EnvironmentI&mational, Vol. 22, No. 5, pp.

507-517,1996
Pergamon
Copytight 01996 Elsevis Science Ltd
Printedin the USA. All rights reserved
0160-4120/96 S15.00+.00

PII S160-4120(96)00040-2

INFLUENCE OF HUMIC SUBSTANCES ON THE

AQUATIC ADSORPTION OF HEAVY METALS ON
DEFINED MlNElRAL PHASES

F.H. Frimmel and L. Huber

Engler-Bunk-lnstitut, University of Karlsruhe, 76128 Karlsruhe, FRG

EI 9507-375 M (Received 26 April 1996; accepted 28 April 1996)

The distribution of heavy metals in the aqueous phase of aquifers is strongly dependent on the kind
of solid phase and on the presence of dissolved organic matter. Batch experiments were performed
to investigate the phase distribution of Cd, Pb, and Cu using Quartz, Potassium Feldspar, Sodium
Feldspar, Kaolinite, Calcite, and Fe&m-coated Quartz sand as mineral phases. Natural organic
matter (NOM) from a brown water lake was used to study its influence on the phase distribution of
the metals. From the experimental results, Freundlich isotherms were calculated. They turned out
to be suited for a description of the heterogeneous systems. Despite the obvious limitations of
mechanistic interpretation, NOM increased the dissolved fraction of Cu and Pb, and decreased that
of Cd.

INTRODUCTION

The protection of groundwater is important for safe-
guarding the drinking water supply. Besides organic
micropollutants, heavy metals are a risk for water
quality. Therefore, it is essential to know about the dis-
tribution and transport of metals in aquifers. Only from
a detailed understanding of these processes can mea-
sures to protect aquatic systems be derived. There have
been numerous attempts to determine and calculate the
interactions of heavy metals with sediments and rock
formations. Adsorption, ion exchange, and precipitation
or surface complexation have been identified as major
mechanisms (Farley et al. 1985; KorJ 1992). Natural soil
and sediment samples, clay minerals and well defined
solid phases like SiO,, Ah%, or goethite have been
used to investigate the distribution of heavy metals and
their dissolved species(Gerthand Brtinner 1983; Zasoski
and Burau 1988; StSver and Roennefahrt 1992).
Model calculations have shown that it is still difficult
to apply the data obtained for well defined systems to
the complex mixtures in nature. This is most obvious for
cases where the ubiquitous natural organic matter
(NOM) is taken into account (Buffle 1989).
The role of protons in complexation reactions of
metals was studied thoroughly by Perdue (1990).
Tipping (1986) focused on the phase distribution of
humic substances caused by iron oxides and hydroxides,
and Marinski and Ephraim (1986) applied the poly-
electrolyte model to explain the interactions of NOM
and metals.
Recently, Benedetti et al. (1995) showed that copper
competes much more effectively with protons bound to
the phenolic type groups of NOM than calcium and
cadmium. An attractive approach to improve the predic-
tive power of models uses phenomenological and
508
mechanistic parts. The application of adsorption iso-
therms seems to be a relatively uncomplicated way
which leads to useful information on the phase distri-
bution of water constituents. This has been demon-
strated mainly for the elimination of organic micro-
pollutants and NOM with activated carbon (Sontheimer
et al. 1988; Johannsen and Worth 1994).
It is promising to gain data sets also on the behaviour
of metals in systems with NOM and defined mineral
components of aquifer material, and by this to set a
basis for the description and simulation of real aquifer
systems. The aim of this work was to:
1) Investigate the phases quartz (Qz), sodium feldspar
(NaFs), potassium feldspar (Kfs), kaolinite (Kaol),
calcite (Cal), and iron/manganese hydroxides (MnFe) as
sorbents for dissolved cadmium, copper, and lead;
2) Determine the phase distributions in the presence and
absence of NOM; and,
3) Describe the equilibria by means of Freundlich iso-
therms.
THEORETICAL BASIS
For the description of phase distribution equilibria in
solid/liquid systems, isotherms according to Freundlich
and Langmuir, or distribution coefficients are suited
(Stumm 1992). For aqueous systems, the Freundlich
approach is commonly used (Sontheimer et al. 1988).
The adsorption equilibrium between the adsorbed and
dissolved part of the sorbate is described by Eq. 1:
q=K&’ (1)
where,
q is the amount of sorbate adsorbed on solid phase (e.g.,
umol/g); and,
K, is the Freundlich constant; the better the adsorption
the higher the number.
In the double logarithmic plot of q versus c, the
isotherm is linear for n = 1; for n > 1 adsorption iso-
therms are called ‘unfavourable’ according to the low
load of the solid phase at low equilibrium concentra-
tions in solution. The experimental determination of the
isotherms can be done by using the initial concentration
(c,,)and the concentration remaining in the solution after
equilibration (c). According to Eq. 2, the amount of
adsorbed matter can be calculated:
q = 2 (co-c) (2)
mS
F.H. Frimmel and L. Huber
Fig. 1. Competing reactions between metals and ligands (L) in a
solid/liquid system (X, Y, Z different sites on the solid phase surface).
where m, is the mass of sorbent (e.g., in g); and, V, is the
volume of the solution (e.g., in L).
The complex formation in solution is described as a
1: 1 metal/ligand interaction according to Eq. 3 :
iv+ L &ML (3)
where,
L is an independent functional group of NOM.
The different reactions in the heterogeneous system of
an aquifer can be expressed as shown in Fig. 1. The
complexity of the interactions of metals, ligands, and
sites in a heterogeneous system is obvious.
The kinetics of the adsorption of NOM on the mineral
phases can be calculated (Stumm 1992; Cosovic 1990).
For the calculation of the adsorption isotherms, the
analysis developed for activated carbon adsorption was
used (Frick and Sontheimer 1983; Johannsen and
Worth 1994). The method is applied to undefined
mixtures (e.g., NOM) and uses fictitious components
with different adsorption properties. It is based on the
ideal adsorption solution theory (IAS-theory) (Radke
and Prausnitz 1972). The relative amount (initial con-
centration) of the fictitious components and their two
Freundlich parameters (Kr and n) are varied until a
good fit with the experimental data is obtained. Nor-
mally only a few (e.g., 3) components are necessary and
a constant number for the Freundlich value n is selected.
EXPERIMENTAL DETAILS
Mineralphases
Quartz (SiOd was obtained from Quarzwerke Freenen
as type F 32. According to the classification system of
DIN 18123 (1983) it is a middle sand which has
practically no basic or other impurities. Sodium feldspar
(albite, NaAlSisO,) and potassium feldspar (orthoclase,
Influence of humic substances on aquatic adsorption of heavy metals on defined mineral phases 509

Table 1. Composition of mineral phases (main components), numbers as weight %, average values from x = 5 determinations.
L.O.I. = loss of ignition; n.m. = not measured.

Qz KFs NaFS Kaol Calc MnFe

SO, 98.3 66.0 68.8 57.7 1.00 4.73

-4’203
0.13 18.4 18.8 28.6 0.32 7.01

Fe*03 0.06 0.30 0.13 0.18 0.71 18.4

TiO, 0.02 0.02 0.01 0.58 0.01 n.m.

K,O 0.01 11.6 2.45 3.76 0.38 1.52

Na& co.01 2.82 7.11 0.19 0.06 1.02

CaO 0.05 0.46 1.75 0.01 53.9 5.22

MgO 0.06 0.02 0.02 0.02 0.35 1.25

MnO, co.01 co.01 co.01 co.01 0.02 53.4

PA 0.01 0.01 0.02 0.05 0.01 n.m.

BaO 8.63

L.O.I. 0.2 0.2 0.30 0.09 42.3

W 5.2

sum 99.2 99.8 99.4 99.8 99.1 97.8

KAlSi,O,) were from SCR Sibelco (Antwerpen,
Netherlands) and had the size of middle and fine sand.
Kaolinite (Al,(OI-I)@i,O,) was from Gebriider Dorfher
GmbH & Co of the type Dorfner-Kaolin ZY. The
particle size equals that of a clay-fraction. Calcite
(CaCO,) was from Mathis GmbH & Co, and its origin
was a quarry of the Tuniberg near Freiburg. Manganese
and iron oxihydrate was supplied by Stadtwerke
Karlsruhe (water works Rheinwald) and originated from
a quartz sand filter which functioned for elimination of
iron and manganese from raw water. The chemial
composition of kaolinite and calcite are given in Tables
1 and 2. Numbers are averaged from x = 5 deter-
minations. Some characteristic properties of the solid
phases are given in Table 3 (Hodel et al. 1995).
Particle size distribution was obtained by dry sieving
according to DIN 18 123. Most of the particles had a
fairly homogeneous chemical composition. The excep-
tion was Mn/Fe-oxihydrate. A core of quartz-particles
(d ca. 0.5 mm) was coated with Mn(IV)- and Fe(III)-
hydroxide (ca. 120 pm).
To increase the yield in the isotherm experiments,
mineral phases (except MnFe) were ground which led
to specific surfaces of 670 (Qz), 940 (KFs), 750 (NaFs),
860 (Kaol), and 1250 (Calc) m2/kg.
Solutions and chemicals
Experiments were performed with deionized water
(dw) and synthetic water. The synthetic water had an
ion balance similar to groundwater from calcite
dominated aquifers, with electric conductivity K (20°C)
= 375 @/cm and ionic strength u = 5.18 mmol/L.
Concentration of Ca” was 0.76 mmol/L and of Mg2+
510 F.H. Frimmel and L. Huber

Table 2. Composition of mineral phases (trace elements), numbers in mgkg, average values from x = 5 determinations.
n.d.: < limit of detection.

Qz KFs NaFs Kaol Calc MnPe

Ni n.d. n.d. n.d. 8.26 9.97 26.0

CU 9.30 9.31 9.68 73.3 11.6 3.84

Zn 5.78 7.14 7.98 9.68 15.9 52.4

Pb 4.91 55.7 33.3 192 8.87 2.05

As n.d. n.d. n.d. n.d. 9.66 301

Sr 8.25 52.7 52.5 115 335 165

Zr 22.8 2.98 7.98 161 6.71 n.d.

Nb n.d. n.d. n.d. 16.8 n.d. n.d.

Th n.d. n.d. n.d. 16.7 n.d. 0.2

Rb n.d. 392 111 121 2.91 n.d.

Cd 0.02 n.d. 0.02 n.d. 0.04 0.53

Tl n.d. 0.71 0.25 n.d. 0.02 0.22

Bi 0.02 0.15 1.27 n.d. n.d. n.d.

co n.d. n.d. n.d. n.d. n.d. 26.0

The analytically determined concentrations of the trace metals show standard deviations of up to 10%.

Table 3. Characteristic properties of the mineral phases.

Properties Qz KFS NaFs Kaol talc MnEe

density, g/cm3 2.65 2.57 2.605 2.6 2.72 2.68

Co,, weight % 0.07 0.17 0.22 0.15 0.29

cation exchange capacity, 6.0 11 19 20 101 493

mmol(eq)/kg

specific surface, m2/kg 9.2 16.5 12.6 860 6.5 63 000

average particle size 300 220 300 14 550 1300

(based on 50 % sample
ma=), Pm
Influence of humic substances on aquatic adsorption of heavy metals on defined mineral phases
i
cKaoIinite
0.4- ‘i
1
0.3 ! I I I
0 100 200
Fig. 2 Kinetics of the adsorption of brown water NOM on the mineral phases calcite, kaolinite, sod iron/manganese
was 0.31 mmol/L; pH (20°C) was 8.4. The water
was in thermodynamic equilibrium with respect to
CaCO,. In addition, water from a brown water lake
(Hohlohsee, Black Forest) was used as source for
NOM in its original composition or diluted with dw.
The original brown water had an electric conductivity
K (20°C) = 48 @/cm. Concentrations of Ca*’ and
M$+ were 0.05 mmol/L; concentration of DQC was
20 mg/L; A (254 nm) / DOC was 4 L/(mg - m); pH
(20°C) was 4. The water was filtered through a
0.45 urn membrane filter (cellulose-nitrate, Sartorius)
after sampling. All chemicals used for pH adjustment
and analytical purposes were of analytical grade
(puriss. p. a.).
Analytical methods
Metal concentrations were determined with atomic
absorption spectrometry (AAS) and inductively coupled
plasma atomic emission spectrometry (ICP-AES) ac-
cording to standardized methods (Fachgruppe Wasser-
chemie 1995). The detection limits were 0.1 ug/L
(Cd), 0.6 ug/L (Cu), and 0.6 ug/L (Pb). The con-
centrations of dissolved organic carbon (DOC) were
determined with a Carbon-analyzer (Dohrmann). The
detection limit was 0.2 mg/L. Standard deviations
for the methods were three times the detection
limits. The pH-values were determined with a glass
electrode.
511
I I I
300 400 s60 600 700
time [h]
oxihydrate.
Reactions
Batch experiments were performed in PTFE bottles
(volume V = 250 mL). First, 2 g (Qz, NaFs, KFs, Kaol),
1.5 g (Calc), or 1 g (MnFe) of the mineral phases were
defined metal concentrations (about 1 umol/L each)
and NOM (mostly 4.7 mg/L) for 2 d at 100 rpm.
After this, samples were taken, filtered (0.45 pm
membranes) and analysed. Most experiments were run
five times. All experiments were performed at ambient
temperature.
RESULTS
Selection of time scale
In multicomponent systems, equilibration time has to
be watched carefully. Metal adsorption is a relatively
fast process. Adsorption of NOM generally requires
more time. Figure 2 shows the kinetics of DOC
adsorption on selected mineral phases. In the case of
calcite and kaolinite, equilibrium is nearly reached after
a few days. Equilibration in the MnFe system takes
longer. From a pragmatic view, 2 d was chosen as the
reaction time for equilibration.
Stability of pH-values
Adsorption processes, especially those with metals,
are strongly influenced by the proton concentration (pH-
values). Figure 3 shows the change of pH-value in the
512 F.H. Frimmel and L. Huber

[Kaolinitef
-
7- [lb

1. initial pH

- 3.6
3- * 4.8
+ 6.5
* 7.3

2 1 I I I I I I I
0 10 20 30 40 SO 60 70 80

m (solid) I V (solution) [g/L]

Fig. 3. Plot of pH-values of aqueous systems containing solutions with different initial pH-values and different kaolinite/solution ratios.

Table 4. Initial concentrations of NOM and Freundlich parameters for the fictitious components adsorbed on kaolinite.

Fraction KF n D°CO,i

(mg/g)-(wCY m@- %

non adsorbable 0 3.3 17.5

1 0.5 0.16 2.0 10.5
2 1.0 0.16 5.6 30.2
3 2.0 0.16 7.8 41.8

sum: 18.7 100

averaged error (Oh): 1.15

solution as a function of the relative amount of kaolinite
as the solid phase. Kaolinite acts like a solid pH-buffer
with a pK value around 4.6. As a consequence, the pH-
value of heterogeneous phases will be significantly
dependant on the component with the highest buffer
capacity.
The other mineral phases showed equilibrated pH-
values of 5.8 f 0.2 for Qz, KFs, and NaFs; 8.6 f 0.1 for
calcite, and 7.2 f 0.2 for MnFe.
Adsorption isotherms for NOM
The adsorption of NOM onto mineral phases can be
described according to the model of Johannsen and
Worth (1994). It was originally developed for the de-
scription of NOM adsorption on activated carbon and
is based on the assumption of fictitious components
with different adsorbability given as Freundlich para-
meters (I& n). Table 4 shows a set of data for the
fictitious components of brown water NOM which de-
scribes the experimental results of its adsorption on
kaolinite.
The four components include a non-adsorbable frac-
tion (17.5% of DOC) and fractions with increasing
afIinity to kaolinite. It turned out that a single value for
the Freundlich exponent n was well suited to match the
experimental data points with the model calculation
(Fig. 4).
Influence of humic substances on aquatic adsorption of heavy metals on defined mineral phases 513

pH = 3.6

Dilution 1:l

Isotherm for original brownwater

0.0 0 2 4I 6I I I I
8 10 12 1
Remaining concentration in solution (DOC mg/L)
Fig. 4. Adsorption isotherms (experimental data points and model calculation) for aquatic NOM on kaolinite.

L.”

1.8-
I
1.6- .

8 1.2-
e
$.g 1.4-
1.0-
/
,o
d DH-values:
‘;; 0.8-
initial1 final
-S
2 0.6- hY -3.7 1

= 7.6 1 93~0.07

I I I I I I I I I
2 4 6 8 10 12 14 16 18 20
Remainingconcentrationin solution (DOC mg/L)

Fig.5. Adsorption of aquatic NOM on calcite starting from different initial pH-values.

Good agreement is obtained for the adsorption of
NOM in the original brown water and in its dilutions
(1: 1, 1:2, 1:3). Isotherms for the adsorption of aquatic
NOM on calcite are given in Fig. 5. Even though there
was quite a difference in the initial pH-value of the
solutions, there is only a narrow range in the final sus-
pensions. Obviously there is an increase in the adsorbed
NOM as the pH-value decreases. This is in good
agreement with observations from the adsorption of
humic substances on XAD-resins (Abbt-Braun and
Frimmel 199 1).
The comparison of the isotherms for the NOM ad-
sorption on kaolinite, on calcite, and on the iron/ man-
ganese oxihydrate is given in Fig. 6.
514 F.H. Frimmel and L. Huber

Kaolinite (2 days)
DO&= 18.7 mg/L.

a, = 1250 m*/kg

s, = 63000 m*ikg
0.001 ! I I I I ‘ I I I I
0 2 4 6 8 10 12 14 16 18 1
Remaining concentration in solution (DOC mg/L)

Fig. 6. Adsorption isotherms(experimentaldata points and model calculations)for aquaticNOM on kaoliite, calcite, and iron/manganeseoxihydrate.

Table 5. Dissolved and adsorbed amounts of NOM in aqueous suspensions of selected mineral phases.
c,(DOC) = 20 mgIL; solid/solution-ratio m/v, = 20 a.

Solid phase DOCIDOC, DOC Load

% mg/L DOC mg/n DOC me/m2

Qz 100 19.4 0.00 0.00

NaFs 93 17.3 0.06 0.08

MnFe 41 8.5 0.61 0.01

Calc 20 3.7 0.72 0.57

Kaol 21 3.8 0.74 0.86

Note that theNOM load on mineral phases is based on
their surface area. It has to be kept in mind that the pH-
values in the experiments were different according to the
buffer capacities of the mineral phases. For the Fe/Mn
oxihydrate containing system, the reaction time was 40 d.
This was according to the aim to reach thermodynamic
equilibrium (Fig. 2). Despite the long reaction time, there
was relatively poor coverage of the surface area deter-
mined via a nitrogen-adsorption-isotherm according to
the BET-method of Brunauer et al. (193 8). One might
assume that most of the determined surface is due to a
porous structure and cannot easily be reached by the
NOM. A comparison of kaolinite and calcite shows that
the lower pH-value ofthe kaolinite containing system is
stimulating the adsorption ofNOM and obviously rules
out the effect of the larger surface area of the calcite.
The degree of adsorption of NOM on the different
mineral phases under comparable reaction conditions is
given in Table 5.
There is a clear tendency for better adsorption from
quartz to kaolin&e+if the load is given as mass ratio. The
sequence is somewhat different on the basis of surface
area. Again the special role of the Fe&In-oxihydrate is
obvious.
Influence of humic substances on aquatic adsorption of heavy metals on defined mineral phases 515

IPb/Calci

Dissolved concentration [pmol/L]

Dissolved concentration [pmol/L]

Fig. 7. Adsorption isotherms for the lead/calcite system in absence a) and presence b) of NOM (c,,(DOC) = 4.7 mg/L).

Metal adsorption reason for that may be seen in a partly irreversible

An example of the influence of NOM on the ad-
sorption of heavy metals on mineral phases is given in
Fig. 7.
The experiments were done in model water (ground-
water type). The two lines in Fig. 7a show the isotherms
for the individual application of lead as a single metal
(upper line) and for th e 1ea d application simultaneous
with the other metals (lower line). The isotherms for the
lead adsorption on calcite simultaneously applied with
the other metals and, in the presence of NOM (Fig. 7b),
show the numbers obtaitied from balance calculations
according to Eq. 2 (uptier line), and the data points
gained from extractions of the sorbent (lower line).
There was a smaller yield of the experimental extraction
which led to an apparent smaller load of the calcite. The
fixation of the metal in the calcite. This has to be kept in
mind in case of a quantitative use of the data. The
discrepancy was also shown for the other systems but to
a smaller extent. (More details are given in Hodel et. al.
(1995).) Mineral phase specific effects can be deduced.
Comparison of the isotherms for the lead adsorption
in the presence and absence of NOM (Fig. 7: a, lower
line and b, upper line) proves the vehicle function of
NOM for lead which remains in solution to a higher
extent if NOM is present.
The simultaneous adsorption of the metals Cd, Pb,
and Cu on the different mineral phases is given in Table
6. The influence of the NOM on the phase distribution
of the metals is given in Table. 7. The values of the
Freundlich parameters are only given for correlation
coefficients ? > 0.5 1. The K, values were calculated
516 F.H. Frimmel and L. Huber

Table 6. Freundlich isotherms for the simultaneous adsorption of cadmium, lead, and copper (c,, (each) = 1 urnoIL) on different
mineral phases; K, in (pmobkg) - (umo!/L)-“.

Solid Cd Pb cu

phase KF n 3 K, n rr K, n ?

QZ 6.0 1.8 0.95 140 1.8 0.98 150 1.4 0.99

KFs 22 1.9 0.89 170 1.2 0.98 71 1.2 0.81

NaFs 17 0.2 0.84 270 1.1 0.92 340 1.5 0.94

Kaol 42 1.8 0.96 350 0.8 0.87 1200 2.1 0.92

Calc 230 0.9 0.96 300 1.2 0.92 210 1.5 0.89

MnFe 220 1.0 0.87 0.51 400 1.6 0.83

Table 7. Freundlich isotherms for the adsorption of cadmium, lead, and copper (cO(each) = 1 pmol/L) on different mineral phases in the presence
of NOM (c,(DOC) = 4.7 mg/L); KF in (umolkg) * (umol/L)“.

Solid Cd Pb cu

phase KF n ? K, n 3 K, n rs

Qz 2100 6.5 0.98 390 8.0 0.98 730 23 0.93

KFs 680 3.7 0.98 800 5.6 0.99 0.51

NaPs 3700 7.4 0.94 220 4.3 0.91 140 17 0.86

Kaol 415 2.2 0.96 470 2.4 0.99 510 12 0.76

Calc 590 1.4 0.92 120 1.6 0.97 0.04

MnFe 1200 1.5 0.93 0.05 200 1.9 0.94

according to Eq. 1, using six data points for q in
pmol/kg and c in pmol/L. Note that the linearized iso-
therms are only valid for a fairly narrow concentration
range which, for copper, for example, is 0.79 umol <
c(Cu) < 0.87 pmol. The values of n which are extremely
high in case of copper can only partly be explained by
the kg-based loads. This shows clearly that an inter-
pretation similar to the explanation of phase equilibria
of purely organic compounds cannot be done. Because
of the empirical character of Freundlich isotherms and
the necessity to work in analytically manageable con-
centration ranges, no general mechanistic information
can be derived. In addition, the conditions under which
it is reasonable to work with the different model systems
can vary significantly (e.g., pH, size distribution). This
implies a limitation of the direct comparability of the
results.
Despite those pitfalls, the results show some general
trends. For copper, to a lesser extent, there is a mobili-
zation effect caused by NOM. Cadmium is better ad-
sorbed in the presence of NOM. This was also observed
in breakthrough experiments using columns filled with
the mineral phases (Huber and Frimmell994). The ex-
perimental results so far cannot be fully explained by
the stability of the metal/NOM-complexes which in-
creases from Cd to Cu.
Influence of humic substances on aquatic adsorption of heavy metals on defined mineral phases
CONCLUSIONS
The work shows that under some precautions
Freundlich isotherms are suited to describe the phase
distribution of metals also in the presence of NOM.
From the isotherms of the different mineral phases, it is
possible to deduce-at least for a limited concentration
range-the behaviour of the metals and their species in
aquifers if their mineral composition is known. The
results so far are in good agreement with most data
which were obtained in other investigations.
Model calculations depend strongly on sound experi-
mental data and results. Therefore, it is essential to
identify the key components for the system under in-
vestigation. In addition to the influences of pH, tem-
perature, and ionic strength, competing reactions and
their degree of reversibility have to be investigated.
There is no doubt about the fundamental role of NOM
in the distribution, transport, and reactions of metals in
aquifers. Due to the complicated structure of NOM,
there is little hope for reliable mechanisms to describe
the interaction with other parts of a groundwater system.
Simple approaches like isotherms obtained from batch
experiments in addition to column experiments will help
to obtain the necessary sets of data for model calcula-
tions with greater prognostic power.
Acknowledgment-The hmding of the work by Stithtng Volkswagen-
werk (AZ: 4710,410S) and the experimental help of Brigitte Raue and
Gudrun Grabe is gratefully acknowledged.
REFERENCES
Abbt-Braun, G.; Frimmel, F.H. Spektroskopische Struktur-
aufkl%rung aquatischer Huminstoffe (Spectroscopic investi-
gations of the structure of aquatic humic substances). Vom
Wasser 77: 291-302; 1991.
Benedetti, M.C.; Milue, C.J.; Kimiburgh, D.G.; van Riemsdijk,
W.H.; Koopal, L.K. Metal ion binding to humic substances:
Application of the non-ideal competitive adsorption model.
Environ. Sci. Technol. 29: 446-457; 1995.
Brunauer, S.; Emmett, P.H.; Teller, E. Adsorption of gases in
multimolecular layers. J. Amer. Chem. Sot. 60: 309-319; 1938.
Buffle, J. Complexation reactions in aquatic systems. Chichester:
John Wiley & Sons; 1989.
Cosovic, B. Adsorption kinetics of the complex mixture of organic
solutes at model and natural phase boundaries. In: Stumm, E., ed.
Aquatic chemical kinetics. New York: John Wiley & Sons; 29 l-
310; 1990.
Deutsches Institut fbr Normung e. V.: DIN 18123. Bestimmung der
KomgroRenverteilung (Determination of grain-size distribution).
12 S. Berlin: Beuth Verlag GmbH; 1983.
Fachgruppe Wasserchemie in der Gesellschatt Deutscher Chemiker
(Ed.) Deutsche Einheitsverfahren zur Wasser-, Abwasser- und
Schlammuntersuchung (German standard methods for the
examination of water, waste water and sludge). 32. Auflage.
Weinheim: VCH; 1995.
517
Farley, K.J.; Dzombak, D.A.; Morel, F.M.M. A surface precipi-
tation model for the sorption of cations on metal oxides. J.
Colloid Interface Sci. 106: 226-242; 1985.
Frick, B.; Sontheimer, H. Adsorption equilibria in multisolute
mixtures of known and unknown composition. In: McGuire,
M.J.; Suffett, I.H., eds. Treatment of water by granular activated
carbon. Advances in Chemistry Series 202. Washington, D.C.:
Am. Chem. Sot.; 1983.
Gerth, J.; Brtimmer, G. Adsorption und Festlegung von Nickel,
Zink und Cadmium durch Goethit (a-FeOOH) (Adsorption and
immobilization of nickel, zinc and cadmium by Goethite
(cc-FeOOH)). Fres. Z. Anal. Chem. 316: 616-620; 1983.
Hodel, M.; Huber, L.; Lehmann, M.; Lensing, H.J.; Herrling, B.;
Frimmel, F.H.; Puchelt, H. Quantitative Wechselwirkungen
zwischen Metallen und Mineralphasen unter Berticksichtigung
des Einflusses nattirlicher organischer Wasserinhaltsstoffe und
Modellienmg des StoITtransports im Grundwasser (Quantitative
interactions of metals and mineral phases with respect to the
influence of organic matter and transport modelling in aquifers).
Veroffentlichung des Lehrstuhls fur Wasserchemie und der
DVGW-Forschungsstelle am Engler-Bunte-Institut der Universi-
tat Karlsruhe; 1995.
Huber, L.; Frimmel, F.H. Zum Transportverhalten von Cadmium,
Blei und Kupfer in ausgewtilten mineralischen Phasen von
Grundwasserleitem (Characterization of the transport of
cadmium, lead and copper in selected mineral phases from
natural aquifers). Vom Wasser 83: 9-22; 1994.
Johannsen, K.; Worth, E. Eine mathematische Methode zur Durch-
fiihrung von Adsorptionsanalysen (A mathematical method for
evaluation of adsorption analysis). Acta hydrochim. hydrobiol.
22(5): 225-230; 1994.
KoD, V. Zur Modellierung der Metalladsorption im natiirlichen
Sediment-Grundwasser-System (Metal adsorption modelling in
natural sediment-groundwater systems). Habilitation; TU Berlin;
Fachbereich 2 1; 1992.
Marinsky, J.A.; Ephraim, J. A unified physicochemical description
of the protonation and metal ion complexation equilibria of
natural organic acids (humic and fulvic acids): Analysis of the
influence of polyelectrolyte properties on protonation equilibria
in ionic media: Fundamental concepts. Environ. Sci. Technol. 20:
345-354; 1986.
Perdue, E.M. Modelling the acid-base chemistry of organic acids in
laboratory experiments and in freshwaters. In: Perdue, E.M.;
Gjessing, E.T., eds. Organic acids in aquatic ecosystems.
Chichester: John Wiley & Sons; 1990: 111-126.
Radke, C.J.; Prausnitz, J.M. Adsorption of organic solutes from
dilute aqueous solution on activated carbon. Int. Eng. Chem.
Fund. 11: 445-451; 1972.
Sontheimer, H.; Crittenden, J.C.; Summers, R.S. Activated carbon
for water treatment. DVGW-Forschungsstelle, Engler-Bunte-
Institut, Universitgt Karlsruhe; 1988.
Stbver, T.; Roennefahrt, K. Arsenentfemung aus Trinkwasser
(Arsenic removal from drinking water). Vom Wasser 78: 363-
376; 1992.
Stumm, W. Chemistry of the solid-water interface. New York: John
Wiley & Sons; 1992.
Tipping, E. Some aspects of the interactions between particulate
oxides and aquatic humic substances. Marine Chem. 18: 161-
169; 1986.
Zasoski, R.J.; Burau, R.G. Sorption and sorptive interaction of
cadmium and zink on hydrous manganese oxide. Soil Sci. SOC.
Am. J. 52: 81-87; 1988.