T H E THERMAL DECOMPOSITION OF METHANE BY R . C. CANTELO
That methane could be formed synthetically by direct union of carbon and hydrogen was first demonstrated by Bone and Jerdan in 18g7.l They heated pure sugar carbon in a current of dry hydrogen in a porcelain tube at 1200°C. and obtained about 1% of methane in the products. Later Bone and Coward2, in order to confirm these results, heated purified sugar charcoal as before in a stream of pure hydrogen and from 0.34 g. of carbon heated a t 1020-1070"C., they obtained 14.25 cc. of CH,. This they claimed, confirmed the synthesis of methane at this temperature. Berthelot3 attempted the same synthesis, but could obtain no methane. He attributed the methane obtained by Bone and Jerdan to impurities contained in the carbon used by the FJnglish chemists. Sabatier and Senderens4 by the action of hydrogen upon an intimate mixture of carbon and nickel obtained by reduction of its oxide a t 250-300°C., ascertained that methane was formed a t 250°C., but a t the same time there was a formation of water vapor. After some time, however, the hydrogen ceased to react. The same chemists found no methane when the temperature was raised to 400°C. Mayer, Henseling, Altmayer and Jacoby5, experimenting with the effects of catalysts on the synthesis of methane, found nickel, cobalt and iron efficient in the order given. Pring and Fairlie6found that below IOOO"C.and a t atmospheric pressure, the reaction was extremely slow, even in the presence of a catalyst like platinum. The reaction then began to receive attention from the other side, and several investigations on the decomposition of methane were undertaken. Bone and Coward7, working a t temperatures between 500°C. and IZOO~C. declared that the rate of decomposition was inappreciable at temperatures below 700°C. unless a very large surface was exposed by packing with porous material. They believed decomposition to be largely a surface effect and, as such, affected by the deposit of carbon. Packing with quicklime resulted in a more rapid attainment of equilibrium. An analysis of the products from anlhour's heating a t 103oOC. revealed the decomposition of all but 0.7% of methane. The deposit of carbon was hard and lustrous. J. Chem. SOC.71, 42 (1897). J. Chem. SOC.93, 1975 (1908). Ann. chim. phys. (8),6 , 183 (1905). 4Bull. (4),1, 107 (1907). 5 J. Gasbel, 52, 166, 2.38, 305. 6 J. Chem. SOC.89, 1591 (1906), 99, 1796 (1911). 1 J. Chem. SOC.93, 1197 (rgo8). 2
THERMAL DECOMPOSITION O F METHANE
103 7
Slaterl carried out an investigation to determine whether the rate of dissociation of methane depended upon the physical and chemical nature, as well as on the amount, of the heating surface exposed to the gas. On the basis of the amount of dissociation in an empty tube it was found that silica, magnesia, alumina and baryta slowed up the reaction, while quicklime, copper, carborundum, graphite, charcoal and iron accelerated the decomposition. Heated with powdered iron a t 910’C. for ten minutes, a gas analyzing 99.470 methane gave a product that was 73.1% hydrogen. Slater concluded that each substance exerts a specific influence on the rate of decomposition which therefore depends upon the nature of the substance as well as on the amount of the heating surface. As a resuIt of a later investigation Pring and Fairlie2 found that high pressure hastens the attainment of equilibrium, and that ra.pid cooling prevents the formation of much ethylene. Under a pressure of 30-50 atmospheres, equilibrium was reached at the end of 2 hours with a temperature of 1 2 0 0 13oo0C., and a t the end of 1 5 minutes a t 14oo0C., with or without a catalyst. The reaction was still more rapid a t other pressures. Mayer and Altmayer3 studied the methane equilibrium in the presence of cobalt and nickel catalysts. Using the method of Haber these investigators obtained the equation: 21.1+’8,50~-5.9934
T
logeT-o.oo2936 T
Pca, P2H, They then state : “Durch beliebiges Einsetzen der Temperaturwerte T konnen wir die Gleichgewichtskonstante p berechnen. Hieraus sind mit Hilfe einer quadratischen Gleichung die Gleichgewichtswerte fur CH, und Hz zu ermitteln.” Their calculated equilibrium values for temperatures from 600’c. to 8ooOC. are here reproduced. TABLE I Temperature “C. % CH4 7% Hz 600 31.68 68.32 19.03 80.97 650 11.07 88.93 7 00 6.08 93.92 7 50
= R log, -
800
850
4.41 1.59
95.59 98.41
They attack the observation of Bone and Jerdanl that in their experiments upon the synthesis of methane a t 1200’C. they obtained 1 - 2 7 ~methane. The above results seem to indicate that methane cannot be formed from carbon and hydrogen a t IZOO’C. J. Chem. SOP.109, 160(1916). e J. Chem. SOC.101,91 (1912).
* Ber. 40,2134(1907).
1038
R. C . CANTELO
The following thermodynamical treatment of the equilibrium C+ zHz *CH4 while considerably simplified, is essentially that used by Mayer and Altmayer. The van’t Hoff Isochore is-
d loge K, - - HP dT RT2 Integrated log,, K,= T-R - . dH+const. Hp .fLT
-
dT+ const. RT where C1 is the heat capacity of the reactants and Cz that of the products. Rearranging
Again if a1 is the thermodynamic potential of the gaseous system CH4 and % that of the system zH2 formed from the first one a t constant temperature and pressure, then (@l-@z)TD=RT loge KP-RTZvl log, PI (4) Substituting equation (3) in (4) gives
[(e)
(a’- @ 2 ) T > P -H,-T -
dT+T,const-RTZvJog,
J
For equilibrium +RTZv:loge PI
=
I
az whence T . const. = - H,+T
pl
1
(5)
v ) . d T (6)
Since both C1 and Cz are functions of T, it is possible to express (C1-CZ) in theforma+PT+yT2+ . . . . whence equation (6) becomes T,const= -H,+T
Y’’
TT+ r T 2 . d T + R T log, K,
(7)
or T.const= -H,+aT log, T+P T z + X y T 3 + R T log, K, (8) I n this equation H, represents the number of calories evolved when the products of the reaction are restored to the temperature of the reactants, i.e. H, represents the heat effect that would attend the reaction if it took place isothermally a t the initial temperature. €Ip, the heat of formation of methane is equal to 18,900 calories per gram molecular weight.’ The mean specific heat of carbon per atomic weight between 0 and T absolute is given by the equation2 C, = 2 ,0994+0.001736 T. 1
Hodgman and Lange: “Handbook of Chemistry and Physics,” 9th ed., p. 384. Ann. Physik. (4),14, 328 (1904).
THERMAL DECOMPOSITION O F METHANE
I039
The specific heat per mol of hydrogen is 6.5+0.0006 T.
The specific heat per mol of methane, according to the determined value of Wullner,’ is 9.106. Accordingly ~r=2.0994+13-9.106=5.9934
p=
.001736+,0012=0.002936
Equation (8) above applied to the methane equilibrium may be writtenTconst= - 1 8 , 9 0 0 f 5 . 9 ~ 3 4 T log, T+o,oo2936 T2 + R T log,
-!?.EL
(9)
P2H*. This equation differs from that of Mayer and Altmayer solely in the first term. The writer has used for H,, the heat of the reaction for ordinary temperatures; Mayer and Altmayer in place of this term have used the heat of the reaction calculated for 0°C. There remains the determination of the constant of integration. Mayer and Altmayer determined this constant, from their experimental results. They attempted both the decomposition and synthesis of methane in the presence of the catalysts nickel and cobalt. From their results they obtain the constant = 2 I . I . It is, however, very doubtful that this is the correct value. A critical examination of their results reveals the fact that in no case was equilibrium actually reached. For example a t 536OC., in experiments on the synthesis of methane by nickel as catalyst we find the following divergent values for log K,-o.08061, 0.25239, 0.18290, 0 . 4 5 2 0 0 , 0.28302, 0.17747, 0.13888. Other experimental results for differing temperatures show similar disagreements. The constant, however, calculated from their results varies for all the experiments between 1 8 . 8 and 2 2 . 0 . This is quite possible as an inspection of their final equation shows.PCX, Const. T = - 18, 507+5.9934 T log, T+o.o02936 T 2 + R T log, P2H2
PCE, A variation of rooyo in the value of log, P2H2 produces a difference of only 5% in the value of the constant. It is interesting to note here that von WartenburgZattempted to use the values of K, obtained by Mayer and Altmayer in an application of the Nernst Heat Theorem to the methane equilibrium and he found tha; the calculated and observed values of T were not in agreement. Fortunately the Nernst approximation formula enables one to obtain a “fairly accurate” value for Kp, from which it will be possible to obtain some idea of the order of the constant of integration. Values are those from Mayer and Altmayer’s paper. Z. physik. Chem. 63,269 (1908).
I 040
R. C. CANTELO
The Nernst formula1 is
+
HP Zv I . 75 log T+ZvC 4.571 T where H, is the heat developed a t ordinary temperatures, Zv represents the volume changes and 2vC a summation of constants. logK=
~
If we apply the approximation to the equilibrium a t 600°C. the equation becomes
whence KeOo =0 . 0 7 7 . I n the same wag the results given in Table I1 were calculated.
TABLE I1 Temperature
600 650
CA4 Kp = P7 P =2
0.077
0.039
700
0.021
7 50
0.012
800
0.007
850
0.003 0.003
900 IO00
0.001g
The constant of integration of equation (9) may now be obtained by substituting in it K6o0= 0 . 0 7 7 whence 873Xconst.= -18,goo+5.9934X873 X2.3 log 873 +o.002936 X 873 2 X873 X 2.3 log 0 . 0 7 7 whence const. = 16.
+
O in turn the constants 2 1 and 16. Let us now calculate K ~ Ousing 873 X 2 I = - 18,900+3 5,250+ 2 230+ 2 X873 X 2.3 log K, whence log K,= i.9426 whence K, = . 8 7 where const. = 2 I . Again 873X16= -18,900+35,250+2230+1746X2.3 log K, whence log K, =Z . 8480 whence K,= .07 where cofist. = 16. It is evident, therefore, that the equilibrium concentrations calculated by Mayer and Altmayer (Table I) are in error. Table I11 contains the equilibrium concentrations of CH, and Hz, calculated by the writer using the values for K, given in Table 11. Vide Lewis’ “System of Physical Chemistry,” 1st ed. Vol. 11, p. 388.
1041
THERMAL DECOMPOSITION OF METHANE
TABLE I11 Tempera.ture "C. 600 650
7% CH4
% HZ
6.9 3.5
7 00 7 50
2 . 0
93 * 1 96.5 98.0
800 850 900
0.5
IO00
0.2
I .o
0.4
0.4
99.0 99.5 99.6 99.6 99.8
The method of calculation of the values given in Table I11 is evident from the following example for temperature = 6 0 o 0 c . C+2 H z S C H I Let 2x be number of mols of hydrogen in the equilibrium mixture I -x = number of mols of methane Then
whence
x=o.87
so that the percentage of hydrogen
2x
= I+X ' = 0 0 = 9 3 - I
The values given in Table I1 indicate that even a t as low a temperature as 700OC. methane should be decomposed almost completely into carbon and hydrogen, and that only a trace of methane could be present a t IOOO~C. These conclusions are of considerable interest when applied to the problem of the manufacture of carbon black from natural gas, which is chiefly methane. That an attempt should be made to manufacture carbon black from the thermal decomposition of natural gas is not surprising, for both the products carbon and hydrogen have considerable economic value, while in the production of carbon, in particular, the thermal decomposition process would be competing with an established method of low efficiency. 11. Carbon Black' The various carbons used in the industries are produced by one of the following methods: I. Liberation by direct contact of flame on depositing surface. The product is usually known as carbon black. 2. Production by combustion of oil, tar, or other liquid or solid carbonaceous material with insufficient air. The soot so formed slowly settles on the floors and walls of collecting chambers. The product is usually known a s lampblack. Information included here taken from U. S. Bur. of Mines Bull. 192.
1042
R. C. CANTELO
3. Carbonization of solids, usually refuse, and subsequent mechanical reduction t o a fine state of subdivision. 4. Thermal Decomposition. The blacks yielded by the different processes possess different physical and, sometimes, chemical properties. As a result, each black has its specific use and usually cannot be substituted for another. These substances are not pure carbon, but usually contain hydrocarbons as well as other organic substances and mineral matter. The recovery of carbon black, as manufactured from natural gas under present processes, varies between 2 and 3% and is in direct proportion to the content of ethane. For the gases richer in the higher hydrocarbons, not only is the actual quantity of carbon obtained greater, but the percentage recovery increases as well. The low efficiency has led to the belief that the present processes are wasteful ones, and this fact, taken together with the menace of steadily declining supplies of available natural gas, in turn has produced considerable agitation for laws against the use of natural gas in this manner, for the manufacture of carbon black. Absolute prohibition of the industry in one state and its close regulation in several others have resulted from these conditions. Irrespective of the merits of the question, it is evident that if the thermal decomposition process can offer a greater yield of a similar quality of carbon, these other processes will disappear with the substitution of one which will afford the conservationists less opportunity and desire of attacking. Hydrogen, the other product of the reaction, also has an economic value. The decomposition process for its production seems to be commercially pract,ical, for the only plant a t present operating under this process produces hydrogen alone and does not even collect the carbon formed. A number of methods for the thermal decomposition of natural gas have been patented. In most of them the process takes place a t a temperature of of the theoretical yield has been obI , Z O O ~or C .over, While as high as 4 0 7 ~ tained in this way, a. salable grade has not yet been made by this means. If the carbon after formation is allowed to remain in the heated zone too long, the particles become agglomerated and a grey variety of amorphous carbon results. I n one method this is overcome by providing a cold surface for the collection of the carbon immediately after its liberation, while other processes rely upon the rapid expansion of the decomposing gases to sweep the carbon out of the heated zone. At any rate, the production of a black suitable for use in the rubber industry ought to be capable of attainment by one of these processes, The opinion of an expert on the future of the process may be quoted here. R. 0. Neal of the U. S. Bureau of Mines says: “Thermal decomposition probably offers the most promising method of increasing the quality of black from natural gas. The present methods are destructive to the apparatus-a defect that can undoubtedly be overcome-and the resultant product contains grit or adamantine matter, is grayish and contains some volatile matter.” The above, taken in conjunction with the calculated results given in Table 111, suggests the idea of a low temperature (Le. a t 600-800°C.) decomposition
‘
THERMAL DECOMPOSITION OB’METHANE
I043
of natural gas. The difficulty of such a process lies, of course, in the extremely slow velocity a t which equilibrium is reached between methane and its decomposition products. Obviously, such a low temperature decomposition would require the use of a catalyst; and in the third part of this paper, the results are given of experiments which were carried out in the writer’s laboratory, in the search for a suitable catalyst. One other question may be considered a t this point, viz: does the equilibrium between methane and hydrogen represent the lowest potential energy possible, that is, might there not be also the simultaneous occurrence of other reactions? Methane might decompose in any one of the following ways: (I) CH4 s C + 2 H 2 (2) 2CH4 1 C&4+2H2 (3) 2CH4 SCzH-z+3Hz However, if the equilibrium constants are calculated by the Nernst formula the following values are obtained: K600 K 7 5 0 Ksoa Reaction (I) 13.0 83.4 330 React’ion (2) I X IO-’ I X IO-^ 4X IO-^ Reaction (3) ~ . g X ~ o - ’ IXIO-’ l ~XIO-~ Evidently an entirely negligible amount of ethylene and acetylene can be expected. From 1000 cu.ft. of methane there would be obtained theoretically by decomposition a t 600°C : IOOOX(I.87) = 1870 cu. ft. of mixed hydrogen and methane, containing: 1870X .069=130 cu. ft. of methane. 1870 x .93 I = I 740 cu. ft. of hydrogen, i.e. 870 cu. ft. of methane should be decomposed, and should yield 2 7 pounds of carbon. While these figures would not be reached in any practical operation involving this process, it is interesting to compare the theoretical yield of carbon with those actually obtained in ordinary lampblack and carbon black plants. I n the latter case the value never exceeds 1.5 pounds. Should a successful catalyst for the decomposition of methane into carbon and hydrogen be discovered, there will still remain the difficulty of preventing the carbon from depositing upon the catalyst. This, however, is a mechanical difficulty which human ingenuity can no doubt overcome.
111. Experimental Part The apparatus used in making the experiments which will be described later was of the simplest. The heating unit was a cylindrical electric furnace with a lamp bank in circuit for regulating the current to give any desired temperature. The reaction tube was of fused silica. The gasometer containing
I044
R. C . CANTELO
the methane was connected by means of two large calcium chloride tubes with the reaction tube. The gas, accordingly, passed from the gasometer through the U-tubes and then through the reaction tube, after which it passed through another pair of calcium chloride tubes into a second gasometer. These gasometers were large glass bottles fitted with large separatory funnels and drains, whereby the pressure was regulated. The admission of water through the funnel of the first bottle increased the pressure on the contained gas, and the opening of the drain of the second bottle removed water and lowered the pressure. By proper regulation gas could be transferred from one bottle to the other a t any desired rate. The calcium chloride tubes were provided to keep the gases in the reaction tube free from water. The methane used was obtained by heating together a mixture of sodium acetate and soda lime; the following reaction taking place: CHsCOONa+NaOH
Na2C03+CH4.
Methane prepared by this method is not pure and contains considerable hydrogen. This, however, was of little consequence for our experiments, as we are concerned solely with an equilibrium mixture, one of the constituents of which is hydrogen. The experimental method was as follows: The furnace was brought first up to the required temperature. Nitrogen was passed at a rapid rate through the calcium chloride tubes and the reaction tube containing the catalyst, for 2 0 minutes, and following this the two gasometers were connected in their proper positions and the methane passed through the reaction tube at the desired rate. After about 3 litres of methane had passed through the tube, the experiment was stopped, the gasometers disconnected from the furnace, and the resulting gas mixture was analyzed. The question will arise as to whether the nitrogen will influence the rate of the reaction and the final equilibrium state. Naturally, its presence by decreasing the partial pressures of methane and hydrogen will reduce the velocity a t which equilibrium is reached, but the presence of an efficient catalyst would render this effect negligible. It will, however, affect the final equilibrium st,ate. If we examine the equilibrium expression for this reaction, viz: P2H1 K, = - , it is evident that the presence of nitrogen at an appreciable partial PCH4 pressure tends to decrease the value of the expression on the right hand side of the equation (assuming no reaction takes place). Therefore, to restore the equilibrium value of the ratio K,, a further dissociation of methane must take place. The presence of nitrogen, then, favors the dissociation. The analyses given below have been calculated to a nitrogen-free basis, in order to compare the methane and hydrogen contents with those given in Table I11 for the equilibrium percentages of these gases.
THERMAL DECOMPOSITION OF METHANE
Experimental Results.
I. TEMPERATURE 600°C. Run No. I-No catalyst. Rate-20 cc. per minute. Volume CH, used-2700 cc. Volume of resulting gas-2650
cc.
Analyses Original
CH4
Final
91.2 6.9 0.6 1.3
88.9 8.7 1.2
c02
0.0
0.3
0 2
0.0
0.0
H2
CzH4
co
0.9
Run No. 2-CaO used as catalyst. Space Velocity-58 litres gas per litre catalyst space per hr. Volume of CH4 used-3400 cc. Volume of resulting gas-3400
cc.
Analyses Original
Final
CHI
85.0
83.5
HZ
12.0
14.4
2.1
1.1
CzH4
co
0.6
1.1
COZ
0.0
0.0
0 2
0.3
0.0
Run. No 3-Bone black as catalyst. Space Velocity-42 litres gas per litre of catalyst space per hour Volume of CH4 used-25 jo cc. Volume of resulting gas-2700 cc. Analyses Original
Final
CH,
85.5
80.0
H2
13.1
17.8
CzH4
1.4
0.2
co coz
2.0
0.0
0.0
0.0
0 2
0.0
0.0
Run No. 6-Cu-CuO as catalyst. No increase in volume was obtained. Final gas was not analyzed. Run No. 7-Asbestos impregnated with Ni-NiO as catalyst. Space velocity-41 litres gas per litre of 'catalyst space per hour. Volume of CH, used-3100 cc. Volume of resulting gas-3400 cc.
1046
R . C. CANTELO
Analyses CH4 Hz CZH4
Original
Final
88.0 10.3
81.2
1.7
0.0
0.0
0.0
co coz 0 2
17.2
0.0
1.6
0.0
0.0
Run No. 8-Asbestos impregnated with Ni-NiO as catalyst. Space velocity-46 litres gas per litre of catalyst space per hour. Volume of CHI used-3500 cc. Volume of resulting gas-3800 cc. Analyses CHI CZH4
Original
Final
88.0 10.3 1.7
81.3 17.7
0.0
0.5
0.0
0.5
0.0
0.0
co coz 0 2
2.
0.0
TEMPERATURE 7ooOC.
Run No. 11-No catalyst. Rate-30 cc. per minute. Volume of methane used-4050
cc. Volume of resulting gas-4000 Analyses
Original
Final
CH4 Hz CZH4
95.6 1.9
87.6
1.1
1.0
c02
0.0
0.3
0 2
0.0
0.0
1.4
co
10.4 0.7
Run No. 9-Asbestos impregnated with Ni-NiO as catalyst. Space velocity-43 litres of gas per litre of catalyst space per hour. Volume of CH4 used-3300 cc. Volume of resulting gas-4150 cc. Analyses CH4 H2 CzH4
Original 91 . o
8.4
co coz 0 2
'
Final
56.0 43 ' 2
0.4
0.0
0.2
0.8
0.0
0.0
0.0
0.0
Run No. Io-Catalyst as in Run 9. Space Velocity-43 Volume of CH, used-3500 cc. Volume of resulting gas-4000
cc.
cc.
THERMAL DECOMPOSITION O F METHANE
Analyses CH4 H2 CZH4
co coz 0 2
Original 91 . o
8.4 0.4
Final
62. I 37.9 0.0
0.2
0.0
0.0
0.0
0.0
0.0
3. TEMPERATURE 76oOC. Run No. 12-No catalyst. Rate-30 cc. per minute. Volume of CH4 used-2700 cc. Volume of resulting gas-2900 cc. Analyses
CH4 H2 C A
Original
Final
92.9 6.5
85.7 12.2
0.0
0.6 0.3
GO2
0.6
0.8
0 2
0.0
0.4
co
0.0
Run No. 13-Asbestos impregnated with FezOs-Fe. Space velocity-30 litres gas per litre of catalyst space per hour. Volume of CII, used--g700 cc. Volume of resulting gas-3950 cc. Analyses Original
CH, Hz CZH4 CO
89.5 9.1
= .4
0.0
Final
81.4 13.8 0.0
2.2
coz
0.0
2.5
0 2
0.0
0.0
Run No. 14-Silica gel as catalyst. Space velocity-21 litres gas per litre of catalyst space per hour. Volume of CH4 used-3300 cc. Volume of resulting gas-3400 cc. Analyses CH4 H2 CzH4
co
Original
Final
89.5
83.5
7.0
14.2
3.5
1.6
0.0
0.0
COa
0.0
0.7
0 2
0.0
0.0
I
048
R. C. CANTELO
Run No. I 5-Asbestos impregnated with MnOz Space velocity-zo litres of gas per litre of catalyst space per hour. Volume of CH4used-3400 cc. Volume of resulting gas-3400 cc. Analyses As there was no increase in v'olume during this experiment an analysis of the resulting gas was not made. 4.
TEMPERATURE 78oOC.
Run No. 16-Asbestos impregnated with Ni-NiO as catalyst. Space velocity-52 litres of gas per litre of catalyst space per hour. Volume of CH, used-2700 cc. Volume of gas obtained-4250 cc. Analyses CH4 H2 CzH4
co coz 0 2
Original
Final
83.5
23.6
7.9
71.9
4.0
0.0
2.4 0.9
1.3
1.9 2.5 0.0
IV. Conclusion An examination of the above experimental results shows that the majority of the contact materials used produced some catalytic effect upon the reaction. The effect of the Ni-NiO catalyst is especially noticeable, and the catalytic action of this material is such that it seems highly probable that complete equilibrium may be obtained a t 78ooc., using t,his catalyst and a smaller space velocity. Experiments are now in progress in this laboratory with a view to determining the equilibrium concentrations of methane and hydrogen in the way indicated above. From such equilibrium concentrations, the equilibrium constant, K,, may be determined for the methane equilibrium, and the value so obtained compared with that calculated by means of the Nernst approximation. The writer gratefully acknowledges the assistance of Messrs. A. G. Hewitt, and E. L. Stauffer, in part of the experimental work described in this paper. William H . Chandler Chemical Laboratory Lehiph University, Bethlehem, Fa.