June, 1952
785
THEMECHANISM OF THE CARBON DIOXIDE-CARBON REACTION I
THE MECHANISM OF THE CARBON DIOXIDE-CARBON
IItEACTION'
BYARNOLDE. RE IF^ Cml Research Laboratory, Carnegie Institute of Technology, Pillsburgh 15, PenwdvarLia Received October 86, 1061
-
Ekpcrirnental evidence is presented to show that the reaction of carbon dioxide with a high telnperature coke a t 900" and (0) CO where ( 0 )repreunder approqimately atmospheric pressure is given by the equations CO?F? (0) CO and C sents a molecuIe chemically bonded to the carbon surface. This mechanism ascribes the inhibition of the carbon dioxidecarbon reaction to the chemical combination of carbon monoxide with oxygen deposited on the carbon surface by carbon dioxide.
+
+
Introduction Research on the Lnetic aspects of carbon gasification reactions is progressing concurrently a t a number of laboratories the world over. In the United States, work in this field has been carried on for a number of years a t the Carnegie Institute of the Massachusetts lnstitute of Technology,576the Institute of Gas Technology,' the Brookhaven National Lahora tory,8 the United States Bureau of Minesg and in other laboratories where the work has sometimes been confidential i n nature. In England, significant contributions have come from groups a t Oxford Universit,ylolll alld Lee& UIliversity.12 Russian \\rorkers13 are among the additional research groL1ps active in this field. Examination of theories proposed for the kinetics of the carbon dioxide-carbon reaction gives a first impression that they differ widely. Similarities and differences between theories stand out clearly only when the proposed mechanisms are restated in uniform nomenclature. It is the purpose of this paper to present experimental and theoretical evidence that has a bearing on these theories. Reactions which play a part in the mechanism of the carbon dioxide-carbon reaction may be summarized as kl
(1)
(1) Abstracted from a dissertation by the author, Coal Research Laboratory Fellow in the Department of Chemistry. Carnegie Instituta of Technology, submitted in partial !ulfillnit!nt of the requirements for the degree of Doctor of Science, June, 1950. (2) hIcArdle Memorial Laboratory, Medical School, University of \Visconsin. hladison, Wisconsin. (3) M. A. Mayers, J . A m . Chem. Soc., 61,2053 (1939). (4) A. A. Orning, 5. Mallov and M . Neff. Ind. Eng. Chem., 40, 429 (1948). (5) W. K. Lewis, E. R. Gilliland and G. T. McBride, tbid., 41, 1213 (1949). (6) P. C. Wu. "The Kinetics of the Reaction of Carbon with Carbon Dioxide," So. D. Thesis, Cheni. Eng., hbmachusetts Institute of Technology, 1949. (7) J . D. Parent and 9. Kats, "Equilibrium Compositions and Enthalpy Changes for the Reactions of Carbon, Oxygen and Steam," Research Bulletin, No. 2, Inst. Gas Tech., Chicago 16, Illinois, 1948. (8) F. Bonner and J. Turkevich, J. A m . Chem. Soc., 73, 5Gl (1951). (9) H. R. Barchelder and J. C. Sternberg. Ind. Eng. Chem., 42, 877 (1950). (10) J. Gadsby, C. N. IIiiislielwood arid Iionof (0)by carbon dioxide, or the rate of desorption of (CO) to be much slower than the rate of gasification of the carbon-osygen comples (0)by reaction (2) forward. Evidence to show that the rate of gasification of (CO) is similar to the rate of gasification of the carbowoxygen c,omplex (0)is obtained from a comparison of Experiments B-69, B-70 with Experim?mts D-99, D-100. I n the first case, a carbon surface saturated with carbon monoxide was evacuated for 0.3 hour before carbon monoxide was readmitted, and 0.026 cc./g. carbon was adsorbed in the succeeding 90 minutes. In the second case, a carbon surfacc saturated with carbon dioxide was evacuated for thc same pcriod before carbon inonoxide was admitted to i t , and 0.014 cc./g. carbon was adsorbed in the succeeding 30 minutes. The amount of adsorpt,ion, which is of the s a n e order, is a measure of the desorption of (CO) in one case and the gasification of the carbon-oxygen complex (0)in the other, which took place in the 0.3-hsur evacuation period. Thus, a t this temperature of 900°, a t which gasification occurs, both the assumpt)ions which would make possible retardation through equation (3) appear to be unwarrant,ed. In addition, the fact that approximately only one-fifth of the carbon surface capable of taking up (0)from carbon dioxide was found to adsorb (CO), makes i t difficult to understand how the chemisorption of (CO) could account for the great rct,ardation of gasification caused by carbon monoside. T o make this theory possible, i t would be necessary to assume that gasificat.ion occurred almost exclusively froin sites capnblc of t>akingup (CO) molecules. A study of t>hereactmionof carbon monoxide with a dcy;tsscd coke surface'l revealed that a t the temperatures of 700, 800 and lOOO", monolayer adsorpt,ion of carbon monoxide was, respectively, 49,74 and 204% of its value a t 900'. The increase in monolayer adsorption a t 1000" to twice its value a t 900" is insufficient t.o account. for a major change in reaction mechanism between these temperatures. The smaller monolayer adsorption values at temperatures lower than 900" decrease the probability that equation (3) controls the retardation mechaiiism a t those temperatures. Further evidence to show that the chemisorption of carbon monoxide on carbon by equation (3) is unlikely to be the controlling mechanism is contained in Table 111. The
Final COS, %
...
"0" adfiorbcd cc./g. c
Ratio
VOd 0
. .
0,290 ,011 ,029 ,088
0.44 1.43 0.20 .10
... .,. .,. ...
,075 ,016 ,072 ,022
.08 .66 .10 .32
,299 - .009
.40 -2.78
.,. ... ,
1,260
...
non-exponential factor and the energy of activation of thc direct>lydetermined14 equilibrium constant,k&/ks for equation (3) are both different in magnitude or in sign from the values for Koz and Ezto which they should correspond when the latter are computed from data by other i n v e ~ t , i g a t o r s 5 ~ ~ ~ ~ ~ on the assumption that) equation (3) cont,rols the retardation mechanism. Even allowing for the difference in carbons used, t,here is no agreement between directly and indit,ectly determined equilibrium constants.
Discussion The conclusion drawn from the evideuce presented is that retardation of the carbon dioxidecarbon reaction occurs through equation (1) reverse. This differs from the choice of retardation by equation (3) made by the Oxford group," which was based mainly on two sets of experiments. From six experiments, in which mixtures of carbon dioxide and carbon monoxide in varying proportions were introduced to charcoal a t 750" for a period of 20 minutes, after which the quantity of oxygen adsorbed 011 the surface was determined, they concluded that a wide variation in the final pressure of carbon dioxide was not accompanied by a corresponding change in the amount of oxygen complex adsorbed, which followed more closely the smaller variation in final pressure of carbon monoxide. This led to the conclusion that a large part of the oxygen on the surface a t t8heend of this time interval was probably due to the adsorption of carbon monoxide. A statistical analysis of the data does not support their interpretation. Straight lines drawn by the method of least squares show that the dependence of adsorbed oxygen on the final partial pressure of carbon dioxide is 1.92 times as great as its dependence on the final partial pressure of carbon monoxide. The standard deviation of points from the former line is only 57% of the standard deviation of points from the latter line. It would seem, therefore, that an opposite conclusion to that drawn by the authors is in better agreement with the data. It was also pointed out by these authors in support of their mechanism that if their experimental data were interpreted in terms of the alternative mechanism, the energy of activation of reaction (1) reverse would assume the impossible value of -16.8 kcal. It may be seen from Table 111, which is an extension of one given in the R4.I.T. publication,6 that the energy of activation of reaction (1) re'i'erse is the sum of Ez and ( E , - E 3 ) ,
788
'
ARNOLD E. REIF
RATECONSTANTS FOR
Vol. 50
TABLE I11 DIOXIDE-CARBON llEACTlON
THE CARBON
(KO,k'OJ/ liol/ KO1 g. iiiolc
KO1 p. mole min.-1
atm-1
Group of workern
Carbon aam111e
g.-l
Oxford11 Coconut shell charcoal 108.8 M.I.T.6 NewEnglandcoke IO4.* M.I.T.6 NewEnglandcoke lo7,$ Electrodecarbon 106.0 Reifl4 High temperature coke
Ei kcal.
Iioz atin.-'
h ' z kcal.
58.8 10-7.9 -45.5 47.9 l O - I . 9 -15.4 61.7 10-6.4 -40.3 50.1 10-8.6 -60.6
..
...
....
Koa atln.-I
E,, kcal.
IIIIII.-~
g.-l
KO8
-
EI Ea kcal.
B. 111olc iiiin-1
ntiti.-L g.-l
E L+&I -Ea, kua/lio, kc.rl. atin.-!
30.1 102.3 28.7 10-5.6 -16.8 - 6 . 4 106-25 4 . 3 103.a 38.0 10-l.6 - 6 . 1 lo9.' 6 7 . 8 103,7 2 7 . 5 10-0.8 - 6 . 6 106.8 56.7 -2.1 lO6J
10-0.6
...
..
...
..
...
....
.. .. ..
E&-Es kcal.
... ...
...
..
...
102J
8.0
The rate of reaction of carbon dioxide with carbon may be calculated for any given temperature by the rate equation where K , = Kol e-1000 E i / R T , Kz = KO, e-1000 E d R T and Ka = Koa e-1000 Ea/RT. The rate = KIPCO~
1
+ KZPCO+ Kspco,'
rate constants employed I n two different representations of the mechanism of thc carbon dioxidc-carbon reaction may be evaluated from Table IV in the following manner
c + (0) +
C
ka
+ (0) -2-co
co
A.5
(CO) ke
where ki = K I ,kz = KiKg/Ka, ka = K I / K ~
and includes experimental ' errors made in the separate determination of E2 and (E1 - E 3 ) . Since the probable error of - 16.8 kcal. derived from the data given by the Oxford group" was 5.6 kcal. or one-third of its absolute value, it is possible that a positive value is within the limits of experimental error. The average value of this energy of activation obtained in the work a t M.I.T. was +21.4 kcal., the positive value indicating that equation (1) reverse is feasible on thermodynamic grounds. It may be argued that experiments with one type uf carbonaceous fuel make it impossible to predict the reaction mechanism with another type of fuel. While there is much evidence to show that the rates of gasification reactions vary widely with composition and reactivity of fuel, there appears to be agreement, between experimenters that the same mechanism applies for all types of carbonaceous fuel. Even when different types of carbonaceous fuels were investigated by the same groups of worker^,^*^^^^ only a single mechanism to cover the behavior of all carbonaceous fuels was advanced. Nevertheless, only further experimental work can disprove the possibility that two competing reactions, expressed by equation (1) reverse as opposed to equation (3), may have different relative magnitudes i n different t
