 14
Flash number Figure 2.
is applied to photogenerate O2 . As the actual instrumental
response time is relatively slow (being about 10 s because
of the diffusion of the O2 gas across a semipermeable
membrane into the inlet line from the sample chamber),
the flash spacings of a subsequent flash sequence that
normalizes the oxygen signals are increased, typically to
20 s. As such, to retard the deactivation reactions of the
higher S states during these flash spacings (see above), the
temperature of the sample is reduced (usually to 10 ◦ C),
which is accomplished by enclosing the sample chamber in
a cooling block. Details of the setup have been published
earlier (63,64) and are given in Fig. 3.
S-STATE DEPENDENCE ON THE RATE OF ISOTOPE
EXCHANGE BETWEEN 18 O-LABELED WATER AND THE
PHOTOGENERATED O2
To determine S-state dependence on the rate of isotope
exchange, the mass measurements are made as a function
of time between the addition and equilibration of 18 O-
labeled water with the sample in a particular S-state
and the first peak yield of the photogenerated O2 . Thus,
for example, for the S3 state, an S1 -enriched sample
suspended in H2 16 O is given two preflashes to advance
the OEC into the S3 state. H2 18 O is then rapidly injected
and followed by a variable delay time (t), which allows
for the time-dependent isotopic exchange to occur between
the solvent water and the substrate water. A single
flash is then given, and the oxygen signals at m/e = 34
and m/e = 36 are measured. The oxygen signals are
then collected on subsequent flashes (in which complete
538 ISOTOPE EXCHANGE IN GAS-WATER REACTIONS

PC control
trigger pulses

H2 18O Injection

Mass spectrometer

Xe flash Magnetic
lamp stirrer

Sample chamber
(window, stirrer,
membrane, support)

Water cooling Water cooling

Figure 3. Temperature probe

exchange has occurred because of the 20 s flash spacings) The basic protocol involves the following procedure.
to normalize the data among the different sample aliquots Before loading, the samples are thawed in the dark at room
measured at different t values. temperature and given one saturating preflash followed by
The complete 18 O exchange kinetics were determined a 10-min dark period at room temperature to enrich the
for the different S states using the injection/flash protocols S1 population. The S1 -enriched sample is then loaded
depicted in Fig. 4. into the sample chamber in the dark, equilibrated to
10 ◦ C for 10–12 min, and then given several activating
flashes (spaced 1 s apart). At a given S state, H2 18 O is
rapidly injected into the chamber (t1/2 (mixing) ∼ 4 ms) and
a variable delay time (4 ms < t <200 s) is provided to
18
O2 Products allow the water bound in the OEC to exchange with the
H218O Injection 18
O-enriched solvent water. After the delay time, several
Exchange time rapid turnover flashes (spaced at 5, 10, and 10 ms apart
∆t for S2 , S1 , and S0 states, respectively) are then applied to
S3 S4 S0
advance the sample through to the S3 to S4 to S0 transition,
where the mass ratio of the photogenerated O2 (YM )
∆t
is determined. After this process, several normalization
S2 S4 S0 flashes (spaced 20 s apart) are given and their separate
S1 O2 yields (Yn(1→4) ) are measured. The YM is corrected for
background O2 contributions, double hits, and the injection
∆t response as described earlier (17,18) to give YC .
S1 S4 S0
The resolvable 18 O exchange kinetics for spinach
thylakoid samples as a function of the S-states are shown
in Fig. 5. Here, the corrected, normalized yields of O2
∆t
S0 S4 S0 (YC ) measured at m/e = 34 for the mixed labeled 16 O18 O
product (left side of the figure) and at m/e = 36 for the
Activating preflash(es) Turnover flash(es) Normalisation flashes
double labeled 18 O18 O product (right side of the figure) are
presented. Each point is a plot of Yc at a particular delay
Figure 4. time (t) between the injection of H2 18 O into the sample
 ISOTOPE EXCHANGE IN GAS-WATER REACTIONS 539

and the turnover flash sequences that photogenerated O2 .
The m/e = 36 data exhibit only a single kinetic phase and
are fit with a simple exponential function:
36
Y = (1 − exp(−36 k t))
In contrast, the m/e = 34 data on the left side in
Fig. 5 reveals two distinct kinetic phases. The insets
show expanded time ordinates for the fast phases. The
two phases, however, are unequal in amplitude, with the
fast phase constituting slightly more than half of the
total signal. The basis for this difference in amplitude
is well explained by the enrichment condition for two
independent, exchanging sites. As the apparent kinetics
of the two phases differ by at least a factor of 10, the
fast phase of exchange is virtually complete before the
slow phase begins. Thus, at short t only one substrate
water molecule is exchanging at the catalytic site, which
means at a typical 18 O enrichment of ε = 12%, the mass
distribution at 32:34:36 for the two oxygen isotopes will
be 88:12:0. On the other hand, at longer t, when the
second substrate water molecule is also exchanging, the
mass distribution will be 77.44:21.12:1.44. Therefore, the
relative contributions of the fast and slow phases will
(1) be unequal, with the fast phase representing ∼57% (i.e.,
12/21.12) of the total amplitude and the slow phase ∼43%.
The 0.57:0.43 distribution between the two phases is found
consistently in the m/e = 34 data. Thus, the exchange
kinetics in this case is fit exactly by the sum of two
exponentials:
34
Y = 0.43(1 − exp(−34 k1 t)) + 0.57(1 − exp(−34 k2 t) (2)
The solid lines in Fig. 5 represent the best fits to the
data points based on Eqs 1 and 2 for the m/e = 36 and
m/e = 34 data, respectively. The exchange rates from the
corresponding kinetic fits at 10 ◦ C are given in Table 1.
The entire S-state results for the thylakoids have been
presented earlier (17,18).

m/e = 34 m/e = 36

1.0 1.0

0.50 S3
0.5 0.5
Normalised O2 yield

0.25

0 .05 .1

1.0 1.0

1.0
S2
0.5 0.5
0.5

0 .05 .1

5 0 5 1.0 5.0 10.0 0.2 0.5 5 1.0 5.0 10.0

0.2 0.5 0.7 0.7

1.0 1.0
1.0

0.5 S1
0.5 0.5
Normalised O2 yield

0 .05 .1

25 50 75 10
0 0 0
15 30 25 50 75 0 0 0
10 15 30

1.0 1.0

1.0
S0
0.5 0.5 0.5

0 .05 .1 .15 .2 .25

5 0 0 0 5 0 0 0
0.2 0.5 15 30 0.2 0.5 15 30
Time (s) Time (s) Figure 5.
540 ISOTOPE EXCHANGE IN GAS-WATER REACTIONS
Table 1. Rate Constants for the 18 O Exchange in Spinach
Thylakoid Samples as a Function of the S States at 10 ◦ C
S-State 36
k (s−1 ) 34
k1 (s−1 ) 34
k2 (s−1 )
S3 2.1 ± 0.2 1.9 ± 0.2 36.8 ± 1.9
S2 2.2 ± 0.1 1.9 ± 0.3 120 ± 6
S1 0.022 ± 0.002 0.021 ± 0.002 >175
S0 18 ± 3 8±2 >175
Adapted from Ref. 18.
IMPLICATIONS OF SUBSTRATE WATER BINDING FOR THE
MECHANISM OF O2 EVOLUTION
Several striking features are found in the rate constants
listed in Table 1. In each S state, the slow phase of
exchange was resolvable, whereas the fast phase of
exchange could be resolved only in the S2 and S3 states. For
the S0 and S1 states, the fast phase of exchange is equal to
or faster than the response time of the currently available
experimental setup (∼100 − 175 s−1 ). Most importantly,
the rate constants for the slow phase kinetics in the
m/e = 34 data are virtually identical to the rate constants
for the monophasic kinetics in the m/e = 36 data. This last
result clearly indicates that the extent of 18 O incorporation
into the photogenerated O2 is limited throughout the S-
state cycle by the substrate water molecule undergoing
the slow isotope exchange process and that the biphasic
behavior in the m/e = 34 data does not originate from PSII
heterogeneity. Rather the two kinetic components in the
m/e = 34 data reflect the binding of the substrate water at
two separate sites.
The most relevant question to ask is at which S state
does the substrate water bind. Based on the kinetic
resolution of our current mass spectrometric setup, a
measurable 18 O exchange kinetic will conclusively show
the existence of bound water in a particular S state, which
is justified by the observations that the O2 release on the
S3 to S4 to S0 transition takes place in ∼2 ms, whereas
the rest of the S-state advancement from S0 to S3 takes at
most another 1 ms (8). The catalytic site thus can make one
complete turnover in ∼3 ms. Therefore, any 18 O exchange
that is slower than 3 ms would be indicative of bound
water. From the rate constants for the slow phase of 18 O
exchange listed in Table 1 for the m/e = 34 data, it is
clear that one substrate water molecule is bound to the
catalytic site through the S-state cycle. Interestingly, the
results for the fast phase of exchange show that the second
substrate water molecule is bound in at least the S2 and S3
states, but they are inconclusive for the S0 and S1 states
because the fast exchange kinetics in these S states are not
resolvable by our current setup (i.e., faster than 6–8 ms).
Even though the measurable exchange rates listed in
Table 1 are slower than the total turnover time of the
catalytic site, the magnitudes of these exchange rates
in general are all remarkably fast, where the slowest
exchange is measured in the S1 state with t1/2 ∼50 s.
Exchange rates in this magnitude range indicate relatively
labile binding sites and can be interpreted for two overall
mechanisms: (1) exchange of water bound to a metal site
or (2) diffusional or isotope equilibration processes, where
the solvent water must be translocated to the binding site
across a physical barrier.
Rates of whole water exchange at a metal site can span
a tremendous range (as much as 18 orders of magnitude),
and examples for various transition metals have been
summarized earlier (19–22). The most significant factors
governing water exchange at a metal site are the charge
and ionic radius of the metal center and any electronic
occupancy of d orbitals. As a general rule, water exchange
rates will decrease (by several orders of magnitude) as
the metal center (Mn ) is oxidized and charge is increased;
i.e., kex (Mn -OH2 ) > kex (Mn+1 -OH2 ). Furthermore, metal-
bound water can undergo deprotonation in response to an
increase in charge (and acidity) at the metal center. Thus,
it may also be expected for water exchange to decrease
at deprotonation for a given oxidation state; i.e., kex (Mn -
OH2 ) > kex (Mn -OH). In the case of the deprotonation
of a neighboring ligand (L), the opposite effect may occur
because the effective charge of the metal center is reduced,
As such the rate of water exchange will tend to increase;
i.e., kex (HL-Mn -OH2 ) < (L-Mn -OH2 ). The protonation state
will thus be defined by the pK of the aquo metal complex.
Any number of possible intermediates in the water
oxidation reaction can exist, and it is important to
appreciate that the 18 O exchange measurements could
reflect not only whole water and hydroxyl exchange but
also the exchange of µ-oxo (–O–), peroxo (–O–O–), and
oxo (=O) intermediates. Indeed various models in the
literature on O2 evolution invoke one or more of these
intermediates (9,10). Although a dearth still exists in
studies of 18 O exchange in model compounds, it is expected
that the isotope exchange in µ-oxo (–O–), perxo (–O–O–),
and oxo (=O) intermediates is much slower (in the order
of hours to days) than whole water exchange because of
tighter bonding and lack of protonation.
The other type of mechanism that could limit the 18 O
exchange could be diffusion or isotope equilibration across
a physical barrier. The catalytic site for O2 evolution in
PSII is located within the protein domain away from
the solvent interface. In this situation, the structural
barriers provided by the protein matrix may be mediated
through specific water chains or water channels (23),
something akin to what has been proposed for the
cytochrome c oxidase (24,25). Such a water channel may
provide an important function in optimizing O2 formation
by controlling the accessibility of the water substrates
and O–O bond formation at the catalytic site, thereby
minimizing side reactions that lead to the formation of
reactive oxygen intermediates (23). The inability to resolve
the fast phase of exchange in the S0 and S1 , states (Table 1)
could indicate that the second substrate water molecule
only enters the reaction sequence at the S2 state. However,
such a possibility can only be clarified by resolution of the
fast phase kinetics or by the progressive improvement in
the resolution of crystal structure of PSII (26–28).
The S-state dependence of the 18 O exchange kinetics for
the substrate water are clearly complex, but these results
place limits on the types of mechanisms that may be
involved in the oxidation of water during photosynthesis.
Although the S-state-dependent changes in the kinetics

may in the first instance be interpreted for metal-
centered oxidations and/or substrate or other ligand
protonation/deprotonation events, second-order effects
mediated by the protein environment, such as H-bonding
networks and structural conformations, will also have
to be considered. To sort out the fine details, additional
measurements will have to be made under different
biochemical conditions and in mutants of the PSII. This
work is currently in progress.
Acknowledgments
The authors gratefully acknowledge the various contributions
to this research by Johannes Messinger, Murray Badger, and
Garth Hendry. This research was supported in part by a Human
Frontier Science Program grant (RGP 29/2002).
BIBLIOGRAPHY
1. Blankenship, R.E. and Hartman, H. (1998). T.I.B.S. 23:
94–97.
2. Wydrzynski, T. and Satoh, K. (Eds.). (In press). Photosystem
II: The Water/Plastoquinone Oxido-Reductase in Photosyn-
thesis. Springer-Kluwer Academic Publishers, Dordrecht, The
Netherlands.
3. Diner, B.A. and Rappaport, F. (2002). Annu. Rev. Plant. Biol.
53: 551–580.
4. Barter, L.M.C., Durrant, J.R., and Klug, D.R. (2003). Proc.
Natl. Acad. Sci. USA. 100: 946–951.
5. Joliot, P., Barbieri, G., and Chabaud, R. (1969). Photochem.
Photobiol. 10: 309–329.
6. Kok, B., Forbush, B., and McGloin, M. (1970). Photochem.
Photobiol. 11: 457–475.
7. Forbush, B., Kok, B., and McGloin, M. (1971). Photochem.
Photobiol. 14: 307–321.
8. Joliot, P. and Kok, B. (1978). Bioenergtics of Photosynthesis.
Govindjee (Ed.). Academic Press, New York.
9. Nugent, J.H.A. (Ed.). (2001). Special Issue: Photosynthetic
Water Oxidation. Biochim. Biophys. Acta 1503.
10. Messinger, J. and Lubitz, W. (Eds.). Special Issue: Photosyn-
thetic water oxidation. Phys. Chem. Chem. Phys. (In press).
11. Radmer, R. and Ollinger, O. (1980). F.E.B.S. Lett. 110: 57–61.
12. Radmer, R. and Ollinger, O. (1985). F.E.B.S. Lett. 195:
285–289.
13. Bader, K.P., Thibault, P., and Schmid, G.H. (1987). Biochim.
Biophys. Acta 893: 564–571.
14. Messinger, J., Badger, M., and Wydrzynski, T. (1995). Proc.
Natl. Acad. Sci. USA 92: 3209–3213.
15. Hillier, W., Messinger, J., and Wydrzynski, T. (1998). Bio-
chemistry 37: 16908–16914.
16. Guy, R.D., Fogel, M.L., and Berry, J.A. (1993). Plant. Phys.
101: 37–47.
17. Hillier, W. and Wydrzynski, T. (2000). Biochemistry 39:
4399–4405.
18. Hillier, W. and Wydrzynski, T. (In press). Phys.Chem. Chem.
Phys.
19. Hillier, W. and Wydrzynski, T. (2001). Biochim. Biophys. Acta
1503: 197–209.
20. Hendry, G. and Wydrzynski, T. (2003). Biochemistry 42:
6209–6217.
21. Lincoln, S.F. and Merbach, A.E. (1995). Adv. Inorg. Chem.
42: 1–88.
RADON IN WATER 541
22. Richens, D.T. (1997). The Chemistry of Aqua Ions. John Wiley,
West Sussex, UK.
23. Wydrzynski, T., Hillier, W., and Messinger, J. (1996). Physiol.
Plant. 96: 342–350.
24. Iwata, S., Ostermeier, C., Ludwig, B., and Michel, H. (1995).
Nature 376: 660–669.
25. Tsukihara, T. et al. (1996). Science 272: 1136–1144.
26. Zouni, A. et al. (2001). Nature 409: 739–743.
27. Kamiya, N. and Shen, J-R. (2003). Proc. Natl. Acad. Sci. USA.
100: 98–103.
28. Ferreira, K.N. et al. (2004). Science 303: 1831–1838.
RADON IN WATER
BABS A. MAKINDE-ODUSOLA
Riverside, California
Some groundwater contains the naturally occurring
radioactive gas, radon-222 (radon). Radon dissolved in
groundwater is released into indoor air when plumbed in
water is agitated indoors. Radon released from domestic
water typically contributes less than 2% of total indoor
radon. This article describes the health risks from radon
in domestic water and the efforts within the United States
to regulate radon in drinking water. The U.S. National
Research Council (NRC) found that the primary health
risks from radon in domestic water are lung cancer
and stomach cancer, primarily to ever-smokers. Radon
concentration in air could reach levels considered unsafe
for underground miners at some water treatment facilities
that aerate water as part of the treatment process.
Inhalation of radon gas and its progeny may result in lung
cancer (inhalation risk). An estimated 11% of the health
risk is from direct ingestion of radon dissolved in drinking
water (ingestion risk). The NRC found that radon is a
greater source of cancer risk than any other regulated
chemical in drinking water. Some ecological studies
contradict using the linear nonthreshold (LNT) model to
assess the health risks from exposure to radon at low levels
typical of indoor air. The U.S. Environmental Protection
Agency (EPA) relied on the LNT model to assess health
risks and propose standards and goal for radon in drinking
water. Some water utilities could opt for treatment to
comply with the radon in drinking water standard. Some
other water utilities might avoid treatment by adopting a
multimedia approach to radon mitigation. The multimedia
approach could be more cost-effective and achieve greater
health risk reduction. Noneconomic considerations may
influence that decision. Health risk reduction and cost
analyses (HRRCA) suggest that most of the benefits would
accrue to ever-smokers and smaller water systems could
incur most of the costs.
INTRODUCTION
Radon, in this article, refers to the specific isotope,
radon-222, the daughter product of radium-226 and a
member of the uranium-238 decay series. The decay
scheme for uranium-238 is illustrated in Drago (1) and
542 RADON IN WATER
NRC (2). Uranium is found in trace amount in soils
everywhere. Radon escapes as a gas from soils into
groundwater, indoor air, and outdoor (ambient) air. The
U.S. Geological Survey (3) reported that soil air within
the United States contains between 200 and 2,000 pCi/L
(1 pCi/L = 37 Bq m−3 ) of radon. Radon is a chemically
inert, colorless, and odorless gas whose half-life is about
3.82 days. Radon emits alpha (α) particles as it decays
to a series of solid short-lived radionucleotides, such as
polonium-218 and lead-214. Those physical properties
are important in understanding the health risks from
exposure to radon and the mitigation of such risks.
The United Nations Scientific Committee on the Effects
of Atomic Radiation (4) estimated that about 50% of the
average annual effective dose from natural sources of
ionizing radiation worldwide is from exposure to the alpha-
particles produced by radon and its progeny. Potential
contributors of radon to indoor air include soil gas, natural
gas, building materials, and to a lesser extent—domestic
water. Radon emanating from soils under homes is the
primary source of radon in indoor air, which can range
from <1 to 3,000 pCi/L (3).
Radon in drinking water is typically of concern to
sources that rely on groundwater, where there had
not been sufficient opportunity for significant decay
nor release into the atmosphere before reaching indoor
plumbing. A portion of the radon in plumbed water
used indoors is released into the air, especially during
agitation (e.g., laundering, dishwashing, humidifying,
toilet flushing, etc.), which increments the level of radon
in indoor air. The NRC (2) estimated that the average
transfer coefficient is about 0.01% (1 in 10,000) of radon
in water; that is 10,000 pCi/L of radon in water would
increase the level of radon in indoor air by 1 pCi/L. The
U.S. EPA (5) estimated that less than 2% of radon in
indoor air emanates from domestic water. This author
prefers to emphasize ‘‘domestic water’’ because of the
widespread use of bottled water for drinking and because
the NRC (2) found that the major part of the health risk is
from inhalation of radon and its progeny.
Some countries have policies to reduce public exposure
to radon, especially in indoor environments. Cole (6)
provides the historical development of radon risk policies
in the United States. The U.S. Indoor Radon Abatement
Act of 1988 established the goal of reducing indoor radon
concentrations to natural ambient level. The NRC (2)
estimated that the arithmetic national ambient average
level of radon was 0.4 pCi/L. The U.S. EPA established
a voluntary action level of 4 pCi/L for radon in indoor
air. The indoor radon action level in other countries
is as high as 27 pCi/L (7). Indoor radon guidelines are
22 pCi/L and 11 pCi/L in Canada and the European Union,
respectively (7). The U.S. EPA (5) reported that more than
2500 lives were ‘‘saved’’ through the year 2000 since the
inception of its voluntary indoor radon abatement program
in 1986.
Many countries are developing policies to reduce expo-
sure to radon in domestic water. The World Health Organi-
zation (WHO) is creating a database of national radon reg-
ulations [http://www.who.int/ionizing radiation/env/
radon/en/]. The rest of this article focuses on efforts within
the United States. The NRC (2) found that radon is a
greater source of cancer risk than any other regulated
chemical in drinking water. Evaluating, and where neces-
sary, mitigating the level of indoor radon is common during
real estate transactions in the northeastern United States,
even though compliance with the action level is voluntary.
That practice could extend to testing for radon in pri-
vately owned wells after the EPA finalizes the maximum
contaminant level (MCL) for radon in drinking water.
Regulating radon in domestic water had been very contro-
versial in the United States and dates back to 1986. In
1986, Congress directed the EPA to establish a manda-
tory standard for radon in drinking water as part of the
amendments to the Safe Drinking Water Act (SDWA).
The EPA (8) proposed the radon in drinking water rule
using the risk management approach typically used for
drinking water contaminants. Amendments to the SDWA
in 1996 directed the EPA to withdraw the proposed rule
of 1991 and repropose after an independent risk assess-
ment by the NRC. The amendments include provisions for
a multimedia approach to achieving an equal or greater
level of health risk reduction. In 1999, the EPA (5) pro-
posed 300 pCi/L and 4,000 pCi/L as the MCL and alternate
MCL, respectively, for radon in drinking water. The EPA is
expected to finalize the radon in drinking water rule after
2004. Key inputs and major milestones in developing the
radon rule are illustrated in Fig. 1 of Article MW-66 (9).
OCCURRENCE OF RADON IN DOMESTIC WATER
Radon levels in water, like those in ambient air and
indoor air, vary temporally and spatially. The EPA (5)
analyzed data from the Safe Drinking Water Information
System (SWDIS) to estimate the nationwide occurrence of
radon in community water systems (CWS). The EPA (10)
found geographic disparity in its occurrence. New England
CWS had the highest arithmetic and geometric mean
concentrations of 2933 pCi/L and 1214 pCi/L, respectively.
Some privately owned wells, not monitored nor regulated
by the EPA, might contain radon at much higher levels.
The USGS (3) reported that the radon level in groundwater
ranges from about 100 pCi/L to nearly 3,000,000 pCi/L.
The EPA (5) estimated the population-weighted national
average concentration of radon in CWS at 213 pCi/L, and
a credible range of 213 to 240 pCi/L. In general, smaller
CWS that rely on groundwater have the highest levels
of radon. Radon levels exceeded 300 pCi/L at more than
51% and 16% of the CWS serving fewer than 100 people
and 10,000 people, respectively (5). Radon levels exceeded
300 pCi/L and 4000 pCi/L in the domestic water of more
than 16 million people and 77,000 people, respectively (5).
The EPA (5) and NRC (2) have additional information
on the occurrence and spatial distribution of radon in
domestic water, indoor air, and ambient air. SWDIS would
contain additional radon occurrence data when the radon
rule is finalized and CWS begin compliance monitoring.
Radon levels at a particular well could fluctuate
significantly. The EPA (5) proposed a monitoring regime
that would require CWS using groundwater to monitor
for radon initially at each entry point to the distribution
system, quarterly for 1 year. Compliance would be based

Figure 1. Estimated fatalities in the U.S. by pathway from radon in drinking water provided by CWS.
on the average of four consecutive quarterly samples.
Zikovsky and Chah (11) found that concentrations of radon
are lognormal in distribution and that the arithmetic
mean of concentrations might create a false impression
about exposure levels. The EPA (5) estimated that
26.8% of sources whose radon concentrations averaged
between 200 and 300 pCi/L during the first four quarterly
measurements would have long-term exceedances of
300 pCi/L. Field et al. (12) found higher levels of radon in
domestic water at the point of use than prior to distribution
because of increments from decay of the radium-226
deposits within the older portions of water distribution
systems. The EPA (5) identified man-made sources of
radon in groundwater to include releases from wastes
from radium or thorium processing and from phosphate
and uranium mining and milling operations.
RISK ASSESSMENT OF RADON IN DOMESTIC WATER
The health risks from radon in domestic water are pri-
marily lung cancer and stomach cancer. Radon is a
human carcinogen based on epidemiological studies of
underground miners, cellular mutagenesis studies, and
experimental research in animals (13). A linear nonthresh-
old (LNT) model is often used to assess the health risks
RADON IN WATER 543
of exposure to radon. The use of the LNT model to assess
the health effects from exposure to radon at typical indoor
concentrations is controversial. The National Council on
Radiation Protection and Measurements (14) evaluated
the evidence for and against the LNT dose–response
hypothesis and found that the LNT model is more plausi-
ble. The NCRP (14) did not exclude other dose–response
relationships because of growing evidence that adaptive
responses, bystander effects, and other variables may
modify the dose–response relationship. The NRC Com-
mittee on the Biological Effects of Ionizing Radiation
(BEIR) VI did not exclude the possibility of a thresh-
old relationship at the very low levels of radon exposure
typically found in indoor air. The NRC (2) supported
using the LNT relationship to derive quantitative esti-
mates of cancer risks in the absence of credible evidence
to the contrary. The NRC (2) recommended reconsidera-
tion of the LNT assumption when the ‘‘probability of an
exposed cell becoming transformed’’ and the ‘‘latent period
before malignant transformation is complete’’ are known.
Lorimore et al. (15) observed the potential for interaction
between irradiated and nonirradiated cells in the produc-
tion of genetic damage and its significant implications for
risk estimation. Brenner and Sachs (16) suggested that
bystander effects from alpha-particle damage could cause
544 RADON IN WATER
downwardly curving dose–response relations and inverse
dose-rate effects.
Parsons (17) presented a biological model of radiation
hormesis that invalidated the LNT. Cohen (18) tested
the LNT model by analyzing lung cancer mortality rates
and average radon concentration in homes in 1601 U.S.
counties while addressing problems typical of ecological
studies. Cohen (18) found a ‘‘strong tendency for lung can-
cer rates to decrease with increasing radon exposure in
sharp contrast to the increase’’ expected from the LNT
model. The NRC (2) recommended investigating the sig-
nificant negative trend of county lung cancer mortality
with increasing home radon concentration established in
the ecological study by Cohen (18). Puskin (19) found sim-
ilar ‘‘strongly negative correlations for cancers strongly
linked to cigarette smoking’’ but did not find ‘‘such cor-
relations’’ for cancers not linked to smoking. Puskin (19)
explained that the ecological correlation (18) ‘‘provides no
substantial evidence for a protective effect of low level
radon exposure.’’ Van Pelt (2003) found that altitude (ele-
vation) explains some of the inverse association between
radon and lung cancer. Van Pelt (2003) postulated that
the decrease in lung cancer is ‘‘caused by the carcino-
genic effect of higher absolute oxygen concentration in the
inspired air at lower.’’
The most direct way to assess the effects of radon
in homes is through case-control studies (20). Krewski
et al. (13) summarizes case-control studies of residential
radon and lung cancer in North America. Field and
Duport (21) presented points and counterpoints regarding
lung cancer risks from exposure to residential radon.
The NRC (2) reviewed available epidemiological studies
of indoor radon and concluded ‘‘it is low and difficult,
if not impossible to detect given the high background
lung-cancer mortality in the populations studied.’’
Inhalation of radon gas, especially its progeny, causes
lung cancer primarily among ever-smokers. BEIR VI (20)
evaluated the risks of domestic radon to ever-smokers and
never-smokers separately. CDC (22) defines ever-smokers
as people who had smoked at least 100 cigarettes in their
entire lifetimes. Alavanja (23) found many similarities
between the pathologic lung-cancer mechanisms observed
among those exposed to tobacco smoke and those exposed
to radon gas.
NRC (2) conducted the most comprehensive review of
the health risks of radon in domestic water. That review
included epidemiological studies, BEIR studies, and
earlier EPA (8,24) risk assessments. Both BEIR VI (20)
and the NRC (2) reviewed the mechanistic information
on alpha-particle-induced carcinogenesis. The NRC (2)
postulated that the biological effects of exposure to radon
at the low level of typical indoor air are initiated by
the passage of single alpha-particles that have very high
linear energy transfer (LET). Hei et al. (25) described the
mutagenesis of a single alpha-particle in mammalian cells.
Genes involved at low dose are different from those
involved at higher doses and adaptive responses in cells
exposed to low-level ionizing radiation allow the cells to
better withstand a much higher dose later (priming effect)
[http://www.llnl.gov/str/JulAug03/Wyrobek.html].
The NRC (2) developed a new model of ingested radon
diffusion because its review of published ingestion risks
found a wide range in the estimates. The uncertainties
were mainly due to the assumed model of radon absorption
in the gastrointestinal tract. The NRC (2) noted the lack
of data on the extent to which radon diffuses through
the stomach wall and recommended research to reduce
the uncertainty regarding ingestion risks. Details of the
mathematical models of radon diffusion in the stomach
and the physiologically based pharmacokinetic (PBPK)
model of radon behavior in the body can be found in
the NRC (2). The NRC (2) estimated the ingestion risks
using estimates of the amount of water ingested, the
fraction of radon lost from water before ingestion, and
calculations of the dose absorbed by tissues exposed. The
NRC (2) estimated the confidence limits on the ingestion
risk but noted, ‘‘the actual risk from ingested radon could
be as low as zero depending on the validity of the LNT
dose–response hypothesis.’’ Yamaoka el al. (2004) found
inhalation of radon gas from radon enriched water induced
a small amount of active oxygen in the blood, decreased
the levels of lipid peroxide and total cholesterol in the
blood of osteoarthritis patients alleviating pain related to
peroxidation reactions and immune depression.
Table 1 is a summary of the risks estimated by the
NRC (2). Table 1 shows that males have higher lifetime
risks than females and that the inhalation risk to ever-
smokers is five times as high as those for never-smokers
of either sex. Notice that the aggregate health risk to the
general population is thrice as high as the risk for never-
smokers due to the inclusion of ever-smokers. Figure 1
shows the relative distribution of cancer fatalities by
exposure pathways. Notice in Fig.1 that 89% of the health
risks are from inhalation of the radioactive by-products
of radon released from water into indoor air. About 11%
of the health risk is from direct ingestion of tap water
(0.6 L/day directly from a tap). An earlier risk assessment
by the EPA (5) estimated that the proportion of ingestion
and inhalation risks were 53% and 47%, respectively.
Both the NRC (2) and the EPA (5) found that the
hazards of exposure to radon during showering are likely to
be small because the buildup of radon progeny is less than
4% of the maximum possible value during a typical shower
that lasts 10 minutes. The NRC (2) found it highly unlikely
that radon in drinking water has substantial teratologic or
reproductive effects. The NRC (2) found that information
was insufficient to determine radon risks to ‘‘sensitive’’
sub-populations, such as children. However, about 30% of
fatal stomach cancers are attributed to exposure before
age 10 (2). The NRC (2) suggested that there might be an
opportunity interval for lung-cancer risk reduction from
exposure during childhood. Bilban and Vaupotic (26) found
higher chromosome aberrations in pupils at elementary
schools that had high indoor radon levels. Fewer than 34%
of lung cancer patients are diagnosed before age 65 (27).
Many researchers have estimated fatalities from radon
in domestic water. Mills (28) estimated 60, 400, and 5000
annual fatalities from radon in surface water, public
groundwater, and private wells sources, respectively. The
NRC (2) used the risk estimates (Table 1) and information
regarding the occurrence of radon in CWS, daily water
 RADON IN WATER 545

Table 1. Lifetime and Unit Risks Posed by Exposure to Radon in Drinking Water at 1 pCi/L (37 Bqm−3 )a
Lifetime Riskb
Unit Risk
Exposure Pathway Male Female U.S. Population U.S. Population

Inhalation (ever-smokers) 1.15 × 10−6 7.40 × 10−7 9.62 × 10−7 1.31 × 10−8
Inhalation (never-smokers) 2.18 × 10−7 1.48 × 10−7 1.85 × 10−7 2.43 × 10−9
Inhalation (population) 7.77 × 10−7 4.44 × 10−7 5.92 × 10−7 7.90 × 10−9
Ingestion* (population) 5.55 × 10−8 8.51 × 10−8 7 × 10−8 9 × 10−10
Total risk (inhalation and ingestion) 8.14 × 10−7 5.18 × 10−7 6.66 × 10−7 8.89 × 10−9

*Breakdown of Ingestion Risks

Stomach 5.92 × 10−8
Colon 2.18 × 10−9
Liver 2.15 × 10−9
Lung 1.26 × 10−9
General tissue 2.92 × 10−9
Total ingestion 7 × 10−8
a
Data sources: Reference 2, pp. 16, 17; Reference 5, Table X11.3.
b
Annual unit risk can be obtained by dividing lifetime unit risks by the expected lifetime for each population category. Expected lifetime in years:
Ever-smokers = 73.7; never-smokers = 76.1; combined population = 74.9.

Table 2. Health Risks from Radon in Drinking Water

a. Estimated Annual U.S. Fatalities Due to Radon in ∗ Drinking Water

Health Risk Source of Estimate Central Estimate Range

Lung cancer NRC (2) 160 25–290

Lung cancer EPA (5) 148 70–263
Stomach cancer NRC (2) 23 5–50
Stomach cancer EPA (5) 18 3–38

b. Occurrence of Lung Cancer due to Radon by Exposure Pathway

Pathway Central Estimate Range

Outdoor air (2) 720 120–300

Indoor air (5) 18200 3,000–33,000
*Drinking Water (2) 160 25–290
*Drinking Water (5) 148 70–263

c. Comparative Health Risks of Radon in *Drinking Water

Annual Fatalities (Central Estimates) % of All Causes

Due to *Drinking
Health Risk All Causes *Drinking Water Water

Lung cancer (2) 160,000 160 0.10%

Stomach cancer (2) 14,000 18 0.13%

Data sources: NRC(2)page 18; EPA(5).

Fatalities based on 1998 data.
∗
Potable water delivered by a community water system (CWS).

intake, and smoking patterns among several factors to
derive estimates of cancer fatalities in the United States.
Table 2 summarizes the estimated U.S. fatalities due to
radon in domestic water relative to other sources. The NRC
estimated that in 1998 radon from soil gas caused about
18,200 lung cancer fatalities in the United States. The
NRC (2) also estimated that 160 lung cancer fatalities were
due to radon in domestic water. Notice that an estimated
720 fatalities were from an unavoidable background level
of radon in outdoor air. The NRC (2) estimated that 23
stomach cancer fatalities were from ingesting radon in
drinking water. The EPA (5) estimated that radon in
domestic water caused 148 lung cancer and 18 stomach
cancer fatalities annually. The EPA (5) also estimated
(screening level) that an additional 36 fatalities annually
were from radon in water delivered outside residences, at
schools, offices, etc.
RISK MANAGEMENT
The EPA (8) first proposed 300 pCi/L and 0 pCi/L as
the maximum contaminant level (MCL) and MCL goal
(MCLG), respectively, for radon in drinking water. Many
546 RADON IN WATER
criticized the proposed rule because it was not considered
cost-effective. In addition, a multimedia approach could
be more cost-effective and achieve greater health risk
reduction. In 1996, as part of the amendments to the
SDWA, Congress directed the EPA to withdraw the
proposed rule and re-propose using an updated risk
assessment provided by the NRC. Congress also directed
the EPA to develop an alternate MCL (AMCL) and
guidelines for multimedia mitigation (MMM) programs
(MMMP). The NRC (2) defined the AMCL as the
concentration of radon in water that would cause an
increase of radon in indoor air that is no greater than
the level of radon naturally present in outdoor air. The
NRC (2) determined that the AMCL should be about
4000 pCi/L based on the national average outdoor air
radon concentration of 0.4 pCi/L and the water to air
transfer factor of 10,000. EPA is only to establish an AMCL
if the MCL is more stringent than the water equivalent of
the national average of radon in outdoor air.
The EPA (5) used NRC risk estimates (Table 1) to
repropose the radon rule (MCL = 300 pCi/L; MCLG =
0 pCi/L, AMCL = 4000 pCi/L). CWS whose radon levels
are within the range of the MCL and the AMCL have
the option of complying with either the MCL or the less
stringent AMCL. CWS that choose to comply with the
AMCL must participate in an approved and verifiable
MMMP that achieves at least an equivalent level of
risk reduction as treatment to the MCL. CWS may
opt to develop individual MMMP in accordance with
EPA guidelines if the primacy state chooses not to
develop an MMMP. The EPA (5) proposed four criteria
required for approval of MMMP: involve the public in
developing MMMP; include quantitative goals to measure
the effectiveness of MMMP; include implementation
plans, strategies and specific activities for achieving
quantitative goals; and include plans for measuring and
reporting results. Examples of MMMP include mitigation
of radon levels in existing construction and use of radon-
resistant techniques for new construction. CWS that
choose the AMCL approach are not required to treat,
but to participate in approved MMMP. The implemented
MMMP might not reduce the ingestion risk, but reduce
the inhalation risks, though not necessarily in the
same community, nor in the same homes. Some CWS
systems may opt to purchase water from other sources
or consolidate with other CWS in lieu of treating or
implementing the MMMP.
HEALTH RISK REDUCTION AND COST–BENEFIT
ANALYSES (HRRCA)
Radon is the first drinking water contaminant for which
the EPA (27) used the new cost–benefit analyses frame-
work. The HRCCA (27) included detailed cost–benefit
analyses by water system size for each considered MCL
and MMMP adoption. In estimating benefits, the EPA (27)
used the value of $5.8 million (1997 dollars) as the value
of statistical life (VSL) for each avoided cancer fatality
and willingness to pay (WTP) of $536,000 for nonfatal
cancers. The EPA (5) estimated that the cost for MMMP
is $0.7 million per avoided lung cancer (the estimated
average costs of indoor radon screening for all homes and
mitigation cost for each home at or above the 4-pCi/L
indoor air action level). The EPA (5) estimated annual-
ized national costs at $408 million and monetized annual
benefits at $362 million if 100% of impacted CWS choose
to comply with the proposed MCL and assuming costs of
capital of 7%. The estimated national annualized costs
and monetized benefits are $60 million and $301 million,
respectively, under a scenario that assumes that 95% of
states implement MMMP and 90% of the CWS in each
state comply with the AMCL. The EPA (5) prefers that
states adopt the AMCL approach and develop MMMP but
the NRC (2) noted that noneconomic considerations, such
as equity, could be deciding factors in whether to under-
take MMMP. The EPA (5) estimated that annual MCL
compliance costs per household are about $270 and $10
for CWS serving 25 to 100 people and more than 100,000
respectively. The HRRCA shows that small CWS would
incur most of the treatment costs but that the major-
ity benefits would go to larger systems under the MCL
approach. Some have argued that the EPA underestimated
the costs of compliance and overstated the benefits. Costs
could be higher because of unique local conditions and
environmental permitting. Hahn and Burnett (29) found
that the proposed radon rule was unjustifiable based on
cost–benefit analyses.
TREATMENT TECHNOLOGIES
CWS may choose to comply with either the MCL or,
under certain circumstances, with the AMCL. Treatment
of radon in domestic water will reduce the health risks
from both direct ingestion and inhalation through a
decrement. Deb (30) reported on the decrements in indoor
radon level after water treatment in three New Hampshire
communities. Deb (30) found that significant reduction in
the indoor radon level, occurred only in homes served by
the CWS where water (radon concentration >20,000 pCi/L
in water) was the major source of radon in indoor air.
The EPA (5) identified aeration as the best available
technology (BAT) for removing radon from domestic water
in large CWS. Article DW-32 (31) contains additional
information regarding radon removal from drinking water.
Aeration treatment methods such as a packed tower,
diffused bubble, spray, slat tray, and cascade aeration
are very efficient in removing radon from water because
of the volatility of radon gas. Liquid-phase granular
activated carbon (GAC) removes radon by adsorption
but requires longer empty bed contact time (EBCT)
ranging from 20 to 130 minutes, and its performance
depends on water quality (1). The EPA (5) identified
point-of-entry (POE) GAC as one of the affordable
small systems compliance technologies (SSCTs). Drago (1)
reported removal efficiencies of 20% to greater than 99% at
34 water treatment facilities and found that construction
costs for either BAT are site specific. Either treatment
technology requires disinfection after treatment, which
EPA believes, will reduce the microbial risk and provide
the additional benefit of complying with the proposed
groundwater rule (32). The EPA (5,27) found that high
levels of dissolved iron and manganese co-occurring

with radon in water may complicate and increase radon
treatment costs.
The EPA (5) reviewed the permitting and other
environmental issues of either technology. Beta-gamma
radiation from radon progeny might pose concern to GAC
operators (1). A major concern of aeration technologies
is off-gas emissions. Fisher et al. (33) found that radon
concentration in air could reach levels considered unsafe
for underground miners at some water treatment facilities
that aerate water as part of the treatment process.
The NRC (2) and EPA (5) recommended monitoring for
radon in air around facilities treating groundwater for
any contaminant and mitigation, if necessary, to reduce
exposure. The EPA (5) modeled risks from potential air
emissions of radon off-gas from aeration facilities based
on data from 20 CWS whose radon levels ranged from
1000 pCi/L to 16,000 pCi/L. The EPA (5) estimated a
maximum lifetime risk of 2 × 10−5 and a crude national
estimate of 0.09 cancer fatalities per year if all CWS
comply with the MCL. An air dispersion modeling study
for the East Valley Water District, San Bernardino,
California, suggests that the PTA (Packed Tower Aeration)
downwind risk estimates could exceed the local regulatory
risk threshold (34). The CDM (34) model predicted that a
maximum individual cancer risk at some locations near the
PTA could exceed 10−5 and 0.2 as the mean cancer burden
for the service area. Preliminary design parameters for
PTA to mitigate risks from radon emissions were not
feasible from design/construct perspectives, for example,
stack height up to 200 feet (34).
CONCLUSIONS
The preferred risk management model for radon in
domestic water differs from those of other regulated
drinking water contaminants because
• Amendments to the SDWA in 1996 require the EPA
Administrator to take into account the costs and
benefits of control programs for radon from other
sources in setting the MCL.
• The major source of radon in indoor air is not
regulated. Domestic water, which contributes on
average less than 2% of radon in indoor air, would be
regulated. Areas that have the highest indoor radon
level may not coincide with service areas of CWS that
have the highest level of radon in domestic water.
• The majority of health risks is not from direct
ingestion of water, but rather from radon released
during domestic use. Drinking water standards are
usually set based on risks primarily due to ingestion.
• The majority of health risks are due to synergy
with voluntary lifestyle choice—smoking. An ongoing
public health antismoking campaign may result in
significant reductions in the population of ever-
smokers, thus reducing overall population unit risks.
Though the EPA did not specifically categorize
ever-smokers as a ‘‘sensitive subpopulation,’’ their
inclusion in risk analyses and benefit analyses
significantly increased the benefits of adopting
the rule.
RADON IN WATER 547
• CWS may choose to comply with the AMCL rather
than the MCL. In some scenarios, health benefits
from risk reduction might accrue to communities
outside the service area of the CWS. Health benefits
might accrue to selected individuals rather than
uniform risk reduction to all consumers.
Ayotte et al. (35) estimated that a 0.05% reduction in the
prevalence of smoking would prevent as many deaths from
lung cancer as mitigation of indoor radon. The NRC (2)
noted ‘‘except in situations where concentrations of radon
in water are very high, the reduction of radon in water
will generally not make a substantial reduction in the
total radon-related health risks to occupants of dwellings
served by the water supply.’’ The NRC (2) expressed
similar concerns about the lack of data regarding the
‘‘overall effects of radon-resistant construction methods
on reducing concentrations of radon in indoor air radon
concentrations.’’
BIBLIOGRAPHY
1. Drago, J.A. (1998). Critical Assessment of Radon Removal
Systems for Drinking Water Supplies. American Water Works
Association Research Foundation, Denver, CO.
2. National Research Council. (1999). Risk Assessment of Radon
in Drinking Water. National Academy Press, Washington,
DC.
3. U.S. Geological Survey. (2000). The Geology of Radon. U.S.
Geological Survey, Denver, CO.
4. United Nations Scientific Committee on the Effects of Atomic
Radiation. (1993). Report to the General Assembly. United
Nations, New York. Available: http://www.unscear.org/
reports/2000 1.html
5. U.S. Environmental Protection Agency. (1999). National
Primary Drinking Water Regulations; Radon-222: Proposed
Rule. 40 CFR Parts 141 and 142. Federal Register (November
2, 1999) 64(211): 59246–59378. [64 FR 59246]. U.S. Govt.
Printing Office, Washington, DC.
6. Cole, L.A. (1993). Element of Risk: The Politics of Radon.
AAAS Press, Washington, DC.
7. World Health Organization. (2002). Radon and Health. Final
Information Sheet. October 2002. Available: http://www.
who.int/ionizing radiation/env/Radon Info sheet.pdf
8. U.S. Environmental Protection Agency. (1991). National Pri-
mary Drinking Water Regulations; Radionuclides: Proposed
Rule. 40 CFR Parts 141 and 142. Federal Register (July 18,
1991) 56(138): 33050–33127. [56 FR 33050]. Govt. Printing
Office, Washington, DC.
9. Makinde-Odusola, B. (2004). MW-66 drinking water quality
standards—United States. Encyclopedia of Water. John
Wiley & Sons, New York.
10. U.S. Environmental Protection Agency. (1999). Methods
Occurrence, and Monitoring Document for Radon, Draft. U.S.
Environmental Protection Agency, Office of Ground Water
and Drinking Water.
11. Zikovsky, L. and Chah, B. (1990). The lognormal distribution
of radon concentration in ground water. Ground Water 28(5):
673–676.
12. Field, R.W., Fisher, E.L., Valentine, R.L., and Kross, B.C.
(1995). Radium-bearing pipe scale deposits: implications for
national waterborne radon sampling methods. Am. J. Public
Health 85: 567–570.
548 SILICA IN NATURAL WATERS
13. Krewski, D. et al. (2002). A combined analysis of North-
American case control studies of residential radon and lung
cancer: An update. Radiat. Res. 158(6): 785–790.
14. National Council on Radiation Protection and Measurements.
(2001). Evaluation of the Linear -Nonthreshold Dose-
Response. Model for Ionizing Radiation NCRP Report No.
136. NCRP, Bethesda, MD. Available: http://www.ncrp.com/
rpt136.html.
15. Lorimore, S.A. et al. (1998). Chromosomal instability in the
descendants of unirradiated surviving cells after a-particle
irradiation. Proc. Natl. Acad. Sci. USA 95: 5730–5733.
16. Brenner, D.J. and Sachs, R.K. (2003). Domestic radon risks
may be dominated by bystander effects—but the risks are
unlikely to be greater than we thought. Health Phys. 85(1):
103.
17. Parsons, P.A. (2002). Radiation hormesis: Challenging LNT
theory via ecological and evolutionary considerations. Health
Phys. 82(4): 513.
18. Cohen, B.L. (1995). Test of the linear-no threshold theory of
radiation carcinogenesis for inhaled radon decay products.
Health Phys. 68(2): 157–174.
19. Puskin, J.S. (2003). Smoking as a confounder in ecologic
correlations of cancer mortality rates with average county
radon levels. Health Phys. 84(4): 526–532.
20. National Research Council. (1999). Health Effects of Exposure
to Radon (BEIR VI). National Academy Press, Washington,
DC.
21. Field, R.W. and Duport, P.J. (2003). Exposure to residen-
tial radon causes lung cancer. W.R. Hendee, Moderator,
Point/Counterpoint. Med. Phys. 30(4): 485–488.
22. Centers for Disease Control. (1995). Morbidity Mortality
Wkly. Rep., Cigarette smoking among adults—United States
1993.
23. Alavanja, M.C. (2002). Biologic damage resulting from
exposure to tobacco smoke and from radon: Implication for
preventive interventions. Oncogene 21(48): 7365–7375.
24. U.S. Environmental Protection Agency. (1994). Report to
the United States Congress on Radon in Drinking Water,
Multimedia Risk and Cost Assessment of Radon. Office of
Drinking Water, Washington, DC.
25. Hei, T.K. et al. (1997). Mutagenic effects of a single and an
exact number of α particles in mammalian cells. Proc. Natl.
Acad. Sci. USA 94: 3765–3770.
26. Bilban, M. and Vaupotic, J. (2001). Chromosome aberrations
study of pupils in high radon level elementary school. Health
Phys. 80(2): 157–163.
27. U.S. Environmental Protection Agency. (1999). Health
Risk Reduction and Cost Analysis for Radon in Drinking
Water Regulations: Notice and Request for Comments and
Announcement of Stakeholder Meeting. 40 CFR Parts 141 and
142. Federal Register (February 26, 1999) 64(38): 9559–9599.
[64 FR 9559], U.S. Govt. Printing Office, Washington, DC.
28. Mills, W.A. (1990). Risk assessment and control management
of radon in drinking water. In: Radon, Radium and Uranium
in Drinking Water. C.R. Cothern and P.A. Rebers (Eds.).
Lewis, Chelsea, MI, pp. 27–37.
29. Hahn, R.W. and Burnett, J.K. (2001). The EPA’s Radon Rule:
A Case Study in How Not to Regulate Risks. Regulatory
Analysis 01-01. AEI-Brookings Joint Center for Regulatory
Studies. Available: http://www.aei.brookings.org
30. Deb, A.K. (1992). Contribution of Waterborne Radon to Home
Air Quality. American Water Works Association Research
Foundation, Denver, CO.
31. DW-32. (2004). Methods of Reducing Radon in Drinking
Water. Encyclopedia of Water. John Wiley & Sons, New York.
32. U.S. Environmental Protection Agency. (2000). National Pri-
mary Drinking Water Regulations: Groundwater Rule; Pro-
posed Rule 40 CFR Parts 141 and 142. Federal Register (FR)
65(91): 30194–30274). Available: http://www.access.gpo.gov/
su docs/aces/aces140.html (accessed Jan 2001).
33. Fisher, E.L., Fuortes, L.J., and Field, R.W. (1996). Occupa-
tional exposure of water-plant operators to high concentra-
tions of radon-222 Gas. J. Occup. Environ. Med. 38: 759–764.
34. CDM. (1999). Radon Removal Preliminary Cost Analysis and
Risk Assessment. Prepared for the East Valley Water District,
San Bernardino, CA.
35. Ayotte, P. et al. (1998). Indoor exposure to 222 Rn: A public
health perspective. Health Phys. 75(3): 297–302.
36. Yamaoka et al. (2004). Study on biologic effects of radon and
thermal therapy on osteoarthritis. Journal of Pain 2004 5:
20–25.
37. Van Pelt, W.R. (2003). Epidemiological associations among
lung cancer, radon exposure and elevation above sea level–a
reassessment of Cohen’s county level radon study. Health
Physics 85(4): 397–403.
Note: All referenced URLs were valid as of July 2003.
SILICA IN NATURAL WATERS
ROBERT Y. NING
King Lee Technologies
San Diego, California
INTRODUCTION
Silica (1–4) and silicates (5–7) are generic names given
to the family of silicon dioxide compounds derived from
the dehydration–polymerization of silicic acid [Si(OH)4 ].
Silicic acid, variously called orthosilicic acid, monosilicic
acid, or silicic acid monomer, is a weak tetravalent acid
that has pKa values of 9.9, 11.8, and 12. Although silicic
acid is prevalent in natural waters and is readily available
from acidification of commercial crystalline sodium silicate
or metasilicate (8), it has a great propensity to polymerize
by dehydration to form Si–O–Si anhydride bonds:
n Si(OH)4 −−−→ (OH)3 Si–O–Si(OH)3 dimer
−−−→ Oligomers −−−→ Colloidal polymers
−−−→ (SiO2 )n
Thus silicon dioxide (SiO2 , silica) is the final and most
stable product in normal conditions of silico-oxygen acid
polymerization. The molecule of silica represented by
the formula SiO2 is polymeric and is more accurately
represented by the formula (SiO2 )n , where n is a
large number.
The rate of silicic acid polymerization is strongly pH-
dependent and is influenced by silicic acid concentration.
It is very fast in neutral and slightly alkaline solutions,
and extremely slow at low pH values of 2 to 3. Dilute
polysilicate solutions that are undersaturated with respect
to amorphous silica depolymerize to monosilicic acid.

Silicates imply more complex forms of silica in which
silicic acid co-polymerizes with hydroxides of other metal
elements covering almost the entire periodic table. This
gives rise to the immense variety of rocks found on earth,
which are copolymers of silicic acid in ultimate dehydrated
forms. In solution, silicic acid readily reacts with alcohols
and other hydroxyl-group containing organic compounds
to form anhydrides. This provides for a great diversity of
products and interactions in the biosphere.
CLASSIFICATION
Silica in natural waters can generally be classified in three
categories: reactive soluble, nonreactive soluble (colloidal),
and nonreactive insoluble (particulate). Reactivity refers
to staining by molybdate reagent in the colorimetric assay
of dissolved silica. The rate of this staining reaction
is quantitatively reproducible only with low molecular
weight polysilicic acids and varies inversely with the
degree of polymerization of silicic acid (9). Monosilicic
acid reacts with molybdic acid under assay conditions
within 2 minutes, and disilicic acid reacts completely in
about 10 minutes. Higher polysilicic acids require much
longer times for substantial but incomplete reactions.
After removal of suspended insoluble silica by filtration
using typically a 0.45 micron filter, total dissolved silica
(reactive + colloidal) can be measured by atomic emission
or absorption spectroscopy. The colloidal silica value is the
difference between the total and reactive silica. Normal
detection limits of reactive silica in 25 mL samples are
0.1 mg/L. A greater sensitivity of 0.001 mg/L is possible in
100 mL sample sizes (9).
Agglomerated silica and silicates can reach a size,
in solution, where they are no longer ions, but are
actually colloidal particles that can remain in suspension
indefinitely and are very difficult to remove by filtration.
GEOCHEMISTRY
Silica (SiO2 ) in nature occurs in the crystalline forms of
quartz, cristobalite, and tridymite, and also in the form
of amorphous silica and glass. In number of atoms as
well as weight, silicon is exceeded only by oxygen in the
earth’s crust. More than 95% of the volume of the earth’s
crust is composed of quartz and a small number of other
rock-forming silicates.
Second only to carbon, silicon forms the largest number
of compounds with other elements. The large number
of carbon compounds is due to the fact that the bond
energies of the C–C, C–O, and C–H bonds are of about
equal magnitude, and they will therefore be formed with
about the same probability. In contrast, the bond energy
of the Si–O bond is considerably higher than that of the
Si–H bond and more than twice that of the Si–Si bond. As
a consequence, instead of the common C–C–C chains of
carbon chemistry, chains of the type Si–O–Si–O are the
skeletons of silicon chemistry (5).
Typical concentrations of dissolved silica in natural
waters are:
Lakes 1–3 ppm
Major rivers 3–15 ppm
SILICA IN NATURAL WATERS 549
Seawater 1–10 ppm
Wells 2–60 ppm
Wells in volcanic and oil fields 50–300 ppm
In river and seawater, silica is not found exclusively in
the colloidal state, but also in molecular solution, because
its saturation concentrations are not attained. Despite the
propensity for monosilicic acid to polymerize, it has been
shown that silica in seawater is nearly exclusively in the
monosilicic acid form in concentrations in the range of
<0.006 to 10.1 ppm (11, 12). The concentration is low at
the surface and increases with depth. Samples of lake and
river water contain only monosilicic acid; there was no
evidence of the presence of dimeric silicic acid (13). Silica
in ground waters also appears to be in low molecular
weight forms rather than the colloidal form (14). The
concentration of particulate silica in seawaters of several
oceans has been determined as 0.0006 to 0.43 ppm SiO2
(12). It is thought that it is mainly of biogenic origin.
Colloidal silica is depolymerized by seawater, which is
considered at least one order of magnitude undersaturated
with respect to amorphous silica.
DISSOLUTION OF SILICATE ROCKS
Pure silica such as quartz is completely dehydrated silicic
acid. Depolymerization by rehydration as represented
below represents dissolution to form soluble silicic acid:
SiO2 (s) + 2H2 O −−−→ Si(OH)4
The dissolution rate depends on temperature and pH, it is
minimum at pH = 2 to 3 and rises rising about 100-fold at
pH = 11. Heat-accelerated hydrolysis would be expected
in volcanic regions.
Silicic acid is also generated from the weathering of
silicate rocks. Silicates are basically dehydrated forms of
metal ion salts of silicic acid, which can be decomposed
by carbonic acid. The following two reactions represent
important dissolution processes of silicates (1):
2NaAlSi3 O8 + 2CO2 + 6H2 O −−−→ Al2 Si4 O10 (OH)2 + 2Na+
+ 2HCO3 − + 2Si(OH)4
KMgFe2 AlSi3 O10 (OH)2 + 1/2O2
+ 3CO2 + 11H2 O −−−→ Al(OH)3 + 2Fe(OH)3 + K+
+ Mg2+ + 3HCO3 − + 3Si(OH)4
Field measurements of the weathering rate on a
per square meter catchment area basis from different
geographic areas are remarkably similar: 10−2 to 10−1
moles m−2 yr−1 . It is estimated that 105 m2 surface area
of mineral grains active in weathering is available per
m2 of geographic area and that only a few monolayers
of rock silicates are dissolved per year (1). The silica
concentrations on porewater increase continuously from
the earth’s surface to a depth of 3 meters. The silica
concentration decreases slightly with depth either by
precipitation or—under the influence of groundwater—by
dilution. The total annual chemical load delivered from the
550 SILICA IN NATURAL WATERS
continents to the oceans is estimated at 39.3 × 1014 g of
which SiO2 amounts to 4.26 × 1014 g (15).
DEPOSITION OF SILICA AND SILICATES
Most natural waters are undersaturated in amorphous
silica and silicates, but unusual conditions found in
geothermal waters (16) and industrial concentration or
uses of waters do lead to deposition of silica and
silicates. Biochemical extraction and deposition of silica
by living organisms (see below) constitute a major
deposition pathway.
Up to a pH of approximately 9, the solubility of
amorphous silica (SiO2 ) is relatively constant. It has been
reported by various investigators as 100 to 150 ppm at
25 ◦ C; the dissolved species is Si(OH)4 . Amorphous silica
is metastable relative to quartz. The rate of crystallization
of quartz is so slow at ordinary temperatures that the
solubility of amorphous silica represents the upper limit
of dissolved aqueous silica.
The precipitation of amorphous silica and silicates
apparently proceeds in a series of steps. Polymerization by
dehydration accompanied by cross-linking reactions and
aggregation by van der Waal forces leads to negatively
charged silica colloids. Further aggregation eventually
leads to soft then hard gels.
At suitably high pHs, if Fe(OH)3 , Al(OH)3 , Ca(OH)2 ,
and Mg(OH)2 are present, these hydroxides readily
participate in the copolymerization resulting in the
incorporation of Fe, Al, Ca, and Mg into complex anhydrous
silicate structures. The strong tendency of magnesium
hydroxide to react with silica, even in the monomeric
state, is shown by the fact that the addition of 300 ppm
of Mg(OH)2 to water will reduce the soluble silica content
from 42 to 0.1 ppm. The order of precipitability starting
with the most readily precipitated metal is copper, zinc,
manganese, cadmium, lead, nickel, silver, magnesium,
and calcium. Metals that react in acidic solutions with
silicates are Fe3+ , Co2+ , Zn2+ , and Cu2+ (4). Aluminum
seems to occupy a unique position in combining with
silica. Aluminosilicate gels (zeolites) can be prepared
that have ion-exchange properties. A gel composition of
5SiO2 :Al2 O3 :Na2 O has an exchange capacity of 6% of its
weight of CaO, and almost all of the sodium ions are
exchangeable. Colloidal silica and silicates are active as
flocculants and precipitate with other natural colloidal
matter in rivers, lakes and the sea.
GEOCHEMICAL SILICA CYCLE
Geologically, the silica cycle is intimately associated
with the carbonate cycle (see Carbonate in Natural
Waters). The atmospheric CO2 concentration of about
0.04% on a timescale of millions of years is controlled
mainly by a combination of degassing from the earth’s
interior balanced by the uptake of carbon dioxide
via the weathering of calcium and magnesium silicate
minerals (17,18). The carbon dioxide in water as carbonic
acid is converted to dissolved calcium and magnesium
bicarbonate on land during weathering. This carbon form
is stored in lakes and aquifers or is carried to the sea by
rivers and precipitated there as calcium and magnesium
carbonate minerals. The overall reactions shown here were
first formulated by Urey (18) in 1952.
CO2 + CaSiO3 ⇐⇒ CaCO3 + SiO2
CO2 + MgSiO3 ⇐⇒ MgCO3 + SiO2
The reactions from left to right represent carbon dioxide
absorption via weathering. The reactions from right to
left represent degassing due to thermal decarbonation
of carbonates after burial at sufficient depths in the
earth’s crust. Added to the Urey reactions should also
be the organic carbon subcycle. It exerts an important
additional control on the atmospheric carbon dioxide
level. This involves the removal of carbon dioxide from
the atmosphere through photosynthesis by organisms, the
burial of organic matter in sediments, and the weathering
of organic matter (or its thermal decomposition) which
returns the carbon dioxide. The overall reaction is
represented as:
6CO2 + 6H2 O ⇐⇒ (CH2 O)6 + 6O2
Together, these reactions regulate carbon dioxide and
silica on a geologic timescale.
BIOCHEMICAL ASPECTS
The importance of silicon for life on the earth results
primarily from the fact that the fertility of soil relies in
large part on the ability of clay minerals to absorb and
release water and several cations that are indispensable
for plant nutrition. This process is fundamental to the life
of plants and based on them, to the life of those animals
which in turn feed upon plants. Data on the physiology of
silicon is limited, except that in a number of species, its
presence is needed for normal growth.
Dissolved silica compounds play a substantial role in
the cells of living organisms. Large amounts of silica are
found in horsetail, rice, feather grass, reed and bamboo,
where it contributes to the strength of the leaves and
stems, and in the skeletons of diatoms which consist of
very pure SiO2 . Silicon compounds are vital components
in the metabolism of many bacteria, especially those
that live in hot springs. The bacterium Proteus mirabilis
even substitutes silicon for phosphorous in phospholipids.
Silicon is also present as a trace element in higher animals
and in humans, where it constitutes about 10 mg per kg
live weight. It is essential in the cells of connective
tissue and is involved in the biosynthesis of collagen, the
substance that forms hair and nails, and in the formation
of bony tissue.
The enrichment of silicon in mineralized tissues of uni-
cellular organisms, such as diatoms and radiolaria, and a
few metaozoa, such as sponges and gastropods is spectac-
ular (19). The mineral phase is always amorphous silica
in various stages of hydration as SiO2 ·nH2 O. Individual
diatom populations require a certain level of monosilicic
acid, and declining ambient silica concentrations in natu-
ral environments may influence the sequence of seasonal

successions. The present day ocean contains a few ppm
(parts per million) of dissolved silica; diatoms are often
cited as the primary agent of silica removal from the
world’s ocean and keeping a low silicic acid level. Others
favor inorganic extraction as the main regulating mech-
anism because common silicates take up dissolved silica
from silica-rich seawater. The data suggest that the world
ocean was low in silicon for most of its history.
BIBLIOGRAPHY
1. Stumm, W. and Morgan, J.J. (1996). Aquatic Chemistry. 3rd
Edn. John Wiley & Sons, New York.
2. Iler, R.K. (1979). The Chemistry of Silica. John Wiley & Sons,
New York.
3. Sosman, R.B. (1965). The Phases of Silica. University
Microfilms International, Ann Arbor, MI (originally from
Rutgers University Press, New Brunswick, NJ).
4. Iler, R.K. (1955). The Colloidal Chemistry of Silica and
Silicates. Cornell University Press, Ithaca, NY.
5. Liebau, F. (1985). Structural Chemistry of Silicates. Springer-
Verlag, New York.
6. Babushkin, V.I., Matveyev, G.M., and Mchedlov-Petrossyan,
O.P. (1985). Thermodynamics of Silicates. Springer-Verlag,
New York.
7. Eitel, W. (1954). The Physical Chemistry of the Silicates.
University of Chicago Press, Chicago, IL.
8. Alexander, G.B. (1953). J. Amer. Chem. Soc. 75: 2887.
9. Alexander, G.B. (1953). J. Amer. Chem. Soc. 75: 5655.
10. Kemmer, F.N. (Ed.). (1988). The Nalco Water Handbook. 2nd
Edn. McGraw-Hill, New York.
11. Isshiki, K., Sohrin, Y., and Nakayama, E. (1991). Marine
Chemistry 32: 1.
12. Kido, K. and Nishimura. (1975). Deep-Sea Research 22: 323.
13. Okada, T. and Kuwamoto, T. (1985). Anal. Chem. 57: 258.
14. George, S., Steinberg, S.M., and Hodge, V. (2000). Chemo-
sphere 40: 57.
15. Matthess, G., Frimmel, F., Hirsch, P., Schulz, H.D., and
Usdowski, H-E. (Eds.). (1992). Progress in Hydrogeochem-
istry. Springer-Verlag, New York.
16. Henley, R.W. (1983). Geothermics 12: 307.
17. Berner, R.A. (1995). Am. J. Science 295: 491.
18. Urey, H.C. (1952). The Planets: Their Origin and Develop-
ment. Yale Univ. Press, New Haven, CT, p. 242.
19. Degens, E.T. (1976). Topics in Current Chemistry. Vol. 64,
Inorganic Biochemistry. Springer-Verlag, New York.
SODIUM IN NATURAL WATERS
NITISH PRIYADARSHI
Ranchi University
Ranchi, Jharkhand, India
INTRODUCTION
Sodium is a light alkali metal which floats on water in
its pure state. However, it is very chemically active and is
rarely found in its pure state. All natural waters contain
SODIUM IN NATURAL WATERS 551
sodium; nearly all sodium compounds readily dissolve in
water, and it naturally leaches from rocks and soils.
The sodium ion, the sixth most abundant metallic ion in
the earth’s crust, is a natural constituent of both food and
water from several sources. Most water supplies contain
less than 20 mg of sodium per liter, but in some countries,
levels can exceed 250 mg/liter. Saline intrusion, mineral
deposits, seawater spray, sewage effluents, industrial
brines, chemicals used in paper processing, soap, glass,
and glass-wool industries, food processing, slaughtering
and meat packing industries and salt used in road
deicing all contribute significant quantities of sodium
to water. In addition, water treatment chemicals, such
as sodium fluoride, sodium bicarbonate, and sodium
hypochlorite can together result in sodium levels as high
as 30 mg/liter. Domestic water softeners can give levels
of over 300 mg/liter. A much more important source of
sodium in fluvial waters is the input from leaching of
halite from sedimentary or evaporite deposits. More than
40% of the sodium in fluvial waters comes from this
source. Sodium is also derived from the weathering of
silicate minerals, principally sodic plagioclase feldspar.
This source accounts for more than 20% of the sodium
in river water. It is also an important element in mica,
amphibole, and pyroxene.
A Na+ restricted diet is recommended to patients
suffering from hypertension or congenital heart disease;
in such cases, the intake of high Na+ through drinking
water may prove critical (1). The current drinking water
standard for sodium is 20 milligrams per liter.
SODIUM IN NATURAL WATER
Groundwaters
The Na content of groundwaters is a function of
(1) weathering of Na plagioclase from bedrock, followed
by (2) exchange of Ca2+ for Na+ on the surfaces of newly
formed clay minerals. Feth et al. (2) considered the weath-
ering reactions which supplies Na+ to groundwater as:
3Na0.66 Ca0.33 Al1.33 Si2.66 O8 + 14H2 O + 4CO2
→ 2H4 Al2 Si2 O9 + 2Na+ + Ca2+ + 4H4 SiO4 + 4HCO−
3
(kaolinite)
and
6Na0.66 Ca0.33 Al1.33 Si2.66 O8 + 6CO2 + 18H2 O
→ 3Na0.66 Al2.66 Si3.33 O10 (OH)2 + 2Ca2+ + 2Na+
+ 6H4 SiO4 + 6HCO3 − (montmorillonite)
These theoretical calculations apply well to the actual
minerals in the rocks and soils of the area studied
(western U.S.)
Garrels (3) suggested that the Na+ content of ground-
water would rise due to weathering of plagioclase
feldspars. If the system is closed, the waters eventually
proceed through kaolinite stability. If calcite is precipitat-
ing, this montmorillonite will be sodic and hold the Na+
552 SODIUM IN NATURAL WATERS
content of the water relatively steady. If calcite is not pre-
cipitating, Ca2+ will be used by the montmorillonite, and
the Na+ content of the groundwater will continue to rise.
Low concentrations of Na+ ions are associated with the
minerals, kaolinite and quartz (4).
Sathyanarayanan and Periakali (5), in their study
of Na concentration in groundwater in Ultrabasic and
Peninsular Gneissic Rocks of Tamil Nadu State of India,
suggested that weathering of plagioclase feldspar and
biotite in the Peninsular gneiss could be the source of
Na in groundwater.
In short, the Na+ content of groundwater is a func-
tion of weatherable minerals, pH, reaction time versus
drainage time, the supply of Ca2+ to form Ca montmoril-
lonite, sluggish drainage conditions, intensive irrigation
practices, extensive use of chemical fertilizers, inadequate
sanitary facilities, and precipitation dominance control-
ling water chemistry (6–9). In addition, groundwater can
contain large amounts of Na+ from the dissolution of evap-
orites, from saltwater intrusion, and smaller amounts from
cyclic sea salts in original precipitation.
Rivers
The concentration of elements in river water varies with
time at any single sampling site as a function of discharge,
tributary supply, and groundwater discharge. In areas
where rivers are draining evaporite deposits, the Na
content can be very high. For example, streams in the
Brazos River basin (Texas) have as much as 99,800 ppm
Na and an average of 35,900 ppm (10).
The unweighted average Na content of 639 rivers and
√ In addition to water treatment, another method of
streams is 38.6 ± 5.9 ppm (s/ n). The range is 0.1 to
1.540 ppm, and the median is 6.1 ppm Na (11,12). The
mean content may be used as a worldwide estimate,
but the population is actually polymodal with a mean
for humid and temperate weathering zones, a mean for
arid climates, and an extremely high mean for rivers
draining evaporite deposits (13). Comparison shows that
river water and seawater are opposite in character for
Na concentration. In seawater, Na>Mg>Ca; in average
river water, Ca>Na>Mg (14). Livingstone’s (11) weighted
mean Na+ content of world rivers was 6.3 ppm.
Lake Waters
The sodium content of lake water is a function of
river and groundwater supply followed by increase due
to evaporation where this process exceeds input. The
evaporation may reach a stage of Na mineral precipitation,
which occurs at a high level of Na+ .
Seawater
The sodium released into solution during weathering is
transported by rivers to seas/oceans and represents a
primary cyclic reservoir. Early knowledge of the average
composition of seawater was provided by Dittmar in 1884;
he carefully analyzed 77 water samples, representative of
all oceans and taken both from the surface and from the
depths. These samples had been collected on the voyage
around the world of H.M.S. Challenger (1872–1876).
He determined halides, sulfate, magnesium, calcium,
and potassium. Sodium was calculated by difference by
subtracting the sums of the chemical equivalents of the
negative and positive ions. This procedure was followed
because Dittmar was unable to achieve satisfactory direct
determinations for sodium. He estimated 30.61% sodium
in sea water.
In the valuable treatise, The Data of Geochemistry,
(1924), F. W. Clarke calculated that the 300,000,000 cubic
miles of oceanic waters owed their salinity to 4,800,000
cubic miles of dissolved salts which had been derived from
the igneous rocks of the earth’s crust. He confined his
computations to the sodium (Na) percentages, reckoning
an average of 1.14% in seawater and 2.90% of this element
in igneous rocks. The average concentration of sodium
in seawater is 1.1 × 107 g/L (14). Earlier, Goldberg (15)
estimated 10,500 ppm Na+ in seawater. Calculations of
Garrels and Thompson (16) indicated that 99% of the Na
in seawater exists as uncomplexed Na+ .
TREATMENT
Sodium may be removed from a water supply by ion
exchange. The ion exchange resin in the water softener is
enriched with calcium and removes the sodium from the
water. Other demineralization techniques, such as reverse
osmosis, distillation, electrodialysis and freeze–thaw
processes, should be effective for removing sodium from
water. If a large supply of water is required, it may be
necessary, if feasible, to blend it with another water source
low in sodium.
removing sodium from drinking water would require
the construction or reconstruction of a water well. This
frequently involves installing additional casing beyond
the depth normally required by regulations.
BIBLIOGRAPHY
1. Holden, W.S. (1970). Water Treatment and Examination.
J & Churchill, London.
2. Feth, J.H., Roberson, C.E., and Polzer, W.L. (1964). Sources
of Mineral Constituents in Water from Granite Rocks of the
Sierra Nevada. U.S. Geological Survey Water Supply Papers,
1535-I, pp. 1–70.
3. Garrels, R.M. (1967). Genesis of some groundwaters from
igneous rocks. In: Researches in Geochemistry, 2nd Edn.
P.H. Abelson (Ed.). John Wiley & Sons, New York, pp.
405–420.
4. Garrels, R.M. (1976). A survey of low temperature water-
mineral relations. In: Interpretation of Environmental
Isotope and Hydrochemical Data in Groundwater Hydrology.
International Atomic Energy Agency, Vienna, Austria.
5. Sathyanarayanan, M. and Periakali, P. (2003). Geochemistry
of groundwater in Ultrabasic and Peninsular gneissic rocks,
Salem District, Tamil Nadu. J. Geol. Soc. India 62(1): 63–73.
6. Gibbs, R.J. (1970). Mechanisms controlling world water
chemistry. Science. 170: 1088–1090.
7. Prakasa Rao, J. (1997). Environmental Evaluation of Ground-
water Quality of a Developing Urban Area of Andhra
Pradesh, India. Unpublished. Ph.D. Thesis, Andhra Univer-
sity, Visakhapatnam, p. 104.
 SOFT WATER 553

8. Subba Rao, N. (1998). Groundwater quality in crystalline 4. Temporary hardness: Equals alkalinity, if alkalinity
terrain of Guntur District, Andhra Pradesh. Visakha Sci. J. is less than total hardness
2: 51–54.
5. Permanent hardness: The amount of hardness
9. Ballukrya, P.N. and Ravi, R. (1999) Characterization of
greater than temporary hardness
groundwater in the unconfined aquifers of Chennai city,
India: Part I- Hydrogeochemistry. J. Geol. Soc. India 54: 6. Noncarbonate hardness: Permanent hardness
1–11.
10. Baker, R.C., Hughes, L.S., and Yeast, I.D. (1964). Natural Of the six different forms of hardness, the most important
Sources of Salinity in the Brazos River, Texas. U.S. Geological and the one that is commonly known as hardness,
Survey Water Supply Papers. 1669-cc, pp. 1–81. is total hardness. It is calculated by the following
11. Livingstone, D.A. (1963) Chemical Composition of Rivers simplified equation:
and Lakes. U.S. Geological Survey Professor. Papers 440-G,
pp. 1–64.
Hardness = 2.5 Ca + 4.1 Mg (units in mg/l)
12. Anderson, P.W. and George, J.R. (1966). Water-Quality Char-
acteristics of New Jersey Streams. U.S. Geological Survey
Water Supply Papers 1819-G, pp. 1–48.
Note that there are equations to calculate each of the six
different forms of hardness.
13. Wedepohl, K.H. (1970). Handbook of Geochemistry. Vol. II/2,
Springer-Verlag, Berlin.
Example: The Ca and Mg concentrations in a water sample
are 20 and 15 mg/l, respectively. What is its hardness?
14. Mason, B. and Moore, C.B. (1985). Principles of Geochemistry,
4th Edn. Wiley Eastern Limited, New Delhi, p. 242.
15. Goldberg, E.D. (1957). Biogeochemistry of trace metals. In: Hardness = 2.5 × 20 + 4.1 × 15 = 111.5 mg/l
Ecology, vol.1. Treatise on Marine Ecology and Paleocology.
J.W. Hedgpeth (Ed.). Geological Society America Member 67, As stated, each of the six forms of hardness can be
pp. 345–358. calculated by analyzing a water sample for the relevant
16. Garrels, R.M. and Thompson, M.E. (1962). A chemical model ion(s), followed by using the relevant equation. However,
for sea water at 25 ◦ C and one atmosphere total pressure. Am. there are techniques to measure total hardness directly
J. Sci. 260: 57–66. such as (1) a standard solution of soap, that was in use for
many years (2); and (2) Schwarzenbach et al’s. method of
colorimetric titration (5), which is sometime referred to as
SOFT WATER the EDTA method.
The author believes that earlier, measuring hardness
GHOLAM A. KAZEMI
was a common and an inseparable practice in water
Shahrood University of
analysis. To agree with this statement, one needs
Technology
Shahrood, Iran to look at a relatively old handbook and read, ‘‘The
common denominator of the majority of water problems is
hardness’’ (2, p. 4.1, first published in 1979). Somehow, in
INTRODUCTION contrast, recent hydrochemical analysis of water samples
does not usually include a measurement of hardness.
For instance, there is not a single report of hardness
At the outset, it should be pointed out that soft and
in the hundreds of hydrochemistry papers presented at
hard waters are two sides of a coin. Hence, the respective
the International Association of Hydrogeolgists (IAH)
articles, this one and the Hard Water article are alike in
conferences in Prague (2003) and in Munich (2001) and
many aspects. In defining soft water, it is necessary first to
in the recent past (1997–2001) annual symposia of the
describe hardness, a closely related chemical property of
Geological Society of Iran. This occurred, perhaps, because
water. Hardness is defined by some as a property of water
other aspects of water chemistry and water biology,
that prevents lather formation with soap and produces
such as nutrients, heavy metals, bacteriological species,
scale in hot water systems, (1); by others, as the soap
chlorinated hydrocarbons, and many contaminants have
consuming capacity of water (2); and in broader terms
become so widespread that they are considered more
by Freeze and Cherry (3), as the ‘‘metallic ion content of
important and more hazardous than hardness. This is to
water which reacts with sodium soaps to produce solid
say that rejecting water for drinking or other uses, these
soaps or scummy residue and which reacts with negative
days, is many times more likely to be due to its harmful
ions, when the water is evaporated in boilers, to produce
DDT level than because of its high hardness value. One
solid boiler scale.’’
other possible reason is that hardness and soft water topics
The ions which result in, or produce, hardness include
are important only in some narrow fields of water, e.g., hot
Mg+2 , Ca+2 , Sr+2 , Fe+2 , and Mn+2 , and to a lesser extent,
water boilers.
Ba+2 and Zn+2 and other divalent ions. The first two of
The second point is that industrialization, urbaniza-
these, are the most influential. There are different forms
tion, agricultural practices, and most environmentally
of hardness (4):
unfriendly activities are not likely to result in an increase
in the hardness of water resources because these activities
1. Total hardness: Ca and Mg expressed as CaCO3
do not usually generate hardness-producing ions such as
2. Calcium hardness: Ca expressed as CaCO3 Ca2+ or Mg2+ , though, they may produce other trivial (in
3. Magnesium hardness: Mg expressed as CaCO3 terms of hardness) ions such as Mn2+ and Fe2+ .
554 SOFT WATER
The author, here, wishes to draw the reader’s attention
to one other point and pose a question as well. He believes
that ‘‘softness’’ would have had an identical application
and a similar meaning to hardness if it would have been
there. Therefore, it is interesting to speculate as to what
was the rationale for use of the term ‘‘hardness’’, not
softness. One reason might be that hard water is generally
easier to notice because of its unwanted and disappointing
characteristics and our predecessors looked for something
to describe it. In other words, the term hardness was
created to explain easily noticeable hard water, not to
describe soft water. This is why there is no such practice
as water hardening, changing soft water to hard water, in
contrast to the well-established exercise of water softening
which changes hard waters to soft waters.
DEFINITION AND SPECIFICATIONS OF SOFT WATER
Todd (6, p. 282) reports that Hippocrates (460–354 B.C.)
was the first to use the terms hard and soft for water, when
in his treatises on public hygiene, he wrote, ‘‘consider the
waters which the inhabitants use, whether they be marshy
and soft, or hard and running from elevated and rocky
situations, and then if saltish and unfit for cooking. . . for
water contributes much to health.’’ However, it is probably
safe to say that hard and soft water in Hippocrates’
writings, are likely to mean different things from what
we today call hard and soft water.
The most frequently used classification system for
water based on hardness is that of Sawyer and
McCarty (7), presented in Table 1. There is another
slightly different classification system by Hem (8), which
is referred to in some publications as the classification of
the U.S. Geological Survey (9).
The definition of soft water varies slightly among
researchers. Based on Table 1, soft water is defined as
water whose hardness is less than 60 or 75 mg CaCO3 /l.
This means that waters with less than 15 mg Ca2+ /l and
10 mg Mg2+ /l are soft; higher concentrations in waters
belong to the other categories such as hard water or very
hard water. However, some researchers argue that there
is no exact definition for determining soft or hard water
because it depends, at least partially, on what one is
accustomed to (9). What is soft water for one person, may
be moderately hard for another and vice versa. The very
hard water category in Table 1 can be grouped into the
hard water category as it has the specifications of hard
water. It is, however, unclear to which category one can
assign moderately hard water, if hard water and soft water
are the only categories available.
Table 1. Different Types of Water
Based on Hardness
Water Type Hardness mg CaCO3 /l
Soft below 75a Below 60b
Moderately hard 75–150a 60–120b
Hard 150–300a 120–180b
Very hard Over 300a Over 180b
a
Reference 7.
b
Reference 8.
Manning (9), discredits soft water and states that very
soft water when used for bathing feels slick on the body,
as though the soap is still there. He also explains that
most people prefer slightly hard water because of its taste
and its ability to wash the soap off the body. Soft waters
are not very good for irrigation either because they may
lead to a lower infiltration rate compared with slightly
hard water (9). In contrast to hard water, soft water
produces a foam or lather easily but does not produce
scale in hot water pipes, heaters, boilers, and other high
water temperature units. Therefore, using soft water is
an economically worthwhile practice because it leads to
a decrease in soap consumption and lowers fuel costs for
boilers due to good heat conduction.
The recommended water for domestic use should have
a hardness of 80 mg/l; i.e., it should be soft (10). This is
why water softening is a common practice where the water
supply has a hardness greater than 80–100 mg/l (3). Soft
water is also required for some industrial processes such
as food equipment washing, confectionery, food canning
and freezing, food processing, and laundering (6).
SOFT WATER AND HEALTH
There is a general perception among the public that
soft water is healthier to drink. Studies, though not
satisfactorily conclusive, suggest the opposite, especially
with regard to artificially softened water. For instance,
Tebbutt (1995) (1) reports that there is some statistical
evidence to suggest that artificially softened waters may
increase the incidence of some forms of heart disease.
Also, a recent study in Taiwan (11) shows that there is a
significant negative relationship between drinking water
hardness and colon cancer mortality, a finding considered
important for the Taiwanese water industry and human
health. All these confirm Muss’s (12) observation that in
the United States, death rates from heart and circulatory
diseases are lower in states where public water supplies
are harder. Furthermore, artificially softened water may
taste salty because its sodium content increases in the
softening process.
AVAILABILITY OF NATURALLY SOFT WATER
Soft water may exist naturally or may be produced
artificially, as stated before. Atmospheric precipitations
(snow and rain) are soft waters. Similar to these are
waters that emanate from upland catchments with an
average total hardness of 10 mg/l (1). In contrast, most
groundwaters, excluding those that occur in fractured
igneous rocks, are hard. All ground water and surface
water that pass through, or are in contact with, carbonate
and other Ca–Mg bearing rocks are hard. Lowland rivers
with a hardness of 200 mg/l (1) are also hard. Therefore,
only a limited percentage of freshwater resources of the
world, groundwaters and river waters, is soft.
Examples of naturally occurring soft waters include the
Vilnius well fields groundwater (intermorainic and allu-
vial deposit aquifer), Vilnius District, Lithuania, that have
unbelievably low hardness values of 3.9–8.5 mg/l (13),

groundwater in Akwa Ibom State of Nigeria (fluvio-
volcanic rocks) with a hardness of 1.2–65 mg/l (Akujieze,
2003) (14), and shallow basement groundwater of Lithem-
bwe dambo, Malawi, with 1–14 mg/l of calcium and
1–7 mg/l of magnesium (15).
BIBLIOGRAPHY
1. Tebbutt, T.H.Y. (1995). Principles of Water Quality Control.
Reprint of 4th Edn. Butterworth Heinemann, Oxford, UK.
2. Kemmer, F.N. (1988). The NALCO Water Handbook, 2nd
Edn. McGraw-Hill, New York.
3. Freeze, R.A. and Cherry, J.A. (1979). Groundwater. Prentice-
Hall, Englewood Cliffs, NJ.
4. Hounslow, A.W. (1995). Water Quality Data: Analysis and
Interpretation. Lewis, New York.
5. McCoy, J.W. (1981). The Chemical Treatment of Boiler. Water
Chemical Publishing, New York.
6. Todd, D.K. (1980). Groundwater Hydrology, 2nd Edn. John
Wiley & Sons, New York.
7. Sawyer, C.N. and McCarthy, P.L. (1967). Chemistry for
Sanitary Engineers, 2nd Edn. McGraw-Hill, New York.
8. Hem, J.D. (1985). Study and Interpretation of the Chemical
Characteristics of Natural Waters, 3rd Edn. U.S. Geological
Survey Water Supply Paper 2254, Denver, CO.
9. Manning, J.C. (1997). Applied Principles of Hydrology.
Prentice Hall, Upper Saddle River, NJ.
10. Bell, F.G. (1998). Environmental Geology, Principles and
Practice. Blackwell Science, Cambridge, UK.
11. Yang, C.Y. and Hung, C.F. (1998). Colon cancer mortality
and total hardness levels in Taiwan’s drinking water. Arch.
Environ. Contamination Toxicol. 35(1): 148–151.
12. Muss, D.L. (1962). Relationship between water quality and
deaths from cardiovascular disease. J. Am. Water Works
Assoc. 54: 1371–1378.
13. Klimas, A.A. (1995). Impacts of urbanization and protection of
water resources in the Vilnius District, Lithuania. Hydrogeol.
J. 3(1): 24–35.
14. Akujieze, C.N., Coker, S.J.L., and Oteze, G.E. (2003). Ground-
water in Nigeria—a millennium experience—distribution,
practice, problems and solutions. Hydrogeol. J. 11(2):
259–274.
15. Chilton, P.J. and Foster, S.D. (1995). Hydrogeological char-
acterization and water-supply potential of basement aquifers
in tropical Africa. Hydrogeol. J. 3(1): 36–49.
SOLUBILITY OF CHEMICALS IN WATER
CARL L. YAWS
PRASAD K. NARASIMHAN
HELEN H. LOU
Lamar University
Beaumont, Texas
RALPH W. PIKE
Louisiana State University
Baton Rouge, Louisiana
Correlations are presented for the solubility of chemicals in
water. A wide variety of compounds are covered—alcohols,
SOLUBILITY OF CHEMICALS IN WATER 555
ketones, ethers, aldehydes, acids, acetates, amines,
monochloroalkanes, and monobromoalkanes. The new
correlation provides reliable solubility values down to
very low concentrations (parts per million range). The
correlation is based on the boiling point of the compound.
The correlation and experimental data are in favorable
agreement. The results are useful in process engineering
for wastewater minimization.
IMPORTANCE OF WATER SOLUBILITY
The solubility of compounds in water is very important.
This importance will increase in the future because of
heath, safety, and environmental considerations. The
following brief discussion illustrates the importance even
at very low concentrations.
For health involving human exposure to substances
in air, the threshold limit value (TLV) for sec-butanol in
air is given as 100 ppm (parts per million, vol) by the
Occupational Safety and Health Act (1). A concentration
of only 0.0005 mol fraction of sec-butanol in water provides
about 236 ppm of sec-butanol in air at the air–water
interface. This concentration of 236 ppm in air exceeds
the threshold limit value of 100 ppm.
For safety, the lower explosion limit (LEL) for sec-
butanol in air is given as 1.7% by Yaws (2). A concentration
of only 0.04 mol fraction of sec-butanol in water provides
about 1.9% of sec-butanol in air at the air–water interface.
This concentration of 1.9% in air exceeds the lower
explosion limit of 1.7%.
For environmental, consider a spill of sec-butanol in
contact with water. The water will become saturated with
sec-butanol. At saturation, the solubility of sec-butanol
in water is about 0.181 weight fraction or 0.0508 mol
fraction (3). This concentration of only 0.0508 mol fraction
at saturation will provide about 24,000 ppm or 2.5%
of sec-butanol in air at the air–water interface. This
concentration of 24,000 ppm or 2.4% greatly exceeds both
the threshold limit value of 100 ppm and the lower
explosion limit of 1.7%.
This brief discussion (using sec-butanol as an example)
indicates that very low concentrations of a chemical in
water can provide concentrations in air at the air–water
interface that exceed the threshold limit value for human
exposure and the lower explosion limit for flammability.
CORRELATION FOR WATER SOLUBILITY
In earlier work by Yaws and co-workers (2), the water
solubility of hydrocarbons was correlated as a function of
the boiling point of the compound. In this article, it was
determined that the boiling point method is also applicable
to the correlation of the water solubility of chemicals:
log10 (S) = A + B TB + C TB 2 + D TB 3 (1)
where S = solubility in water at 25 ◦ C, ppm
by weight,
TB = boiling point of compound, K
A, B, C, and D = regression coefficients
556 SOLUBILITY OF CHEMICALS IN WATER

Table 1. Solubility in Watera

log10 S = A + B TB + C TB 2 + DTB 3 (S, ppm(wt), TB , K)

Chemical Family A B C D TBMIN TBMAX Formula Example Compound

Alcohols 45.6398 −2.3859E−01 4.8739E−04 −3.7160E−07 370 620 Cn H2n+2 O n-Pentanol

Ketones 45.2000 −2.3859E−01 4.8739E−04 −3.7160E−07 350 590 Cn H2n O Diethyl ketone
Ethers 11.4000 −2.2000E−02 0.0000E+00 0.0000E+00 300 560 Cn H2n+2 O Dipropyl ether
Aldehydes 12.4200 −2.2000E−02 0.0000E+00 0.0000E+00 320 560 Cn H2n O 1-Hexanal
Acids 14.3000 −2.2000E−02 0.0000E+00 0.0000E+00 440 560 Cn H2n O2 n-Hexanoic acid
Acetates 13.0000 −2.3382E−02 0.0000E+00 0.0000E+00 330 600 Cn H2n O2 Ethyl acetate
Primary amines 11.4300 −1.8000E−02 0.0000E+00 0.0000E+00 400 560 Cn H2n+3 N n-Hexyl amine
Secondary amines 11.3600 −1.8000E−02 0.0000E+00 0.0000E+00 360 560 Cn H2n+3 N Dipropyl amine
Tertiary amines 11.8200 −2.0000E−02 0.0000E+00 0.0000E+00 360 560 Cn H2n+3 N Triethyl amine
Monochloroalkanes −7.4500 1.0050E−01 −2.7288E−04 1.9987E−07 300 590 Cn H2n+1 Cl 2-Chlorobutane
Monobromoalkanes −7.1700 1.0050E−01 −2.7288E−04 1.9987E−07 300 590 Cn H2n+1 Br 1-Bromobutane
a
S-solubility in water at 25 C, parts per million by weight, ppm(wt).
A, B, C, and D-regression coefficients.
TB -boiling point temperature of compound, K.
TBMAX -maximum temperature, K.
TBMIN -minimum temperature, K.

The range of boiling point depends on the chemical family. Data-alcohols Equation
The coefficients (A, B, C, and D) are given in Table 1 1000000
for a wide variety of chemicals—alcohols, ketones, ethers,
100000
aldehydes, acids, acetates, amines, monochloroalkanes,
Solubility in water, ppm (wt)

10000
and monobromoalkanes. The table also provides the
boiling point range (TBMIN , TBMAX ) for each chemical 1000
family. The correlation is applicable to compounds whose 100
boiling points are in this range. The correlation should not 10
be used for compounds whose boiling points are outside 1
the boiling point range.
0.1
The coefficients were determined from regression of
0.01
available data. In preparing the correlation, a literature
search was conducted to identify data source publica- 0.001
tions (2–13). The excellent compilations by Horvath (5–6); 0.0001
Howard and Meylan (7); Mackay, Shiu, and Ma (9); Ver- 300 350 400 450 500 550 600 650
schueren (11); Yalkowsky (3); and Yaws (2) were used Boiling point, K
extensively. The publications were screened and copies Figure 1. Solubility of alcohols in water.
of appropriate data were made. These data were then
keyed in to the computer to provide a database for which
experimental data are available. The database also served Data-alcohols Data-ketones
as a basis to check the accuracy of the correlation. Equation Equation
Graphs of water solubility versus boiling point are 1000000
shown in Figs. 1–11 for the chemical families. The graphs
100000
indicate favorable agreement of correlation values and
Solubility in water, ppm (wt)

experimental data. 10000

For ketones, the graphs disclose that more data are 1000
available for alcohols. The graphs, also disclose that the
100
data for ketones appear to be below and approximately
parallel to the more extensive data for alcohols. This 10
suggests that the solubility of ketones is lower and 1
approximately parallel to the water solubility of alcohols. 0.1
For the other oxygenated compounds (ethers, aldehy-
des, acids, and acetates), the graphs disclose that the data 0.01
for oxygenated compounds appear to be above and approx- 0.001
imately parallel to the data for benzenes. This suggests 0.0001
that the solubility of oxygenated compounds is higher and 300 350 400 450 500 550 600
approximately parallel to the water solubility of benzenes. Boiling point, K
For the nitrogen compounds (primary amines, sec-
Figure 2. Solubility of ketones in water.
ondary amines, and tertiary amines), the graphs disclose
 SOLUBILITY OF CHEMICALS IN WATER 557

Equation Data-ethers Equation Equation

Equation Data-benzenes Data-acids Data-benzenes
100000 1000000
10000 100000
Solubility in water, ppm (wt)

Solubility in water, ppm (wt)

1000 10000
1000
100
100
10
10
1
1
0.1 0.1
0.01 0.01

0.001 0.001
300 350 400 450 500 550 600 300 350 400 450 500 550 600 650
Boiling point, K Boiling point, K

Figure 3. Solubility of ethers in water. Figure 5. Solubility of acids in water.

Data-acetates Data-benzenes
Equation Data-aldehtdes
Equation Equation
Equation Data-benzenes
1000000 1000000

100000 100000
Solubility in water, ppm (wt)
Solubility in water, ppm (wt)

10000
10000
1000
1000
100
10 100

1 10
0.1
1
0.01
0.1
0.001 300 350 400 450 500 550 600
300 350 400 450 500 550 600 Boiling point, K
Boiling point, K
Figure 6. Solubility of acetates in water.
Figure 4. Solubility of aldehydes in water.

Equation Equation
that the data for amines appear to be above and approx- Data-amines-primary Data-benzenes
imately parallel to the data for benzenes. This suggests 100000
that the solubility of amines is higher and approximately
10000
Solubility in water, ppm (wt)

parallel to the water solubility of benzenes.

For monochloro and monobromoalkanes, the graphs 1000
disclose that the data appear to be lower and approxi-
100
mately parallel to the data for benzenes. This suggests
that the solubility of monochloro- and monobromoalkanes 10
is lower and approximately parallel to the water solubility 1
of benzenes.
0.1

0.01
EXAMPLES
0.001
300 350 400 450 500 550 600
The correlations for water solubility are useful in Boiling point, K
engineering applications involving process wastewater.
Examples are shown below. Figure 7. Solubility of primary amines in water.
558 SOLUBILITY OF CHEMICALS IN WATER

Equation Equation Monobromo alkanes Benzenes

Data-amines-secondary Data-benzenes Equation Equation
100000 100000

10000 10000

Solubility in water, ppm (wt)

Solubility in water, ppm (wt)

1000 1000

100 100

10
10
1
1
0.1
0.1
0.01
0.01
0.001
0.001 300 350 400 450 500 550 600
300 350 400 450 500 550 600 Boiling point, K
Boiling point, K
Figure 11. Solubility of monobromoalkanes in water.
Figure 8. Solubility of secondary amines in water.
Example. A chemical spill of dodecanol occurs into
a body of water at ambient conditions. Estimate the
Equation Equation concentration in the water at saturation.
Data-amines-tertiary Data-benzenes The correlation for alcohols maybe used to determine
100000 the solubility in water. Substitution of the coefficients and
boiling point of dodecanol in the correlation equation yields
10000
Solubility in water, ppm (wt)

1000 log10 (S) = A + B TB + C TB 2 + D TB 3

100
= 45.6398 − 2.3859E−01(535.00)
+ 4.8739E−04(535.00)2 − 3.7160E−07(535.00)3
10
S = 3.93 ppm (wt).
1
The calculated value and data compare favorably
0.1
(3.93 vs. 4.00; deviation = 0.07/5 = 1.8%).
0.01
Example. A chemical spill of diethyl ketone occurs into
0.001 a body of water at ambient conditions. Estimate the
300 350 400 450 500 550 600 concentration in the water at saturation.
Boiling point, K The correlation for ketones maybe used to determine
the solubility in water. Substitution of the coefficients
Figure 9. Solubility of tertiary amines in water.
and boiling point of diethyl ketone in the correlation
equation yields
Monochloro alkanes Benzenes log10 (S) = A + B TB + C TB 2 + D TB 3
Equation Equation
100000 = 45.2000 − 2.3859E−01 (375.14)
10000 + 4.8739E−04(375.14)2 − 3.7160E−07 (375.14)3
Solubility in water, ppm (wt)

1000 S = 45, 257 ppm (wt).

100
The calculated value and data compare favorably
10 (45,257 vs. 48,235; deviation = 1,758/48,235 = 3.6%).
1 Acknowledgment
0.1 The Texas Hazardous Waste Research Center (Lamar University,
Beaumont, Texas) provided partial support for this work.
0.01

0.001 BIBLIOGRAPHY
300 350 400 450 500 550 600
Boiling point, K 1. Guide to Occupational Exposure Values. (1997). American
Conference of Governmental Industrial Hygienists, Cincin-
Figure 10. Solubility of monochloroalkanes in water. nati, OH.

2. Yaws, C.L. (1999). Chemical Properties Handbook. McGraw-
Hill, New York.
3. Yalkowsky, S.H. (1990–2004). Aquasol Database. University
of Arizona, Tucson, AZ.
4. Hine, J. and Mukherjee, P.K. (1975). J. Org. Chem. 40: 292.
5. Horvath, A.L. and Getzen, F.W. (1999). J. Phys. Chem. Ref.
Data 28(3): 649–777.
6. Horvath, A.L. (1982). Halogenated Hydrocarbons: Solubility-
Miscibility with Water. Marcel Dekker, New York.
7. Howard, P.H. and Meylan, W.M. (Eds.). (1997). Handbook of
Physical Properties of Organic Chemicals. CRC Press, Boca
Raton, FL.
8. Liquid–Liquid Equilibrium Data Collection. (1979). Dechema
Chemistry Data Series, Frankfurt/Main, Germany, Vol. V,
part 1.
9. Mackay, D., Shiu, W.Y., and Ma, K.C. (1992, 1992, 1993,
1995, 1997). Illustrated Handbook of Physical-Chemical
Properties and Environmental Fate for Organic Chemicals.
Vols. 1–5, Lewis, New York.
10. Suzuki, T. (1991). J. Computer-Aided Molecular Design 5:
149.
11. Verschueren, K. (1996, 2001). Handbook of Environmental
Data on Organic Chemicals, 3rd and 4th Edn. Van Nostrand
Reinhold, New York.
12. Donald, M., Shiu, W-Y., and Ma, K-C. (1995). Illustrated
Handbook of Physical-Chemical Properties and Environmen-
tal Fate for Organic Chemicals. Vol. 4, Lewis, New York.
13. (2001). Handbook of Environmental Data on Organic
Chemicals, 4th Edn. John Wiley & Sons, New York.
SOLUBILITY OF HYDROCARBONS IN SALT
WATER
CARL L. YAWS
PRASAD K. NARASIMHAN
Lamar University
Beaumont, Texas
INTRODUCTION
Physical and thermodynamic property data are required
to design and operate industrial processes. In particular,
the water solubility of hydrocarbons in the sea and ocean
is becoming increasingly important because of more and
more stringent regulations regarding the environment.
In this article, results are presented for the solubility
of hydrocarbons in salt water. The results are applicable
to the complete range of salt concentrations, including
water without salt to water saturated with salt. The
results are intended for use in engineering applications.
As an example of such usage, solubility values from the
correlation are useful in determining the distribution
of a hydrocarbon spill upon contact with seawater.
Solubility values at other salt concentrations may also
be ascertained.
CORRELATION OF SALT WATER SOLUBILITY
The correlation of the solubility of hydrocarbons in salt
water containing salt is based on a series expansion in salt
SOLUBILITY OF HYDROCARBONS IN SALT WATER 559
concentration:
log10 S = A + B X + C X 2 (1)
where S = solubility in salt water at 25 ◦ C,
parts per million by weight, ppm(wt)
X = concentration of salt (NaCl) in water,
parts per million by weight, ppm(wt)
A, B, and C = regression coefficients
The coefficients (A, B, and C) are given in Table 1.
The values in the table were determined from regression
of the data sources (1–10) for water solubility. The
compilations by Price (7) and Sutton and Calder (8) were
especially helpful. The values presented are applicable to
a wide variety of hydrocarbons (alkanes, cyclopentanes,
cyclohexanes, benzenes (no and single substitution), and
benzenes (double and triple substitution). The range
of application for the correlations is about C5 to
C16 compounds.
Graphs of water solubility are shown in Figs. 1–3 for
representative hydrocarbons (pentane, methylcyclopen-
tane, and benzene). In the figures, solubility values are
plotted at salt concentrations ranging from water without
salt to water saturated with salt. The graphs discloses
favorable agreement of correlation values and experimen-
tal data.
Examples
The correlations for solubility are useful in engineering
applications involving hydrocarbons in contact with water
containing salt. Examples are shown below.
Example 1. A chemical spill of n-pentane occurs in
a body of saltwater [salt = 10,000 ppm(wt)] at ambient
conditions. Estimate the concentration in the saltwater.
The values in Table 1 are used for the coefficients.
Introducing the boiling point of n-pentane to calculate A
and using the tabulated values for B and C provides
A = −17.7916 + 1.7865E−01 TB − 5.0255E−04
TB 2 + 4.1218E−07 TB 3
= −17.7916 + 1.7865E−01(309.22)
− 5.0255E−04(309.22)2 + 4.1218E−07(309.22)3
= 1.5850
B = −4.5956E−06
C = 2.2978E−12
Substitution of these coefficients and the salt con-
centration in the correlation equation for solubility in
saltwater yields
log10 S = A + B X + C X 2
= 1.5850 − 4.5956E−06(10,000)
+ 2.2978E−12(10,000)2
= 1.5393
S = 34.62 ppm (wt)
560 SOLUBILITY OF HYDROCARBONS IN SALT WATER

Table 1. Solubility in Salt Watera

log10 S = A + BX + CX 2 S-solubility, ppm (wt), X(salt) − ppm(wt)

Chemical Family A, B, and C Formula Example Compound

Alkanes A = −17.7916 + 1.7865E−01TB − Cn H2n+2 n-pentane

5.0255E−04 TB 2 + 4.1218E−07 TB 3
B = −4.5956E−06
C = 2.2978E−12
Cyclopentanes A = 9.55 − 2.2987E−02TB Cn H2n Methyl cyclopentane
B = −4.5987E−06
C = 2.2993E−12
Cyclohexanes A = 9.75 − 2.2987E−02TB Cn H2n Cyclohexane
B = −4.5987E−06
C = 2.2993E−12
Benzenes (no and single A = 5.725 + 3.6778E−03 TB − Cn+6 H2n+6 Benzene
substitution) 3.0346E−05 TB 2
B = −3.2714E−06
C = 1.6357E−16
Benzenes (double and A = 5.375 + 3.6778E−03 TB − Cn+6 H2n+6 Benzene
triple substitution) 3.0346E−05 TB 2
B = −3.2714E−06
C = 1.6357E−16
a
S-solubility in salt water at 25 C, parts per million by weight, ppm(wt).
X-salt (NaCl) concentration in water, ppm(wt).
Salt concentration range is X = 0 to 358,700 (saturated).
Salt in seawater is 34,472 ppm(wt).
A, B, and C—regression coefficients.
TB —boiling point of compound, K.
TBMAX —maximum temperature, K = 560.
TBMIN —minimum temperature, K = 300.
Correlation should not be used for compounds having boiling points outside TB MAX and TB MIN.
Range of application for correlation is about C5 to C16 .

Data Equation Data Equation

45 45
40 40
Solubility in salt water, ppm(wt)

Solubility in salt water, ppm(wt)

35 35
30 30
25 25
20 20
15 15
10 10
5 5
0 0
0 50 100 150 200 250 300 350 400 0 50 100 150 200 250 300 350 400
Salt concentration in water, thousands of ppm(wt) Salt concentration in water, thousands of ppm(wt)

Figure 1. Solubility of n-pentane in salt water. Figure 2. Solubility of methyl cyclopentane in salt water.

using the tabulated values for B and C provides

The calculated value and data compare favorably (34.62
vs. 34.50; deviation = 0.12/34.50 = 0.3%). A = 5.725 + 3.6778E−03 TB − 3.0346E−05 TB 2
= 5.725 + 3.6778E−03 (353.31)
Example 2. A chemical spill of benzene occurs in
seawater [salt = 34,472 ppm(wt)] at ambient conditions. − 3.0346E−05 (353.31)2
Estimate the concentration in the seawater.
= 3.2364
The values in Table 1 are used for the coefficients.
Introducing the boiling point of benzene to calculate A and B = −3.2714E−06
 SOLUBILITY OF HYDROCARBONS AND SULFUR COMPOUNDS IN WATER 561

Data Equation 9. Verschueren, K. (1996, 2001). Handbook of Environmental

Data on Organic Chemicals, 3rd and 4th Edn. Van Nostrand
2000
Reinhold, New York.
1800
Solubility in salt water, ppm(wt)

10. Yalkowsky, S.H. (1990–2004). Aquasol Database. University

1600 of Arizona, Tucson, AZ.
1400 11. Yaws, C.L. (1999). Chemical Properties Handbook. McGraw-
1200 Hill, New York.

1000
800 SOLUBILITY OF HYDROCARBONS AND
600 SULFUR COMPOUNDS IN WATER
400
200
CARL L. YAWS
PRASAD K. NARASIMHAN
0
0 50 100 150 200 250 300 350 400 HELEN H. LOU
Salt concentration in water, thousands of ppm(wt) Lamar University
Beaumont, Texas
Figure 3. Solubility of benzene in salt water.
RALPH W. PIKE
Louisiana State University
C = 1.6357E−16. Baton Rouge, Louisiana

Substitution of these coefficients and the salt con-

centration in the correlation equation for solubility in Correlations are presented for the solubility of hydrocar-
saltwater yields bons and organic sulfur compounds in water. A wide vari-
ety of compounds are covered—alkanes, olefins, diolefins,
log10 S = A + B X + C X 2 acetylenes, cyclopentanes, cyclohexanes, benzenes, mer-
captans, thiophenes, and sulfides. The new correlation
= 3.2364 − 3.2714E−06(34,472) provides reliable solubility values down to very low con-
centrations (parts per million range). The correlation is
+ 1.6357E−16(34,472)2
based on the boiling point of the compound. Correlation
= 3.1236 and experimental data are in favorable agreement. The
results are useful in process engineering for wastewater
S = 1329.24 ppm (wt) minimization.

The calculated value and data compare favorably

(1329.24 vs. 1391.00; deviation = 61.76/1391.00 = 4.4%). IMPORTANCE OF WATER SOLUBILITY

The solubility of compounds in water is very important.

Acknowledgment
This importance will increase in the future because of
The Texas Hazardous Waste Research Center (Lamar University,
Beaumont, Texas) provided partial support for this work.
heath, safety, and environmental considerations. The
following brief discussion illustrates the importance even
at very low concentrations.
BIBLIOGRAPHY For health involving human exposure to substances
in air, the threshold limit value (TLV) for n-pentane in
1. Donald, M., Shiu, W-Y., and Ma, K-C. (1995). Illustrated air is given as 600 ppm (parts per million, vol) by the
Handbook of Physical-Chemical Properties and Environmen- Occupational Safety and Health Act (1). A concentration of
tal Fate for Organic Chemicals. Vol. 4, Lewis, New York. only 0.0000001 mol fraction of n-pentane in water provides
2. (2001). Handbook of Environmental Data on Organic about 7000 ppm of n-pentane in air at the air–water
Chemicals, 4th Edn. John Wiley & Sons, New York. interface. This concentration of 7000 ppm in air exceeds
3. Hine, J. and Mukherjee, P.K. (1975). J. Org. Chem. 40: 292. the threshold limit value of 600 ppm.
4. Howard, P.H. and Meylan, W.M. (Eds.). (1997). Handbook of For safety, the lower explosion limit (LEL) for n-
Physical Properties of Organic Chemicals. CRC Press, Boca pentane in air is given as 1.4% by Yaws (2). A
Raton, FL. concentration of only 0.000001 mol fraction of n-pentane
5. Mackay, D., Shiu, W-Y., and Ma, K-C. (1992, 1992, 1993, in water provides about 7% of n-pentane in air at the
1995, 1997). Illustrated Handbook of Physical-Chemical air–water interface. This concentration of 7% in air
Properties and Environmental Fate for Organic Chemicals. exceeds the lower explosion limit of 1.7%.
Vols. 1–5, Lewis, New York. For environmental, consider a spill of n-pentane in
6. Price, L.C. (1976). Bull. Am. Assoc. Pet. Geologists 60(2): 213. contact with water. The water will become saturated. At
7. Sutton, C. and Calder, J.A. (1974). Environ. Sci. Technol. 8(7): saturation, the solubility of n-pentane in water is about
654. 0.0000385 weight fraction or 0.00000916 mol fraction,
8. Suzuki, T. (1991). J. Comput.-Aided Molecular Design 5: 149. as given by Yalkowsky (3). This concentration of only
562 SOLUBILITY OF HYDROCARBONS AND SULFUR COMPOUNDS IN WATER

0.00000916 mol fraction at saturation will provide about
644,000 ppm or 64.4% of n-pentane in air at the air–water
interface. This concentration of 644,000 ppm or 64.46%
greatly exceeds both the threshold limit value of 600 ppm
and the lower explosion limit of 1.4%.
This brief discussion (using n-pentane as an example)
indicates that very low concentrations (ppm or less) of
compounds in water can provide concentrations in air at
the air–water interface that exceed the threshold limit
value for human exposure and the lower explosion limit
for flammability.
CORRELATION FOR WATER SOLUBILITY
In earlier work by Yaws and coworkers (2), the water
solubility of various chemical types was correlated as a
function of the boiling point of the compound. In this
article it was determined that the boiling point method
is also applicable to correlation of the water solubility of
hydrocarbons and organic sulfur compounds:
search was conducted to identify data source publica-
tions (2–10). The excellent compilations by Howard and
Meylan (5); Mackay, Shiu, and Ma (6); Verschueren (8);
Yalkowsky (3); and Yaws (2) were used extensively. The
publications were screened and copies of appropriate data
were made. These data were then keyed in to the com-
puter to provide a database for which experimental data
are available. The database also served as a basis to check
the accuracy of the correlation.
Graphs of water solubility versus boiling point are
shown in Figs. 1–8 for the compound families. The graphs
indicate favorable agreement of correlation values and
experimental data.
For olefins and diolefins, the graphs disclose that more
data are available for alkanes. The graphs also disclose
that the data for olefins and diolefins appear to be above
and approximately parallel to the more extensive data for
alkanes. This suggests that the solubility of olefins and
diolefins is higher and approximately parallel to the water
solubility of alkanes.

log10 (S) = A + BTB + CTB 2 + DTB 3 (1) data-alkanes equation

◦
100
where S = solubility in water at 25 C, ppm by
weight,
Solubility in water, ppm (wt)

10
TB = boiling point of compound, K
A, B, C, and D = regression coefficients 1
The range for boiling point is 298 K to about 560 K.
0.1
The coefficients (A, B, C, and D) are given in Table 1
for a wide variety of hydrocarbons and sulfur com-
0.01
pounds—alkanes, olefins, diolefins, acetylenes, cyclopen-
tanes, cyclohexanes, benzenes, mercaptans, thiophenes,
0.001
and sulfides. The table also provides the boiling point
range (TBMIN , TBMAX ) for which the correlation is appli-
0.0001
cable. The correlation should not be used for compounds
250 300 350 400 450 500 550 600
whose boiling points are outside the boiling point range.
Boiling point, K
The coefficients were determined from regression of
available data. In preparing the correlation, a literature Figure 1. Solubility of alkanes in water.

Table 1. Solubility in Watera

log10 S = A + BTB + CTB + DTB 3

2
(S - ppm(wt), TB − K)

Chemical family A B C D TBMIN TBMAX Formula Example Compound

Alkanes −17.6520 1.77811E−01 −5.00907E−04 4.11124E−07 298 560 Cn H2n+2 n-Hexane

Olefins −17.0300 1.77811E−01 −5.00907E−04 4.11124E−07 298 560 Cn H2n 1-Hexene
Diolefins −16.5610 1.77811E−01 −5.00907E−04 4.11124E−07 298 560 Cn H2n−2 1,5-Hexadiene
Acetylenes −15.8350 1.77811E−01 −5.00907E−04 4.11124E−07 298 560 Cn+2 H2n+2 1-Hexyne
Cyclohexanes −16.7000 1.77811E−01 −5.00907E−04 4.11124E−07 298 560 Cn H2n Methyl cyclohexane
Cyclopentanes −16.9000 1.77811E−01 −5.00907E−04 4.11124E−07 298 560 Cn H2n Methyl cyclopentane
Benzenes (single −24.0080 2.21196E−01 −5.55632E−04 4.18830E−07 298 560 Cn+6 H2n+6 Toluene
substitution)
Benzenes (multiple −23.6500 2.21196E−01 −5.55632E−04 4.18830E−07 298 560 Cn+6 H2n+6 p-Xylene
substitution)
Mercaptans −6.9000 1.00500E−01 −2.72880E−04 1.99870E−07 298 560 Cn H2n+2 S Propyl mercaptan
Thiophenes −6.8500 1.00500E−01 −2.72880E−04 1.99870E−07 298 560 Cn H2n−4 S 2-Methyl thiophene
Sulfides −6.5390 1.00500E−01 −2.72880E−04 1.99870E−07 298 560 Cn H2n+2 S Ethyl methyl sulfide
a
S = solubility in water at 25 C, parts per million by weight, ppm(wt).
A, B, C, and D = regression coefficients.
TB = boiling point of compound, K.
TBMAX -maximum temperature, K.
TBMIN -minimum temperature, K.
 SOLUBILITY OF HYDROCARBONS AND SULFUR COMPOUNDS IN WATER 563

data-cyclopentanes equation Data-alkanes Data-diolefins

1000.0000 Equation Equation
1000
Solubility in water, ppm (wt)

100.0000
100

Solubility in water, ppm (wt)

10.0000
10

1.0000 1

0.1000 0.1

0.01
0.0100
0.001
0.0010
250 300 350 400 450 500 550 600 0.0001
Boiling point, K 250 300 350 400 450 500 550 600
Boiling point, K
Figure 2. Solubility of cyclopentanes in water.
Figure 5. Solubility of diolefins in water.
data-benzenes equation
10000
Equation Equation
Data-mercaptans Data-benzenes
Solubility in water, ppm (wt)

1000 100000

10000
Solubility in water, ppm (wt)

100
1000

10 100

10
1 1

0.1
0.1
250 300 350 400 450 500 550 600 0.01
Boiling point, K
0.001
Figure 3. Solubility of benzenes in water. 250 300 350 400 450 500 550 600
Boiling point, K

Data-alkanes Figure 6. Solubility of mercaptans in water.

Data-olefins
Equation Equation
1000
Equation Data-benzenes
100 Data-thiophenes Equation
Solubility in water, ppm (wt)

10000
10
Solubility in water, ppm (wt)

1000
1
100
0.1

0.01 10

0.001 1

0.0001 0.1
250 300 350 400 450 500 550 600
Boiling point, K 0.01
250 300 350 400 450 500 550 600
Figure 4. Solubility of olefins in water.
Boiling point, K

Figure 7. Solubility of thiophenes in water.

For sulfur compounds (mercaptans, thiopheolefins, and
diolefins), the graphs disclose that more data are available
for benzenes. The graphs also, disclose that the data approximately parallel to the more extensive data for
for sulfur compounds appear to be slightly above and benzenes. This suggests that the solubility of sulfur
564 SORPTION KINETICS

Equation Equation The calculated value and data compare favorably

Data-benzenes Data-sulfides (524.68 vs 542.4; deviation = 17.72/542.4 = 3.3%).
100000
Acknowledgments
10000 The Texas Hazardous Waste Research Center (Lamar University,
Solubility in water, ppm (wt)

1000 Beaumont, Texas) provided partial support for this work. Portions
of this material appeared in Oil & Gas Journal 89; 79 (April 8,
100 1991), Chem. Eng., 100; 122 (Oct., 1993), Oil & Gas Journal 89:
86 (Sept. 16, 1991), Oil & Gas Journal 95 (35): 80 (Aug. 29, 1994),
10
and Chem. Eng. 110: 60 (Aug., 2003). These portions are reprinted
1 by special permission.

0.1
BIBLIOGRAPHY
0.01

0.001 1. Guide to Occupational Exposure Values. (1997). American

250 300 350 400 450 500 550 600 Conference of Governmental Industrial Hygienists, Cincin-
Boiling point, K nati, OH.
2. Yaws, C.L. (1999). Chemical Properties Handbook. McGraw-
Figure 8. Solubility of sulfides in water. Hill, New York.
3. Yalkowsky, S.H. (1990–2004). Aquasol Database. University
compounds is higher and approximately parallel to the of Arizona, Tucson, AZ.
water solubility of benzenes. 4. Hine, J. and Mukherjee, P.K. (1975). J. Org. Chem. 40: 292.
5. Howard, P.H. and Meylan, W.M. (Eds.). (1997). Handbook of
Examples Physical Properties of Organic Chemicals. CRC Press, Boca
Raton, FL.
The correlations for water solubility are useful in
6. Mackay, D., Shiu, W-Y., and Ma, K-C. (1992, 1992, 1993,
engineering applications involving process wastewater. 1995, 1997). Illustrated Handbook of Physical-Chemical
Examples are shown below. Properties and Environmental Fate for Organic Chemicals.
Vols. 1–5, Lewis, New York.
Example 1. A chemical spill of n-hexane occurs into 7. Suzuki, T. (1991). J. Comput.-Aided Molecular Design 5: 149.
a body of water at ambient conditions. Estimate the 8. Verschueren, K. (1996, 2001). Handbook of Environmental
concentration in the water at saturation. Data on Organic Chemicals, 3rd and 4th Edn. Van Nostrand
The correlation for alkanes maybe used to determine Reinhold, New York.
the solubility in water. Substitution of the coefficients and 9. Donald, M., Shiu, W-Y., and Ma, K-C. (1995). Illustrated
boiling point of n-hexane in the correlation equation yields: Handbook of Physical–Chemical Properties and Environmen-
tal Fate for Organic Chemicals. Vol. 4, Lewis Publishers, New
log10 (S) = A + BTB + CTB 2 + DTB 3 York.
10. (2001). Handbook of Environmental Data on Organic
= −17.6520 + 1.77811E−01(341.88)
Chemicals, 4th Edn. John Wiley & Sons, New York.
− 5.00907E−04(341.88)2
+ 4.11124E−07(341.88)3
SORPTION KINETICS
S = 10.46 ppm (wt)
TH.D. KARAPANTSIOS
The calculated value and data compare favorably (10.46 M.X. LOUKIDOU
vs 9.47; deviation = 0.99/9.47 = 10.4%). K.A. MATIS
Aristotle University
Example 2. A chemical spill of toluene occurs into a body Thessaloniki, Greece
of water at ambient conditions. Estimate the concentration
in the water at saturation.
The correlation for benzenes (single substitution) INTRODUCTION
maybe used to determine the solubility in water.
Substitution of the coefficients and boiling point of toluene Sorption is, by definition, a general term describing
in the correlation equation yields the attachment of charged species (such as toxic metal
ions) from a solution to a coexisting solid surface.
log10 (S) = A + BTB + CTB 2 + DTB 3 Sorption kinetics may be controlled by several independent
processes that can act in series or in parallel (see Fig. 1).
= −24.0080 + 2.21196E−01 ∗ (383.78)
These processes fall in one of the following general
− 5.55632E−04 ∗ (383.78)2 categories:
+ 4.18830E−07 ∗ (383.78)3 1. bulk diffusion,
S = 524.68 ppm (wt) 2. external mass transfer (film diffusion),

Intraparticle diffusion
Reaction
Mass transfer
Figure 1. Illustration of the mechanism during sorption to
describe suitably and model the respective process kinetics.
3. chemical reaction (chemisorption), and
4. intraparticle diffusion.
Kinetic analyses allow estimating sorption rates and also
lead to suitable rate expressions characteristic of possible
reaction mechanisms (1–3).
At sufficiently high agitation speed in a reaction vessel,
the bulk diffusion step can be safely ignored because
then sorption onto sorbent particles is decoupled from
mass transfer in the bulk mixture. Apart from that,
it is quite common that more than one process can
contribute to system performance at the same time. In this
case, the extensive interrelationships between the various
equations make the overall kinetic model exceedingly
complicated to evaluate. A rather simplifying approach
to circumvent this problem is to assume that each one of
the co-current processes dominate over the others (i.e., the
rate-controlling step) at specific time regimes of the process
and then study them independently (4); this approach was
also adopted in this article.
Many studies engaged so far to examine sorption
phenomena involved analysis of batch experiments where
data were sampled at even time intervals over the entire
course of the process. As a result, fast changing kinetic
data characteristic of the phenomena just after the onset of
sorption could not be accurately depicted on an adequately
short timescale. Thus, a primary objective of this article
is to investigate the kinetic mechanism of metal sorption
on sorbent particles, putting more emphasis on samples
collected at short times after the initiation of the process,
where the major part of the adsorption occurs.
To identify the most appropriate mechanism for a
process, several models must be checked for suitability
and consistency in a broad range of system parameters.
The model selection criteria proposed by Ho et al. (5) for
sorption of pollutants in aqueous systems were used
herein as a general guideline. According to this, several
reaction-based and diffusion-based models were tested in
simulating our data. The finally chosen kinetic models are
those that fit the data closely and also represent reasonable
sorption mechanisms.
EQUILIBRIUM AND KINETIC MODELING
An overall mass balance of the sorbate across the sorbent
surface can be written as
(Co − Cst )
X qt + Cst = Co ⇔ qt = (1)
X
SORPTION KINETICS 565
where q is the specific metal uptake (mg of metal
per g of sorbent), C is the metal bulk concentration
(mg/L) and X is the sorbent feed per unit volume of
solution (g/L). Subscripts o and t denote conditions at
the beginning and any other instant of the process,
respectively. The superscript s denotes conditions at
the sorbent interface. For adequately long adsorption
times: qt ≈ qe and Cst ≈ Cse ≈ Ce , where the subscript e
denotes equilibrium conditions. The known Langmuir and
Freundlich adsorption models have usually been employed
to describe equilibrium metal uptake:
qmax bCe
qe = (2a)
1 + bCe
qe = KF C1/n
e (2b)
qmax (mg/g), b (L/mg), and KF [mg [g (mg/L)−1/n ]−1 ] and n
(dimensionless) are constants.
To compare measurements from various experiments
for a kinetics investigation, it is necessary to introduce a
dimensionless degree of conversion. Thus, by normalizing
the remaining ion concentration, Ct , with respect to some
reference value, an index of sorption is defined. Taking
advantage of the values of Ct before the onset of sorption,
Co , and for completed sorption, C∞ = Ce , the following
degree of conversion is proposed:
Co − C t
α= (3)
C o − Ce
Kinetic studies customarily use the basic conversion
rate equation:
dα
= k(T)f (α) (4)
dt
where f (α) is a conversion-dependent function and k(T) is
the reaction rate constant. Sorption in heterogeneous
liquid–solid systems has usually been considered to follow
Arrhenius kinetics, which means that the temperature
dependence of the rate constant k may be described by the
well-known expression:
k(T) = A exp(−E/RT) (5)
However, to apply the model in such systems, the
morphology of the reactive surface and the final sorption
capacity must not vary with temperature, conditions
commonly overlooked by many workers (6).
The example used in this article deals with an
investigation of the removal and depletion of cadmium
and chromates from aqueous solutions by biosorption
on Aeromonas caviae (the sorbent). This microorganism
is often present in groundwater and generally in
aquatic environments. Biosorption, the uptake of heavy
metals by dead biomass, has gained credibility during
recent years as it offers a technically feasible and
economical approach (7–10); the latter author illustrated
that adsorption is different from ion exchange. Below,
evidence is also provided that the system examined is a
complex process. More than one sorption model were often
reported to describe a case study correctly (11). Attention
should be paid to any possible experimental error.
566 SORPTION KINETICS

The Langmuir isotherm equation was often found
to fit quite satisfactorily the experimental data (12,13).
Experimental adsorption isotherms of metal cations or
oxyanions, obtained with different sorbent concentrations,
temperatures, etc. are usually presented as figures and
then, tables display the results of fitting the Langmuir
and/or the Freundlich model to the data obtained; the
values of the respective correlation coefficient (r2 ), if high,
are often used to demonstrate a satisfactory description.
The assumptions involved in these models were well
described, among others, by Volesky (9). The Langmuir
isotherms obtained during the equilibrium study lend
further support to the notion of a monolayer (14).
On the other hand, the Freundlich isotherm is heavily
used in industry. This model is often accurate for higher
concentrations, but is inaccurate at low concentrations due
to its power-law nature. The equilibrium properties do not
change proportionally with the sorbent addition, possibly
due to an aggregation of solids at higher loads and hence
reducing the effective adsorption area (15).
From the chemical reaction category (chemisorption),
the best fit for the data sets of this study is achieved
by second-order chemical reactions (16). The solution of
the standard second-order reaction, based on constant
stoichiometry of one metal ion per binding site, is (e.g., 4):
Co
Ct =
Co
1− exp(−k1 Ce t)
Ce
Equations 7 and 8 provided quite a suitable description
of data for advancing time (see Fig. 2). It is noteworthy
that both models adequately capture the rapid rate of
adsorption during the first minutes of the experiments.
This implies that the metal uptake by the sorbent is a
satisfactory rate-controlling parameter under a second-
order reaction mechanism. The numerically best fit values
of the rate parameters of these equations were displayed in
a table (16). The predicted equilibrium sorption capacities
were quite close to the experimental values for both
models. Nevertheless, the rate constant of the pseudo-
second-order model, km , is monotonically correlated with
changes in the biomass load and in the bulk concentration,
features that have been encountered in the past in
biosorption (18).
In contrast, the rate constant of the Ritchie second-
order equation, k2 , fluctuated, beyond any physical
reasoning. In addition, Equation 8 exhibited better fitting
statistics. Despite the goodness of fit for sorption at 40 and
60 ◦ C, the reaction rate constant of both models varied
randomly with temperature. Preliminary calculations,
(a) 7
6
(6)
5
4
q, mg/g

X :1 g/L
where k1 is the reaction rate constant [L*(mg−1 of
metal)*min−1 ]. This adsorption model has been very 3 X :2 g/L
effective in describing the kinetics of adsorption of gases X :0.5 g/L
on solids. Nevertheless, when the data were plotted, it 2 T :40°C
was shown that Eq. 6 clearly fails to capture the steep
concentration gradient of the early removal stage (16). T :60°C
1
This was a direct indication that adsorption on solids
from a liquid phase is a process different from adsorption 0
from a gas phase, where traditionally the remaining bulk 0 20 40 60 80 100 120 140
concentration dictates the kinetics. t, min
If the rate of sorption depends not on bulk concentration (b) 6
but on uptake by the sorbent, this can be described by the
so-called Ritchie second-order equation according to which X :1 g/L
5
one metal ion occupies two binding sites (17): X :0.5 g/L
  
1 4 X :2 g/L
qt = qe 1 − (7)
1 + k2 t T :40°C
t /q

3 T :60°C
where k2 is the reaction rate constant (min−1 ). When in the
above treatment, qe does not dictate sorbate uptake, then a
pseudo-second-order rate expression is more appropriate: 2

t 1 1 1
= + t (8)
qt km q2m qm
0
where km is the reaction rate constant [g of sorbent*(mg−1 0 20 40 60 80 100 120
of metal)*min−1 ] and qm is a numerically determined t, min
parameter which, under ideal second-order rate control, Figure 2. Comparison of experimental removal curves vs.
corresponds to qe . In the literature (5), various other theoretical predictions based on (a) the Ritchie second-order
kinetic equations have been attempted: zero, first (forward equation (at an initial cadmium concentration of 5 mg/L) and
or reversible)-order, Langmuir–Hinshelwood, and the (b) the pseudo-second-order equation (at an initial cadmium
Elovich-type. concentration of 50 mg/L). Reprinted from Reference 16.

using the Arrhenius model between two temperatures
every time, always gave activation energies below
10 kJ/mol, which is far less than expected for reaction-
controlled sorption processes. The morphological changes
in the biomass (sorbent) surface at different temperatures
and the dependence of sorption capacity on temperature
may be blamed for this irregularity.
DIFFUSION KINETICS AND MODELING
For intraparticle diffusion, Crank (19) proposed a model
that takes into account the continuously decreasing bulk
concentration due to sorbate removal. This gives rise to a
time-dependent boundary condition for the concentration
at the surface of the sorbent particle. The solution of the
diffusion equation for such a boundary condition and a
concentration independent diffusivity is
∞ where Km is the external mass transfer coefficient (m/s)
 exp(−ξ p2n t)
α =1−6 (9)
9/(1 − ) + (1 − )p2n
n=1
where pn is given by the nonzero roots of
3pn
tan(pn ) = (10)
3 + p2n /(1 − )
and  ≡ (Co − C∞ )/Co is the fraction of metal ultimately
adsorbed by the sorbent.
Equation 9 can be solved numerically to determine ξ ,
the effective diffusional time constant, which for particle
(also called micropore) diffusion control equals Dc /R2c ;
Dc and Rc are the intraparticle diffusion coefficient
(m2 /s) and mean particle radius (m), respectively. The
same expression is the solution of the diffusion equation
for (macro)pore diffusion control but only where the
equilibrium isotherm is linear (qe ≈ K*Ce ) for the
concentration range under investigation. Then ξ equals
(Dp /R2p )/[1 + (1 − εp )K/εp ], where εp Dp /[1 + (1 − εp )K] is
the effective macropore diffusivity, K is the equilibrium
constant, εp is the void volume fraction, and Rp is the
radius of the particle (14).
The experimental procedure in this article is based
on measuring the remaining metal concentration in the
bulk, so  is always above 0.5. For  greater than about
0.1, the effect of diminishing bulk concentration becomes
significant, and under these conditions, the assumption of
constant metal concentration at the surface of the sorbent
leads to an erroneously high diffusivity. Provided that
the concentration step employed in a sorption experiment
corresponds to a linear section of the equilibrium isotherm,
the constant K can be replaced by the local slope of the
isotherm, dqe /dCe . This is approximately valid for the
concentrations employed in this work, as can be seen
from the experimental isotherms. Our interest is chiefly
focused on the short time region ( Dc,p t/R2c,p < 0.2) where
Eq. 9 converges slowly, so at least 200 terms are used in
the summation to achieve satisfactory accuracy.
External mass transfer has been customarily analyzed
in literature by adopting a pseudo-first-order reaction
model (5,20). This approach assumes that the sorbate
SORPTION KINETICS 567
concentration at the sorbent surface is zero at all times.
However, this is not true, particularly where a significant
quantity of sorbate is adsorbed rapidly at the beginning
of the process. A more realistic model should consider
instead that a rapid equilibrium is established between
the sorbate at the interface and that present on the sorbent
surface (21), and this concept is adopted in this article.
If one combines the mass balance across the sorbent
surface, Equation 1, the Langmuir adsorption isotherm,
Equation 2a, and the rate of change equation in the bulk
concentration, one ends up with (21)
dCt
= −Km S(Ct − Cst ) (11)
dt
 
dCst Km S
= [(Ct − Cst )(1 + bCst )2 ] (12)
dt Xqmax b
and S is the specific surface area of the sorbent particles
per unit volume of the reactor (m2 /m3 ). Using the following
dimensionless variables: C∗ = Ct /Co , C∗s = Cst /Cso , and t∗ =
t/τ , where Cso is determined from Eqs. 1 and 2a and τ is
the total adsorption time, Eqs. 11 and 12 convert to
dC∗
= −Km Sτ (C∗ − C∗s ) (13)
dt∗
 
dC∗s Km Sτ
= (C∗ − C∗s )(1 + bCo C∗s )2 (14)
dt∗ Xqmax b
which is a system of two first-order ordinary differential
equations that must be solved simultaneously. The initial
conditions are C∗ = 1 and C∗s = 0 at t∗ = 0. Note that
normalizing Cst with respect to Cso —and not Co , as Puranik
et al. (21) did —and also t with respect to τ , improved the
stability and convergence characteristics of the solution
markedly due to comparable spreading of all variables
over the computational domain.
Equations 9–10 and 13–14 are solved numerically to
determine ξ and Km S, respectively. The nonlinear numer-
ical regression to fit experimental data to those equations
is performed by the Levenberg–Marquardt method, which
gradually shifts the search for the minimum of the sum
of the errors squared (SSE), from the steepest descent to
quadratic minimization—Gauss–Newton (22):
 (qexp,i − qcal,i )2
SSE = (15)
i
q2exp,i
Figure 3a presents the results of fitting Eq. 9 to biosorp-
tion data obtained with different initial concentrations,
solids loads, and temperatures. It is apparent that, despite
some scatter in measurements, the finite volume diffusion
model can fairly well describe the entire range of data, also
including the steep concentration gradient at short times.
The values of , ξ , and the computed values of Dc were
displayed in a table (14). Such behavior has been custom-
ary as a consequence of the decreasing slope of a nonlinear
equilibrium curve, example, a Langmuir isotherm, which
568 SORPTION KINETICS

(a) end of sorption (for α higher than 0.9), a much slower

process, intraparticle diffusion, gradually becomes the
1
rate-controlling step. Yet, this is a regime of no practical
significance.
The computed mass transfer rate constant, Km S, was
0.8
presented in a table (14). As regards the effect of changing
the solids load, it appears that Km S is virtually not
0.6 affected. To assess the statistical significance of the
determination, a derivative time series analysis has been
a

performed where dα/dt is plotted versus t and then,

0.4 Co = 5 mg/L, s = 0.5 g/L, T = 20°C a Hanning low-pass filter was applied to flatten out
the signal undulations until dα/dt versus α became
Co = 5 mg/L, s = 1 g/L, T = 20°C
a reasonably smooth curve (23). Next, the integrated
0.2 Co = 5 mg/L, s = 2 g/L, T = 20°C smoothed signal α(t) was processed as before and resulted
Diffusion model in Km S values closely comparable to the values obtained
from the unfiltered data, with no preferential trend
0 in biomass load. This is an additional positive sign
0 20 40 60 80 100 120 140
t, min
that external mass transfer may be the predominant
mechanism of the sorption process after the initial fast
(b) 1.2 chromium removal.
As a conclusion in the example thoroughly studied, we
1 may say that the analysis conducted has not been capable
of providing strong evidence in favor of any of the examined
mechanisms because several diverse kinetic models were
0.8
successful in fitting the experimental data. Perhaps,
the questioning of surface kinetics by Levenspiel (24) is
0.6 known, concluding that it is good enough to use the
a

simplest available correlating rate expression.

0.4 Co = 5 mg/L, s = 1 g/L, T = 20°C
Co = 5 mg/L, s = 1 g/L, T = 40°C NOMENCLATURE
0.2 Co = 5 mg/L, s = 1 g/L, T = 60°C
External mass transfer q: specific metal uptake (mg/g)
0 C: metal bulk concentration (mg/L)
0 0.2 0.4 0.6 0.8 1 1.2 X: sorbent loading per unit volume of solution (g/L)
t qmax :
Langmuir constant (mg/g)
Figure 3. Experimental degree of conversion, α, vs. predictions b: Langmuir constant (L/mg)
based on the solution of (a) the diffusion equation, for various KF : Freundlich constant {mg [g (mg/L)−1/n ]−1 }
adsorbent loads and (b) the mass transfer equation, for various n: Freundlich constant (dimensionless)
temperatures (both at an initial chromium concentration of E: apparent activation energy (J/mol)
5 mg/L). Reprinted with permission from Reference 14; copyright
A: preexponential factor (min−1 )
(2004) American Chemical Society.
n: apparent reaction order
R: gas constant (8.3136 J/mol/K)
T: temperature (K)
causes the diffusivity to increase rapidly as concentration
t: time (min)
increases (11).
k: rate constant
A way to check further on the possibility of a pore
k1 : second-order chemical reaction rate constant
diffusion-controlled mechanism is to perform desorption
[L∗ (mg−1 of metal)∗ min−1 ]
kinetic tests with adsorbent previously used for sorption.
k2 : second-order Ritchie reaction rate constant (min−1 )
For a nonporous sorbent particle, transport of solute inside
km : pseudo-second-order reaction rate constant [g of
the particle may be neglected, and it can be assumed that
sorbent∗ (mg−1 of metal)∗ min−1 ]
biosorption occurs mainly at the particle surface. This
Dc : intraparticle diffusion coefficient (m2 /s)
idea can be effectively extended to cases of relatively large
Rc : mean particle radius (m)
macropores, where the metal ions may have ready access
Rp : radius of a particle
to react with internal surface sites.
Km : external mass transfer coefficient (m/s)
It was argued that, at the beginning of the process,
S: specific surface area of sorbent particles per unit
chromium is sorbed according to quite a fast and highly
volume of the reactor (m2 /m3 )
favorable chemical mechanism, such as ion exchange, but
soon external film diffusion comes into play (14). On this
Greek Letters
account, if one ignores the very first minute of sorption,
the remaining curves were fitted pretty well by the model α: degree of conversion
Eqs. 13 and 14. Figure 3b displays these results. Near the : fraction of metal ultimately adsorbed by the sorbent

ξ: numerically determined parameter
εp : void volume fraction
BIBLIOGRAPHY
1. Tien, C. (1994). Adsorption Calculations and Modelling.
Butterworth-Heinemann, Boston.
2. Yiacoumi, S. and Tien, C. (1995). Kinetics of Metal Ion
Adsorption from Aqueous Solutions. Kluwer Academic,
Boston.
2+
3. Xu, Y. and Schwartz, F.W. (1994). Sorption of Zn and
Cd2+ on hydroxyapatite surfaces. Environ. Sci. Technol. 28:
1472–1480.
4. Hill, C.G. (1977). An Introduction to Chemical Engineering
Kinetics and Reactor Design. John Wiley & Sons, New York,
pp. 24–166.
5. Ho, Y.S., Ng, J.C.Y., and McKay, G. (2000). Kinetics of
pollutant sorption by biosorbents: review. Sep. Purific.
Methods 29: 189–232.
6. Smith, J.M. (1981). Chemical Engineering Kinetics, 3rd Edn.
McGraw-Hill, Singapore, pp. 37–92, 636–645.
7. Aksu, Z. (2001). Equilibrium and kinetic modeling of
cadmium(II) biosorption by C. vulgaris in a batch system:
Effect of temperature. Sep. Purific. Tech. 21: 285–294.
8. Butter, T.J., Evison, L.M., Hancock, I.C., Holland, F.C.,
Matis, K.A., Philipson, A., Sheikh, A.I., and Zouboulis, A.I.
(1998). The removal and recovery of cadmium from dilute
aqueous solutions by biosorption and electrolysis at labora-
tory scale. Water Res. 32: 400–406.
9. Volesky, B. (1999). Biosorption of Heavy Metals. CRC Press,
Boca Raton, FL, Chap. 6. Available: http://ww2.mcgill.ca/
biosorption/publication/index.htm.
10. Zouboulis, A.I., Matis, K.A., and Hancock, I.C. (1997). Bio-
sorption of metals from dilute aqueous solutions. Sep. Purif.
Methods 26: 255–295.
11. Smith, E.H. (1996). Uptake of heavy metals in batch
systems by a recycled iron-bearing material. Water Res. 30:
2424–2434.
12. Mandjiny, S., Zouboulis, A.I., and Matis, K.A. (1995). Re-
moval of cadmium from dilute solutions by hydroxyapatite.
Part I. Sorption studies. Sep. Sci. Tech. 30: 2963–2978.
13. Kefala, M.I., Zouboulis, A.I., and Matis, K.A. (1999). Biosorp-
tion of cadmium ions by Actinomycetes and separation by
flotation. Environ. Pollut. 104: 283–293.
14. Loukidou, M.X., Karapantsios, Th.D., Zouboulis, A.I., and
Matis, K.A. (2004). Diffusion kinetic study of chromium(VI)
biosorption by Aeromonas caviae. Ind. Eng. Chem. Res. 43:
1748–1755.
15. Esposito, A., Pagnanelli, F., Lodi, A., Solisio, C., and Vegliò,
F. (2001). Biosorption of heavy metals by Sphaerotilus
natans: An equilibrium study at different pH and biomass
concentration. Hydrometallurgy 60: 129–141.
16. Loukidou, M.X., Karapantsios, Th.D., Zouboulis, A.I., and
Matis, K.A. (2003). Cadmium(II) biosorption by Aeromonas
caviae: Kinetic modelling. 15th Intl. Biohydrometallurgy
Symp., Athens.
17. Ritchie, A.G. (1977). Alternative to the Elovich equation for
the kinetics of adsorption of gases on solids. J. Chem. Soc.-
Farad. Trans. 73: 1650–1653.
18. Cheung, C.W., Porter, J.F., and McKay, G. (2001). Sorption
kinetic analysis for the removal of cadmium ions from
effluents using bone char. Water Res. 35: 605–612.
SOUND IN WATER 569
19. Crank, J. (1975). Mathematics of Diffusion. Oxford University
Press, London.
20. Dzul Erosa, M.S., Saucedo, T.I., Navarro Mendoza, R., Avila
Rodriquez, M., and Guibal, E. (2001). Cadmium sorption
on chitosan sorbents: Kinetic and equilibrium studies.
Hydrometallurgy 61: 157–167.
21. Puranik, P.R., Modak, J.M., and Paknikar, K.M. (1999). A
comparative study of the mass transfer kinetics of metal
biosorption by microbial biomass. Hydrometallurgy 52:
189–197.
22. Bates, D.M. and Watts, D.G. (1988). Nonlinear Regression
Analysis and its Applications. John Wiley & Sons, New York.
23. Bendar, J.S. and Piersol, A.G. (1986). Random Data: Analysis
and Measurement Procedures. John Wiley & Sons, New York.
24. Levenspiel, O. (1999). Chemical Reaction Engineering, 3rd
Edn. John Wiley & Sons, New York, p. 380.
SOUND IN WATER
D.L. MARRIN
Hanalei, Hawaii
As sound travels through an elastic medium such as water,
a wave is created that imparts energy to the individual
molecules—causing them to compress and then to relax
along the path of the wave. Because the medium is elastic,
the distance separating individual water molecules is
permitted to vary just slightly, so that the propagation
of sound may be defined as a periodic variation in
pressure that is transmitted via adjacent molecules. Sound
emanating from a point source in water usually (but not
always) results in a three-dimensional waveform that
spreads spherically through the medium (1). Similarly,
a planar source of sound in water can be considered
multiple point sources that create a wave consisting of
all the individual waves.
Sound waves are longitudinal, meaning that molecules
(e.g., water or air) comprising the elastic medium move
in a direction parallel to the propagating wave. Because
water has a significantly higher density than air, water
transmits sound more rapidly and efficiently than air. The
water molecules oscillate, or move back and forth over
a very short longitudinal distance; hence, sound waves
are classified as mechanical. By contrast, light waves and
microwaves are classified as electromagnetic and do not
involve the movement of water molecules for propagation.
The speed of sound in water is about 1500 meters per
second or 3300 miles per hour (i.e., approximately five
times greater than that in air), which varies slightly with
temperature, pressure, and salinity.
FREQUENCY AND AMPLITUDE
The frequency of a sound wave is the number of waves
that pass by a given point during a specified period of
time. A single wave consists of a cycle that includes
both compression (squeezing together) and rarefaction
(rebounding apart); the distance over which a cycle is
continually repeated is known as the wavelength. The
570 SOUND IN WATER
period of time over which waves are normally counted is
one second; therefore, sound frequencies are expressed as
cycles per second or hertz. Whereas high-frequency sounds
possess short wavelengths, low-frequency sounds possess
long wavelengths. Frequency ranges and estimated
intensity levels for common underwater sounds are
listed in Table 1. Frequency-dependent differences in
underwater sound attenuation result in low-pitched
sounds that travel farther than high-pitched sounds
generated at similar intensities.
The amplitude of a sound wave is the maximum
longitudinal displacement (i.e., during one cycle of
compression and rarefaction) of water molecules relative
to their resting, or equilibrium, position. The intensity
of a sound wave is proportional to its amplitude and is
often quantified in units of decibels. The decibel scale is
based on a ratio between the intensities (expressed as
a power) of a measured sound and a reference sound
that corresponds to a standard pressure. Many of the
man-made sounds listed in Table 1 are actually more
intense than naturally produced sounds, contributing, at
least in part, to the controversy regarding the possible
effects of anthropogenic noise on marine organisms. Sound
intensities are commonly presented for a distance of only 1
meter from the source and, as such, are rarely indicative of
ambient underwater noise (e.g., the cumulative intensity
of sounds measured at greater distances from their
respective sources).
FATE OF OCEANIC SOUND
The propagation of sounds in the ocean is a rather com-
plex matter owing to differences in water density as a
function of depth, latitude, seafloor topography, and many
other factors. Changes in seawater temperature, salinity,
and pressure contribute to changes in density that, in
turn, affect sound waves traveling through the medium.
Generally, sound is attenuated by spreading (proportional
to distance), reflection (due to solid structures or bound-
aries between water of different densities), scattering (due
to rough surfaces), absorption (conversion of acoustic to
thermal energy), and refraction (deflection of sound waves
Table 1. Approximate Ranges for the Frequencies and Intensity Levels of Common Underwater Sounds. Reported Ranges
were Complied from Data Presented in Various Sourcesa
Sound Source
Ship engines and propellers
Navigation and profiling sonars
Explosive devices/air guns
Military surveillance sonars
Icebreaking/drilling operations
Whale songs and moans
Dolphin clicks and whistles
Cetacean echolocation
Snapping shrimp colony
Lightning strikes/undersea volcanoes/earthquakes
a
Reference 2–4.
b
Intensity levels are presented in units of decibels relative to a reference pressure of 1 micropascal at a distance of 1 meter from the sound source. Sound
level attenuation is a function of distance from the source, frequency range, and various environmental factors.
from a straight path) (5). Ambient noise in a shipping
channel can exceed that outside a channel by as much as
45 decibels (re 1 µPa), but noise in a channel is attenuated
about 100 decibels from that measured directly beneath
ships (2).
Due to the vertical stratification of oceans, sound
behaves differently in turbulent shallow waters from
the way it does in either more isothermal middepth
waters or waters immediately overlying the bottom (see
Table 2). Sound transmission at the surface is highly
dependent on local conditions (e.g., wind, precipitation,
swells, bubbles) and is usually restricted to low-frequency
sounds. However, most of the ocean lies between shallow
and bottom waters, creating a relatively thick layer that
transmits a spectrum of sound waves that is far less
attenuated than that in either shallower or deeper waters.
Both humans and whales have taken advantage of this
sofar channel to broadcast signals and songs, respectively,
over vast oceanic distances (6). Sounds generated within
the sofar channel itself or at shallower depths alongside
seamounts are entrained and transmitted within this
deep isothermal layer as a result of many contributing
physical processes.
UNDERWATER DETECTION
Detecting and identifying an underwater sound from a
particular source is often difficult due to the aforemen-
tioned effects of the environment and to the myriad of
background noises from which any single sound must be
distinguished. Sea sounds are detected by hydrophones
(i.e., the underwater equivalent of a microphone) that are
often positioned in specific arrays, thus permitting the rel-
ative position of a sound source to be determined. This type
of passive detection may be used to locate oceanic precip-
itation, breaking surface waves, underwater explosions,
undersea volcanoes, shipping traffic, and marine organ-
isms that emit audible sounds (e.g., crustaceans, fishes,
mammals) (7).
Active detection techniques, such as sonar, are routinely
used to locate underwater objects because the objects
themselves need not emit sound and because the results
Intensity Level, Decibels
Frequency, Hertz re 1 µPa at 1 mb
10–5000 160–190
100–3000 180–230
1000–17,000 190–260
1000–10,000 190–235
20–1000 100–150
10–8000 120–190
500–25,000 100–180
10,000–150,000 130–230
2000–15,000 180–190
0.1–20,000 Up to 260

Table 2. Approximate Depths and Representative Sound Speeds for Various Water Layers in the Ocean. Reported Ranges
were Complied from Data Presented in Various Sourcesa
Depth, Speed,
Vertical Zone Meters m/s
Overlying air NAb 330
Surface water layer 0 to 50 1515
Seasonal thermal layer 50 to 800 1485–1505
Deep isothermal layer 800 to 4000+c 1485–1525
Bottom water layer 4000+ to seafloorc 1525
Underlying rock NAb 2000–6000
a
References 2, 6, and 7.
b
Not applicable.
c
Indicates that the transition between deep isothermal layers and bottom water layers may exceed 4000 meters, depending on the exact location within an
ocean basin.
are generally more interpretable. The most basic type of
sonar is echolocation, whereby a sound pulse (acoustic
energy) is introduced into the water. Depending on the
sonic wavelength and the size of the objects, a portion of
that acoustic energy is reflected back to the source (e.g.,
a ship) where it is detected, transduced, and analyzed.
The distance, movement, size, and even shape of some
objects may be determined using sophisticated sonars (e.g.,
Doppler, multibeam) and data analysis techniques (8).
Several marine organisms—most notably cetaceans—are
believed to use echolocation as a means of finding or
assessing prey, predators, conspecifics, seabed features,
and environmental conditions (2).
ULTRASOUND AND INFRASOUND
Most of the underwater sounds discussed thus far lie
within the sonic range of humans (i.e., frequencies of 20
to 20,000 hertz), but water also transmits mechanical
waves of ultrasonic (>20,000 hertz) and infrasonic (<20
hertz) frequencies. Dolphins emit whistles at frequencies
as high as 30,000 hertz and complex echolocation clicks
at frequencies exceeding 300,000 hertz (8). Obviously,
the upper limit of hearing of dolphins far exceeds that
of humans. In addition to marine organisms, ambient
oceanic noise associated with bubbles (i.e., trapped gases),
surface winds, waves, and sea spray can reach ultrasonic
frequencies. Man-made ultrasounds include everything
from complex signals for communication to guided waves
for inspecting underwater pipelines (9).
The moans of baleen whales are the most common
source of biologically produced infrasound, which has been
documented down to about 10 hertz (8). By contrast,
a number of geologic and meteorologic events (e.g.,
volcanoes, earthquakes, sea ice cracking, hurricanes)
produce infrasonic waves that propagate through the
ocean at frequencies of approximately 0.1 to 10 hertz.
Even the breaking of large surf, either onto the shore or
SOUND IN WATER 571
Comments
Significantly lower acoustic impedance than water; minimal
transfer of sound energy
Influenced by weather (e.g., wind, rain, temp.) and turbulence
that affect sound waves
Sound channel (mostly for low frequencies); limited vertical
thickness; refracts sound waves originating shallower and
deeper
Optimal sound channel (for all frequencies); extensive vertical
thickness and horizontal extent; minimal sound attenuation
Influenced by seafloor topographic features that reflect and
scatter sound waves
Higher acoustic impedance than water; moderate transfer of
sound energy
on the water itself, creates infrasound at frequencies of
1 to 5 hertz (10). Anthropogenic infrasound in the ocean
is created primarily via shipping traffic, low-frequency
active sonar (LFAS), and acoustic thermography. A
thermography technique known as ATOC generates tones
down to about 1 hertz, whereas LFAS sweeps multiple
tones of 100 to 500 hertz across each other and produces
resulting tones as low as 0.1 hertz (2). Both ATOC and
LFAS generate peak intensity sounds that are greater
than those of ships.
BIBLIOGRAPHY
1. Leighton, T.G. (1994). The Acoustic Bubble. Academic, San
Diego, Chap. 1.
2. Stocker, M. (2003). Ocean bio-acoustics and noise pollution.
J. Acoust. Ecol. 4: 16–29.
3. Science of Sound in the Sea. (2004). Office of Marine Programs.
University of Rhode Island, HTML format.
4. Greene, C.R. and Moore, S.E. (1995). Man-made noise. In:
Marine Mammals and Noise. W.J. Richardson et al. (Eds.).
Academic, San Diego, Chap. 6.
5. Haines, G. (1974). Sound Underwater. Crane Russak, New
York, pp. 22–25.
6. Payne, R. (1995). Among Whales. Scribner, New York,
Appendix.
7. Medwin, H. and Clay, C.S. (1998). Fundamentals of Acousti-
cal Oceanography. Academic, San Diego, Chap. 1 and 10.
8. Thompson, D.H. and Richardson, W.J. (1995). Marine mam-
mal sounds. In: Marine Mammals and Noise. W.J. Richardson
et al. (Eds.). Academic, San Diego, Chap. 7.
9. Na, W.B. and Kundu, T. (2002). Underwater pipeline inspec-
tion using guided waves. J. Pressure Vessel Technol. 124:
196–200.
10. Garcés, M., Hetzer, C., Marrifield, M., Willis, M., and
Aucan, J. (2003). Observations of surf infrasound in Hawai’i.
Geophys. Res. Lett. 30: 2264–2266.
572 WATER ON THE SPACE STATION
WATER ON THE SPACE STATION
PATRICK L. BARRY
TONY PHILLIPS
NASA
Rationing and recycling will be an essential part of
life on the International Space Station. In this article,
Science@NASA explores where the crew will get their
water and how they will (re)use it.
November 2, 2000—Future astronauts poised to blast
off for an extended stay on the International Space Station
(ISS) might first consider dashing to the restroom for a
quick splash at the lavatory, or better yet, a luxurious
hot shower. Once on board the ISS, spacefarers are in
for a steady diet of sponge baths using water distilled
from—among other places—their crewmates breath!
If you’re squeamish, read no farther, because the
crew will eventually include lab rodents—and they’ll be
breathing, too. All of the denizens of the space station lose
water when they exhale or sweat. Such vapors add to the
ambient cabin humidity, which is eventually condensed
and returned to the general water supply.
Sometimes it’s better not to think about where your
next glass of water is coming from!
Rationing and recycling will be an essential part of
daily life on the ISS. In orbit, where Earth’s natural life
support system is missing, the Space Station itself has to
provide abundant power, clean water, and breathable air
at the right temperature and humidity—24 hours a day,
7 days a week, indefinitely. Nothing can go to waste.
In this article, the first of a series about the practical
challenges of living in space, Science@NASA will examine
how the Space Station’s Environmental Control and Life
Support System (ECLSS), under continuing development
at the Marshall Space Flight Center, will help astronauts
use and re-use their precious supplies of water. Future
This article is a US Government work and, as such, is in the
public domain in the United States of America.
installments will explore air management, thermal control
and fire suppression—in short, all of the things that will
make the Space Station comfortable and safe.
MAKING A SPLASH IN SPACE
Before recycling can begin, there has to be some water to
start with.
‘‘We have plenty of water on the Space Station now,’’
says Jim Reuter, leader of the ECLSS group at the
Marshall Space Flight Center. ‘‘The Russian module Zarya
is packed with contingency water containers (CWCs) that
were carried over from the Space Shuttle during assembly
missions earlier this year. They look like duffle bags and
each one holds about 90 lbs.’’
‘‘But it’s expensive to ferry water from Earth,’’ he added.
‘‘We have to recycle. There’s already a Russian-built water
processor in orbit that collects humidity from the air. Here
at Marshall we’re building a regenerative system that will
be able to recycle almost every drop of water on the station
and support a crew of seven with minimal resupplies.’’
The ECLSS Water Recycling System (WRS), developed
at the MSFC, will reclaim waste waters from the Space
Shuttle’s fuel cells, from urine, from oral hygiene and
hand washing, and by condensing humidity from the air.
Without such careful recycling 40,000 pounds per year
of water from Earth would be required to resupply a
minimum of four crewmembers for the life of the station.
Not even research animals are excused from
the program.
‘‘Lab animals on the ISS breath and urinate, too, and
we plan to reclaim their waste products along with the
Shuttle pilot Terry Wilcutt with 7 contingency water containers
destined for the space station Mir

crew’s. A full complement of 72 rats would equal about one
human in terms of water reclamation,’’ says Layne Carter,
a water-processing specialist at the MSFC.
It might sound disgusting, but water leaving the space
station’s purification machines will be cleaner than what
most of us drink on Earth.
‘‘The water that we generate is much cleaner than
anything you’ll ever get out of any tap in the United
States,’’ says Carter. ‘‘We certainly do a much more
aggressive treatment process (than municipal waste water
treatment plants). We have practically ultra-pure water
by the time our water’s finished.’’
MIMICKING MOTHER EARTH
On Earth, water that passes through animals’ bodies is
made fresh again by natural processes. Microbes in the soil
break down urea and convert it to a form that plants can
absorb and use to build new plant tissue. The granular
soil also acts as a physical filter. Bits of clay cling to
nutrients in urine electrostatically, purifying the water
and providing nutrients for plants.
Water excreted by animals also evaporates into the
atmosphere and rains back down to the Earth as fresh
water—a natural form of distillation.
Water purification machines on the ISS partly mimic
these processes, but they do not rely on microbes or any
other living things.
‘‘While you try to mimic what’s happening on
Earth—which is so complicated if you really think
about it—we have to use systems that we can control
100 percent,’’ said Monsi Roman, chief microbiologist
When water evaporates from the ocean and surface waters, it leaves behind impurities. In the
absence of air pollution, nearly pure water falls back to the ground as precipitation
WATER ON THE SPACE STATION 573
for the ECLSS project at MSFC. ECLSS depends on
machines—not microbes—because, ‘‘if a machine breaks,
you can fix it.’’
The water purification machines on the ISS will cleanse
wastewater in a three-step process.
The first step is a filter that removes particles and
debris. Then the water passes through the ‘‘multi-filtration
beds,’’ which contain substances that remove organic and
inorganic impurities. And finally, the ‘‘catalytic oxidation
reactor’’ removes volatile organic compounds and kills
bacteria and viruses.
EVERY DROP COUNTS
Once the water is purified, astronauts will do everything
possible to use it efficiently. ‘‘On the ground, people flick
on the faucet and they probably waste a couple of liters
of water just because it’s free and the water pressure is
high,’’ notes Carter.
‘‘On the ISS, the water pressure will be about half what
you might experience in a typical household,’’ Carter said.
‘‘We don’t use faucets on the ISS, we use a wash cloth. It’s
much more efficient. If you’re an astronaut, you’ll wet the
wash cloth with a spray nozzle and then use the cloth to
wash your hands.’’
On the space station, people will wash their hands with
less than one-tenth the water that people typically use on
Earth. Instead of consuming 50 liters to take a shower,
which is typical on Earth, denizens of the ISS will use less
than 4 liters to bathe.
Even with intense conservation and recycling efforts,
the Space Station will gradually lose water because of
inefficiencies in the life support system.
574 STRONTIUM ISOTOPES IN WATER AND ROCK
‘‘We will always need resupply, because none of the
water reprocessing technology that is available right now
for space flight . . . is 100 percent efficient. So there’s
always some minimal loss,’’ said Marybeth Edeen, deputy
assistant manager of environmental control and life
support at NASA’s Johnson Space Center.
Water is lost by the Space Station in several ways:
the water recycling systems produce a small amount of
unusable brine; the oxygen-generating system consumes
water; air that’s lost in the air locks takes humidity with
it; and the CO2 removal systems leach some water out of
the air, to name a few.
Lost water will be replaced by carrying it over from the
Shuttle or from the Russian Progress rocket. The Shuttle
produces water as its fuel cells combine hydrogen and
oxygen to create electricity, and the Progress rocket can
be outfitted to carry large containers of water.
NASA scientists will continue to look for ways to
improve the life support systems of the Space Station,
reducing water losses and finding ways to reuse other
waste products. If the water recycling systems can be
improved to an efficiency of greater than about 95 percent,
then the water contained in the Station’s food supply
would be enough to replace the lost water, Edeen said.
One of the ‘‘nodes’’ that will become a part of the Space Station.
The ECLSS life support equipment will be housed in Node 3,
which is scheduled to be attached to the station in October 2005
Sign up for our EXPRESS SCIENCE NEWS delivery
‘‘It takes processes that are slightly more efficient than
we have developed for the space station to do that,’’ Edeen
said. ‘‘Those are the next generation water processing
systems. Those are being developed now, but they’re not
ready for space flight yet.’’
The ECLSS life support system will join the Space
Station as part of Node 3, which is scheduled to launch in
October 2005. Until then, the environment inside the ISS
will be maintained primarily by life support systems on
the Russian Zvezda Service Module.
WEB LINKS
International Space Station—NASA’s Web page for the Interna-
tional Space Station
Wheels in the Sky—Science@NASA article about humanity’s
dreams of a space station from the science fiction fantasies
of the Nineteenth Century to Wernher von Braun’s catalytic
vision in the 1950s
Advanced Life Support Web Page—from the Johnson Space
Flight Center
Environmental Control and Life Support Systems—describes
the life support systems being developed at Marshall Space
Flight Center
STRONTIUM ISOTOPES IN WATER AND ROCK
GHOLAM A. KAZEMI
Shahrood University of
Technology
Shahrood, Iran
INTRODUCTION
A brief summary of strontium and strontium isotopes,
their concentrations in natural materials, and their
evolution in the geologic system is given in the following
paragraph. These introductory notes have been taken
from a benchmark paper by Capo et al. (1). Anyone
intending to deal with strontium isotopes should consult
that paper as well as the paper by Clow et al., (2).
Textbooks such as those by Faure (3) and Faure (4) provide
excellent coverage of strontium isotopes in rocks. Being
an encyclopedia of water, the emphasis here is more on
strontium isotopes in water.
Strontium has four naturally occurring isotopes, whose
approximate abundances are 84 Sr, 0.56%; 86 Sr, 9.87%;
87
Sr, 7.04%; and 88 Sr, 82.53%. All four isotopes are
stable, although the fraction of 87 Sr varies due to
radioactive decay of 87 Rb. The ionic radius of strontium
(1.18 Å) is similar to that of calcium (1.00 Å), and,
hence, Sr substitutes for Ca in minerals, including
plagioclase feldspar, apatite, sulfates such as gypsum
and anhydrite, and carbonates (calcite, dolomite, and
especially aragonite). The average concentration of Sr in
soils is 240 ppm but may fall below 10 ppm or exceed
1000 ppm in some cases. Weathering of bed rock or
sediments can be a significant source of strontium in
soil. Soil parent materials often have distinct Sr isotopic
signatures. The average concentration of Sr in crustal
rocks is 370 ppm but can vary from 1 ppm in some

ultramafic rocks to several percent in some aragonitic
corals. If Rb and Sr are incorporated into a mineral
or rock at its formation and the system remains closed
with respect to those elements, then the amount of
87
Sr increases over time as radioactive 87 Rb decays; the
amounts of 84 Sr, 86 Sr, and 88 Sr remain constant. Therefore,
older rocks in general have higher 87 Sr/86 Sr ratios than
younger ones that have the same initial Rb/Sr ratio.
Over geologic time, rocks of a given age, composed of
minerals that have a high Rb/Sr ratio (e.g., granites in
the continental crust), develop a higher 87 Sr/86 Sr than
rocks that have a lower Rb/Sr ratio (e.g., oceanic basalts).
Thus, 87 Sr/86 Sr ratios in geologic materials are indicators
of both age and geochemical origin. The strontium in rocks
is released by weathering, cycled through vegetation and
animals, and eventually enters the oceans, primarily via
rivers. Minerals within a crystalline rock or sediment
generally have variable Rb/Sr ratios and therefore a range
of 87 Sr/86 Sr ratios. River water has a Sr concentration that
varies from ∼6 to 800 ppb and averages ∼60 ppb. The Sr
concentration in seawater is about 8 ppm. Precipitation
near ocean basins generally has a strontium isotopic
composition similar to that of seawater (0.709). However,
the concentration of Sr in rainfall is generally lower
than that in seawater by several orders of magnitude
(usually <1 ppb). Strontium leaves the oceans, the largest
reservoir of dissolved Sr, primarily by deposition in marine
carbonate. A small amount of Sr is also transferred directly
from the oceans to the atmosphere and hence transferred
to the continents in precipitation.
SR ISOTOPES IN WATER AND ROCK: SOURCES
AND IMPORTANCE
The Sr isotopic ratio in natural water reflects the
contribution made by various minerals as they dissolve
in water. Rivers are the major provider of strontium
to the oceans and define, to a great extent, the marine
87
Sr/86 Sr ratio. Strontium in rivers is mainly derived from
two end members: limestone and evaporites as one end
member (high Sr and low 87 Sr/86 Sr ratio of 0.706–0.709)
and silicate rocks as the other end member that are low
in Sr and have a higher 87 Sr/86 Sr ratio. Processes such as
atmospheric deposition, cation exchange, and biological
cycling, too, influence the Sr isotope ratio in water.
However, unless there have been recent anthropogenic
or natural disturbances, annual gains or losses of Sr from
cation exchange and biological pool should be minimal.
Sr isotopes in water samples are used (1) as a
chemical tracer to determine the sources of atmospheric
contributions to soil cation reservoirs over time (1);
(2) to determine the contribution of various components
to the water balance of a lake (5); (3) to distinguish
weathering reactions from cation exchange processes (6);
(4) to investigate the origin of brine, brine migration, and
the relationship between brine and diagenetic phases (7);
(5) to characterize weathering reactions and identify
sources of dissolved Ca in stream water (2); (6) as a tracer
of cation sources in stream water (8); (7) to correlate
and date marine sequences of late Neogene age (9);
(8) to identify the origin of salts in groundwater (10);
STRONTIUM ISOTOPES IN WATER AND ROCK 575
(9) to investigate the effect of carbonate dissolution
on the chemical composition of groundwater (11); and
(10) to indicate the influence of seawater on fresh
groundwater (12).
Analysis of Sr Isotopes in a water sample
The strontium isotopes of a water sample are measured
under clean laboratory conditions using a mass spectrom-
eter, such as a Sector 54 type (13), a Finnigan Mat 261
multicollector model (2), or a VG354 thermal ionization
brand (9). To analyze water for a Sr isotope ratio using a
Sector 54 Type Mass Spectrometer, an initial aliquot of
the sample is placed in a microwave oven to be evaporated
to dryness. For water that has a total Sr content of more
than 0.5 ppm, the aliquot is 2 mL, and evaporation time
is about 1 hour; for water that has low concentrations
such as rainwater, 500 mL is required which takes much
longer to evaporate to dryness. Then 3 mL of 6 molar HCl
is added to the residue to dissolve it and left to stabilize
for a few hours. The following steps are then taken one
after another (13):
1. Add 1 mL of 2.5 molar HCl, and evaporate to
dryness by heating slowly.
2. Add 1 mL of 6 molar HCl, and leave to dissolve.
3. Transfer to centrifuge vessels, and centrifuge for
around 5 minutes.
4. Pour 1 mL of the sample into the column (the
column detail is AG W50-X 8).
5. Add 1 mL of 2.5 molar HCl to the column.
6. Add 1 mL of 2.5 molar HCl to the column.
7. add 1 +30 mL of 2.5 molar HCl to the column in
two stages, respectively.
8. Add 10 mL of 2.5 molar HCl to the column.
9. Collect Sr by placing a container beneath the
column and letting it drain.
10. Add 1 mL of HNO3 , and slowly heat to dryness.
11. Add 1 mL of HNO3 , and slowly heat to dryness.
12. Add 1 mL of Milli-Q water to the sample.
13. Place a small volume of the sample in the
holding filament.
14. Heat sample slowly.
15. Add a small volume of phosphoric acid to the sample
to facilitate ionization.
16. Heat the sample rapidly to a very high temperature
(to burn) to remove any Rb.
17. Place the sample in the sample holder.
18. Transfer sample/s to the mass spectrometer.
19. Start running sample with mass spectrometer.
Analysis of Sr Isotopes in a rock sample
Rocks Other Than Limestone. To analyze a rock sample
for Sr isotopes ratio using a Sector 54 Type Mass
Spectrometer under clean laboratory condition, first it
has to be ground to a 1 micron size. After grinding, for
all rock types but limestone, the following steps should be
undertaken one after another:
576 STRONTIUM ISOTOPES IN WATER AND ROCK
1. 0.1 gram of sample to accuracy of 0.0001 gram
is selected.
2. 2 mL of HF acid and 10 drops of HNO3 are added
to each sample.
3. Samples are heated to dryness.
4. 2 mL of HF is again added to each sample.
5. 10 drops of HClO4 are added to each sample, and
sample is placed on a hot plate at 150 ◦ C.
6. 5 drops of HClO4 are added to each sample.
7. 5 more drops of HClO4 are added to each sample.
8. Samples are placed on a hot plate at 180 ◦ C.
9. Samples are then placed on a hot plate at 130 ◦ C
on the next day, and 1 mL 2.5 M HCl is added to
each sample.
10. Samples are transferred to centrifuge tube to be
centrifuged for 5 minutes.
11. From this step forward, all the steps are exactly
the same as those for water samples from Step
5 onward.
Limestone. Due to its high solubility compared to other
rock types, limestone is prepared for analysis using
different techniques and the following steps:
1. 0.1 mg of sample is selected.
2. 1 mL of 1 M acetic acid is added to the sample.
3. 2 mL of 1 M acetic acid is added.
4. After reaction subsides, sample is centrifuged, and
the residue is decanted to a clean beaker and heated
to dryness at 150 ◦ C.
5. 3 mL of 6.25 M HCl is added and evaporated to
dryness, and 1 mL more is added again and heated
to dryness.
6. 1 mL of 2.5 M HCl is added.
7. The sample is then loaded into columns (like other
rock types).
8. The Sr is then collected and dried.
9. The columns are flushed and cleaned.
10. The dry sample (Sr) is again dissolved with 1 mL
of 2.5 M HCl and loaded into the columns again.
In contrast to the other samples, carbonates have
a higher calcium content. Therefore, they need to
be cleaned up twice before they are loaded into the
mass spectrometer.
11. The rest of the procedure is like that for other
rock types.
REPORTING THE RESULTS OF SR ISOTOPIC ANALYSIS
Before reporting the results of the strontium isotopic anal-
ysis, they should first be corrected for mass fractionation
to 88 Sr/86 Sr = 8.37521. Also, the value of isotopic ratio
(88 Sr/86 Sr) of the NBS (National Bureau of Standards) 987
Sr metal used as a reference by the spectrometer should be
included in the report. Measured strontium isotopic val-
ues are expressed as both ratios (ratio of 87 Sr to 86 Sr) and
standard delta notation (δ 87 Sr) relative to NBS standard
987. To calculate δ 87 Sr, the following formula is used:
 87 
( Sr/86 Sr)sample
δ 87 Sr = 87 − 1 × 105
( Sr/86 Sr)seawater
where (87 Sr/86 Sr)sample is the measured ratio of 87 Sr/86 Sr in
a sample that has been normalized to 86 Sr/88 Sr = 0.1194
and (87 Sr/86 Sr)seawater is the ratio of 87 Sr/86 Sr in modern
seawater which is about 0.709172 (9). In most analysis,
the precision of the 87 Sr/86 Sr ratios is ±0.0001 or better,
and higher precision is of little value because seasonal
variations in 87 Sr/86 Sr ratios in rivers are commonly in the
third figure (14).
In a majority of cases, the Sr isotopic composition of
a water or rock sample is a mixture of two or more end
members. To calculate the relative proportion of each
end member of a two component mixture, the following
equation applies:
(87 Sr/total Sr)m = x(87 Sr/total Sr)a + (1 − x)(87 Sr/total Sr)b
(1)
where the subscript ‘‘m’’ is mixture, ‘‘a’’ is the first
end member, ‘‘b’’ is the second end member, x is the
proportion of end member a, and (1−x) is the proportion
of end member b. To solve this equation, 87 Sr/total Sr can
be calculated from the equation given by (Reference 2)
as follows:
87
Sr/total Sr = (87 Sr/86 Sr)/(3 +87 Sr/86 Sr) (2)
Example (from Reference 13). In Table 1, the isotopic
composition of groundwater (GW) is considered a mixture
of soil and rainwater isotopic compositions. The relative
proportion of Sr from rainfall and soil to groundwater
in Piezometer 40, P40, is computed here. In Equation 1,
‘‘m’’ = groundwater; ‘‘a’’ = rainfall; and ‘‘b’’ = soil; x =
proportion of end member a, rainfall; and 1−x = proportion
of end member b, soil. To calculate the proportion of
each end member, first the 87 Sr/total Sr ratio (Eq. 2) must
be computed:
87
Sr/total Sr for rainfall = (0.7089)/(3 + 0.7089)
= 0.191135
87 total
Sr/ Sr for soil = (0.72068)/(3 + 0.72068)
= 0.19368
87 total
Sr/ Sr for groundwater = (0.7105)/(3 + 0.7105)
= 0.191091
Substituting these values in Eq. 1, the relative contribu-
tion of rainfall to the Sr budget of groundwater in P40 is
calculated as 86%, and only 14% of the Sr is from soils.
Case Studies
Kazemi (13) and Kazemi and Milne-Home (15), through
a strontium isotope study of rainwater, surface water,
groundwater (shallow and deep), soil, and rocks of a
salinized watershed in Eastern Australia (Buckinbah
Creek watershed) found that the ratio of 87 Sr/86 Sr of
shallow groundwater was similar to that of soil and
 STRONTIUM ISOTOPES IN WATER AND ROCK 577

Table 1. The 87 Sr/86 Sr of Various Materials in the Buckinbah Creek Watershed in Eastern Australiaa
GW (Shallow) GW (Deep) Surface Water Rocksb Soils

B7 0.71197 B4 0.7084 S6 0.7099 Kabadah Fmc 0.70555 P45-Top 0.709

P19 0.7058 B3 0.708 S2 0.7087 Wansey Fmd 0.70813 P41 0.7373
P37 0.71 B2 0.7083 S5 0.7076 CV 0.72547 P42-Top 0.7135
P45 0.7078 B1 0.7084 Kabadah Fme 0.7051 P40 0.72068
P40 0.7105 Burrawong Ls 0.70812 P40B 0.72298
B5 0.7099 0.71047 0.720692
P41 0.7169 Rain 0.7089
P39 0.7092 Leaf 0.7115
P42 0.7125
P15 0.71023
a
Source: References 13.
b
CV: Canowindra Volcanics, Burrawong Ls: Burrawong limestone, Fm: Formation.
c
Rock type- shale.
d
Rock type- volcanoclastic sandstone.
e
Rock type- volcanoclastic.

rainwater (Table 1). They showed that between 20 to
90% of the Sr budget of shallow groundwater and surface
waters is derived from rainwater. They further concluded
that the deep groundwater is the source of baseflow at the
watershed outlet (due to identical 87 Sr/86 Sr ratios of deep
groundwater and baseflow).
Hunt et al. (6) used water and strontium isotopes to
investigate the hydrology of a natural and a constructed
wetland. They showed that the porewater from the natural
wetland had a 87 Sr content ranging from 0.00 to 0.86 per
mil (87 Sr/86 Sr = 0.71026 to 0.71087) and the 87 Sr of pore
water from the constructed wetland ranged from −2.35
to +1.24 per mil (87 Sr/86 Sr = 0.70859 to 0.71114). They
attributed this difference to the presence of a thick peat
layer in the natural wetland.
The 87 Sr/86 Sr ratio of 32 brine samples from the Upper
Jurassic Smackover Formation in southern Arkansas was
measured at 0.7071–0.7101, more radiogenic than the Sr
in Late Jurassic seawater because of the contribution of
Sr from detrital sources (7).
The linear negative correlation between the Sr and
87
Sr/86 Sr ratio (when plotted against each other) has
been used by Faure (3), Dogramaci et al. (16), and Singh
et al. (17) as proof of the mixing of two waterbodies.
Similarly, the negative correlation between the molar
ratio of Mg/Ca and Sr isotopes is an indication of
the incongruent dissolution of calcitic minerals in the
aquifer (11). Incongruent dissolution of carbonates occurs
progressively through two mechanisms:
1. Dissolution of calcitic minerals, causing the addition
of Sr to groundwater and modification of the
87
Sr/86 Sr ratio; and
2. Reprecipitation of calcitic minerals, thereby remov-
ing some of the Sr from the groundwater without
changing the Sr isotope ratio, approaching eventu-
ally the ratio in calcitic minerals (11).
A large data set of Sr isotopic ratios of the rivers of the
Ganges, Orinoco, and Amazon basins is found in Palmer
and Edmond (14).
The study by Lyons et al. (18) focuses on the strontium
isotopic signature of some streams and lakes in Antarctica.
They found that the ratios of 87 Sr/86 Sr of Lake Fryxell,
Lake Hoare, and Lake Bonney were 0.70895, 0.71057, and
0.71187–0.71204, respectively. They also found that in
Lake Fryxell and Lake Hoare, the 87 Sr/86 Sr became less
radiogenic with depth, whereas in Lake Bonney, the ratios
increase slightly with depth.
Sr isotopes together with isotopes of Nd and Pb of
Cenozoic basalts in eastern Australia were studied by
Zhang et al. (19) to show that the isotopic signatures
of the Pacific Ocean midoceanic ridge basalts (MORB)
type characterize the lava-field basalts in southeastern
Australia, whereas the mantle isotopic signatures of
Indian Ocean MORB type characterize younger basalts
from northeastern Australia.
BIBLIOGRAPHY
1. Capo, R.C., Stewart, B.W., and Chadwick, O.A. (1998). Stron-
tium isotopes as tracers of ecosystem processes: Theory and
methods. Geoderma 82: 197–225.
2. Clow, D.W., Alisa Mast, M., Bullen, T.D., and Turk, J.T.
(1997). Strontium 87/strontium 86 as a tracer of min-
eral weathering reactions and calcium sources in an
alpine/subalpine watershed, Loch Vale, Colorado. Water
Resour. Res. 33(6): 1335–1351.
3. Faure, G. (1986). Principles of Isotope Geology. John Wiley &
Sons, New York.
4. Faure, G. (2001). Origin of Igneous Rocks: The Isotopic
Evidence. Springer, Heidelberg.
5. Clauer, N., Chauduhuri, S., Toulkeridis, T., and Blanc, G.
(2000). Fluctuations of caspian sea level: Beyond climatic
variations. Geology 28(11): 1015–1018.
6. Hunt, R.J., Bullen, T.D., Krabbenhoft, D.P., and Kendall, C.
(1998). Using stable isotopes of water and strontium to
investigate the hydrology of a natural and a constructed
wetland. Groundwater 36(3): 434–443.
7. Stueber, A.M., Pushkar, P., and Hetherington, E.A. (1984). A
strontium isotopic study of Smackover brines and associated
solids, southern Arkansas. Geochim. Cosmochim. Acta 48:
1637–1649.
8. Miller, E.K., Blum, J.D., and Friedland, A.J. (1993). Deter-
mining of soil exchangeable-cation loss and weathering rates
using Sr isotopes. Nature 362: 438–441.
9. Hodell, D.A., Mead, G.A., and Mueller, P.A. (1990). Variation
in the strontium isotopic composition of seawater (8 Ma to
578 TECHNETIUM IN WATER

present): Implications for chemical weathering rates and cyclotron by E. Lawrence and then sent to Perrier and
dissolved fluxes to the ocean. Chem. Geol. 80: 291–307. Segre. Technetium was the first element to be produced
10. Morgan, K. and Jankowski, J. (2003). Origin of salts in a artificially.
fractured bedrock aquifer in the Central West Region, NSW, Twenty-two isotopes of technetium whose masses range
Australia. Proc. Int. Conf. Groundwater Fractured Rocks, from 90 to 111 are known, and all are radioactive (Table 1).
Prague, Krasny-Hrkal-Bruthans (Eds.). IHP-VI, series on
Technetium has three long-lived radioactive isotopes:
groundwater no. 7, pp. 221–222.
technetium-97 [97 Tc] (half-life {T1/2 } = 2.6 × 106 years),
11. Dogramaci, S.S., Herczeg, A.L., and Bone, Y. (1998). 87 Sr/86 Sr 98
Tc (T1/2 = 4.2 × 106 years), and 99 Tc (T1/2 = 2.13 × 105
in Groundwater as Indicators of Carbonate Dissolution.
years). An important isotope is 95m Tc (‘‘m’’ stands for
Proc. 9th Int. Symp. on water–rock Interaction, Taupo, New
Zealand, March–April, 1998, G.B. Arehart and J.R. Hulston metastable state) [T1/2 = 61 days], which is used in tracer
(Eds.). A.A. Balkema, Rotterdam, pp. 211–214. work. However, the most useful isotope of technetium
12. Jorgensen, N.O. and Heinemeier, J. (2003). Isotope studies is 99m Tc (T1/2 = 6.01 hours), which is used in many
of saline groundwater in a sandstone aquifer on Bornholm, medical radioisotope tests because of its short half-life,
Denmark. Proc. Int. Conf. Groundwater Fractured Rocks, easy detection of the gamma ray it emits, and the ability
Prague, Krasny-Hrkal-Bruthans (Eds.). IHP-VI, series on of technetium to bind chemically to many biologically
groundwater no. 7, pp. 205–206. active molecules.
13. Kazemi, G.A. (1999). Groundwater factors in the management Technetium-99 is a weak, beta-emitting radionuclide,
of dryland and stream salinity in the upper Macquarie Valley, 0.292 million-electron volts (MeV), produced during
New South Wales, Australia. Unpublished Ph.D. Thesis, nuclear fission, which decays to the stable isotope
University of Technology, Sydney. ruthenium-99. It has a specific activity of 3.78 × 1010
14. Palmer, M.R. and Edmond, J.M. (1992). Controls over the transmutations per minute per gram, which corresponds
strontium isotope composition of river water. Geochim. to 17 microcuries per milligram (µCi/mg) (4). Technetium
Cosmochim. Acta 56: 2099–2111.
metal is a silver-gray color, tarnishes slowly in the
15. Kazemi, G.A. and Milne-Home, W.A. (1999). The application air, and has properties similar to those of rhenium.
of strontium, oxygen 18 and deuterium isotopes in the
However, as mentioned earlier, technetium does not occur
estimation of the solute budget and hydrology of a small
naturally.
watershed in Eastern Australia. Eos Trans. AGU 80(17)
Spring Meet. Suppl. S139, 1999. However, 99 Tc in terrestrial material has been evalu-
ated since its discovery. In 1961, 99 Tc was isolated and
16. Dogramaci, S.S., Herczeg, A.L., and Bone, Y. (1998). Chemi-
cal evolution of groundwater in the Murray Basin: Geochem- identified in African pitchblende (a uranium-rich ore) in
istry and mass transfer modelling. Abstract No. 49, Geol. Soc. extremely minute quantities (5). It was theorized that
99
Aust., 14th Aust. Geol. Convention. Tc was produced as a result of the spontaneous fission of
17. Singh, K.S. et al. (1998). Chemical and strontium, oxygen, uranium-238 (238 U).
and carbon isotopic compositions of carbonates from the Technetium-99 is produced primarily through anthro-
Lesser Himalaya: Implications to the strontium isotope pogenic processes such as detonation of nuclear weapons,
composition of the source waters of the Ganga, Ghaghara, and reprocessing of spent nuclear reactor fuel rods contain-
the Indus rivers. Geochim. Cosmochim. Acta 62: 743–755. ing uranium, and disposal of nuclear waste. Reprocessing
18. Lyons, W.B. et al. (2002). Strontium isotopic signatures of the involves dissolving the spent nuclear fuel in acid to sepa-
streams and lakes of Taylor Valley, Southern Victoria Land, rate the potentially reusable components of the spent fuel
Antarctica: Chemical weathering in a polar climate. Aquatic
Geochem. 8(2): 75–95.
19. Zhang, M., O’Reilly, S.Y., and Chen, D. (1999). Location of Table 1. Isotopes, Half-Lives, and Decay Processes for
Pacific and Indian mid-ocean ridge-type mantle in two time Technetiuma
slices: Evidence from Pb, Sr, and Nd isotopes for Cenozoic
Australian basalts. Geology 27(1): 39–42. Type of Type of
Nuclide Half-Lifeb Decay Nuclide Half-Life Decay
91
Tc 3.2 m β+, γ 99
Tc 2.13 × 105 y B−
TECHNETIUM IN WATER 92
Tc 4.4 m β+, γ 100
Tc 15.8 s B− , γ
93m
Tc 43.5 m ε, γ 101
Tc 14 m B− , γ
JAY CLAUSEN
93
Tc 2.7 h ε, β + , γ 102m
Tc 4.3 m B− , γ
94
Tc 53 m β+, γ 102
Tc 6.3 m B− , γ
AMEC Earth and 94
Tc 4.9 h ε, β + , γ 103
Tc 50 s B− , γ
Environmental, Inc. 95m
Tc 60 d ε, β + , γ 104
Tc 18.0 m B− , γ
Westford, Massachusetts 95
Tc 20 h ε, γ 105
Tc 7.6 m B− , γ
96m
Tc 52 m ε, γ 106
Tc 36 s B− , γ
INTRODUCTION 96
Tc 4.3 d ε, γ 107
Tc 21 s B− , γ
97m
Tc 91 d γ 108
Tc 5 s B− , γ
Technetium, element 43, was predicted on the basis of its 97
Tc 2.6 × 106 y ε 109
Tc 1 s B−
location in the periodic table. Technetium was erroneously 98
Tc 4.2 × 106 y B− , γ 110
Tc 0.83 s B− , γ
reported as having been discovered in 1925 by Noddack 99m
Tc 6 h γ , B− 111
Tc ? ?
and Tacke (1), who called it masurium. However, Element a
Reference 3.
43 was actually discovered in Italy (2). It was found in b
d: days; h: hours; m: minutes; s: seconds; y: years.
a sample of molybdenum, which had been bombarded c +
β : positron; β − : beta particle; ε: internal electron conversion; γ :
with deuterons in the University of California—Berkeley gamma ray.
 TECHNETIUM IN WATER 579

(the uranium and plutonium) from the waste component Table 2. Thermodynamic
(the fission products). The fission products, which are of Data for Technetiuma
no use, must be disposed of properly. In a 1000-megawatt G0 f
nuclear reactor (thermal fission yield of approximately Speciesb kcal/gfwc
6%), approximately 28 g of 99 Tc is produced daily from
the slow neutron-induced fission of 235 U or plutonium-239 TcOH(s) −56.1
Tc(OH)2 (s) −110.2
[239 Pu] (6). The long half-life and mobility of 99 Tc makes
Tc3 O4 (s) −206.3
it an important element to the nuclear fuel industry in TcO4 (s) − −149.1
terms of its potential environmental impact. This is a par- TcS2 (s) −51.0
ticular concern for designing underground nuclear waste TcO2 (s) −88.3
storage facilities because 99 Tc is the most mobile of all
a
radionuclides present. Reference 19.
b
s: solid.
There are no known uses of technetium other than c
kcal/gfw: kilocalories per gram
99m
Tc for medical purposes. Technetium has supercon- formula weight
ducting properties and is a good corrosion preventer (7).
However, its unstable nature and production of radioactive
particles negate its practical use. 1.2
1 PO =
ANALYSIS 0.8 TcO2 2 1 bar
0.6
Analysis for 99 Tc typically consists of methyl ethyl ketone 0.4 TcO4−
Eh (v)
(MEK) extraction followed by liquid scintillation counting 0.2
using United States Environmental Protection Agency Tc3O4
0 Tc
(USEPA) Method 900.0. The typical quantification limit
−0.2
is 1 becquerel per liter (Bq/L), which is equivalent to 25
−0.4 Tc(OH)2 PH2 =
picocuries per liter (pCi/L). 1 ba r
−0.6
Chemistry −0.8
0 2 4 6 8 10 12 14
The oxidation states most commonly found for Tc are99 pH
Tc0 , Tc+4 , and Tc+7 . The most stable chemical species Figure 1. Eh -pH diagram for the Tc–O–H system at a concen-
of 99 Tc in aqueous solutions is the pertechnetate anion, tration of 90 µg/L under standard temperature and pressure.
TcO4 − , in the Tc+7 valence state, which is highly mobile
and nonreactive (8–12). However, 99 Tc is very sensitive
to changes in redox conditions (13). Under reducing The solubility of sodium pertechnetate is reportedly 11
conditions, 99 Tc can be precipitated by various cations moles per liter [mol/L] (13). The solubility of 99 Tc in a
to form anhydrous technetium dioxide, TcO2 , (as Tc+4 ) reducing environment varies from 1.32 × 10−8 to 2.56 ×
or technetium heptoxide, Tc2 O7 , or other insoluble oxides 10−9 mol/L and is controlled by the Tc species present (14).
and hydroxides such as Tc2 S7 or TcS2 or incorporated into Redox thus is a key variable affecting solubility (8). The
minerals (14–17): solubility of technetium under reducing conditions in
carbonate groundwater is pH dependent. At pH levels
TcO4 − + 4H+ + 3e− = TcO2 + 2H2 O; E0 = +0.738 volts between 5 and 7.5, the solubility of Tc+4 is less than
(1) 10−8 mol/L. Pilkington (22) suggests that the solubility of
In the presence of oxygen, the Tc+4 oxide or oxyhydroxide TcO2 is closer to 10−7 mol/L under reducing conditions.
can be reoxidized (14,18): At higher pHs of 8 to 11, solubility increases, and the
dominant ionic complex is TcO(OH)3 CO− 3 .
[TcO2 (s) or TcO(OH2 (s) ] → TcO4 − (2) Equally important are the fate-and-transport proper-
ties of 99 Tc. Under aerobic soil conditions, 99 Tc behaves as
The pertechnetate anion also forms complexes with an inert tracer and readily passes through soil unless
organic compounds containing –C –O and –COO func- there is a high content of organic matter or metal
tional groups, especially under reducing conditions. The oxide (10,23–30). As Gu and Dowlen (31) report, this is
thermodynamic data for technetium (Table 2) can be used not surprising because soil mineral surfaces are nega-
to develop a phase diagram to identify which species are tively charged as is TcO4 − and therefore the two repel
likely to be present at a given set of Eh and pH conditions each other. Once introduced to surface water or ground-
(Fig. 1). water, it can be transported significant distances due to
The pertechnetate ion is mobile under oxidizing its recalcitrant nature and lack of sorption. Given the
conditions (Fig. 1). Bidoglio et al. (20,21) found in column concern over 99 Tc mobility, extensive studies have been
experiments that the oxidized species persists, even under conducted in various geologic media to assess suitable
slightly reducing or anoxic conditions. Technetium can nuclear waste disposal storage facilities. Adsorption coef-
undergo a number of surface reactions on various minerals ficients for various minerals are presented in Sheppard
under reduced conditions. et al. (23), Wildung et al. (27), Palmer and Meyer (32), and
580 TECHNETIUM IN WATER
Strickert et al. (33). Sheppard et al. (23) indicate that soil
redox conditions and organic matter content are the key
variables influencing 99 Tc mobility in soil.
REGULATORY STANDARDS
The USEPA has set an effective dose equivalent of 4
millirem (mrem)/year for exposure, which they incorrectly
calculated as 900 pCi/L (36 Bq/L). The correct activity
equivalent to 4 mrem/year is 3790 pCi/L [152 Bq/L] (34).
The U. S. Department of Energy (DOE) Order 5400.5 sets
a permissible ingested-water limit of 100 mrem/year or
100,000 pCi/L (4000 Bq/L), whereas the U. S. Nuclear
Regulatory Agency (NRC) has set a value of 50 mrem/year
or 60,000 pCi/L (2400 Bq/L). The NRC has set the
maximum activity of 99 Tc that can be released to a sanitary
sewer at 500 mrem/year or 600,000 pCi/L (24,000 Bq/L).
The proposed action level for groundwater cleanup at the
Paducah Gaseous Diffusion Plant (PGDP) site is 25 pCi/L
(1 Bq/L), which is the quantification limit.
Rainwater
Technetium has been released to the atmosphere in the
form of Tc2 O7 and HTcO4 as a consequence of aboveground
detonation of nuclear weapons through the fission of
235
U and 239 Pu. The nuclear industry also releases
TcF6 in addition to Tc2 O7 and HTcO4 during uranium
enrichment. During a nuclear detonation, elemental 99 Tc
burns to form Tc2 O7 , which upon cooling and contact
with water vapor forms HTcO4 . The activity of 99 Tc
measured in rainwater reportedly ranges from 8.2 × 10−8
to 1.0 × 10−6 µg/L (35,36). Work by Leon (37) and Attrep
et al. (36) suggest the concentration of 99 Tc in rainwater
has increased from the early 1960s through the 1980s.
Both papers suggest that the source of this increase in
99
Tc in rainwater is from the reprocessing of spent nuclear
fuel rather than nuclear weapons testing.
Surface Water
Technetium-99 is a surface water and groundwater
contaminant at a number of DOE sites such as the PGDP
in Paducah, Kentucky; the Portsmouth Gaseous Diffusion
Plant in Piketon, Ohio; Oak Ridge National Laboratory
Y-12 Plant, Savannah River Laboratory in Aiken, South
Carolina; the Pacific Northwest National Laboratory 200
west area; and the Nevada Test Site (30,38–40).
Currently, the largest annual release of 99 Tc to the
environment occurs at the two nuclear fuel-reprocessing
facilities in Sellafield, United Kingdom. Technetium-99
laden effluent water from these facilities is discharged
directly into the Irish Sea.
Contamination has been detected as far west as
Svalbard and along the Norwegian coastline, distances
of several hundred miles. The total discharge from these
two plants amounts to between 70 and 90 tera (TBq)/year
of 99 Tc (41).
Techentium-99 is reported in the Savannah River as
a result of discharges from the U. S. Department of
Energy Savannah River Plant located near Aiken, South
Carolina (42,43). Technetium-99 is present at Oak Ridge
National Laboratory where radioactive liquids generated
from routine operations were disposed of in shallow waste
pits (44). Nearby seeps have 99 Tc activity levels of up to
1185 pCi/L [3,200 Bq/L].
Blaylock and Frank (45) conducted a study in which
TcO4 − was added to a freshwater pond. Their study
indicated that 84% of the TcO4 − added remained in
solution; very little was taken up by biota or by settling out
into sediments. The vast majority of the 99 Tc was removed
through drainage, indicating that the ultimate reservoir
for technetium released into the environment is the ocean.
Similarly, in a study of the Rhone River, it was found that
the 99 Tc activity varied from 2.5 to 53 × 10−3 pCi/L (0.1 to
2.1 × 10−3 Bq/L) in seven samples collected in 1985 and
1986 (46).
Finally, significant quantities of 99 Tc laden effluent
water were released to a settling lagoon as well as to
several small streams surrounding the PGDP. Activity
levels as high as 100,000 pCi/L (4000 Bq/L) were routinely
released to surface water. Ferrous iron sulfate was used
to precipitate metals in the lagoon before discharge
into a nearby stream. The addition of this material
shifted the redox state of the lagoon such that TcO4 −
was precipitated from solution. The likely form of the
technetium precipitate was Tc2 S7 . In addition, the iron-
rich clay and organic-rich stream sediments retained much
of the 99 Tc, limiting its downgradient migration from the
lagoon. Therefore, the 99 Tc migrated only several thousand
feet from the point of release. However, downgradient of
the site, groundwater discharges to the surface water
streams and 99 Tc-rich water has been measured in several
seeps (47).
Ground Water
The most extensive groundwater distribution of 99 Tc is
at the PGDP where a 2- 21 mile long plume of 99 Tc
contaminated groundwater is commingled with the solvent
trichloroethene. The 99 Tc was introduced to the site
during the 1970s when fissioned uranium was reprocessed.
99
Tc became entrained in processing equipment at the
PGDP, which was later cleaned with organic solvents and
released to the environment. Gu et al. (48) determined
conclusively that the species of 99 Tc mobile in groundwater
at the PGDP was the pertechnetate anion, TcO4 − . This
later was supported through the development of phase
diagrams using site-specific Eh and pH groundwater
data (38,39). In addition, Gu and Dowlen (31) and Gu
et al. (48) conclusively ruled out colloidal and natural
organic matter facilitated transport as viable mechanisms
of 99 Tc mobilization. Previous studies of PGDP, other sites,
and laboratory studies assumed that TcO4 − was the form
of 99 Tc in most environmental settings. The estimated
partitioning coefficient of 99 Tc exceeded 27,000 milliliter
per gram [mL/g] (48). Technetium-99 activities as high as
179 Bq/L (4800 pCi/L) were observed in the groundwater
plume (38,39). The estimated mass of 99 Tc dissolved in the
groundwater plume was 4 to 6 lb, which equates to an
activity of 1.0 to 1.6 × 1012 Bq/L (28 to 42 × 1012 pCi)
of 99 Tc.
At the Hanford Site in Washington, 99 Tc has been
found in groundwater in the Hanford 200 West Area

where spent nuclear fuel was reprocessed and then the
waste material was stored in underground tanks (30).
Technetium-99 levels between 900 and 27,000 pCi/L (36
and 1080 Bq/L) have been found in groundwater.
Technetium-99 activity in shallow groundwater at the
Portsmouth Gaseous Diffusion Plant has been observed
at the 1000 pCi/L (40 Bq/L) level (49). The causes of
contamination are similar to those discussed for the
Paducah site. However, because the Portsmouth site
received feed material from the Paducah site, the quantity
of 99 Tc introduced into the site was much lower.
A unique study was conducted by the Lawrence
Livermore National Laboratory at the Nevada Test site
to assess the distribution of radionuclides in caverns
produced by underground nuclear testing. The work by
Buddemeier et al. (50) and Schroeder et al. (40) found that
99
Tc was present in groundwater associated with two
separate nuclear weapons tests. The 99 Tc appeared to
be associated and transported with colloidal material, but
its mobility was reduced upon contact with zones of high
levels of iron oxyhydroxides.
There are several known nongovernmental sites
containing 99 Tc. One such facility is the Fields Brook
National Priority Listed (NPL) site in Ashtabula, Ohio.
The site was formerly a storage site for liquid waste,
which contained 99 Tc as well as other radionuclides and
solvents. Technetium-99 activities as high as 100,000
pCi/L (4000 Bq/L) have been detected in groundwater.
A second site is the uranium fuel plant in Hematite,
Montana. Recycled uranium, containing 99 Tc, was used to
fabricate metallic fuel rods. Low-levels of 99 Tc have been
observed in groundwater.
REMEDIATION
The technologies explored for remediating 99 Tc in water
include ion exchange, liquid–liquid extraction, precipi-
tation with various forms of iron, redox manipulation,
the FORAGER sponge, granular activated carbon (GAC),
and natural attenuation (10,16,18,31–33,48,51–65). The
report by Bostick et al. (53) provides an excellent reference
list of the many studies conducted on 99 Tc removal from
aqueous waste streams. At the Portsmouth and Paducah
Gaseous Diffusion plants, ion exchange is being used for
treating 99 Tc-contaminated groundwater. Although, Gu
et al. (31) found that GAC was superior to ion exchange
due to its low cost, efficiency, and sorption capacity, GAC
has not been employed at these sites.
CONCLUSION
Technetium is an element not typically found in nature
and has environmental consequence at facilities that
enrich or reprocess uranium fissioned material. These
facilities are typically government operated. The high
solubility of TcO4 − and its stability across a wide spectrum
of redox conditions enables the ion to be extremely
mobile once introduced into water. As a consequence, at
the few sites where 99 Tc has been introduced into the
natural environment, it has produced very large plumes
of contaminated surface and groundwater.
TECHNETIUM IN WATER 581
BIBLIOGRAPHY
1. Noddack, W. and Tacke, I. (1925). Die Ekamangane. Natur-
wissenschaften 13: 567.
2. Perrier, C. and Segre, E. (1937). Nature (London). 140: 193.
3. Seelmann-Eggebert, W., Pfennig, G., and Munzel, H. (1974).
Chart of the Nuclides. Gersbach Verlag, Munich, Germany.
4. Schwochau, K. (1981). The analytical chemistry of tech-
netium. Top. Curr. Chem. 96: 109–147.
5. Kenna, B.T. and Kuroda, P.K. (1961). Technetium 99. J.
Inorg. Nucl. Chem. Lett. 23: 142.
6. De Alleluia, T.B. (1983). Tc: Technetium, System No. 69. In:
Gmelin Handbook of Chemistry. Supplement Vol. 1, 8th Edn.
Springer-Verlag, Berlin.
7. Cartledge, G.H. (1973). Twenty-year inhibition of corrosion
by the pertechnetate ion. Corrosion 29: 361–362.
8. Eriksen, T.E., Ndalamba, P., Bruno, J., and Caceci, M.
(1992). The solubility of TcO2 ·nH2 O in neutral to alkaline
solutions under constant pCO2 . Radiochimica Acta 58/59:
67–70.
9. Bruhl, H., Bock, W., Trapp, C., and Winkler, A. (1989). Aus-
breitungsverhalten von Technetium (Selen) in geochemisch
unterschiedlichen Grundwasserleitern. Institut fur Ange-
andte Geologie Freie Unversitat Berlin. Final report 02 U
5695 A9.
10. Vandergraaf, T.T., Ticknor, K.V., and George, I.M. (1984).
Reactions Between Technetium in Solution and Iron-
Containing Minerals Under Oxic and Anoxic Conditions.
Atomic Energy of Canada. Limited Report AECL-1957.
Pinawa, Manitoba, Canada, pp. 25–43.
11. Paquette, J., Reid, J.A.K., and Rosinger, E.L.J. (1980).
Review of Technetium Behavior in Relation to Nuclear Waste
Disposal. Whiteshell Nuclear Research Establishment. Tech-
nical Record TR-25. Pinawa, Manitoba, Canada.
12. Allard, B., Kigatsi, H., and Torstenfelt, B. (1979). Tech-
netium: Reduction and sorption in granitic bedrock.
Radiochem. Radioanalytical Lett. 37: 223–230.
13. Gu, B. and Schulz, R.K. (1991). Anion Retention in Soil: Pos-
sible Application to Reduce Migration of Buried Technetium
and Iodine. U.S. Nuclear Regulatory Commission, Washing-
ton, DC.
14. Meyer, R.E., Arnold, W.D., and Case, F.I. (1987). The Solu-
bility of Electrodeposited Tc(IV) Oxides. Oak Ridge National
Laboratory, Oak Ridge, TN.
15. Lieser, K.H. and Bauscher, C.H. (1987). Technetium in the
hydrosphere and in the geosphere 1. Chemistry of technetium
and iron in natural waters and influence of the redox
potential on the sorption of technetium. Radiochimica Acta
42: 205–213.
16. Lee, S.K. and Bondietti, E.A. (1983). Technetium behavior in
sulfide and ferrous iron solutions. Mater. Res. Soc. Symp.
Proc. 15: 315–321.
17. Wildung, R.E., Routson, R.C., Serne, R.J., and Garland, T.R.
(1974). Pertechnetate, Iodide, and Methyl Iodide Retention
by Surface Soils. In Pacific Northwest Laboratory Annual
Report for 1974 to the USAEC Division of Biomedical
and Environmental Research. B.E. Vaughan (Ed.). Pacific
Northwest National Laboratory, Richland, WA, pp. 37–40.
18. Meyer, R.E., Arnold, W.D., and Case, F.I. (1984). Valence
Effects on the Sorption of Nuclide on Rocks and Minerals.
Oak Ridge National Laboratory, Oak Ridge, TN.
19. Rard, J.A. (1983). Critical Review of the Chemistry and
Thermodynamics of Technetium and Some of its Inorganic
582 TECHNETIUM IN WATER
Compounds and Aqueous Species. Lawrence Livermore
National Laboratory, Livermore, CA.
20. Bidoglio, G., Avogadro, A., and De Plano, A. (1985). Influence
of redox environments on the geochemical behavior of
radionuclides. Mater. Res. Symp. Proc. 50: 709–716.
21. Bidoglio, G., Chatt, A., De Plano, A., and Zorn, F. (1983).
Investigation on technetium species in ground water and
migration through clay columns. J. Radioanalytical Chem.
79: 153–164.
22. Pilkington, N.J. (1990). The solubility of technetium in the
near-field environment of a radioactive waste repository. J.
Less-Common Met. 161: 203–212.
23. Sheppard, S.C., Sheppard, M.I., and Evenden, W.G. (1990). A
novel method used to examine variation in Tc sorption among
34 soils, aerated and anoxic. J. Environ. Radioactivity 11:
215–233.
24. Henrot, J., Garten, C.T., Jr., and Blaylock, B.G. (1989).
Behavior of Technetium in Soil: Sorption-Desorption Pro-
cesses. Oak Ridge National Laboratory, Oak Ridge, TN.
25. Schulte, E.H. and Scoppa, P. (1987). Sources and behavior
of technetium in the environment. Sci. Total Environ. 64:
163–179.
26. Wolfrum, C. and Bunzl, K. (1986). Sorption and desorption
of technetium by humic substances under oxic and anoxic
conditions. J. Radioanal. Nucl. Chem. 99: 315–323.
27. Wildung, R.E., Garland, T.R., McFadden, K.M., and Cowan,
C.E. (1986). Technetium sorption in surface soils. In: Tech-
netium in the Environment. G. Desmet and C. Myttenaere
(Eds.). Elsevier Applied Science, London, pp. 115–129.
28. Wildung, R.E., Garland, T.R., and Cataldo, D.A. (1977).
Accumulation of technetium by plants. Health Phys. 32:
314–317.
29. Bondietti, E.A. and Francis, C.W. (1979). Geologic migration
potentials of technetium-99 and neptunium-237. Science 203:
1337–1340.
30. Brown, D.J. (1967). Migration Characteristics of Radionu-
clides Through Sediments Underlying the Hanford Reserva-
tion. In: Disposal of Radioactive Wastes in to the Ground.
International Atomic Energy Agency Symposium. Vienna,
Austria.
31. Gu, B., Dowlen, K.E., and Clausen, J.L. (1996). Efficient
separation and recovery of technetium-99 from contaminated
groundwater. Efficient Sep. 6: 123–132.
32. Palmer, D.A. and Meyer, R.E. (1981). Adsorption of tech-
netium on selected inorganic ion-exchange materials and on a
range of naturally occurring minerals under oxic conditions.
J. Inorg. Nucl. Chem. 43: 2979–2984.
33. Strickert, R.G., Friedman, A.M., and Fried, S. (1980). The
sorption of technetium and iodine radioisotopes by various
minerals. Nucl. Technol. 49: 253–266.
34. Davenport, D. (1994). Interoffice Memorandum to L. Stinton.
Martine Marietta Energy Systems, Piketon, OH.
35. Erhardt, K.C. and Attrep, M., Jr. (1978). Technetium-99 in
the atmosphere. Environ. Sci. Technol. 12: 55–57.
36. Attrep, M., Jr., Enochs, J.A., and Broz, L.D. (1971). Atmo-
spheric technetium-99. Environ. Sci. Technol. 5: 344–345.
37. Leon, M. (1989). Determination and levels of 99 Tc in
environmental and biological samples. In: Technetium in the
Environment. G. Desmet and C. Myttenaere (Eds.). Elsevier
Applied Science, London, pp. 171–178.
38. Clausen, J.L., Zutman, J.L., Pickering, D.A., and Farrow,
N.D. (1995). Final Report on Drive-Point Profiling of the
Northwest Plume and Analysis of Related Data. Lockheed
Martin Energy Systems, Inc., Kevil, KY.
39. Clausen, J.L., Richards, W.L., Korte, N.E., and Liang, L.
(1995). ORNL/MMES research into remedial applications of
zero-valence metals, 3: Removal of TCE, cis-1,2-DCE, vinyl
chloride and technetium. Am. Chem. Soc. Symp. Reductive
Dehalogenation Proc., April 2–7, Anaheim, CA, pp. 755–758.
40. Schroeder, N.S., Morgan, D., Rokop, D.J., and Fabryka-
Martin, J. (1992). Migration of Technetium-99 in the Alluvial
Aquifer at the Nevada Test Site. Los Alamos National
Laboratory, LA-UR-92-463, Los Alamos, NM.
41. Brown, J., Kolstad, A.K., Lind, B., Rudjord, A.L., and
Strand, P. (1998). Technetium-99 contamination in the North
Sea and in Norwegian Coastal areas 1996 and 1997.
Stralevern Report. 1998:3. Oslo, Norway.
42. Turcotte, M.S. (1982). Environmental Behavior of Technetium-
99. Savannah River Plant and Laboratory Report DP-1644.
Aiken, SC.
43. Anderson, T.J. (1979). Development of chemical isolation and
concentration techniques for Tc-99 analysis by resin—bead
mass spectrometry. In: Radioelement Analysis. Savannah
River Laboratory, Aiken, SC.
44. Olsen, C.R. et al. (1986). Geochemical and environmental
processes affecting radionuclide migration from a formerly
used seepage trench. Geochim. Cosmochim. Acta 50: 593–607.
45. Blaylock, B.G. and Frank, M.L. (1981). Bioaccumulation and
distribution of Tc-95m in an experimental freshwater pond.
In: Int. Symp. Migration Terrestrial Environ. Long-Lived
Radionuclides Nucl. Fuel Cycle. Savannah River Laboratory,
Aiken, SC.
46. Ballestra, S. et al. (1987). J. Radiochem. 115: 51.
47. Fryar, A. and Lasage, D.M. (2001). Technetium-99 as an
in-situ tracer of ground-water/stream interactions and
organic co-contaminant fate. Geol. Soc. Am. Annu. Meet.,
November 5–8, Boston, MA.
48. Gu, B., Clausen, J.L., McDonald, J.A., and McCarthy, J.F.
(1994). Assessment of the Influences of Groundwater Colloids
on the Migration of 99 Tc at the PGDP Site, Paducah, KY. Oak
Ridge National Laboratory, Oak Ridge, TN.
49. Muck, M.T., Korte, N.E., Liang, L., Marsh, D., Zutman, J.L.,
and Clausen, J.L. (1995). Technology Screening Report
for Tc Removal from Groundwater Portsmouth Uranium
Enrichment Plant, Piketon, Ohio. Martin Marietta Energy
Systems, Inc., Piketon, OH.
50. Buddemeier, R.W., Finkel, R.C., Marsh, K.V., Ruggieri, M.R.,
Rego, J.H., and Silva, R.J. (1992). Radionuclide migration
at the Cheshire and Cambric event sites. In: Hydrology
and Radionuclide Migration Program 1989 Progress Report.
K.V. Marsh (Ed.). Lawrence Livermore National Laboratory,
Albuquerque, NM, pp. 5–20.
51. Clausen, J.L., Sturchio, N.C., Heraty, L., Abrajano, T., and
Huang, L. (1997). Evaluation of Natural Attenuation Pro-
cesses for Trichloroethene and Technetium-99 in the North-
west and Northeast Plumes at the Paducah Gaseous Diffusion
Plant Paducah, Kentucky. Lockheed Martin Energy Systems,
Inc., Kevil, KY.
52. Clausen, J.L. and Richards, W.L. (1994). Treatability Studies
Using Iron Filings to Remediate Trichloroethene and
Technetium. Martin Marietta Energy Systems. KY/ER-51.
Paducah, KY.
53. Bostick, W.D., Osberne, P.E., Del Cul, G.D., and Sim-
mons, D.W. (1995). Treatment of Aqueous Solutions Con-
taminated with Technetium-99 Originating from Uranium
Enrichment Activities: Final Report. Martin Marietta Energy
Systems, Inc., Oak Ridge, TN.
 WATER—NATURE’S MAGICIAN 583

54. Ishanullah. (1994). Optimization of various factors for the

separation of technetium using anion-exchange resins. Sep.
Sci. Technol. 29: 239–247.
55. Del Cul, G.D., Bostick, W.D., Trotter, D.R., and Osberne, P.E.
(1993). Technetium-99 removal from process solutions and
contaminated groundwater. Sep. Sci. Technol. 28: 551–564.
56. Kawasaki, M., Amori, T., and Hasegawa, K. (1993). Adsorp-
tion of pertechnetate on an anion exchange resin. Radiochim-
ica Acta 63: 53–56.
57. Takeuchi, M., Tanaka, S., and Yamawaki, M. (1993). Solvent
extraction of Tc(vii) and Re(vii) from nitric acid solution
with the mixture of CMPO and TBP. Radiochimica Acta 63:
97–100.
58. Yamagishi, I. and Kubota, M. (1993). Recovery of technetium
with active carbon column in partitioning process of high-level
liquid waste. J. Nucl. Sci. Technol. 30: 717–719.
59. Karalova, Z.K., Lavrinovich, E.A., and Myasoedov, B.F.
(1991). Separation of technetium from accompanying
elements in basic solutions by extraction with
triphenyltetrazolium chloride. Sov. Radiochem. 33: 266–269.
60. Chen, Q.J., Dahlgaard, H., Hansen, H.J.M., and Aarkrog, A.
(1990). Determination of Tc-99 in environmental samples by
anion exchange and liquid extraction at controlled valency.
Anal. Chim. Acta 228: 163–167.
61. Emery, M.F. and Lim, C.K. (1989). Separation of cationic
technetium-99m amine complexes on porous graphiti carbon.
J. Chromatogr. 479: 212–215. • Canada has about 7% of the world’s renewable
62. Bostick, W.D. and Evans-Brown, B.S. (1988). Sorptive freshwater supply, compared with 18% for Brazil,
Removal of Technetium from Heavy Metals Sludge Filtrate 9% for China, and 8% for the United States.
Containing Nitrate Ion. Oak Ridge National Laboratory, Oak
Ridge, TN. • The ice we skate on in winter is water in its solid
form. Unlike most substances, which are densest in
63. Yamagishi, I. and Kubota, M. (1989). Separation of tech-
netium with active carbon. J. Nucl. Sci. Technol. 26: their solid state, ice is less dense than water and
1038–1044. thus floats. If this were not the case, lakes and rivers
64. Walton, F.B., Paquette, J., Ross, J.P.M., and Lawrence, W.E. would freeze from the bottom up. Fish could not
(1986). Tc(IV) and Tc(VII) interactions with oxyhydroxides. survive, and it is unlikely that rivers and lakes in
Nucl. Chem. Waste Manage. 6: 121–126. northern countries would ever completely thaw.
65. Fried, S., Friedman, A.M., Cohen, D., Hines, J.J., and Strick- • Water vapour forms a kind of global ‘‘blanket’’ which
ert, R.G. (1978). The Migration of Radioactive Processing helps to keep the earth warm. Heat radiated from
Wastes in Selected Rocks. Argonne National Laboratory, the sun-warmed surface of the earth is absorbed and
Argonne, IL. held by the vapour.

WATER’S MAGICAL PROPERTIES

WATER—NATURE’S MAGICIAN
Environment Canada
Water is the solvent, the medium and the participant
in most of the chemical reactions occurring in our
environment.
Water is at once simple and complex. A water molecule
itself is simple, made up of three atoms: two hydrogen and
one oxygen, H2 0. The configuration of these building blocks
produces a molecule with almost magical properties.
WATER—VISIBLE AND INVISIBLE
On earth, water is found as a liquid, as a solid (ice) or as a
gas (water vapour).
This article is produced by Environment Canada and is
reproduced with permission.
Water molecules are attracted to each other, creating
hydrogen bonds. These strong bonds determine almost
every physical property of water and many of its chemical
properties too.
Boiling and Freezing
Pure water at sea level boils at 100 ◦ C and freezes at
0 ◦ C. At higher elevations (lower atmospheric pressures)
water’s boiling temperature decreases. This is why it takes
longer to boil an egg at higher altitudes. The temperature
does not get high enough to cook the egg properly. If a
substance is dissolved in water, then the freezing point is
lowered. That is why we spread salt on streets in winter
to prevent ice formation.
Thermal Properties
Water absorbs or releases more heat than many
substances for each degree of temperature increase or
584 WATER—NATURE’S MAGICIAN
decrease. Because of this, it is widely used for cooling
and for transferring heat in thermal and chemical
processes.
Differences in temperature between lakes and rivers
and the surrounding air may have a variety of effects. For
example, local fog or mist is likely to occur if a lake cools the
surrounding air enough to cause saturation; consequently
small water droplets are suspended in the air
Large bodies of water, such as the oceans or the Great
Lakes, have a profound influence on climate. They are the
world’s great heat reservoirs and heat exchangers and the
source of much of the moisture that falls as rain and snow
over adjacent land masses.
When water is colder than the air, precipitation is
curbed, winds are reduced, and fog banks are formed.
Surface Tension
Surface tension is a measure of the strength of the water’s
surface film. The attraction between the water molecules
creates a strong film, which among other common liquids
is only surpassed by that of mercury.
This surface tension permits water to hold up
substances heavier and denser than itself. A steel needle
carefully placed on the surface of a glass of water will
float. Some aquatic insects such as the water strider rely
on surface tension to walk on water.
Surface tension is essential for the transfer of energy
from wind to water to create waves. Waves are necessary
for rapid oxygen diffusion in lakes and seas.
Molecules in Motion
Water molecules as well as binding to each other,
bind to many other substances such as glass, cotton,
plant tissues, and soils. This is called adhesion. For
example, in a thin glass tube, when the molecules
at the edge reach for and adhere to the molecules
of glass just above them, they at the same time tow
other water molecules along with them. The water
surface, in turn, pulls the entire body of water to a
new level until the downward force of gravity is too
great to be overcome. This process is called capillary
action.
Thus water readily wets many materials. Capillary
action allows a paper towel or a sponge to be used to soak
up spilled water. Without this property, the nutrients
needed by plants and trees would remain in the soil.
The Universal Solvent
An extraordinary property of water is its ability to dissolve
other substances. There is hardly a substance known
which has not been identified in solution in the earth’s
waters. Were it not for the solvent property of water, life
could not exist because water transfers nutrients vital to
life in animals and plants.
A drop of rain water falling through the air dissolves
atmospheric gases. When rain reaches the earth, it affects
the quality of the land, lakes and rivers.
Did You Know
• Raindrops are not tear-shaped. Scientists, using
high-speed cameras, have discovered that rain-
drops resemble the shape of a small ham-
burger bun.
• About 70% of the human body is water.
• Life on earth probably originated in water.
• More than half of the world’s animal and plant
species live in the water.
• Almost 75% of the earth is covered in water.
• The human body needs 2 liters of water a day
in our climate; we can last only a few days
without water.
• Most of our food is water: tomatoes (95%), spinach
(91%), milk (90%), apples (85%), potatoes (80%),
beef (61%), hot dogs (56%).
THE SUN-POWERED CYCLE
The endless circulation of water from the atmosphere
to the earth and its return to the atmosphere through
condensation, precipitation, evaporation and transpiration
is called the hydrologic cycle.
Heating of the ocean water by the sun is the key
process that keeps the hydrologic cycle in motion. Water
evaporates, then falls as precipitation in the form of rain,
hail, snow, sleet, drizzle or fog. On its way to earth some
precipitation may evaporate or, when it falls over land, be
intercepted by vegetation before reaching the ground. The
cycle continues in three different ways:
• Evaporation/Transpiration— On average, as much
as 40% of precipitation in Canada is evaporated or
transpired.
• Percolation Into the Ground— Water moves down-
ward through cracks and pores in soil and rocks to
the water table. Water can move back up by capil-
lary action or it can move vertically or horizontally
under the earth’s surface until it re-enters a surface
water system.
• Surface Runoff— Water runs overland into nearby
streams and lakes; the steeper the land and the less
porous the soil, the greater the runoff. Overland flow
is particularly visible in urban areas. Rivers join
each other and eventually form one major river that
carries all of the subbasins’ runoff into the ocean.
Although the hydrologic cycle balances what goes up with
what comes down, one phase of the cycle is ‘‘frozen’’ in
the colder regions during the winter season. During the
Canadian winter, for example, most of the precipitation is
simply stored as snow or ice on the ground. Later, during
the spring melt, huge quantities of water are released
quickly, which results in heavy spring runoff and flooding.

FRESHWATER SERIES A-1
Note: A resource guide, entitled Let’s Not Take Water For
Granted, is available to help classroom teachers of grades
5–7 use the information from the Water Fact Sheets.
FREEZING AND SUPERCOOLING OF WATER
KARL ERIK ZACHARIASSEN
Norwegian University of
Science and Technology
Trondheim, Norway
Water in the crystalline solid state is known as ice, and
for pure water at normal atmospheric pressure, ice is the
physically stable form at temperatures below 0 ◦ C.
Ice has a specific gravity of 0.92 g/ml, which is about 8%
below the corresponding value of water, reflecting the fact
that water expands by about 8% when it enters the solid
state. The specific heat capacity of ice is 0.502 cal/g.deg,
which is about half the corresponding value of water. The
amount of heat liberated during the solidification (heat of
fusion) is 79.71 cal/g.
The transition of water from the liquid to the solid
state is known as freezing, and pure water has a so-
called freezing point of 0 ◦ C, which does not imply that
pure water will freeze when cooled to this temperature. A
small sample of pure water can be cooled to as much as
−40 ◦ C before it freezes spontaneously. For this reason, the
term freezing point is misleading, and it is often replaced
by the more precise terms melting point or equilibrium
freezing point.
A substance that exists in the liquid state at a
temperature below the equilibrium freezing point is said to
be supercooled or undercooled, and the temperature where
freezing is initiated is referred to as the supercooling point
or the temperature of crystallization.
When a sample of supercooled water freezes, the
released heat of fusion will immediately heat the system
to the melting point. As a result of the high heat of fusion
of water, the melting point temperature will be reached
following the freezing of only a small fraction of the sample.
From this point, the freezing can proceed only as fast as the
released heat of fusion can be removed from the system.
When pure water freezes, the temperature will remain at
0 ◦ C until all water is frozen.
Initiation of freezing requires that the water sample
contains an ice-like structure, which can act as a
nucleus for ice formation, from which ice can grow.
Liquid water is assumed to contain aggregates of water
molecules organized in an ice-like pattern, but which
usually are too unstable to act as nucleators. As the
temperature drops, these aggregates become more stable
and increase in size. On sufficient cooling, one of these
aggregates will eventually reach the size critical for
ice nucleation, and freezing will be initiated from this
aggregate. When nucleation is initiated in this manner
from the water molecules themselves, it is known as
homogeneous nucleation.
FREEZING AND SUPERCOOLING OF WATER 585
Water samples often contain other substances that
organize water molecules in an ice-like pattern and that
cause freezing to be initiated at a higher temperature than
would otherwise be the case. Such substances are known as
ice nucleators, and freezing catalyzed by an ice nucleator is
referred to as heterogeneous freezing. Ice nucleator activity
is seen in various types of crystals and dust particles,
and biological organisms may synthesize nucleating
lipoproteins, which are physiologically important as ice
nucleators, and which may have a high nucleating activity.
Certain bacteria produce ice nucleators that can nucleate
ice formations at above −1 ◦ C.
FACTORS AFFECTING FREEZING
The equilibrium freezing point is influenced by factors
such as solute concentration (Fig. 1) and hydrostatic
pressure (Fig. 2). The osmolal melting point depression is
1.86 ◦ C/Osm, whereas pressure depresses melting points
by about 1 ◦ C/100 bar.
The effect of these parameters on supercooling points
depends on whether nucleation is homogeneous or
heterogeneous. Homogeneous nucleation temperature is
depressed 1.5–3 times more than melting points by
increased osmolality and pressure, whereas, at least
for the more active nucleators, nucleation temperature
depression seems to be strictly equivalent to the
corresponding melting point depression.
Nucleation temperature of purified water samples (no
potent ice nucleators) also depends on sample volume
as shown by Bigg in 1953 (1) (Fig. 3). When potent
ice nucleators are present, the nucleation temperature
depends only on the activity of the most active nucleator,
and it does not matter if this nucleator occurs in a large or
small water volume.
0
Melting point
−5
−10
Nucleator-induced
supercooling point
−15
Temperature (°C)
−20
−25
−30
Supercooling point without
potent nucleators
−35
−40
0 500 1000 1500 2000
Osmolality (mOsm)
Figure 1.
586 CHEMICAL PRECIPITATION

2 CHEMICAL PRECIPITATION
0 Melting point
−2 BRADLEY A. STRIEBIG
Temperature (°C)

−4 Gonzaga University
−6 Spokane, Washington
−8
−10
−12 Chemical precipitation is the removal of dissolved and
−14 Nucleation temperature of
suspended solids from aqueous solutions through chemical
−16 different ice nucleators means. Chemical precipitation includes the removal of
−18 dissolved solids by altering the aqueous chemistry to
0 100 200 300 400 500 600 700 800 900 100011001200 make the dissolved solids less soluble. It also includes
Pressure (bar) removing suspended solids by increasing their size in a
process referred to as chemical flocculation. The larger the
Figure 2.
particle size of the suspended solids, the easier it is to
remove them, usually through gravimetric settling in a
0 process called clarification.

−5
USES OF CHEMICAL PRECIPITATION
Nucleation temperature (°C)

−10
−15
Curve of Bigg (1953)
−20
−25
−30
−35
−40
−45
−9 −8 −7 −6 −5 −4 −3 −2 −1
Log microlitre water volume
Figure 3.
FREEZING OF SOLUTIONS
When a solution freezes, the solutes are not incorporated
in the ice structure, but remain in the liquid fraction
surrounding the ice. Here, they become concentrated
precisely until the solution reaches a melting point equal
to the actual temperature. In this situation, vapor pressure
equilibrium will exist between the solution and the ice, and
no further freezing will occur until the released heat of
fusion is removed from the system. As the heat is removed
and the freezing proceeds, a gradual reduction of the
temperature of the system will occur as the fluid fraction
becomes more concentrated and the melting temperature
of the fluid fraction drops. The solute concentration will
increase until the eutectic point is reached. At the eutectic
point, water and solutes freeze together.
BIBLIOGRAPHY
1. Bigg, E.K. (1953). The supercooling of water. Proc. Roy. Soc.
London. B 66: 688–694.
READING LIST
Zachariassen, K.E. and Kristiansen, E. (2000). Ice nucleation and
antinucleation in nature. A review. Cryobiology 41: 257–279.
Chemical precipitation processes are utilized in water
and wastewater treatment due to their effectiveness for
pollutant removal with relatively low capital investment.
The chemical reaction rates are typically very fast
compared to biological reactions, so residence times in
treatment processes or the reactor size may be significantly
smaller than in other treatment techniques. Chemical
precipitation processes can also process very toxic
wastewaters. However, chemical precipitation processes
often require the on-going addition of chemicals to
control the pH and polymers to increase the size of
0 1 2 3 4 5
the particles. The residuals collected from chemical
precipitation processes may also be quite toxic and
expensive to dispose of as sludge. Chemical precipitation
is an effective tool for water treatment, and the most
common uses are briefly described.
Industrial wastewaters often require pretreatment
before being released to the sewer or receiving waters.
High strength industrial wastewaters, such as those used
in electroplating, contain high concentrations of dissolved
metals. These metals can be removed from the wastewater
by altering the pH of the water to reduce the solubility of
the metal ions. The insoluble fractions of the metal ions
form solid particles or precipitates, which are removed
through filtration or gravimetric clarification. This allows
industrial wastewaters to significantly reduce their metals
concentrations in order to meet RCRA standards and also
reduce the toxicity of the wastewater.
Municipal wastewaters may contain dissolved ions and
a large amount of cell mass from biological treatment
processes. The focus of chemical precipitation on municipal
wastewaters is to remove the cell mass or suspended solids
prior to and after biological treatment. The suspended
solids are removed by adding chemical polymers, which
attach to the existing suspended solids and help combine
the individual particles into larger particles called flocs.
This process, called coagulation, is accomplished by the
chemical destabilization of particles and the subsequent
aggregation. The process of gentle mixing and increasing
the size of the suspended solids particles is called
flocculation. The large floc particles are removed through
 CHEMICAL PRECIPITATION 587

gravimetric settling in a process called clarification. commonly estimated from measurements of the total
Thus a large fraction of the suspended solids and dissolved solids (TDS) and the electrical conductivity:
also the remaining BOD (in the form of cell mass)
Langelier approximation I = 2.5 × 10−5 (TDS)[mg/L]
can be removed with the aid of chemical precipitation.
Russell approximation I = 1.6 × 10−5 (conductivity)
Chemical precipitation is also commonly used when
[µ mho]
phosphorus removal is required at municipal wastewater
treatment plants. One estimate of the activity coefficient is the Debye–
Industrial process water and recycled wastewater use Hückel limiting law for solutions with ionic strength less
chemical precipitation to remove hardness and dissolved than 0.005 mol/L (I < 0.005 M):
salts from water. This allows a higher quality of water
needed for specific industrial processes. The limits of γA ∼
= −Az2 I1/2
hardness and salt removal are governed by the solubility
where A is a function of the dielectric constant of the
of the dissolved ions in solution. Therefore, chemical
solution. For water at 25 ◦ C, A = 0.51. The term z is the
precipitation has limited uses in industries that require
ionic charge of the ion of interest. For more concentrated
highly purified water. However, chemical precipitation can
solutions (I > 0.005 M), other estimates for the activity
be used as an inexpensive pretreatment step prior to ion
coefficient, such as the extended Debye–Hückel model
exchange or reverse osmosis processes, which produce very
and the Davies approximation can be used.
high-quality water.
The potential to precipitate multiple ions from solution
can be approximated from the activity of the ions in
DISSOLVED SOLIDS solution and the solubility product. Often, the pH of
the solution is used as a ‘‘master variable.’’ Solids lines
Dissolved solids are defined as the solid residuals in water
on a graph of this type indicate the solids equilibrium
that pass through a filter with a nominal pore size of
concentration. For most hydroxide species, a minimum
2.0 µm. These solids include dissolved ions and small
soluble concentration appears between a pH of 8 and 12.
colloidal solids. The dissolved ions in solution can be
Similar graphical representations can be made for sulfide
predicted based on the solubility product constant for
and carbonate species. This minimum solubility can serve
the dissolved species. The solubility product constant,
as a starting point for estimating and optimizing chemical
or solubility product, is an equilibrium constant that
precipitation of various ions in solution.
describes the reaction by which ions dissolve and solids
The removal of hexavalent chromium (Cr6+ ) from solu-
precipitate:
tion is one illustration of the potential of chemical pre-
Az By(s) −−
−−
y+
−− zA + yB
z−
cipitation processes to reduce the toxicity of wastewater.
For ions A and B, the generic equilibrium constant is Hexavalent chromium can be converted to the less toxic
trivalent form by reaction at low pH with a reducing agent
{Ay+ }z {Bz− }y such as ferrous bisulfite. Figure 1 shows a solution con-
Ks =
{Az By }(s) taining hexavalent chromium, which when dissolved in
water results in a yellow color. Figure 2 shows the change
Many salts will form and precipitate as either hydroxides, in color associated with the change in the oxidation state of
with the hydroxide ion (OH− ), or carbonates, with the the chromium from the 6+ state to the 3+ state. When the
carbonate ion (CO3 2− ). The concentration of these ions pH of the solution is raised to 8 or 9, the trivalent chrome
can often be predicted through an equilibrium analysis becomes much less soluble in solution than the hexavalent
with the aid of electron or pH diagrams. Since the activity form of chromium as shown in Fig. 3. The resulting solids
of the solid substance is defined as 1, and the solubility precipitate out of solution to form a concentrated sludge
product is a constant value, the activity of the anion of as shown in Fig. 4.
interest (A) can be predicted in the aqueous solution of
interest. Solubility constants for a wide array of dissolved SUSPENDED SOLIDS
ions of interest in water and wastewater are found in the
referenced literature (1–3). Suspended solids are composed of floating matter,
In a dilute solution under ideal conditions the activity settleable matter, and large colloidal matter. This matter
of the ions of interest can be closely approximated by their
concentration:
{A} ≈ [A]

However, many wastewaters have high concentrations

of dissolved ions and do not behave ideally. An activity
coefficient, γA , must be used to establish an accurate
relationship between the activity of a compound in solution
and the concentration:

{A} ≈ γA [A]

The activity coefficient can be estimated from the ionic

strength of the solution. The ionic strength is also Figure 1. Hexavalent chromium dissolved in water.
588 CHEMICAL PRECIPITATION
Figure 2. Reduction of hexavalent chromium to trivalent
chromium.
Figure 3. Formation of insoluble chromium oxide salts.
Figure 4. Precipitation of solid insoluble chromium into a
sludge blanket.
may be organic or inorganic in nature. In municipal
wastewater treatment plants, a significant fraction of
the suspended solids consist of biological cell mass. The
amount of suspended solids in water is an important
parameter in determining water quality. Suspended solids
are typically removed through a process of coagulation,
flocculation, and gravimetric settling (or clarification).
Chemical precipitation or coagulation of the suspended
particles involves the destabilization of colloidal particles.
Coagulation is usually followed by flocculation, which is
the addition of a chemical, typically an organic polymer,
to increase particle aggregation.
Suspended solids in water often have a charge that acts
at the surface of the particle. Oil droplets, gas bubbles,
and other inert substances may have a negative charge
due to the preferential absorption of anions. Cellular
materials such as microorganisms and proteins develop
a charge through the ionization of carboxyl and amino
acids functional groups. These charges create a force or
zeta potential around the particles, which prevent them
from increasing in size through aggregation as illustrated
in Fig. 5. Coagulants are added to overcome these surface
forces by the addition of potential determining ions or
electrolytes. Once these charges have been neutralized,
flocculants can be added to increase the speed of
aggregation (Fig. 6).
Flocculant polymers that are anionic and nonionic
become attached to the particles. These long particles
increase the rate of aggregation and become intertwined
Particle
Particle
Repulsive
forces
Particle
Figure 5. Repulsive forces acting on the surface of sus-
pended particles.
Particle
Particle
Slow mixing
brings particles
Coagulants overcome into contact
repulsive forces with one another
Particle
Figure 6. Chemical coagulants overcome repulsive forces and
the solution is mixed.

Particle
Particle
Flocculants
increase
Particle
the rate of
aggregation
and particle stability Floc
Figure 7. Flocculants increase the rate of aggregation to form
large flocs.
during flocculation. This results in a significant increase
in the particle size and mass (Fig. 7). The large particles
called flocs are stabilized by the polymers, which prevent
them from breaking up during the mixing process. The
large flocs can then be removed with gravimetric settling
in the clarification process. The resultant solids collected
in the clarifier are the wastewater solids, or sludge.
Chemical precipitation by coagulation and flocculation
has been used in the past to enhance removal of suspended
solids and BOD. These processes are widely used to
improve the performance of primary clarifiers, in basic
physical–chemical processing of industrial wastewaters,
for phosphorus removal, and for heavy metal removal.
BIBLIOGRAPHY
1. Tchobanoglous, G., Burton, F.L., and Stensel, H.D. (2003).
Wastewater Engineering Treatment and Reuse, 4th Edn.
Metcalf and Eddy, Inc/McGraw Hill, Boston, pp. 42–52,
475–517.
2. Stumm, W. and Morgan, J.J. (1996). Aquatic Chemistry
Chemical Equilibria and Rates in Natural Waters, 3rd Edn.
John Wiley & Sons, Hoboken, NJ, pp. 252–398.
3. Jensen, J.N. (2003). A Problem-Solving Approach to Aquatic
Chemistry. John Wiley & Sons, Hoboken, NJ, pp. 418–450.
ANTIMONY IN AQUATIC SYSTEMS
MONTSERRAT FILELLA
University of Geneva
Geneva, Switzerland
INTRODUCTION
Antimony is a naturally occurring element. It belongs to
group 15 of the periodic table of the elements. Antimony
can exist in a variety of oxidation states (−III, 0, III,
V) but it is mainly found in two oxidation states (III
and V) in environmental, biological, and geochemical
ANTIMONY IN AQUATIC SYSTEMS 589
samples. Although antimony was already known to the
ancients, it is still often overlooked, both as an element of
environmental concern and as a subject for study, probably
because of its lower abundance and the relative insolubility
of most of its compounds. This is reflected in the poor
standard of existing data on the behavior of antimony
in natural waters. However, interest in the study of this
element seems to be growing and an increasing number of
papers on the subject are being published. A useful series
of comprehensive reviews on antimony in the environment
has recently been published (1,2).
Antimony has no known biological function and, like
arsenic, is toxic. Trivalent species are reported to be more
toxic than pentavalent forms (3). The U.S. Environmental
Protection Agency (EPA) (4) and the Commission of the
European Communities (5) both consider antimony and
its compounds to be pollutants of prime interest. The U.S.
EPA drinking water standards are: MCLG (maximum
contaminant level goal) and MCL (maximum contaminant
level), both 6 µg/L (6). The European Union’s maximum
admissible concentration of antimony in drinking water is
5 µg/L (7). Antimony is on the list of hazardous substances
under the Basel convention on the restriction of the
transfer of hazardous waste across borders (8).
ABUNDANCE
The relative abundance of antimony in different systems
is shown in Table 1. According to Goldschmidt’s classifica-
tion, antimony is a strong chalcophile element of which the
most common natural form is the sulfide mineral stibnite,
Sb2 S3 . Antimony is occasionally found as a native metal or
as an intermediate compound with arsenic. More than 113
different sulfide, antimonide, arsenide, and telluride min-
erals have been identified. Sulfosalts represent the most
common antimony-bearing minerals, usually in some com-
bination with silver, copper, arsenic, mercury, and lead.
More than 35 oxide, hydroxide, and oxyhalide mineral
species are known. Antimony is also found in seven sili-
cate minerals, all of which are rare (9). Antimony is also a
common component of coal and petroleum.
Antimony is present in the aquatic environment as a
result of rock weathering, soil runoff, and anthropogenic
activities. Typically, concentrations of dissolved antimony
in unpolluted waters are less than 1 µg/L. The mean
antimony concentration in surface marine waters is 184 ±
45 ng/L (1). However, in the proximity of anthropogenic
sources, concentrations can reach up to 100 times
natural levels.
Airborne supply to aquatic and terrestrial systems
is important for the environmental fate of antimony
in systems located far from direct sources of pollution.
Atmospheric emission values for antimony are shown in
Table 2. The values in this table should be viewed with
some caution because the data used comes from relatively
old studies. For instance, antimony production was
considered to be only 80,000 metric tons/yr (142,000 metric
tons in 2003) and, more importantly, the uses of antimony
have changed significantly over the past few years. Pacyna
and Pacyna (15) estimated natural and anthropogenic
emissions into the atmosphere to be 2400 tons/yr and
590 ANTIMONY IN AQUATIC SYSTEMS

Table 1. Antimony Geochemical element have exceeded natural fluxes for more than 2000
Abundance years. The present day enrichment factor (relative to the
Material Sb Abundancea element/Sc ratios of typical crust rocks) is on the order of
70 times (as compared to 20 for As and 130 for Pb) (18).
Cosmic abundance 0.246b The chronology and intensity of the antimony enrichments
Chondrites 0.1
in peat are similar in many ways to those of lead, which
Mean crustal average 0.2
Upper continental crust 0.31 indicates that the extent of the human impact on the
Lower continental crust 0.30 geochemical cycle of antimony is comparable to that of
Terrestrial abundance 0.7 lead (16).
Basic rocks (basalt) 0.15 The production and use of antimony have steadily
Granitic rocks 0.2 increased. World reserves of antimony are in excess of
Sedimentary rocks 2 million tons and are located principally in China,
Shales 1–2 South Africa, Bolivia, Russia, and Tajikistan (19). The
Carbonates 0.2
world’s mines produced 142,000 tons of antimony in
Deep sea clays 1
2003 (19). Table 3 shows the main uses of primary
a
All values in ppm (mg/g), except for the antimony in the United States in 2003, which can be
cosmic abundance. seen as representative of the use of antimony across the
b
Atomic abundance relative to Si (=1,000,000)
deduced from the following mean composition
developed world. Uses of antimony have changed over
of meteorites: silicate = 100, sulfide = 7, metal = the years. Traditionally, it was used in lead–antimony
10.6 parts per weight. alloys. Bulk-secondary antimony could be recovered as
Source: Adapted from Reference 1. antimonial lead, most of which was regenerated and then
consumed by the battery industry. Today, the main use of
antimony is as Sb2 O3 in flame-retardants, which have
1600 tons/yr, respectively. Very recently, Shotyk and applications in, among others, children’s clothing and
co-workers (16) dismissed all previous values of natural toys and aircraft and automobile seat covers. Antimony
antimony releases as being grossly overestimated. Based trioxide cannot be recycled and is released into the
on antimony concentrations in ombrotrophic peat from environment.
Switzerland and Scotland, these authors suggested that
90–154 metric tons/yr would be more appropriate values.

USES Table 3. Reported Industrial

Consumption (20) of Primary Antimony in
The first recorded use of antimony as a pigment to make the United States in 2002 (Metric Tons of
mascara is found in an Egyptian papyrus dating from 1600 Antimony Content)a
B.C. According to Emsley (17), the first uses of antimony Uses Sb Amountb
date back even further; part of a 5000-year-old vase now
Metal products
in the Louvre is made of almost pure antimony. A peat
Antimonial lead 910
core from an ombrotrophic Swiss bog revealed significant
Bearing metal and bearings 43
enrichments of antimony dating back to Roman times, Solder 85
which indicates that the anthropogenic fluxes of this Ammunition, cable 1,370
covering, castings, sheet
and pipe, and type metal
Table 2. Global Emissionsa of Antimony into the Total 2,410
Atmosphere, 103 metric tons/yr Nonmetal products
Ammunition primers W
Sb Anthropogenic Sb
Ceramics and glass 487
Natural Sources Emission Sources Emission
Pigments 597
Windborne dust 0.8 Energy production 1.3 Plastics 532
(includes industrial (coal, oil, gas) Fireworks and rubber products 488
sources of dust) Total 2,100
Sea salt spray 0.6 Mining 0.1 Flameretardants
Volcanic activity 0.7 Smelting and 1.4 Adhesives W
refining Plastics 3,680
Forest fires 0.2 Waste incineration 0.7 Rubber 126
Biogenic sources 0.3 Textiles 150
Total 2.6 Total 3.5 Paper and pigments 767
Total 4,720
a
Adapted from Reference 1. Original estimations come from References Grand Total 9,230
10–13. A completely different estimation, which is often cited in the
a
literature, was published by Lantzy and Mackenzie in 1979 (14). They Secondary antimonial lead was 5600 metric tons in
estimated natural and anthropogenic emissions of antimony into the 2003.
atmosphere to be 0.98 × 103 metric tons/yr and 38 × 103 metric tons/yr, b
W = Withheld to avoid disclosing company propri-
respectively. etary data.

ANTIMONY SOLUTION CHEMISTRY
Both Sb(III) and Sb(V) ions hydrolyze easily in aqueous
solution, thus making it difficult to keep antimony ions
stable in solution except in highly acidic media. Since most
of the techniques used for the study of solution equilibria
(solubility measurements, potentiometry, voltammetry,
spectroscopy) necessitate the use of macroamounts of
reactants, relatively little information is available on
the chemical species of antimony that exist in aqueous
solution, and on their stability constants in particular.
Moreover, most of the published studies are relatively old.
Only the system antimony–sulfur has been the object of
recent studies, mainly in relation to the understanding of
hydrothermal systems.
The following main points can be made regard-
ing the antimony inorganic chemistry in solution (2):
(1) According to thermodynamic data, antimony should
exist as Sb(V) in oxic systems and as Sb(III) in anoxic
ones. (2) At environmental concentration levels, at the
pH range commonly found in natural waters and in the
absence of sulfur, antimony exists as a soluble species
regardless of its oxidation state. Antimony(V) is present
as a negative species in solution [Sb(OH)6 − , also repre-
sented as SbO3 − ] and Sb(III) as a neutral species [Sb(OH)3 ,
also often quoted as HSbO2 ]. (3) It is not clear whether
Sb(V) and Sb(III) form any chloride-containing species in
seawater, but hydroxychloro species would be the most
probable entities. (4) Antimony oxides exhibit low water
solubility, which may affect the fate of antimony in soils
and water, particularly since today antimony is mostly
used as Sb2 O3 (s) in flame-retardant applications. More-
over, it is thought that, when released into the atmosphere
as an aerosol, antimony is oxidized to antimony trioxide
through reactions with atmospheric oxidants. (5) In anoxic
systems, and in the presence of sulfur, antimony forms
insoluble stibnite, Sb2 S3 (s), and soluble SbS2 2− , depending
on the pH. The formation of polymerized species at natural
concentration levels (e.g., Sb2 S4 2− , Sb2 S5 4− , Sb4 Sb7 2− ) is
not proved. The formation of an Sb(V)-containing species
(SbS4 3− ) has been observed by EXAFS.
Eh–pH diagrams can illustrate the redox behavior of
elements like antimony. Figure 1 shows one such diagram
for the system Sb–S–H2 O at environmentally realistic
concentrations of antimony and dissolved sulfur.
A comprehensive, thermodynamically consistent anti-
mony database of equilibrium constants of antimony with
inorganic ligands has recently been published (21).
Barely any data have been reported on Sb(V) binding
by low molecular weight organic ligands. More extensive
studies have been carried out on the binding of Sb(III),
which behaves like a borderline metal. The extent of
antimony binding by natural organic matter (NOM) is
a point of some debate. A few studies report significant
Sb–NOM interactions, but most do not, and observed
environmental behavior seems to suggest that Sb–NOM
interactions are not relevant in most aquatic systems.
Hydrous oxides of manganese and iron have been
reported to significantly sorb Sb(III) but very few
systematic studies have been carried out on antimony
sorption by pure metal oxides. These oxides will also
ANTIMONY IN AQUATIC SYSTEMS 591
1.3
1.0
O
2 (g
)
Sb(OH)2+
0.5
Sb(OH)6−
Eh (V)
Sb(OH)4−
0.0 Sb(OH)3
Sb2S3(s)
H (SbS2)−
2 (g
)
−0.5
−0.9
0 2 4 6 8 10 12
pH
Figure 1. Eh–pH diagram of antimony in the Sb–S–H2 O
system at a dissolved antimony concentration of 10−8 mol/L and
a dissolved sulfur concentration of 10−3 mol/L. Adapted from
Reference 2.
oxidize the trivalent antimony to pentavalent antimony.
Several studies have found a correlation between the
adsorption of antimony in soil and oxic sediments and
the presence of manganese and iron oxides (22,23).
ANALYTICAL METHODS
The total concentrations of antimony in aqueous and
solid environmental samples can readily be determined
using current analytical techniques, primarily the result
of the development of microwave digestion techniques
for solid samples and NAA and plasma-based analyte
detection systems. Both ICP-AES and ICP-MS techniques
are in widespread use, although ICP-MS has less spectral
interference and a lower detection limit (ng/L). An
excellent critical review of the methodologies currently
available for the environmental analysis of antimony
has recently been published (24). However, determining
antimony speciation still presents a significant analytical
challenge (25). Recent data are almost exclusively based
on hydride generation methods and often only oxidation
state data are obtained. A lack of stability in the antimony
species being investigated, particularly in trivalent forms,
throughout the entire analytical process may result in an
overestimation of pentavalent species concentrations.
Determining methylated antimony species is hampered
by severe molecular rearrangements originating from the
demethylation of trimethylantimony during the analysis.
This problem has been the subject of extensive literary
discussion but a solution has yet to be found (26).
592 ANTIMONY IN AQUATIC SYSTEMS
ANTIMONY CONCENTRATIONS IN THE DIFFERENT
AQUATIC COMPARTMENTS
The distribution and behavior of antimony in freshwater
systems have not been studied extensively. Published
concentrations range from a few nanograms/liter to a few
micrograms/liter depending on location. They reflect the
wide range of physical and chemical conditions existing in
freshwater bodies and are very sensitive to the proximity
of sources of pollution.
The concentration of antimony in ocean surface waters
is about 200 ng/L. The behavior of antimony in the open
seas is not considered to be highly reactive. According
to some authors, it behaves as a conservative element
and to others as a mildly scavenged element with surface
(atmospheric) input. The comparison between antimony
contents in deep waters of different oceans indicates
that no antimony accumulation is produced during deep-
water oceanic circulation (27). Antimony concentrations
seem to reflect regional geochemical characteristics in
coastal environments. Antimony behavior in estuaries
varies somewhat according to the characteristics of the
estuary. In some cases, it behaves conservatively and in
others it shows a midestuary maximum. Behavior has
even been seen to vary in the same estuary from one year
to another.
Antimony concentrations in sediments are on the
order of a few micrograms/gram of sediment and rarely
exceed 1 mg/g. Higher concentrations are directly related
to anthropogenic sources, mainly proximity to smelting
plants and outfalls of sewage and fertilizer facilities.
ANTIMONY SPECIATION IN NATURAL WATERS
Today it is widely accepted that an understanding
of biogeochemical processes requires knowledge of the
chemical forms, or species, that are present in the natural
environment. Despite this, the speciation of antimony in
natural aquatic systems has yet to form the subject of
many in-depth studies.
Dissolved Versus Particulate Fractions
Weak interaction between anionic species and suspended
particles negatively charged in natural waters is to be
expected and it is generally accepted that antimony is
present as ‘‘dissolved’’ in aquatic systems. Experimental
data (2) exist to support this. Fractionation of the so-called
dissolved phase into several classes by ultrafiltration
showed that most of the dissolved antimony is present
in the smallest molecular weight region (2,28).
Redox
Antimony occurs in two oxidation states in natural waters
and, thus, its behavior and toxicity can be affected by
changes in the redox status of the aquatic environment.
Antimony speciation studies have focused mainly on the
separation and identification of Sb(III) and Sb(V) species.
Antimony (V) is the predominant species found in oxy-
genated systems but thermodynamically unstable Sb(III)
has also been detected in different oxic marine water,
fresh waters, groundwaters, and rain waters. This con-
trasts with thermodynamic equilibrium predictions, which
suggest that Sb(V) concentration should exceed Sb(III)
by several orders of magnitude. For thermodynamically
unstable species to be present in water, mechanisms for
their production and slow rates of interconversion must
exist. The presence of Sb(III) has often been attributed
to phytoplankton activity. However, although possible, a
detailed analysis of published data has so far failed to pro-
duce conclusive proof to validate this hypothesis. Recently,
the photochemical reduction of Sb(V) to Sb(III) has been
suggested as an alternative to the phytoplankton source.
The true speciation of antimony under anoxic conditions
remains unclear. According to thermodynamic calcula-
tions, in the absence of oxygen, antimony should only be
present in the trivalent form. However, the occurrence of
pentavalent forms of antimony has been reported in dif-
ferent anoxic systems. A variety of mechanisms have been
suggested to explain these results, including delivery of
Sb(V) on sinking detritus from oxic waters, formation of
thiocomplexes by the pentavalent element, and advection
of surface waters containing high concentrations of anti-
monate. All these mechanisms must be accompanied by
relatively slow rates of reduction.
Methylated Species
Methylated species have been detected in surface sea-
waters, in estuaries, and in some rivers. They usually
account for 10% or less of the total dissolved antimony. The
monomethyl species is more abundant than the dimethyl
form and close to the surface values tend to be higher.
Monomethyl, dimethyl, trimethyl, and triethyl antimony
derivatives have been detected in sediments from rivers
and harbors. However, the identification of methylated
antimony species has been questioned because some of
these compounds undergo rearrangement reactions when
subjected to the hydride generation procedures typically
used for aquatic speciation.
The organoantimony compounds detected by Dodd
et al. (26) in a freshwater plant (Potamogetan pectinatus)
and a stibnolipid isolated from a marine diatom by Ben-
son (29) are, as yet, the only organoantimony compounds
with unequivocal biological origin to be detected in aquatic
environments.
KINETIC ISSUES
As mentioned, the biogeochemical cycle of antimony is
more complicated than equilibrium thermodynamic calcu-
lations predict (1). Several authors have cited the kinetic
stabilization of thermodynamically unstable species to
explain their presence in natural systems. Several pro-
cesses may influence oxidation and reduction rates, includ-
ing chemical and photochemical transformations and bio-
logical mediated reactions. It is well known that, when a
redox process is accompanied by hydrolysis reactions, as
is likely to be the case for antimony, the overall process
can be much slower than expected. At present, kinetic
information for redox reactions of antimony in natural
waters is limited and rate constants are, for the most

part, unknown. Cutter estimated an overall Sb(III) oxida-
tion rate of 0.008 d−1 in the 65–85 m layer of the Black
Sea from depth profiles (30). However, antimony residence
times in seawater are probably longer because this rate
included all forms of removal; that is, Sb(III) oxidation and
also Sb(III) scavenging by sedimenting particles. More-
over, it was calculated at the oxic–suboxic interface of
the Black Sea, where the presence of manganese and iron
oxides is likely to increase the Sb(III) oxidation rate (31).
Indeed, Belzile and co-workers (32), in a study on the oxi-
dation of Sb(III) in the presence of natural and synthetic
iron and manganese oxyhydroxides, showed that Sb(III)
can be rapidly oxidized to Sb(V) by both compounds. Iron
oxides also proved to play an important role in antimony
redox chemistry in lake sediments (23). However, iron
and manganese oxides are not likely to be present in
oxic waters in significant quantities and thus are unlikely
to influence antimony chemistry in oxic systems. It has
recently been shown that Sb(III) oxidation by hydrogen
peroxide and iodate is strongly pH dependent: the neutral
Sb(OH)3 species is unreactive to hydrogen peroxide and
the presence of Sb(OH)4 − is needed for the oxidation to
occur (33,34).
The addition of organic substances such as tartaric,
lactic, citric, and ascorbic acids to natural or synthetic
solutions has been shown to have a stabilizing effect
on Sb(III). Solutions of Sb(III) prepared from potassium
antimonyl tartrate have been reported to remain stable
for long periods. The organic matter present in natural
waters probably has a similar stabilizing effect. Ligand
stabilization has been used to explain the apparent
presence of Fe(II) in oxic solutions that contain relatively
high levels of organic matter.
Little information is available on Sb(V) reduction
kinetics in anoxic media. The reduction rate for Sb(V) in
water at the bottom of the Black Sea was estimated to be
1.1 × 10−6 d−1 (35).
ANTIMONY AND AQUATIC BIOTA
Reported concentrations for antimony in fresh water and
marine algae range from 0.02 to 1 µg/g dry weight (1).
Although algae bioaccumulation and detoxifying mecha-
nisms may play a key role in antimony redox speciation
in natural waters, few studies have been carried out.
Kantin (36) collected three marine algae near San Diego
Bay in the United States and Sb(V) was the dominant
species in all of them. Only Sargassum sp. contained
up to 30% of Sb(III), thus demonstrating an ability to
form reduced antimony species. Andreae and Froelich (37)
analyzed the chemical speciation of antimony in phyto-
plankton (largely diatoms) from the eastern North Pacific
and found that the trivalent species formed a significant
part of the inorganic antimony. Both studies have since
been used to justify the frequently cited biological origin
of the Sb(III) found in oxic waters. Although no detailed
studies exist, antimonate is expected to behave differently
in bioavailability than arsenate and phosphate because
of the weaker Lewis acidity and larger ionic radius of
the antimonate ion in solution. More recently, the algae
ANTIMONY IN AQUATIC SYSTEMS 593
Chlorella vulgaris, isolated from an arsenic-polluted envi-
ronment, excreted 40% Sb(V) and 60% Sb(III) on exposure
to Sb(III), suggesting that a change in oxidation state is
used by this alga as a detoxifying mechanism (38).
BIBLIOGRAPHY
1. Filella, M., Belzile, N., and Chen, Y-W. (2002). Antimony in
the environment: a review focused on natural waters. I.
Occurrence. Earth Sci. Rev. 57: 125–168.
2. Filella, M., Belzile, N., and Chen, Y-W. (2002). Antimony in
the environment: a review focused on natural waters. II.
Relevant solution chemistry. Earth Sci. Rev. 59: 265–285.
3. Bencze, K. (1994). Antimony. In: Handbook on Metals in
Clinical and Analytical Chemistry. H.G. Seiler, A. Sigel, and
H. Sigel (Eds.). Marcel Dekker, New York, pp. 227–236.
4. United States Environmental Protection Agency. (1979).
Water Related Fate of the 129 Priority Pollutants. Vol. 1,
EP-440/4-79-029A, U.S. EPA, Washington, DC.
5. Council of the European Communities. (1976). Council
Directive 76/464/EEC of 4 May 1976 on pollution caused by
certain dangerous substances discharged into the aquatic
environment of the Community. Official Journal L 129,
18/05/1976, pp. 23–29.
6. United States Environmental Protection Agency. (1999).
National Primary Drinking Water Standards. Doc. 810-F-
94-001, U.S. EPA, Washington, DC.
7. Council of the European Union. (1998). Council Directive
98/83/EC of 3 November 1998 on the quality of water intended
for human consumption. Official Journal L 330, 05/12/1998,
pp. 32–54.
8. United Nations Environmental Program. (1999). Basel
Convention on the Control of Transboundary Movements of
Hazardous Wastes and Their Disposal (with amended Annex
I and two additional Annexes VIII and IX, adopted at the
fourth meeting of the Conference of the Parties in 1998), SBC
No. 99/001. UNEP, Geneva.
9. Lueth, V.W. (1999). Antimony: element and geochem-
istry. In: Encyclopedia of Geochemistry. C.P. Marshall and
R.W. Fairbridge (Eds.). Kluwer Academic Publishers, Dor-
drecht, The Netherlands, pp. 15–16.
10. Nriagu, J.O. and Pacyna, J.M. (1988). Quantitative assess-
ment of worldwide contamination of air, water and soils by
trace metals. Nature 333: 134–139.
11. Nriagu, J.O. (1989). A global assessment of natural sources
of atmospheric trace metals. Nature 338: 47–49.
12. Nriagu, J.O. (1990). Global metal pollution. Poisoning the
biosphere? Environment 32: 7–11, 28–33.
13. Maeda, S. (1994). Safety and environmental effects. In:
The Chemistry of Organic Arsenic, Antimony and Bismuth
Compounds. S. Patai (Ed.). Wiley, Hoboken, NJ, pp. 725–759.
14. Lantzy, R.J. and Mackenzie, F.T. (1979). Atmospheric trace
metals: global cycles and assessment of man’s impact.
Geochim. Cosmochim. Acta 43: 511–525.
15. Pacyna, J.M. and Pacyna, E.G. (2001). An assessment of
global and regional emissions of trace metals to the
atmosphere from anthropogenic sources worldwide. Environ.
Rev. 9: 169–298.
16. Shotyk, W. et al. (1996). Two thousand years of atmospheric
arsenic, antimony, and lead deposition recorded in an
ombrotrophic peat bog profile, Jura Mountains, Switzerland.
Earth Planet. Sci. Lett. 145: E1–E7.
17. Emsley, J. (2001). Nature’s Building Blocks. An A–Z Guide
to the Elements. Oxford University Press, New York.
594 ANTIMONY IN AQUATIC SYSTEMS
18. Shotyk, W., Krachler, M., and Chen, B. (2004). Antimony in
recent, ombrotrophic peat from Switzerland and Scotland:
comparison with natural background values (5,320 to
8,02014 C yr BP) and implications for the global Sb cycle.
Global Biogeochem. Cycles 18: GB1016.
19. Carlin, J.F., Jr. (2004). Antimony. In: U.S. Geological Survey
Mineral Commodity Summaries.
20. Carlin, J.F. (2003). U.S. Geological Survey Minerals Year-
book.
21. Filella, M. and May, P.M. (2003). Computer simulation of
the low-molecular-weight inorganic species distribution of
antimony(III) and antimony(V) in natural waters. Geochim.
Cosmochim. Acta 67: 4013–4031.
22. Müller, B. et al. (2002). As, Sb, Mo, and other elements in
sedimentary Fe/Mn layers of Lake Baikal. Environ. Sci.
Technol. 36: 411–420.
23. Chen, Y-W., Deng, T-L., Filella, M., and Belzile, N. (2003).
Distribution and early diagenesis of antimony species in
sediments and porewaters of freshwater lakes. Environ. Sci.
Technol. 37: 1163–1168.
24. Nash, M.J., Maskall, J.E., and Hill, S.J. (2000). Methodolo-
gies for determination of antimony in terrestrial environmen-
tal samples. J. Environ. Monit. 2: 97–109.
25. Krachler, M., Emons, H., and Zheng, J. (2001). Speciation of
antimony for the 21st century: promises and pitfalls. Trends
Anal. Chem. 20: 79–90.
26. Dodd, X. et al. (1996). Antimony speciation in freshwater
plant extracts by using hydride generation–gas chromatog-
raphy–mass spectrometry. Analyst. 121: 223–228.
27. Cutter, G.A. and Cutter, L.S. (1998). Metalloids in the high
latitude North Atlantic Ocean: sources and internal cycling.
Mar. Chem. 61: 25–36.
28. Stolpe, X., Hassellöv, M., Andersson, K., and Turner, D.
(2004). High resolution ICPMS as an on-line detector for
flow-field-flow fractionation; multi-element determination of
colloidal size distributions in a natural water sample. Anal.
Chim. Acta. in press.
29. Benson, A. (1988). Antimony metabolites in marine algae. In:
The Biological Alkylation of Heavy Metals. P.J. Craig and
F. Glocking (Eds.). Special Publication, Vol. 66, Royal Society
of Chemistry, London, pp. 135–137.
30. Cutter, G.A. (1992). Kinetic controls on metalloid speciation
in seawater. Mar. Chem. 40: 65–80.
31. Cutter, L.S., Cutter, G.A., and San Diego-McGlone, M.L.C.
(1991). Simultaneous determination of inorganic arsenic
and antimony species in natural waters using selective
hydride generation with gas chromatography/photoionization
detection. Anal. Chem. 63: 1138–1142.
32. Belzile, N., Chen, Y-W., and Wang, Z. (2001). Oxidation of
antimony(III) by amorphous iron and manganese oxyhydrox-
ides. Chem. Geol. 174: 379–387.
33. Quentel, F., Filella, M., Elleouet, C., and Madec, C-L. (2004).
Kinetic studies on Sb(III) oxidation by hydrogen peroxide
in natural waters. Environ. Sci. Technol. 38: 2843–
2848.
34. Quentel, F., Filella, M., Elleouet, C., and Madec, C-L. (2004).
Sb(III) oxidation by iodate in seawater: a cautionary tale.
(Submitted).
35. Cutter, G.A. (1991). Dissolved arsenic and antimony in the
Black Sea. Deep-Sea Res. 38: S825–S843.
36. Kantin, R. (1983). Chemical speciation of antimony in marine
algae. Limnol. Oceanogr. 28: 165–168.
37. Andreae, M.O. and Froelich, P.N. (1984). Arsenic, antimony,
and germanium biogeochemistry in the Baltic Sea. Tellus 36:
101–117.
38. Maeda, S. et al. (1997). Bioaccumulation of antimony by
Chlorella vulgaris and the association mode of antimony
in the cell. Appl. Organomet. Chem. 11: 393–396.