CONCLUSIONS i. Rate constants have been measured for the recombinational desorption of the ~ and B forms of nitrogen from the platinum surface. 2. A mechanism for the recombination of the ~ form of nitrogen has been proposed on the assumption that excited nitrogen molecules will form on the Pt surface. 3. The method of isotopic displacement of 14N and IsN atoms has been used to show that the platinum surface is nonuniform with respect to the adsorption of the ~ form of nitrogen. LITERATURE CITED i. 2. 3. 4. 5.
6. 7. 8.
A . V . Sklyarov, T. M. Dangyan, and I. I. Tret'yakov, Izv. Akad. Nauk SSSR, Ser. Khimo~ 1975, 514. M . U . Kislyuk, A. V. Sklyarov, and T. M. Dangyan, Izv. Akad. Nauk SSSR, Ser. Khim., 1975, 2161. O . V . Krylov, M. U. Kislyuk, B. R. Shub, A. A~ Gezalov, N. D. Maksimova, and Yu. N. Rufov, Kinet. Katal., 13, 598 (1972). M.Y.D. Low and H. Inane, Anal. Chem., 36, 2397 (1964). Yu. M. Gershenzon, S. A. Kovalevskii, V. B. Rozenshtein, and B. R. Shub, Proceedings, All-Union Conference on the Mechanism of Heterogeneous Catalytic Reactions, Moscow, 1974, Preprint 23. Handbook of Chemistry [in Russian], Goskhimizdat (1962). G.W.C. Kaye and T. H. Laby, Tables of Physical and Chemical Constants, 12th ed., Wiley (1959). E . K . Rideal, Concepts in Catalysis, Academic Press (1968).
KINETICS AND MECHANISM OF HYDROGENOLYSIS OF n-PENTANE ON AN AL~.IOPLATINUM CATALYST E. M. Davydov, M. S. Kharson, and S. Lo Kiperman
There has been little study of the kinetics of hydrogenolysis of aliphatic hydrocarbons, and even that rather inexact, having been based on measurement in flow systems where the parameters could be varied over only narrow intervals. Most of the work carried out so far in this field has centered around the C2-C4 hydrocarbons [i, 2]. There has also been little study of the kinetics of hydrogenolysis of pentane [3-5], and the kinetics of reforming in gradient-free systems have been investigated only in the case of heptane . We have studied the hydrogenolysis of n-pentane on an alumoplatinum catalyst in a circulating-flow system, our aim being to obtain exact information on the kinetics of aliphatic hydrocarbon hydrogenolysis. EXPERIMENTAL Experiments were carried out in a sealed, glass circulating-flow reactor similar to that described in , working with systems diluted with H2, or occasionally N2, at 1 atm pressure. The gaseous reaction products were analyzed on an LKhM-8M chromatograph equipped with a flame-ionization detector (3 m • 3 mm column, packed with a 30% solution of tetrabutyrate pentaerythrite on a diatomite which had been treated with HN03 and calcined at 850~ Measurements were carried out on n-pentane whose constants agreed with those reported in the literature; the H2 was prepared electrolytically and the N2 had been carefully purified by the method of . The reactor was loaded with 3.5 cm 3 of commercial AP-56 catalyst containing 0.5% Pt on y-A1203 (TU-38-I-251-69), the specific surface area being 180-200 m2/g and N. D. Zelinskii Institute of Organic Chemistry, Academy of Sciences of the USSR, Moscow. Translated from Izvestiya Akademii Nauk SSSR, Seriya Khimicheskaya, No. 12, pp. 2687-2694, December, 1977. Original article submitted November 18, 1976~ 0568-5230/77/2612- 2483507.50
9 1978 Plenum Publishing Corporation
TABLE i. Composition of the Products from the Reactions of n-Pentane and Isopentane on an Alumoplatinum Catalyst (P~3Hx2 : 0.07 atm; P~176 : 13.3; Vo= 231.2 h -x
Composition, mole % Compound
460 I 2,83 48o I r
440 I t,72 460 I 2,72 48O I 4195
I 2,0i 8,46 4,98 0,87 3,26 5,31
1,98 3,37 5,20 0,06 3,19 5,i5
1 0~6 0,37 0,t8 0,30 0,56
0,95 126,41 62,68
2,37 4,05 t,38 2,44 4,28
130,75 I 04,97 I 76,00 t 63,01 I 51,80
5t,52 38,50 t6,08 20,50 21,85
4,2i 4,62 5,76 2,81 4,38 5,50
0,70 1,02 1,26
TABLE 2. Comparison of the Rates of Hydrogenolysis of the O .pO C~ --C5 Hydrocarbons (PH2" CnHm13.3; Vo 231.2 h -l)
Rate of reaction mmole/h- g
I 0,32 0,15
the predominant pore radius 80-100 A. The catalyst was reduced in the reactor by heating with H2 for 6 h at 470~ and further reduced at the end of each experiment for 30 min. Constancy of catalytic activity was checked by returning to standard conditions at the end of each experiment. Experiments were carried out over the 420-480~ interval, varying the space velocity of pentane flow (Vo) from 47.3 to 604.3 h -I, the initial partial pressure of pentane (P~ from i0 to 125 mm, and the initial partial pressure of H2 (P~ from 283 tO 707 mm. The catalyst granule size was varied from 0.25-0.5 to 2-3 mm. The reaction rate was calculated from the equation r =
where U0 is the rate of pentane input into the cycle, x is the degree of pentane conversion, and W is the weight of catalyst. DISCUSSION OF RESULTS The catalysts proved to be stable under extended operation at the conditions of our experiments. Experiments showed that pentane did not undergo reaction in the empty reactor, and isomerized to no more than 3% in the A1203 loaded reactor at 480~ Olefin cracking could not be detected in experiments with 2-methyl-2-butene on y-A1203 at 455~ From this it was concluded that the principal contribution to the catalytic activity in one hydrogenolysis reactions was from the Pt content of the system. Experiments showed that variation of the size of the catalyst granule had essentially no effect on the reaction rate, at least to within the limits in question here. Inhibition from internal diffusion was thus ruled out of account. Inhibition from external diffusion was not a factor here since the reaction proceeded slowly at a high linear circulation rate (0.6 m/sec). All told, the conditions of reaction assured that measurements would be made in the kinetic region. The composition of the products obtained in typical experiments with n-pentane and isopentane are illustrated by the data of Table 1 (isopentane was studied for comparison since 2484
TABLE 3. Variation of the Rate of Buildup of Hydrogenolysis (r') and Isomerization (r") Products with the Mixture Composition
d 420 I0,05.~ 0,07( 440 l0,01.~ 0,04(~ O,07t. O, 11,~ O, 145 O, 148 O, 128 O, 164 0,070 0,070 0,070 0,070 0,070 450 0,053 0,'070 460 0,053
there was extensive n-pentane isomerization under the conditions in question here)~ It can be seen from the data of Table 1 that the reaction of n-pentane in the presence of an excess H2 was chiefly in the direction of isomerization, hydrogenolysis, dehydrogenation, and autoalky!ation being of lesser significance. The alumoplatinum catalyst was most active in the hydrogenolysis reaction, pentane undergoing statistical breakdown at the C-C bond. In order to determine the effect of the pentane reaction products on the reaction rate, experiments were carried out in which isopentane, n-butane, and 2-methy!-2-butene (B-isoamylene) were introduced into the system in amounts considerably in excess of those met in the course of reaction. None of these products had any appreciable effect on the reaction rate. Since only insignificant amounts of the C~-C3 hydrocarbons were formed here, and the inhibiting effect of these hydrocarbons is considerably less than that of the C~-C5 hydrocarbons , it was concluded that the lower paraffins would not affect the reaction rate. The catalytic activity remained constant as the hydrogen-to-hydrocarbon ratio was reduced from 64 to 3, and then began to fall off irreversibly as this ratio was reduced still further. The fact that the catalyst proved to be carbonized at the end of some runs indicated that an excess H2 was required to maintain constancy of catalytic activity and reproducibility of results. It can be seen from Table 2 that the rates of hydrogenolysis of the Cs hydrocarbons were comparable in value; the rate of hydrogenolysis fell off on passing to the C~ hydrocarbons, just as reported in . Table 3 sums the principle results of experiments carried out with mixture of various compositions at various temperatures. Preliminary treatment of the kinetic data involved the construction of graphs showing r ~, the rate of buildup of hydrogeno!ysis products, and r", the rate of buildup of isomerization products, plotted as functions, first of the partial pressure of pentane (PCsH~2) in excess H2 (Fig. 1), and then of the partial pressure of H2 (PH2) at essentially fixed pentane concentration (Fig. 2). It is seen from the figures that r' increased non!inearly with increasing PCsH~2, while r" was proportional to PCsH12. Increasing the partial pressure of H~ sharply increased the value of r" and reduced the value of r'. The rate of buildup of products from the n-pentane breakdown could be formally described through the equation
A satisfactory description of the experimental results was obtained through Eq.
r', rnmole/h- g r %mmole / h. goAr 8
r', mmole/h, g
r', mmole/h, g
Fig. i. Variation of the rate of buildup of the products of hydrogenolysis (i), and the rate of buildup of isopentane (2), with the partial pressure of n-pentane, at 440~ Fig. 2. Variation of the rate of buildup of the products of hydrogenolysis (i), and the rate of buildup of isopentane (2), with the partial pressure of H2, at 440~ setting m' = 0.75 and n' = 0.5. It is clear that this is only an empirical equation devoid of physical meaning. Analysis showed that the relation in question here could be more adequately described by a kinetic equation of the type
(k' , Pc, + k4. P~':)~," with
Pcs = PC,H,, -~- Pi-C.Hu
P~~ - - ~r5 ~z u i 2 being the partial pressure of the isopentane formed in the course of the reaction. Calculations on a BESM-6 computer using the VTs MGU program [i0] showed that the best description of the experimental data through this equation could be attained with m1' = 1 and n~' = 1 (the values of mz' and n~' were varied from 0.5 to 2), the latter then being equivalent to an equation of the form
kl* "-Pc, k**Pc, -[- P~,
Table 3 shows values of r', experimentally determined and calculated from Eq.
(5) (5) with mean
constant values k ~ = IAS.10~3.~RT mmole/h-g and ~ = 2.00.eNT-; the mean-square deviation being 14.7% at a maximum deviation of 26%. A somewhat greater deviation resulted from treatment of these same data through the exponential Eq. 2. The rate of buildup of the isomerization products was best described by an equation of the form ru ~
mH ~u Pc,u,,P~,
with m" = i and n" = 0.5. The value of m" was obtained from the r" vs PCsH12 relation of Fig. i, and the value of n" from an analysis of the data of Fig. 2, varying n" over the interval from 0 to 2, inclusive. The description of the r" kinetics give n by rationalfractional kinetic equations was even less consistent with the experimental data. The
TABLE 4. Stepwise Scheme for the Reactions of n-Pentane on an Alumoplatinum Catalyst
N gm" 1 ber
2 3 4 5 6 7 t0 It 12 t3 t4 t5 t6 17 18 19 20 2! 22 23
tStoichiometric number of the step along the _ r e a c t i o n path*
IC~Ht~ + H~ =CH4--P. C~Ha~; II C~Ifaz+ II~ =CzH~+CaH~; III C~HL~=C~H,o+H~; IV C~II1z=i-CsHtz; V i-C~Ht2-F.]:I~=CH4;-~I-C,H~o; VI i-C~HI~.+H~_=CH4+C4Hto; VII i-C~II,2q- H2 = C~H~ + CsH,; VIII ~-CsH~ = ~-C~H,t,+ g~.
buildup of isopentane under the conditions kinetic equation of the form
of our experiments was therefore described by a
r" = k".Pc~H,~.P~ Table 3 shows r" values,
obtained experimentally and calculated from this last equation using 23000
the mean value k,'= 83.8 "]07"eRT, mmole/h.g.atmZ.S. Here, the mean-square scription of the rate process was 10.5% and the maximum deviation 23.0%.
deviation in the de-
Drawing oil the data of the literature and the results of our own experiments, we have developed a successive reaction scheme for an assumed process proceeding along eight independent reaction paths (Table 4). Paths I and II lead to pentane breakdown into methane-butane and ethane--propane, respectively, path III to dehydrogenation, path IV to isomerization to isopentane, and paths V-VIII to conversion of the product isopentane to analogous hydrogenolysis and dehydrogenation products. This scheme does not account for the formation of those small amounts of isohexane which appear in the course of reaction, obviously as the result of processes occuring on A1203 acid impurity centers which arise in the course of preparing the industrial catalyst. This scheme is consistent with the data of [ii, 12] insofar as it shows each process, with exception of isomerization, to occur on active ZI sectors of the metallic component of the catalyst. Isomerization is supposed to be realized through migration of intermediate C5HII fragments formed on ZI section to the carrier acid centers in boundary Z2 sectors. Hydrogenolysis proceeds through the rapid dissociative adsorption (step 2) with elimination of two hydrogen atoms to form the intermediate CsHII fragment, followed by subsequent slow cleavage through reactions 7 and 8. Either one or the other, or both, of the hydrogenation steps 3 and 15 (17 and 23 in the case of isopentane) may proceed slowly. The slow step in the isomerization is probably the transformation of the adsorbed intermediate C5H~I fragments on the acid boundary centers of the carrier under the effect of the gaseous H2, a process serving to facilitate product desorption (step 5). The H atoms formed in the course of this reaction migrate from Z2 centers to the platinum Zi centers . The product isopentane can obviously once more adsorb on the ZI (rapid step 16), undergoing slow cleavage in subsequent
steps to contribute to the hydrogenolysis products. All told then, this scheme shows the simultaneous consumption of n-pentane and formation of isopentane through the rapid, steps 2, 4, 9-14, 16, 21, 22 and the slow steps 3, 5, 7, 8, 15, 17-20, 23. Derivation of the kinetic equations must take account of the fact that each of the reactions in question here will be essentially irreversible under our working conditions, the concentration of even the isomerization products being considerably lower than that required for maintaining equilibrium. The rate of buildup of the hydrogenolysis products will here be determined by the sum of the rates of reaction along paths I, II and V-VII, i.e., by r' = r x A- r H -6 r v -4- r w "4- r vH
The theory of steady state reactions  shows that r' = r ~ + r s + rls -6 r19 ~- r,o
the subscripts denoting the numbers of the various steps in the reaction mechanism. For the case of moderate surface coverage of the nonuniform catalyst surface, and uniform alteration in the heat of formation of each of the surface compounds, the reaction rates involved here Can be expressed through the equations 
where D = Eaip i being the sum of the products of the constants for the adsorptional equilibrium a i on the most firmly adsorbing sections by the volatility Pi of the i-th fragment adsorbed layer. The subscripts attached to the k's indicate the number of the corresponding step in the reaction scheme. Since the CsHI~Z~, i - C s H ~ Z~, and HZ: are the principal fragments present on the Z~ surface centers
alpc,n. ~ al'N~m,, + a~p~
Assuming the CsHIIZI and i-CsHIIZ~ surface bonding to equally strong, and setting a 1 ~ one has
D = a1 (Pc,~,, -t- P+c,H,,) -4- aHPH
PH being the volatility of the adsorbed hydrogen atom layer. The volatilities of these various fragments can be expressed in terms of the equilibrium constants K~, K=, and K~6 for the rapid steps i, 2, and 16 through the equations
PH = K ~ P ~ $
PCtH,,-- Kz PC,Ha
of t h e s e e x p r e s s i o n r' =
Pi-CtH,, = K16 Pi-CtH,, 0,5 PH, i n t o Eqs. ( 9 ) - ( i 3 ) g i v e s
Since the rates of hydrogenolysis of n-pentane and isopentane are very nearly identical (cf. Table 2), and the concentration of isopentane in the reaction mixture never more than 35%, one can write, for the conditions in question here 2488
Drawing on Eq. (4), and using the generally accepted m* value of 0.5, one finally has r' = kt*P~ ks*Pc, q- P ~ J
This equation gives a good description of the experimental results. It can be anticipated that isomerization proceeds on poorly covered Z2 centers. is indeed the case, the rate of step 5 would be given by the expression r5 = k~pc,m,PH ,
PCsH,I being the volatility of the surface layer of C5H11 fragments on the Z2 centers. Since t step 4 is assumed to be essentially reversible, P~sH~I = PCsH~I and Eq. (15) then leads to (6) which, in turn, gives a good description Of the experimental results. We would like to thank V. L. Krasnopol'skii for his aid in carrying out the computations. CONCLUSIONS i. Study has been made of the kinetics of hydrogenolysis and isomerization of n-pentane on an alumoplatinum catalyst at atmospheric pressure. The hydrogenolysis reaction follows a rational-fractional equation, and isomerization an exponential equation. 2. A stepwise reaction scheme is proposed in which the limiting step for hydrogenolysis is the C-C bond rupture and that for isomerization transformation of the surface C5H11 fragments on carrier boundary acid centers through the action of the gas phase hydrogen. LITERATURE CITED i. 2. 3. 4. 5. 6. 7. 8. 9. I0. ii. 12. 13. 14. 15.
J. H. Sinfelt, Adv. Catal., 23, 91 (1973). R. S. Dowie, D. A. Whan, and C. Kemball, J. Chem. Soc. Faraday Trans. i, 68, 2150 (1972). E. Kikuchi, M. Tsurumi, and Y~ Morita, J. Catal., 22, 226 (1971). L. Forni, Mao Le Van, and V. Ragaini, J. Catal., 33, 153 (1974). J.-P. Brunelle, A. Sugier, and J.-F. Le Page, J. Catal., 43, 273 (1976). V. P. Sokolov, V. I. Shport, and N. M. Zaidman, React. Kinet. Catal. Lett., ~, 389 (1975). N. E. Zlotina and S. L. Kiperman, Kinet. Katal., 8, 393 (1967). L. Guczi, K. Matusek, P. Tetenyi, and A. Sarkany, React. Kinet. Catal. Lett., ~, 363 (1975). J. R. Anderson and B. G. Baker, Proc. Roy. Soc., A271, 402 (1963). V. A. Skokov, Algorithms for Solving Linear and Nonlinear Problems by Least-Squares Methods [in Russian], Computing Center, Moscow University, No. 25, 1972. P, Weis, Catalysis. Polyfunctional Catalysts and Complex Reactions [Russian translation], Mir (1965), p. 9. J. C. Schlatter and M. Boudart, J. Catal., 25, 93 (1972). S. Khoobiar, J. Phys. Chem., 68, 411 (1964). M. I. Temkin, in: Scientific Principles for the Selection and Operation of Catalysts [in Russian], Nauka (1964), p. 46. S. L. Kiperman, Introduction to the Kinetics of Heterogeneous Catalytic Reactions [in Russian], Nauka (1964), p. 180.