654

THE COLLIERY GUARDIAN.

September 27, 1918.

CARBONISATION REACTIONS.*

Founded in 1911, the first William Young Memorial
. Lecture was delivered by Dr. Harold G-. Colman in
1912, the second by Dr. W. B. Davidson in 1913, and
the third by Dr. R. Lessing in 1914. After a lapse
of four years, the long interVai being mainly attribu-
table to war conditions, the fourth lecture was delivered
by Prof. J. W. Cobb, Livesey Professor of Coal Gas
and Fuel Industries in the University of Leeds, who
took for his subject “ Carbonisation Reactions.” Prof.
Cobb said that in taking up a complex study such as
that of the reactions occurring during the carbonisation
of coal, there seemed to be two main roads to follow.
One might take the original substance in all its com-
plexity, carbonise it under various conditions, and, by
the careful consideration of exactly recorded results,
attempt to arrive at general conclusions, the validity of
which could be determined by further observation, or
by modifying practice in the direction indicated. This
method was followed more or less by every good gas
engineer, and was what one understood by intelligent
and progressive works practice. Alternatively, one
might try to arrive at a better understanding of the
highly complex by a preliminary study of the simple,
determining, for instance, the mode of formation and
decomposition of some one constituent of coal gas under
a variety of conditions, including, of course, as of
primary importance, those conditions to which it would
be subjected in carbonising practice. This was par-
ticularly a method for the laboratory. When one
considered the number of substances involved in a
survey of carbonisation on such a plan, it was plainthat
the amount of work waiting to be done was enormous,
and left plenty of room for judicious discrimination
between the more and the less important.

In connection with the first method of investigation,
there might be some temptation to believe at first sight
that so much coal had been carbonised under so many
conditions in the past that data must be to hand already
for the solution of all the principal elementary problems.
This was not so. In order to trace exactly what happens
in the process, coal should be heated slowly, and the
decomposition at each temperature should be completed,
or practically completed, before the temperature was
further raised. This could not be done in ordinary
practice, for commercial and other reasons. It w. s
possible, however, to slow down the operation of car-
bonisation to any desired extent—raising the tempera-
ture with extreme slowness, and collecting the products.
This had been done in the laboratory by a number of
workers, and the results obtained were useful and
interesting, but incomplete, as the quantity of coal
which could be handled was not sufficient to allow of a
thorough quantitative examination of the products,
For the solution of this particular problem, a larger
scale of apparatus was necessary—something in the
pature of a small gas plant, designed primarily to allow
of the most effective control of temperature and every
other determining condition.

A Study of Benzene,

The second method of investigation—the study of the
decomposition of single constituents—was one which he
(Prof. Cobb) had applied recently, in collaboration with
Dr. S. F Dufton, to two constituents of coal gas which
had assumed dominating importance—benzene and
toluene. In developing a process for the production of
the latter from other constituents of coal tar, they set
to wrork (starting with benzene) to find out where and
how these substances decomposed under the action of
heat, but bearing in mind the special conditions
attaching to the carbonisation process.

For this purpose they passed benzene vapour through
a silica tube which contained a 5 in. column of red-hot
coke maintained at the desired temperature, and the
rate at which the benzene passed was so adjusted that
its time of contact with the red-hot coke was of the
same order as would occur in carbonising practice—
some 10 to 15 seconds. The benzene was not simply
boiled through the tube, but was first vaporised in a
“ carburetting ” flask by bubbling gps through it, The
gas used was nitrogen when they wished it to be
inactive, but a potentially active gas—usually hydrogen
—^was employed when they wished to determine its
influence on the reaction. By raising the temperature
of the carburetter, the richness of the gas stream in
the benzene could be varied at will. Thus, saturating
at air temperature (15 degs. Cent.), the percentage of
benzene present by volume would be 8, while at 50 degs.
Cent, the percentage would be 36. The stream of gas,
after passing through the coke, came away to a flask
which was simply air-cooled or immersed in ice and salt,
and there easily condensible products were deposited.
It then passed through a second condenser cooled by
solid CO2 or liquid air to complete condensation, and
the remaining gas was collected, measured and analysed.

They first determined that, in a stream of nitrogen
rich in benzene, decomposition had just started at a
temperature of 550 degs. Cent. The benzene then
passing the tube deposited a small quantity of a solid
resembling naphthalene in appearance. It was identified
as diphenyl, and was produced in greater quantities,
along with hydrogen gas, as the temperature rose.
The decomposition indicated was a molecular condensa-
tion of the one-ring compound benzene to a two-ring
compound diphenyl. Eliminating hydrogen, and in the
material deposited at higher temperatures, they found
a three-ring compound, diphenyl benzene, the molecular
condensation having evidently gone a step further.

At 750 degs. Cent, (a red heat) 32 grammes of these
solid highly condensed compounds accompanied 115
grammes of unchanged benzene. This decomposition
of benzene to produce diphenyl by passing through a
red-hot tube has been known since the time of Berthelot,
and is a recognised method of preparing diphenyl. On
the other hand, although benzene is found in some
quantity in coal gas and tar, diphenyl is only present

* William Young Memorial Lecture, delivered by Prof.
J. W. Cobb before the North British Association of Gas
Managers at Glasgow on September 6.

in minute quantities which can hardly be separated.
They were here faced with one of those so-called
differences between theory and practice, or between
laboratory and large-scale results, which indicated that
important factors in the operation had not been taken
into account. In this case the reason was suggested in
the results of experiments made by Mr. Hol lings and
the lecturer, and published in the Transactions of the
Institution of Gas Engineers in 1914. It was there
noted that benzene was not decomposed at 800 degs. Cent,
under conditions similar to those described above,
excepting that the benzene was present in a state of
high dilution with hydrogen and methane, to repr» sent,
approximately the condition of carbonising piacti e.
Dr. Dufton and he, therefore, repeated their experi-
ments—in the first place using benzene highly diluted
with nitrogen. Decomposition was lessened, and ti e
yield of such solid condensates as diphenyl lowered.

They then tried substituting hydrogen for the
nitrogen. Since the decomposition of benzene to
diphenyl liberated hydrogen, the presence of this gas in
quantity from the beginnir g should tend to counter-
balance the reaction, and so limit the decomposition.
They found it to do so. By gradually diluting further
and further with hydrogen, the amount of molecular
condensates—such as diphenyl—produced at 750 degs.
Cent, gradually diminished, until when hydrogen was
simply bubbled through benzene, and the latter, there-
fore, became only about 8 per cent, by volume of the
gas mixture, only traces of diphenyl were produced, the
benzene being practically undecomposed by the treat-
ment. What this indicated was that the condensation
of benzene with the elimination of hydrogen might well
be reversible; and an experiment was therefore made
in which diphenyl was placed in the carburetter and
hydrogen was bubbh-d through it—being saturated at
90 degs. Cent. A small quantity of liquid was found in the
condensing train at the end of the experiment and
identified as benzene. The condensation of benzene to
diphenyl was so proved to be reversible.

The reason was now clearly established for the
stability of benzene in carbonisation, at such tempera-
tures as 800 degs. Cent., and the almost complete absence
of diphenyl, although the latter is so easily produced
from benzene in an ordinary laboratory experiment at
the same temperature.

The solid condensates which had been obtained from
the benzene—nitrogen mixture as 750 degs. Cent, con-
tained no naphthalene and no unsaturated compounds,
such as consumed sulphuric acid in their treatment; nor
did they yield any pitch on distillation. Moreover, very
little carbon was deposited in the hot silica tube.

So far their experiments had been limited to
750 or 800 degs. Cent—a medium carbonis’ng tempera-
ture. At the higher temperature of 900 degs. Cent,,
even in hiyh dilution with hydrogen, benzene was
found to be freely decomposed, with the deposition
of grey methane carbon in the hot-coke layer, the
particles being cemented together by it, and the
separation of softer black sooty carbon at the cooler
end of the tube and in the issuing gas stream.
The process of decomposition had become more drastic
Mr, Rollings and he had previously found that benzene,
even when highly diluted with methane and hydrogen
to imitate carbonising conditions, disappeared com-
pletely at 1,100 degs. Cent,

Experiments on Toluene.

The next experiments were made on toluene (C7 H8),
which may be regarded as a benzene ring with an
attached — CH3 group. The first decomposition of this
compound when vaporised and carried through the coke
by a stream of nitrogen was noted at 550 degs. Cent,
(a dull red heat), and at 600 degs. Cent, sufficient of the
new solid compound was produced to admit of its
examination. It was found to be stilbene (C1%H12 or
C6H5CH=CH— C6H5), which ivsembles the diphenyl
formed from benzene in containing two benzene rings,
but differs from it in being an unsaturated compound.
At the same time an oily substance was formed. Thus
the first decomposition of toluene, like that of benzene,
was found to be one of condensation from a single-ring
to a double-ring compound. But, as is readily com-
prehensible on chemical grounds, it differed in taking
place in more than one direction, as indicated by the
presence of the oil, and in forming an unsaturated
compound, such as would take up sulphuric acid in a
washing process. At higher temperatures increasing
decomposition was noted, and at 750 degs. Cent, it was
quite extensive. The solid products of decomposition
or molecular condensation, on examination, yielded both
naphthalene (double-ring compound) and anthracene,
and consisted partially of unsaturated substances, as was
plainly shown by their high consumption of sulphuric
acid on washing—resulting in a black, viscous liquid.
The plainest distinction from the behaviour of benzene
was, however, yet to come. On substituting hydrogen
for nitrogen as the carrier gas, it was found that
decomposition of the toluene was very much
accelerated, with the production of large quantities of
oenzene and smaller amounts of solid condensates. The
gas collected at the end of the train contained
considerable quantities (such as 50 per cent.) of
methane. There was no difficulty in interpreting these
results.

The formation of molecular condensates, such as
stilbene, takes place with the liberation of hydrogen.
If, then, a large quantity of hydrogen is present
initially, the action is checked, as explained in connec-
tion with the formation of diphenyl from benzene. The
hydrogen atmosphere can, however, act directly on thi
toluene, reducing it to benzene by combining with the
CH3 group to form methane. This reaction, too, is
reversible. On passing a mixture of methane and
benzene through red hot coke at 750 degs. Cent., and
examining the products, they were found to consist
mainly of undecomposed benzene with some diphenyl,
but to contain also a small quantity of toluene. The
formation of toluene from benzene is, therefore, possible
in this way, but is effected only very slowly and under
special conditions. The main difference between the
decomposition of the two substances, benzene and
toluene, lies, therefore, in the action of hydrogen. The

decomposition of benzene is prevented by hydrogen,
because it counterbalances the tendency to form
diphenyl, which would require the elimination of
hydrogen. It facilitate^ the decomposition of toluene by
combining with its CH3 group, and so reducing the
original compound, toluene, consisting of a ring and an
attached group to benzene—a simple ring, forming
methane at the same time. These considerutions are
probably of wide application in the complex process of
carbonisation.

The hydrocarbon next to toluene in the same chemical
series is xylene, or rather the three xylenes ortho-,
meta-, and para-, boiling from 143 degs. to 138 de^S?
Cent. The decomposition of a sample boiling at 138
degs. to 140 degs. Cent, was very similar to that of
toluene, yielding in an atmosphere of hydrogen at 750
degs. Cent, toluene and benzene by reduction, and a
limited quantity of solid condensates. These three
substances, benzene, toluene, and xylene, boiling
respectively at 80 degs. Cent., 110 degs. Cent., and 140
degs. Cent, (average), constitute the most important
ingredients of the light oil fraction in the tar from
medium temperature carbonisation (800 degs. Cent).
Cresol, the characteristic constituent of the “cresol”
fraction in tar distillation, is also reduced extensively
at 750 degs. Cent, in a stream of hydrogen to toluene,
and necessarily to benzene also.

A Study of Carbonisation.

There is clearly indicated a tendency for hydrogen
^always present in quantity in the products of carbonisa-
tion) to reduce the single-ring benzene compounds to
benzene, and at the same time to preserve benzene
itself. A similar action of hydrogen may be inferr- d
as in every way probable on the attached groups of
more complicated ring compounds, resulting in the
formation of naphthalene and anthracene. Hence it is
not surprising that William Young and Thomas
Glover pointed out in 1897 that carbonising at high
temperatures made for the predominance of benzol and
naphthalene in the products and scantiness of the
higher benzols, phenols, and paraffins. The s'udy of
the decompositions undergone by benzene and toluene
may not only be regarded as useful in itself, because of
the great importance of these substances at the present
time, but may be of service in other ways. We see
clearly why the process of carbonisation can never be
regarded as a simple effect of heat independent of the
gaseous atmosphere in which it is conducted, and the
way in which we may hope to modify the results of
carbonisation in various directions by the deliberate
control of that atmosphere. The experiments on
benzene and toluene lead to a quite definite conception
as to how hydrogen may exert its influence in preventing
or minimising those decompositicns which are of the
nature of molecular condens itions—such as the passage
of a one-ring into a two-ring or three-ring molecule—
and, on the ether band, how hydrogen may promote
decompositions tending to simplification of molecular
structure by combining with attached groups and pro-
moting their detachment. Both effects will tend to set
and keep a course for the reactions in carbonisat on
which will lead to the minimum production of heavy
molecules, such as may be taken to constitute the
bulk of the heaviest tar fractions and solids.

An a'tempt had been made to treat coal under con-
ditions which-may be taken as the extreme logical
application of this effect of hydrogen in determining
the nature of the products. Bergius has described a
process in which coal is heated with hydrogen or
hydrogen mixtures under high pressure (preferably
above 100 atmospheres), and he claims that the
hydrogen reacts readily, giving liquid products or
easily melting solids. The process has been descril ed
as “ the liquefaction of coal.” Bergius claims, also, the
application of the same process to such products of
carbonisation as tar and pitch, stating (as exemplifying
the results obtainable) that 1 kilog. of tar heated with
hydrogen to 400 degs. Gent, at 100 atmospheres pressure
for four hours, and afterwards distilled to 250 degs.
Cent., gave about 60 per cent, of distillate resembling
petroleum oil. The general reaction relied upon is no
doubt the same as the conversion of solid diphenyl
into liquid benzene by hydrogen, previously descril ed,
the effect being intensify d by pressure. This method
of treating coal had just reached the stage of works
trial in Germany when the war broke out. The
treatment was rather heroic, and one would be
interested to know what are the various costs of
installation, operation and depreciation.

Turning to more normal practice, very considerable
experimental work bad been done by gas engineers
during the last few years in which this principle of the
importance of atmosphere had been applied, but it was
on y the advent of the vertical retort which had made it
possible to subject the descending charge during carbo-
nisation to the really effective contact, penetration, and
action of an ascending stream of gas: Hydrogen in
quantity, although not unmixed, could be readily
brought into the retort, either as coal-gas stripped of
its by-products or as water gas, while it could also be
supplied indirectly by blowing in sham and depending
on the water-gas reaction which necessarily occurs
when this steam impinges on red-hot coke at the
bottom of the retort.

The action of a gas in the retort might be regarded
in two ways. Even an inert gas, such as nitrogen,
would not be without effect, because it would wash the
volatile products of decomposition out of the pores of
the coke, assist their volatilisation by lowering their
concentration in the vapour phase, and hurry them
away from the region of decomposition. A gas like
hydrogen, being so light, had greater diffusive power,
its molecules travel their zigzag paths at a higher speed,
and it is, therefore, even better able to penetrate the
pores and perform these duties. But the all-important
action of hydrogen is chemical, as has been already
sufficiently emphasised; and other gases cannot be
expected to exercise the same chemical influence. Their
physical action is no doubt similar, but rendered less
effective by their higher density and lower diffusive
power. The general effect of the ascending gas stream
to be anticipated is that of lowering the average mole-