Pro#. Ener,qy Combust. Sci., Vol. 6, pp. 223 232 0360 1285/80/0901 0223505.

(10/0
© Pergamon Press Ltd., 1980. Printed in Great Britain.

THE HYDROPYROLYSIS OF COAL TO BTX

G. F,CSES, W. R. LADNER a n d J. O. H. NEWMAN
National Coal Board, Coal Research Establishment,
Stoke Orchard, Cheltenham, Gloucestershire, GL52 4RZ, U.K.

Abstract--The importance of hydropyrolysis as a coal-to-liquid route is considered. Yields of BTX of up to

15% w/w have been obtained from bituminous coal and lignite by using very rapid heating (~ 1000Ksec- ~)
to 1000-1100K in hydrogen pressures in the range 100-200 bar at optimum vapour residence times. The
yieldsobtained by various workers and the mechanismsof the formation of BTX from coal by hydropyrolysis
are discussed.

I. INTRODUCTION 2. PUBLISHED WORK RELATING

The hydropyrolysis of coal is defined broadly as the
reaction between coal and hydrogen at elevated tem-
peratures and pressures.1 As the nature of the products
depends to a large extent on the operating conditions,
it is convenient to distinguish between different types
of hydropyrolysis. The term "hydrogasification" is
used when methane is the desired product, and
"hydropyrolysis" is used when the principal products
are liquids and gas. "Flash hydropyrolysis" is a more
recent development of hydropyrolysis which uses very
high heating rates (1000 K sec-1) with the object of
producing benzene, toluene and xylenes (BTX).
The main aim of the early hydropyrolysis studies in
the United Kingdom by Dent and others 2-7 over the
period 1937-67 and, later, in 1972, at the United States
Bureau of Mines by Feldman and co-workers s was to
make methane. More recently, interest has been direc-
ted to the production of liquid products, particularly
BTX, by flash hydropyrolysis. This review is mainly
confined to work on flash hydropyrolysis: emphasis
is on non-catalytic studies, although some work on
catalytic hydropyrolysis is discussed. No attempt
is made to review the published work on hydro-
gasification; however, some relevant references are
included.
Until the 1960s, nearly all BTX was a byproduct of
coal carbonization, but since then the proportion of
BTX derived from petroleum (see Table 1) has steadily
increased. 9 However, the cost of petroleum has risen
steeply since 1973 and, if this trend continues, the
production of aromatics from coal may again become
economic. In the long term, as oil reserves are depleted,
it will be necessary to use coal as a chemical feedstock.
TABLE 1. U.K. production of benzene since 1965
(1000s tons yr-t)9
Year Coal-derived Petroleum-derived
1965 159 30
1969 136 417
1971 147 550
1975 167 878
1977 127 845
TO BTX AND OIL FORMATION
A number of different laboratories in the U.S.A. and
one in the U.K. have been involved in hydropyrolysis
research with the objective of making either oil or BTX
from coal. For convenience, the work has been re-
viewed according to the various laboratories con-
cerned: the highest oil and BTX yields obtained and
the conditions used to achieve these yields are sum-
marized in Table 2. The term "oil" which is used widely
by American workers can cover all liquid products
from water to tars, and sometimes even includes
material solvent-extracted from the char.
2.1. U.S. Bureau of Mines
The U.S. Bureau of Mines has undertaken a number
of studies on the hydropyrolysis of coals with various
objectives. Hiteshue and co-workers lO-14 developed a
simple apparatus shown in Fig. l a : the essential
feature was the containment of 8 g of coal in a steel
tube, through which pressurized hydrogen was flow-
ing; the tube could be rapidly heated (up to 20 K sec- 1)
to a definite temperature by passing a large electric
current through it. The residence time of the coal in the
hot zone was controlled by the duration of heating,
and that of the volatile products by the hydrogen flow-
rate. The effects of various parameters on the yield and
composition of the products were determined.
Although many of the experiments were carried out
using a catalyst (1% molybdenum), it was only effective
when the coal residence-time at the final temperature
was as short as 1 min; for a coal residence-time of
15min the same yields were obtained whether the
catalyst was present or not.
Using a sub-bituminous New Mexico coal, 14 the
effects of coal residence-time (0-15 min), pressure (35-
415bar) and temperature (873-1173 K) on the yield
and composition of the liquid and gaseous products
were investigated systematically. In general, the con-
version of the coal to gas increased as each of these
parameters was increased, but the yield of oil was
independent of the coal residence-time, showing that

, ,,~.~ s. ~,3 ,, 223

224 G. FYNES,W. R. LADNEgand J. O. H. NEWMAN

TABLE2. A summary of hydropyrolysis results

Type of Heating rate Temperature Pressure Yield ~ w/w d.a.f, coal

reactor Type of coal (K sec- 1) (K) (bar) oil BTX benzene Ref.

Lignite rapid 973 104 7 10 27

Lignite rapid 998 139 7 7.5 7 29
Sub-bituminous 0.5 x 104 1063 200 46a 35
Lignite 2.7-8.3 x 104 894-1130 340-207 16 15 36
Entrained
Lignite >5 x 104 973 153 13 5 39
tube
Bituminous rapid 1255 104 25 9 43, 44
Lignite 315~,70 1061 139 7.5 3 45
Bituminous 650 973 173 20b 48
Sub-bituminous >7 773 139 30 56
Bituminous 0.1 873 70 27 15
Fluidized Bituminous rapid > 873 > 22 32° 25
bed Lignite 1 x 10" 1023 103 7 26
Sub-bituminous rapid 840 66 29 0.03 0.02 51
Sub-bituminous rapid 813 346 26 1
Bituminous 7 1073 415 9 3.5 3.2 10
Bituminous 20 753 415 36 2 11
Bituminous (Eastern U.S.A.) 7 1073 415 19
Tube Bituminous(Western U.S.A.) 7 1073 415 26 I 12
Lignite 7 1073 415 33
Bituminous 7 1273 70 19 13
Sub-bituminous 7 1073 415 40 <6 d 14
Bituminous 650 1073 100 11 10 22
Bituminous (U.K.) 5 1123 150 12 12 57

a, postulated ; b, benzene-soluble oil ; c, postulated ; d, total single-ring aromatics.

Note. A typical BTX yield by high temperature carbonization is about I ~.

its formation was confined to the initial stages of
decomposition. At 873K, the oil yield was also
independent of pressure, but at 1173K it decreased
with increasing pressure and was lower than the
corresponding yield at 873 K. This decrease in yield at
the higher temperature was probably due to thermal
cracking of the oil to carbon, methane and other
products.
The maximum yield of oil obtained was about 4 0 ~
of the coal, and the yields of single-ring aromatics
(including phenols) ranged from 2 to 6~o of coal.
Although it was not possible to correlate the com-
position of the oil with all the operating conditions, it
was observed that at 1073 K the residence time of the
volatile products has a marked effect on both the yield
and composition. When the residence time of the
volatiles was 5 sec or longer, the condensed liquids
were light coloured, had a low viscosity and contained
a large quantity of distillable oil. In contrast, a
residence time of 2.3sec gave a dark, viscous oil
containing an appreciable amount of asphaltene. This
difference was attributed to hydrocracking of the
volatile products at the longer residence-times.
Earlier, Weller and co-workers 15-x7 investigated
the mechanisms of decomposition and yields from
heating bituminous coal in hydrogen either in auto-
claves or fluidized beds. The coal and its products were
heated at about 7 K m i n -1. Up to 76~o oil was
obtained at 713 K and 415 bar with tin catalyst in an
autoclave, and 27~o oil at 873 K and 70bar without
catalyst in a fluidized bed.
Later, Karr et al. ~s carbonized coal in hydrogen at
up to 783K and hydrorefined the tar vapours by
passing them at pressure over a catalyst. Increase in
hydrogen pressure from 22 to 63 bar led to a reduction
in oil yield from 10 to 7.5~, while a maximum BTX
yield of 3.5 ~o was obtained at a pressure of 42 bar and a
temperature of 672 K.
Hiteshue et al. 13 used a semi-continuous, entrained
flow, pressure reactor shown in Fig. lb to obtain data
for a hydrogasifier. A rotary feeder was developed with
downward flow of the coal to overcome the agglomer-
ation experienced with high-volatile bituminous coals.
Long residence-times in a 1073K hot-zone led to
increased gasification at the expense of oil production,
which was up to 19~o at 70bar and 1273 K.
2.2. University of New York
Squires 19 included coal hydropyrolysis, in the form
of flash hydropyrolysis, in his Coalplex scheme to
convert coal into chemicals. The experimental
studies 19-22 described by Squires and co-workers
were on a very small scale and were aimed at
establishing the effect of very high heating rates and
controlled vapour residence-times on the gas and
liquid yields from a bituminous coal in hydrogen at
100 bar. The importance of BTX products was
stressed.
The reactor was a small, resistively-heated tube in
which 10 mg of finely ground coal were deposited as a
ring of particles on the inner wall of the tube and is
shown in Fig. lc. The vapour residence-time was
varied by altering the position of the coal layer along
the tube. The yield and composition of the volatile
products were determined by gas chromatography. To
 The hydropyrolysis of coal to BTX 225

~VOLATILEPRODUCTS COAL
RESERVOIR
-~I~COA L
COAI~ ~,.7 FEEDER
HZF " ' - - ~ - I HZ~
l
-" J
Q
I VOLATILE
°
PRODUCTS
H2 FURNACE/ I b C COA/RING

CHAR L _

RECEIVER

H2 =
~ COALFEEDER
AND RESERVOIR COAL
RESERVOIR
HOT~ H2
H2
~02
RESERVOIR L COAL -
FURNACE

1
d • f
HOT FURNACE
VOLATILE
CHARTREAP "--"~RODUCTS
FURNACE WATER~ IQUENCH

LATILE ~ PRO S

I PRODUCTS
] ICHAR
I I TRAP
VOLATILE -'-I
PRODUCTS ~ ]

I~r--'--~VO LATILE ~ilz COAL

PRODUCTS
g ~ C H A R TRAP h ~H2
FIG. 1. Various hydropyrolysis reactors.

obtain a carbon balance, the residual char was com-
busted and the carbon dioxide determined. Initial tests
showed that heating rate was an influential parameter;
20 K sec- 1 gave only 3.5% BTX, whereas 650 K sec- 1
gave a higher yield of 9.7%, similar to the 10.2~
obtained at 1400Ksec-I. Use of a nitrogen atmos-
phere reduced the BTX to one fifth of that obtained
under hydrogen.
Using the same apparatus, Dobner et al. 22 in-
vestigated the effect of temperature and vapour
residence-time with the solids residence-time fixed at
10sec, the hydrogen pressure at 100bar and the
heating rate at 650 K sec- 1. A maximum yield of 10%
benzene was obtained at 1073K and 0.6sec vapour
residence-time, little other aromatics being formed. By
varying the vapour residence-time at 973 K, a lower
maximum of 8% BTX was found at a vapour
residence-time of 2.5 sec. This clearly showed that the
effects of temperature and vapour residence-time on
the BTX yield were interdependent.
Graft and co-workers 2a'24 investigated the effect of
the carbon content of American coals by determining
the optimum yield of BTX from each coal; the reactor
temperature and hydrogen flow at 100 bar were varied
to give a single vapour residence time of 0.6 sec. Coals
containing 75-78 ~o carbon showed a broad maximum
of 10% BTX on hydropyrolysis to an optimum
temperature which increased from 1033 to 1123 K with
increase in the carbon content of the coal. The total
liquid yield increased with carbon content to a 35~o
carbon conversion (at 85% carbon). There was some
evidence that aerial oxidation reduced the BTX yield.
For possible commercial application of hydro-
pyrolysis, Squires 25 proposed a fast fluidized bed to
hydropyrolyse coal rapidly at 873 K and pressures
above 22 bar. Coal would be fed into a fluidized bed of
coke operated at very high gas velocity to keep the
vapour residence-time of the volatiles in the cracking
zone between 5 and 20 sec. A high yield was claimed
(32 ~o) of a light, distillable aromatic tar.
2.3. Brookhaven National Laboratory
Brookhaven National Laboratory has flash hydro-
pyrolysed coal in a variety of ways. Growcock and
Mackenzie 26 dropped 3 mg batches of lignite into a
preheated bed of stainless steel or alumina particles
purged with high-pressure hydrogen; the heating rate
226 G. FYNES, W. R. LADNER and J. O. H. NEWMAN
was calculated as 104K sec 1. Temperature, pressure
and solid and vapour residence-times were varied, and
it was found that above 1023 K the liquid product was
essentially benzene. At 1023 K and 103 bar the benzene
yields changed little with vapour residence-times of
0.3-10sec (in contrast to the results obtained in
Squires' apparatus) but they increased linearly with
pressure and passed through a maximum of 7~0 at
1023 K.
Steinberg and Fallon 27 flash hydropyrolysed North
Dakota lignite in preheated hydrogen by dropping
150/~m particles through a 2.4m long reactor tube
shown in Fig. ld. Oxidation of the coal was minimized
by crushing in inert gas and drying in vacuum. The
highest yield of benzene was 10~o and was obtained
under the most extreme conditions available, 973 K
and 104bar. The yield of oil (60~ of which was
benzene) increased with decreasing solids residence-
time and feed rate. The use of argon instead of
hydrogen considerably reduced the oil yield, and
addition of ammonium molybdate to the coal gave no
enhancement of liquid yields.
A slightly larger reactor with more sophisticated
instrumentation was commissioned in 1976. 28 In this,
0.25 g sec- 1 of dried, 50--150 #m coal could be reacted
with 1.3gsec -~ of hydrogen at 173 bar up to 1173K
and up to 277 bar at lower temperatures with a coal
residence-time of 10 sec or less. Liquid yields of 7.5~o
BTX (>90~o benzene) plus 7~o heavy oil were ob-
tained at 998 K and 139 bar with 7 sec solids residence
time. 29 The lignite agglomerated at pressures above
208 bar.
The costs of the production of petrol and substitute
natural gas were estimated assuming 13~ yields of
benzene and light oil, and a 28~o yield of hydrocarbon
gases. It was concluded that the economics came close
to favouring investment under U.S.A. conditions. 3°'31
In addition, an independent analysis of experimental
data (part provided by Brookhaven) found 32 that the
economics of production of substitute natural gas
(with BTX as a byproduct) by flash hydropyrolysis
appeared to be superior to other conversion processes
such as Hygas, Lurgi and Fischer-Tropsch.
In 1973, Sheehan et al. 33 proposed a rotating
fluidized-bed unit to hydropyrolyse coal at 810 K and
70bar pressure. Centrifugal force would throw fine
coal onto porous walls through which heated hydro-
gen would pass both to pyrolyse and to fluidize the
bed. The unit was designed to generate and sweep BTX
out of the reactor in less than 1 sec. This was patented
by Manowitz et al., 34 who claimed the need for coal
residence-times of less than 5sec, and vapour
residence-times of less than 0.2 sec to produce BTX.
2.4. C i t i e s S e r v i c e R & D
Rosen et al. 35 patented the hydropyrolysis of coal in
disperse phase. Coal was rapidly heated for a pre-
determined time by injecting into preheated hydrogen
and the products were rapidly quenched with cold
hydrogen. Very high yields, up to 46% benzene, were
claimed (but not proved) using a sub-bituminous coal
at 200bar and 1063 K.
In a later publication, Pelofsky et al., 36 when
describing work carried out on a 0.28 g sec-1 scale,
gave the yield of light oil (which contained 94~o
benzene) as only 16~o.
By 1977 Cities Service had built and tested the
following reactors : a 1.2 m x 28 mm i.d. free-fall type
shown in Fig. le, a 6-18 m × 6ram i.d. entrained-flow
type, and a 1.2 m long, vertical, entrained-flow type. 37
A range of coals, oil shale, tar sand and coal tar was
hydropyrolysed, although few results have been pub-
lished. Most of the tests were with North Dakota
lignite which was crushed, dried and pulverized in
nitrogen before final drying in vacuum. Solids
residence-times of 0.5~sec with vapour residence-
times of 4-40 sec were obtained in the free-fall reactor,
whereas the entrained-flow systems gave solids
residence-times of 0.08 30sec with vapour residence-
times of 0.08-8.6sec; the 1.2m upward, entrained-
flow reactor gave the shortest residence-time.38
Hydropyrolysis of lignite for 2 sec at 973 K and 153 bar
gave 4.7~o BTX plus 13~o light oil and 12.5~o methane;
the light oil was said to be easily and economically
hydrocracked to benzene. 39 The coal heating rate was
estimated as > 5 x 104Ksec-1 and the hydrogen
consumption was 5~o w/w on lignite. Rapid heating of
coal in the presence of high-pressure hydrogen was
said 38 to lead to high yields of volatiles because the
vaporized tar was stabilized. The quality of the
secondary products could be controlled by altering the
pyrolysis conditions, low severity pyrolysis yielding a
light oil containing polynuclear aromatics, whereas
high severity pyrolysis (up to 1300 K) gave little light
oil but mainly benzene, methane and ethane. A
sequential reaction mechanism was proposed to ex-
plain the results, coal pyrolysing to tar, which reacted
with hydrogen to give benzene; gas was produced at
each stage. Overcracking of the products yielded gas
(methane and ethane) and coke, the latter by catalytic
decomposition of benzene by hot char. Only after the
coal had been devolatilized was there significant
reaction between char and hydrogen to form methane,
the rate of hydrogenation then being dependent on
pressure.
Means of scaling-up the Cities Service process to
produce BTX and byproduct substitute natural gas
(SNG) have been described; 38-4° lignite could be
dried by distilling a benzene-water azeotrope from a
benzene-lignite slurry, which would be pumped up to
pressure before flashing off the excess benzene. The
coal would be fed through separate injectors into a 36-
tube reactor, through which it would be transported
upwards under fast fluidization in hot hydrogen, part
of which would be combusted with oxygen to raise its
temperature further. The exothermic heat of reaction
would be removed from the tubes by a high-pressure
water boiler around the tubes, and molten metal would
be used to quench the products and recover useful
heat.
An evaluation of the economics of SNG production
 The hydropyrolysis of coal to BTX
by Cities Service and the Foster Wheeler Energy
Corporation 4° concluded that flash hydropyrolysis
SNG (BTX was a major byproduct) could cost 80~o of
Lurgi-produced SNG. The major difference between
plants designed for SNG and BTX production was the
higher temperature reactor used to form SNG and the
use of molten-metal cooling of the reactor tubes.
An independent study of the economics of the Cities
Service process by Oak Ridge National Laboratory 41
concluded that the cost of flash hydropyrolysis tech-
nology is comparable with, if not better than, the cost
of other coal conversion processes. Sufficient data for a
firm process design were required before a detailed
engineering evaluation of the process could be under-
taken to determine where further development was
required.
2.5. Rockwell International,
Rocketdyne Division
Oberg et al. 42 described the initial trials of a
hydropyrolysis reactor based on a rocket-motor injec-
tor. Hydrogen was preheated and further heated by
partial combustion with oxygen before being rapidly
mixed with coal and injected into a disperse-phase
reactor. The reactor was used at pressures of 35-
104 bar with a coal feed-rate of 50 g sec- 1.
The products were quenched with water sprays.
Adhesion of the coal to the walls of the reactor was
reduced by heating them to above 1144 K. Reaction
times of 0.01-1 sec at temperatures up to 1420 K were
employed and tar yields of up to 33~o were deduced by
difference from the yields of gas and char.
Later reports 43"44 described the scale-up of this
technique and the rapid pyrolysis of up to 270 g sec- 1
of caking, bituminous Kentucky coal in the unit shown
in Fig. lf. About one third of the coal could be
converted into liquids by injecting coal into hydrogen
preheated in a fired heater, followed by partial com-
bustion with oxygen to 1340K. Up to 9 ~ BTX was
obtained by hydropyrolysis at 70-104 bar for 0.08 sec
at 1255 K. The BTX was recovered by quenching the
reaction with water sprays, removing the char at
811 K, cooling to 311 K to condense the water and tar,
and removing the light aromatics on activated carbon.
Hydrogen equivalent to 3~o w/w of the coal was
calculated to have reacted with the coal. Increasing the
reaction time and temperature increased the methane
yield at the expense of the liquids.
2.6. Institute of Gas Technology
The hydropyrolysis of lignite or non-caking coals in
entrained flow through a heated tube with high-
pressure hydrogen was described by Duncan et al. 45"46
The aim was to convert coal to high yields of gasoline;
the process is known as the Riser Cracking of Coal. A
bench-scale reactor shown in Fig. lg transported 0 . ~
0.5gsec ~ of - 7 4 # m lignite in hydrogen at up to
139bar through a 21m x 3mm i.d. helix heated to
1090K. The reactants were heated to reaction tem-
227
perature in the first 15 m of the helix in 1.7 sec. The char
was separated and tar mist removed before the pres-
sure was reduced. At a pressure of 139 bar and a
residence time of approximately 3 sec, 0.35 g sec- 1 of
lignite were transported by 11 g min- 1 of hydrogen to
yield 10.6~ oil which contained 3.1~o BTX. The
benzene content of the oil increased to 60~ as the coal
outlet temperature and the hydrogen coal ratio were
increased. Generally, 50~o of the carbon in the coal was
converted to gaseous and liquid products, the liquid
yield increasing with total conversion.
The mechanism of reaction was defined as the
pyrolysis of lignite to volatiles which, on entering the
bulk gaseous phase, were fragmented to free radicals
which then reacted with the hydrogen to form stable
species.
Details of the design of a process development unit
to Riser crack 12.5gsec-~ of lignite to high-octane
blending stock, fuel-oil and fuel-gas were published. 47
Lignite and hydrogen at 139 bar would be preheated to
922K before entering the insulated reactor tube
contained in a balanced pressure-vessel. The reactor
would not be directly heated and the energy to raise
the reactants to 1090 K would be obtained by injecting
and reacting oxygen with hydrogen in the reactor.
2.7. F.M.C. Corporation
Eddinger et al. 1 hydropyrolysed sub-bituminous
coals in small, fixed beds with the aim of defining an
economic process for the conversion of coal to a liquid
feedstock for a petroleum refinery, together with
methane and char. Rapid heating was achieved by
passing preheated hydrogen through a bed of coal in a
horizontal reactor, and short vapour residence-times
(seconds) were obtained by using high hydrogen flow-
rates. The pressure was 346 bar. The optimum yield of
26~ of a heavy oil was obtained at 346 bar, 813 K and
2min solids residence-time. The oil yield was in-
dependent of the type of sub-bituminous coal and 1~
molybdenum catalyst had no effect. The yield in-
creased with temperature and pressure and passed
through a maximum with increasing hydrogen flow-
rate. The char was found as an agglomerated mass in
the reaction zone.
2.8. University of Utah
With the objective of producing a liquid fuel of low
sulphur content, Wood and Wiser 4a and Anderson et
al. 49 investigated the catalytic hydropyrolysis of a
bituminous Utah coal. Zinc chloride (5.5~ by weight
of coal) was used as the catalyst. A disperse-phase
reactor consisting of coiled narrow tubes (3-13mm
i.d.) and heated externally was used. The carrier gas
was pressurized hydrogen and, by using high gas
velocities, turbulent flow and rapid heat exchange were
achieved. The catalyst coating the coal particles melted
and, consequently, a liquid annulus was driven
through the tube much more slowly (2-6 sec duration)
than the gas (0.6 sec). To find the optimum conditions
228 G. FVNES,W. R. LADNERand J. O. H. NEWMAN
for high liquid yields, the effects of pressure up to
173bar and of temperature up to 973K were
investigated.
The oil yield, which included material extracted
from the char by toluene, approached a limit of 50%
with increasing temperature. The oil contained typi-
cally: 20% water, 20% benzene insolubles, 15% as-
phaltenes and 40% light oils. The latter contained 60%
neutral oil, 20% bases and 20% tar acids. The zinc
chloride which appeared in the char and oil was
difficult to recover. According to the authors, further
work was necessary on the recovery and recycling of
the catalyst and to find a method to prevent agglomer-
ation of the char to make the process viable.
The results obtained and the techniques used by the
University of Utah were reviewed 5° by Oak Ridge
National Laboratory, with the conclusion that cata-
lyst recovery must be solved if the process is to become
enonomically viable.
2.9. Union Carbide Corporation
With the objective of making chemicals, particularly
phenols, the Union Carbide Corporation 5~ investi-
gated the hydropyrolysis of a sub-bituminous coal fed
at up to 0.13 g sec- 1 into a small continuous, fluidized-
bed reactor, using low hydrogen consumptions with-
out catalyst, and at the relatively low pressures (70 bar)
and temperatures (840K) suitable for commercial
units. The maximum yield ofoil was 28%, two fifths of
which boiled below 533 K and consisted mainly of
phenols, not BTX. The oil yield was halved by
exposing the coal to air during drying or replacing the
hydrogen with nitrogen.
Larger units with throughputs of 1.3 and 158 g sec-
have been operated by Union Carbide. The design of
one of these units is based on an evaluation by Holmes
et al. 52 Problems resulting from agglomeration of
bituminous coals under hydropyrolysis conditions 53
led to this programme being discontinued.
2.10. Schroeder's Studies
Schroeder, a proponent of flash hydropyrolysis, has
published patents on the production of BTX.S4's 5 The
earlier patent 54 dealt with the work performed jointly
by the USBM and the El Paso Natural Gas, and
published later by the USBM. 14 Schroeder based his
claims on the results of experiments especially selected
for their good yields. The later patent 55 describes a
disperse-phase reactor of "raining solids", in which
coal is dropped through a heated reactor with co-
current hydrogen to form a fluidized bed, in which the
char and tar are completely gasified. Schroeder sug-
gested that 30% of liquids could be obtained using
a vapour residence-time of 22 sec at 139 bar and
773 K. 56
2.11. N a t i o n a l Coal Board
The pyrolysis of British coals has been studied at
hydrogen pressures of 50-150 bar and heating rates of
up to 30 K sec-1 in single- and two-stage, tube reac-
tors.57 In the single-stage reactor similar to that shown
in Fig. la, hydropyrolysis of a high-volatile coal gave
up to 4.5% single-ring aromatics; the results suggested
that the evolution of volatiles and their cracking to
benzene were sequential reactions. This was confirmed
using a two-stage reactor shown in Fig. lh, in which
the coal was heated at a controlled rate and the
volatiles cracked in a separate zone at constant
temperature. Yields of up to 12% benzene were
obtained from a high-volatile coal in hydrogen (similar
to the yields obtained by various workers from
American coals by flash pyrolysis) without very rapid
heating of the coal. Hydropyrolysis of model com-
pounds showed that benzene was a major product of
the cracking of polynuclear aromatics under hydrogen
pressure.
3. DISCUSSION
Hydropyrolysis involves the thermal decomposition
of coal together with evolution and cracking of
volatiles under hydrogen pressure. Hydropyrolysis of
coal, like pyrolysis, involves complex mass and energy
transfers and chemical reactions which are dependent
on a wide range of parameters, particularly heating
rate, final temperature, rate of removal of volatiles,
atmosphere and coal rank and size. Many, but not all,
coals melt during devolatilization, so that the physical
structure of the coal and the evolution kinetics are
altered.
The pattern of hydropyrolysis shown in Fig. 2 is a
highly simplified picture based on the results of Finn et
al., 57 but it does illustrate the sequential nature of
reactions and interaction of products in two important
heating regimes. Wolfs et al. derived a pattern of
pyrolysis based on model materials, 58 but pyrolysis
under hydrogen pressure is more complex. Pressurized
inert gases such as nitrogen and helium reduce the ease
of egress of volatiles from coal particles and so increase
the polymerization of tar to char, whereas hydrogen
has a separate effect. Pressurized hydrogen protects
the tar vapours after they have escaped from the coal,
reducing the polymerization of free radicals by react-
ing with them to form stable molecules. The use of
hydrogen pressure encourages the formation of
benzene from the tar. Hydrogen also increases the
agglomeration of coal during pyrolysis so that non-
caking coals can form coherent, even strong, chars. In
addition, once the initial surge of volatiles from the
coal has occurred, hydrogen can penetrate and react
rapidly with fresh, reactive char to produce additional
methane; volatile hydrocarbons may also be hydro-
genated to methane.
Most hydropyrolysis studies have been carried out
with low-rank coals. The amount used per experiment
varied from a single charge of 10 mg or less (Growcock
and Mackenzie ;26 Graft et al. 21), to 283 g sec-1 in the
Rocketdyne reactor. 44 Rapid heating in hydrogen
causes even low-rank coals to agglomerate, although
the effect is less pronounced with lignite than with
 The hydropyrolysis of coal to BTX 229

Cool + H 2

Thermal breakdown of reactive Pyrolysis t o

structures, disproportionation of 873 K under
hydrogen in coal, some H 2 pressure
reaction with H2 gas

Primary tar
I
Primary gas :har Water

I Hydrogenation

Cracking and hydrogenation

_ (~
including demethylation, Thermal cracking to
hydroxylation and functional 1150 K under H2
group elimination pressure

"~ CH Hydrogenation In
_ 4 . ~-
I pyrolysis zone
= C2H6
NH
H~

Benzene Secondary Secondary Char Water

tar gas
FIG. 2. Simple picture of two-stage hydropyrolysis.

bituminous coal. Dilution with inert material and change in hydrogen flow will also change the ratio of
preoxidation have been used to reduce this problem. coal to hydrogen. Consequently, the following re-
marks present an over-simplified picture.

3.1. The Yields of Oil and BTX 3.2.1. Heating rate

The yields of oil, BTX and benzene obtained by
hydropyrolysis are summarized in Table 2. Although
oil yields of up to 40% w/w on dry ash-free coal have
been achieved, those of BTX are much lower (3-15%)
except for one unsubstantiated, and perhaps rather
optimistic, patent claim of 46% by Rosen et al. 35
However, most of these BTX yields are appreciably
greater than the 1% obtained by conventional carboni-
zation at atmospheric pressure in an inert atmosphere.
The effect of rapid heating in an inert atmosphere at
atmospheric pressure is well known. Weight losses can
be at least 30% greater 59 than in the BS method for
volatiles; moreover, the tar yield can be markedly
increased by rapid heating. Most of the best yields of
benzene obtained have also been achieved at very high
heating-rates, as shown (see Table 2). Squires has
shown, 21 however, that for BTX formation no ad-
vantage is gained by increasing the heating rate
beyond about 650 K sec- 1.

3.2. Effect of Reaction Conditions 3.2.2. Temperature and vapour residence-time

The important variables affecting the hydropyro-
lysis of a particular coal are: (I) the rate at which the
coal is heated, (2) the temperature and time histories of
both the coal and the volatiles, and (3) the hydrogen
pressure. Although the main findings on the effects of
these variables are discussed under the headings in the
rest of this section, it must be remembered that
variables are often inter-related in a set of experimental
data; e.g. an alteration of vapour residence-time by a
As might be expected, negligible reaction occurs
when coal is hydropyrolysed below 600 K, but, with
increasing temperature, both the benzene and oil
yields go through a maxima in the range 1000-1100 K.
From the point of view of BTX and oil formation, the
residence-time of the solid appears to be relatively
unimportant, provided that it is adequate to allow the
release of the volatiles, after which the main reaction is
hydrogenation of the residual char. However, there is
230 G. FYNES, W. R. LADNERand J. O. H. NEWMAN
some evidence that the char can catalyse the de-
composition of benzene to coke. 37
On the other hand, the vapour residence-time can
have a marked effect on the yield and composition of
the oil. Friedman et al. 14 found that vapour residence-
times less than 3 sec resulted in the production of dark,
viscous liquids, whereas longer residence-times re-
suited in light distillable oils. The results of Squires and
co-workers 22 showed that there is an optimum
residence-time (depending on temperature) for maxi-
mum BTX yields, which in their apparatus was 3 sec at
973 K. Pelofsky et al., 36 operating a disperse-phase
system, have claimed high benzene yields (15~o) using
vapour residence-times of up to 14 sec.
3.2.3. Pressure
The effect of hydrogen pressure on the yields of BTX
and oil in hydropyrolysis is complex. Application of
pressure alone suppresses the release of volatiles, while
the presence of hydrogen is likely to reduce polymeriz-
ation and increase liquefaction. The overall effect is
that gas yields and, under certain conditions, BTX and
liquid yields, all increase with increase in hydrogen
pressure, as has been shown by Growcock and
Mackenzie. 26 However, the exact form of the relation-
ship between yield and hydrogen pressure is uncertain.
Holmes and colleagues, 52 who have attempted to
correlate the data of other workers on hydrogen
pressure, found that oil yields can be either linear or
square-root functions of the pressure. For example,
their correlation of the data of Albright and Davis 5~
gave a linear function, while that of the data of Kawa 6°
and Eddinger et al. 1 revealed square-root functions. As
the oil yield is also dependent on several other
parameters, it is perhaps not surprising that the
pressure relationship may vary from one apparatus to
another.
3.3. M e c h a n i s m o f H y d r o p y r o l y s i s
Although the main features of the kinetics of coal
pyrolysis (Fitzgerald,61 Chermin and van Krevelen, 62
Pitt 63) and also of hydrogasification have been es-
tablished, it has not yet been possible to resolve in
detail the chemistry of the complex reactions involved
in the formation of BTX and tar. Early work on the
hydrogasification of coal 2 showed that, although the
rate was initially high, it decayed rapidly to a low
value. Some workers considered that the initially high
rate corresponded to the hydrogenation of the vola-
tiles and the lower rate to that of fixed carbon. Other
workers, such as Mosely and Paterson, 5 concluded
that, at sufficiently high pressures, the hydrogenation
of the volatiles and the fixed carbon could not be
treated separately. In a more detailed mathematical
analysis of the kinetics of hydrogasification of a
bituminous coal, Feldman et al. 64 postulated that the
coal contained three types of carbon of different
reactivities; the most reactive was the side chains
which were readily split off.
These kinetic studies of hydrogasification, apart
from showing different regions of reactivity in the
carbon matrix, do not contribute to our understanding
of the formation of liquid products by hydropyrolysis.
In fact, the later stages of gasification are probably not
associated with the formation of liquid products.
However, papers by Anthony and his co_workers65 69
on rapid devolatilization consider the interaction of
hydrogen with the coal during the initial stages of
pyrolysis. They used an apparatus similar to that of
Squires but with an electrically heated wire-mesh,
whereby a thin layer of coal was extremely rapidly
heated. After maintenance of the coal at a definite
temperature for a specified time, it was rapidly cooled
and, from the weight of the residual char, the extent of
devolatilization was determined. The composition of
the volatiles was not investigated.
They found that, for extremely high heating-rates
(750-10,000Ksec 1) and a temperature of 1173K,
most of the volatile loss occurred during the first
second, and was the same whether the atmosphere was
hydrogen or helium. Thus, under these conditions, the
hydrogen had little effect on the initial rapid de-
volatilization of the coal. After this initial stage,
however, devolatilization continued in a hydrogen
atmosphere but not in helium. Moreover, the effect of
pressure on the extent of devolatilization depended on
the atmosphere. In helium, the extent of devolatiliz-
ation decreased with increasing pressure, whereas in
hydrogen it increased with pressure after passing
through an initial minimum.
Anthony and his co-workers, on the basis of their
experimental results, developed a theory of the kinetics
of primary devolatilization of the coal in an inert
atmosphere. Their theoretical treatment, which was
similar to that of Pitt, 63 postulated a set of in-
dependent first-order parallel reactions covering a
range of activation energies. As Fitzgerald61 and,
independently, Chermin and van Krevelen, 62 had
proposed earlier, they postulated that the primary
pyrolysis products formed initially within the coal
particles consisted of two types. These were: (1)
unreactive products which escaped readily into the
surrounding atmosphere, and (2) reactive products,
including free radicals, which could either polymerize
within the coal particle to form coke, or crack on the
hot internal surfaces, thereby depositing solid within
the particle and releasing volatiles into the atmos-
phere. Consequently, factors such as a high heating-
rate and small particle size would reduce the residence
time of the reactive decomposition products within the
coal particle and enhance the yield of volatile products.
Conversely, in an inert atmosphere, high pressure, by
delaying the escape of the reactive molecules from the
particles, would lead to cracking to coke and reduce
the volatile yield. According to their view, hydrogen,
by reacting with the reactive decomposition products
within the coal, would inhibit polymerization and so
increase the volatile yield. The extent to which this
occurred would increase with hydrogen pressure. It
should be stressed that their theory only applies to the
initial devolatilization, i.e. to short residence-times of
 The hydropyrolysis of coal to BTX
the coal. For longer residence-times, the residual
carbon will be partly gasified and, although the extent
will increase with pressure, the type of reaction
involved will be different.
However, Greene 3s suggested that the rapid, initial
release of volafiles from within coal particles undergo-
ing flash hydropyrolysis reduces ingress of hydrogen.
Thus, hydrogen is not available for reaction with
volatiles until they have entered the bulk, gaseous
phase, where the larger, volatile organic molecules
crack to free radicals, which are stabilized by the
hydrogen and cannot polymerize to tar and coke.
Cracking is progressive and sequential via the most
stable molecules such as polynuclear aromatics, sub-
stituted benzenes, benzene and ethane to methane.
Only after the main devolatilization processes have
occurred is there hydrogen counter-diffusion into the
porous char structure to give increased reaction
between hydrogen and carbon; higher hydrogen
pressures strongly influence this counter diffusion.
Anthony and his co-workers investigated the effect
oftemperature on the devolatilization of coal heated in
various atmospheres for short residence times (5-
20 see). The extent of devolatilization, as measured by
the weight loss, increased with increasing temperature
and, above 870 K, was greater in hydrogen than in
helium. They concluded that this effect could not be
explained in terms of kinetics (i.e. that the pyrolysis
and hydrogenation reactions had markedly different
activation energies), but rather that the proportion of
volatile products from the hydrogenation reactions
greatly exceeded that from pyrolysis.
Although these detailed investigations provided
some insight into the way hydrogen affects the de-
volatilization of coal during the initial stages of
hydropyrolysis, they were not concerned with the
proportion and composition of liquid products ob-
tained. These must to some extent be dependent on the
hydrocracking of the volatile products, including
polynuclear aromatics, after they have escaped from
the coal. According to Korosi, Woebcke and Virk, ~°
polynuclear aromatic compounds present in oils pro-
duced from coal must be hydrogenated to their fully
alicyclic analogues in order to provide desirable
pyrolysis products. The extent of hydrocracking will
depend on the temperature and residence time of the
vapour, parameters which Squires has found markedly
influence BTX formation. The combination of tem-
perature and vapour residence-time should be such as
to promote the maximum hydrocracking to BTX and
avoid any subsequent hydrocracking to carbon.
Apart from the possible formation of BTX by
hydrocracking of the volatile products, there are two
other possible mechanisms of formation. The first is by
synthesis from smaller molecules or radicals. The other
possibility is that single benzene rings chemically
bound to the coal matrix are split off by pyrolysis as
benzene radicals and are immediately converted to
benzene by the addition of hydrogen. It is probable
that all three types of reaction contribute to the
formation of BTX.
231
Studies by the authors 57 show that the evolution of
volatiles from high-volatile British coals and their
cracking to benzene during hydropyrolysis are sequen-
tial reactions. This was confirmed using a two-stage
reactor in which the coal was heated at a controlled
rate (5 K sec- 1) and the volatiles cracked in a separate
zone at constant temperature. Hydropyrolysis of
model compounds showed that benzene was a major
product of the cracking of polynuclear aromatics
under hydrogen pressure.
4. CONCLUSIONS
(1) Rapid carbonization of coal in hydrogen at
pressures above 70 bar can yield up to 40% of oil w/w
on coal.
(2) Suitable cracking conditions under hydrogen
pressure, about 1100 K with vapour residence-times of
the order of 1 sec, convert part of the volatiles to BTX,
up to 16% w/w on coal, 94%/0of which is benzene.
(3) The greater part of the benzene is formed by the
fragmentation of larger, volatile molecules in the gas
phase to free radicals, which are stabilized by hydrogen
addition.
(4) Low-rank coals give the highest BTX yields.
(5) Methane, a major product, is formed by:
(a) initial carbonization of coal,
(b) hydrogenation ofvolatiles in bulk hydrogen,
(c) hydrogenation of devolatilized char.
Acknowledgements--We thank the National Coal Board for
permission to publish this review and the European Coal and
Steel Community for financial support. The views expressed
are those of the authors and not necessarily those of the
National Coal Board.
REFERENCES
1. EDDINGER,R. T., JONES,J. F., START,J. F. and SEGLIN,L.,
Paper to 7th Int. Conf. Coal Sci., Prague (1968).
2. DENT, F. J., B.P. 491,453 (1937).
3. DENT, F. J., Gas Research Board Communication, GRB
13 (1944).
4. DENT,F. J., BLACKBURN,W. H. and MILLET,H. C., Inst.
Gas Enars. No. 167 (1937).
5. MOSELY,F. and PATERSON,D., J. Inst. Fuel, 38 (288), 13
(1965).
6. MOSELY,F. and PATERSON,D., d. Inst. Fuel, 38 (2¢~, 378
(1965).
7. MOSELY,F. and PATERSON,D., J. Inst. Fuel, 40 (322), 523
(1967).
8. FELDMAN, H. F., WEN, C. Y., SIMONS, W. H. and
YAVORSKY, P. M., Paper presented at 71st National
A.I.Ch.E. Meeting, Dallas, Texas, 20-23 February (1972).
9. ANON, E. C. N., Chemscope 17 October (1975); British
Carbonisation Research Association communication.
10. HITESHUE, R. W., ANDERSON, R. B. and SCHLESINGER,
M. D., Ind. Engng Chem. 49 (12), 2008 (1957).
11. HITESHUE,R. W., FRIEDMAN,S. and MADDEN, R., USBM
Rep. Invest. 6027 (1962).
12. HITESHUE,R. W., FRIEDMAN,S. and MADDEN,R., USBM
Rep. Invest. 6125 (1962).
13. HITESHUE,R. W., FRIEDMAN,S. and MADDEN, R., USBM
Rep. Invest. 6376 (1964).
14. FRIEDMAN,S., HITESHUE, R. W. and SCHLESINGER,M. D.,
USBM Rep. Invest. 6470 (1964).
15. CLARK,E. L., PELIPETZ,M. G., STORCH,H. H., WELLER,S.
and SCHREmER, S., Ind. Engna Chem. 42 (5), 861 (1950).
232 G. FYNES,W. R. LAONERand J. O. H. NEWMAN
16. WELLER,S., CLARK, E. L. and PELIPETZ,M. G., Ind. Engng
Chem. 42 (2), 334 (1950).
17. WELLER, S., PELIPETZ, M. G. and FRIEDMAN,S., Ind.
Enona Chem. 43 (7), 1575 (1951).
18. KARR JR., C., MAPSTONEJR., J. O., LITTLE JR., L. R.,
LYNCH,R. E. and COMBERIATLJ. R., Am. Chem. Sue., Div.
Fuel Chem. Prepr. 14 (4), 58 (1970).
19. SQUIRm,A. M., Paper to 65th Meeting of A.I.Ch.E., New
York (1972).
20. GRAFF,R. A., DOBNER,S. and SQUIRES,A. M., Am. Chem.
Sue., Paper to Div. Fuel Chem., Chicago (1975).
21. GRAFF,R. A., DOBNER,S. and SQUIRES,A. M., Fuel, 55,
109 (1976).
22. DOBNER,S., GRAFF,R. A. and SQUIRES,A. M., Fuel, 55,
113 (1976).
23. GRAFF, R. A., Flash Hydropyrolysis/Hydrogasification
Cross-Cut Meeting, ERDA (1977).
24. GRAFF,R. A. and LACAVA,A. I., Paper to 176th Meeting
of A.C.S. Div. Petr. Chem. (1978).
25. SQtaRES,A. M., USP 3, 855, 070 (1975).
26. GROWCOCK, R. B. and MACKENZIE, D. R., Fuel, 55,
349 (1976).
27. STEINBERG,M. and FALLON, P., Am. Chem. Soc., Div.
Petr. Chem. Prepr. 20 (2), 542 (1975).
28. FALLON,P. and STEINBERG,M., Brookhaven National
Laboratory Reports BNL 50698 (1977) and BNL 22519
(1977).
29. STEINBERG,M., FALLON, P., DANG, V. D., BnATT, B.,
S1EGLER, E. and LEE, Q., Reaction, Process and Cost
Engineering for the Flash Hydropyrolysis of Coal, 71st
Meeting of A.I.Ch.E, Florida (1978).
30. STEINBERG,M. and SnEEaAN, T. V., Coal Chemistry
Workshop, Stanford Res. Inst., California (1976).
31. STEINBERG,M., SrmEnAN, T. V. and LEE, Q., Paper to
April Meeting of Amer. Chem. Soc., Div. I. and E.C.
(1976).
32. BOESCH, L. P. et al., Flash hydropyrolysis of coal--
techno-economic analysis and comparison with other
processes, Mechanics Research Institute, ERDA Report
FE-2345-1 (1977).
33. SnEEHAN,T. V., STEINBERG,M. and LEE,Q., Brookhaven
National Laboratory, BNL 18268R (1973).
34. MANOWITZ,B., STEINBERG,M., SHEEHAN,T. V., WINSCHE,
W. E. and RASEMAN,C. J., USP 3, 963, 598 (1976).
35. ROSEN,B. H., PELOFKSY,A. H. and GREENE,M. I., USP 3,
960, 700 (1976).
36. PEI.OFSKY,A. H., GREENE,M. 1. and LADE1.FA,C. J., Am.
Chem. Sue., Div. Fuel Chem. Prepr. 21 (5), 154 (1976).
37. GREENE, M. 1., Am. Chem. Soc., Div. Fuel Chem. Prepr.
22 (7), 133 (1977).
38. GREENE,M. I., Fuel Proc. Technol., 1 (3), 169 (1978).
39. GREENE, M. I., LADELFA, C. J. and BIRACCA, S. J.,
Petrochemical feedstocks from coal via the cities service
hydropyrolysis process, Paper 77 to 175th Meeting Am.
Chem. Soc., Div. Ind. Engng Chem. (1978).
40. LADELFA,C. J. and GREENE,M. I., Fuel Proc. Technol., 1
(3), 187 (1978).
41. SINGH, S. P. N., Oak Ridge National Laboratory,
ORNL/TM-6265 (1978).
42. OBEgG,C. L., FALK, A. Y., HOOD,G. A. and GRAY,J. A.,
Am. Chem. Soc., Div. Fuel Chem. Prepr. 22 (2), 185
(1977).
( M a n u s c r i p t received 20 D e c e m b e r 1979)
43. OBERG,C. L., 9th Quarterly Progress Report to ERDA,
Contract EX-76-C-01-2044 (1977).
44. OBERG,C. L., COMBS,L. P. and S1LVERMAN,J., Proc. 15th
lntersociety Enffnff Conf., San Diego, 1,402 (1978).
45. DUNCAN,D. A., BEESON,J. L. and OBERLE,R. D., Am.
Chem. Soc., Div. Fuel Chem. Prepr. 22 (5), 183 (1977).
46. DUNCAN, D. A., BEESON, J. L. and OBERLE, R. D.,
Research and development of rapid hydrogenation for
coal conversion to synthetic motor fuels, Second Quarter
Report, U.S./DOE/FE-2307-25 (1977).
47. DUNCAN,D. A., BEESON,J. U and OBERLE,R. D., Design
of a process development unit, U.S./DOE/FE-2307-29
(1978).
48. WOOD,R. E. and WISER,W. H., Ind. Engnq Chem., Proe.
Des. & Dev. 15 (1), 144 (1976).
49. ANDERSON,L. L., WOOD,R. E. and WISER,W. H., Trans.
Sue. Min. Enffn# 260, 318 (1976).
50. LANTZ,P. M., The University of Utah's continuous coal
hydrogenation process, Oak Ridge National Laboratory,
ORNL/TM-5737 (1977).
51. AI.BRIGHT,C. W. and DAVIS,H. G., Am. Chem. Sue., Div.
Fuel Chem. Prepr. 14 (4) Part l, 99 (1970).
52. HOLMES,J. M., COCHRANJR., H. D., EDWARDS,M. S., JOY,
D. S. and LANTZ,P. M., Oak Ridge National Laboratory,
ORNL/TM-4835 (1975).
53. MCNEESE,L. E., Fossil energy program quarterly prog-
ress report for the period ending 30 June 1978; Oak
Ridge National Laboratory, ORNL-5444 (1978).
54. SCHROEDER,W. C., USP 3, 030, 297 (1962); B.P. 891,971
(1962).
55. SCHROEDER,W. C., STEVENSON,L. G. and STEPHENSON,
T. G., USP 3, 152, 063 (1964).
56. SCHROEDER,W. C., Hydrocarbon Processing, 55 (1), 131
(1976).
57. FINN, M. J., FYNES,G., LADNER,W. R. and NEWMAN,
J. O. H., Am. Chem. Soc., Div. Fuel Chem. Prepr. 24 (3),
99 (1979).
58. WOLFS,P. M. J., VAN KREVELEN,D. W. and WATERMAN,
H. I., Fuel, 39, 25 (1960).
59. ANTHONY,D. B. and HOWARD,J. B., A.I.Ch.E. Journal,
22 (4), 625 (1976).
60. KAWA,W., USBM Bulletin 579 (1958).
61. FITZGERALD,D., Trans. Faraday Soc. 52, 362 (1956).
62. CHERMIN,H. A. G. and VANKREVELEN,D. W., Fuel, 36, 85
(1957).
63. PITt, G. J., Fuel, 41, 267 (1962).
64. FELDMAN, H. F., SIMONS, W. H., MIMA, J. A. and
HITESHUE, W. R., Am. Chem. Sue., Div. Fuel Chem.
Prepr. 14 (4), Part 2, 1 (1970).
65. ANTHONY,n. B., Sc.D. Thesis, M.I.T. (1974).
66. ANTHONY,D. B., HOWARD, J. B., HOTrEL, H. C. and
MEISSNER,H. P., Paper to 15th Symp. on Combustion,
Tokyo, Japan (1974).
67. ANTHONY,D. B., HOWARD, J. B., HOTTEL, H. C. and
MEISSNER,H. P., Rev. Sci. Instr. 45 (8), 992 (1974).
68. ANTHONY, D. B., HOWARD, J. B., HOTTEL, H. C. and
MEISSNER, H. P., Paper to Am. Chem. Soc., Div. Fuel
Chem., Chicago (1975).
69. ANTHONY,D. B., HOWARD,J. B., HOTTEL, H. C. and
MEISSNER,H. P., Fuel, 55 (2), 121 (1976).
70. KOROSl,A., WOEBCKE,H. N. and VIRK, P. S., Am. Chem.
Sue., Div. Fuel Chem. Prepr. 21 (6), 190 (1976).