physicail chemistry - ACS Publications - American Chemical Society

---

THE J O U R N A L O F

PHYSICAIL C H E M I S T R Y 0 Capyright, 1979, by the American Chemical Society

Registered in U.S. Patent Office

VOLUME 83, NUMBER 7

APRIL 5, 1!)79

Computer Simulation Studies of Ethane Pyrolysis in Shock Tubes at 1206 K Wei-Mln Lee and Chuln-Tlh Yeh" Institute of Applied Chemistty, National Tsing Hua University, Hsinchu, Taiwan 300, Republic of China (Received December 12, 1977; Revised Manuscript Received December 8, 1978)

The mechanism of ethane pyrolysis in a shock tube at 1206 K was investigated. A kinetic scheme of tieven elementary steps provides a good description of the formation of ethylene and acetylene. Reaction 9, the combination of hydrogen atoms and methyl radicals to form methane, was the dominant chain termination process in this new kinetic scheme. A computer-simulated experiment at 1206 K using this proposed mechanism explained the product distribution observed in Burcat et al.

Introduction Studies on the thermal decomposition of ethane may be divided into two groups according to temperature. Low temperature experiments (5" < 1000 K) are usually carried out wiith pure ethane in conventional reactors,1-6 while experiments a t high temperatures (T > 1000 K) are conducted with dilute ethane (mixtures of ethane in an inert gas) in shock The thermal decomposition of ethane in a conventional reactor has been widely accepted to proceed through the following Rice-Herzfeld mechanism,l which generates the observed products methane, ethylene, and hydrogen? chain initiation C$& 2CH3 (1) CH, + CzH6 CH4 + CzH5 (2) chain propagation CzH5 H

--

+ CzH6

CzH4 + H

--+

Hz

+ CzHS

(3) (4)

chain termination 2C2H5 -.+ mC4Hlo (54 2CzH5 CzH6 CzH4 (5b) According to this mechanism, hydrogen atoms and ethyl radicals are the chain carriers and methyl radicals are the essential intermediates to initiate the propagating chain. The steady state concentration of this intermediate, attained a t early stages in this thermal decomposition, is +

+

0022-3654/79/2083-077 l$Ol .OO/O

[CH,I,, = 2hl/h2 ( 1) At low conversion, the rate of methane generation has been successfully used to calculate hl, the first-order rate constant of ethane dissociation, according to d[CH41/dt = kz[CH3Iss[C&I = 2 h i [ C & j l (11) The high pressure limit hl of such studies is temperature dependent according to the Arrhenius e q ~ a t i o n ~ 1 ~ J - l ~ log k l = 16.75 - 88000/2.3RT (111) The decomposition of ethane has also been extensively studied in shock tube^.^-^^ Methane, ethylene, and hydrogen were also observed as products in these high temperature experiments. However, relation I1 and the rate of methane formation observed in these high temperature pyrolyses always lead to a hl value that is smaller than the constant calculated from eq 111. Many investigators have proposed6J0-12that this anomalous devi,ation phenomenon may be caused by the occurrence of ,additional radical-termination processes along with the interaction between two ethyl radicals (reactions 5a and 5b) which was the domination termination process a t low temperatures. Although reactions 6-8 have been suggested 2CH3 CzH6 (6) CH3 + C2H5 C3Hs (7) H + CzH5 CzH6 (8) as possible termination steps and the distribution of products is available in the literature over a wide range -+

-

+

6Z 1979 American

Chemical Society

772

The Journal of Physical Chemistry, Vol. 83, No. 7, 1979

W . 4 . Lee and C.-T. Yeh

TABLE I: Mechanism and Rate Constants Used in This Computer Simulation Study log A , or E, L mol-’ s-’ kcal/mol

reaction (1)C,H, 2CH, (2) CH, t C,H, -+ CH, + C,H, (3) C,H5 + H + C,H, H, C,H, (4) H + C,H, (5a) 2C,H5 n-C,H,, C,H, t C,H, (5b) 2C,H, (6) 2CH, C,H, ( 7 ) CH, t C,H, -+ C,H, (8) H + C,H, 2CH, (9) H t CH, -+ CH, (10) 2H t M .+ H, t M ( l l a ) H t C,H, C,H, ( l l b ) H t C,H, H, t C,H, (12a) CH, t C,H, -+ C,H, (12b) CH, t C,H, -+ CH, t C,H, (13) C,H, -+ H t C,H, (14) H t C,H, -+H, t C,H, CH, t C,H, (15) CH, t C,H, -+

+

-+

-+

-+

--f

-+

16.75

k at 1206 K

88.0

13.80 11.00 10.00 9.15 10.40 10.05 11.00 11.30b 9.00 10.20 11.20 8.32 8.30 11.20 10.20 9.00

6.3 (6.5 X 3.6 X 2.2 x 1.0 x 1.3 x 2.5 X 3.0 X 1.0 x (2.7 X 1.0x 2.0 x 4.6 X 1.0x 3.0 x 2.0 x 1.5 X 1.0x

40.0 9.4 0 0 0 0 0 0 0 5.7 14 7.3 10.0 38 .O 0 0

reaction ratea 3.7 x 2.8 x 1.3 x 1.3 X 1.3 X 1.7 X 1.4 x 8.1 x 3.4 x 1.9 x 5.0 X 1.0x 2.3 x 9.1 x 2.7 X 2.1 x 1.4 x 8.0 x

10’)’ 10, 109 10’O 109 10” 10’’ 10” 10”)’ 109 109 10’ 107

10-4 10-4

lo-, lo-’

10-4

lo-, 10-5 lo-,

ref 4,5,11-15 see text 4,5d

e

see, text

f f,g

h 2

see text 10-l’ e 10-3 e,j lo-, k 1@6 1 lo6 lo-? m,n 104 k 10” 10-5 k ,0 109 lo-’ P In M s-l at 700 M S , At infinite pressure. Simulated value of the present study. J. V. Michael and G. N. Suess, J. Chem. Phys., 58,2807 (1973). e W. E. Johns, S. D. Macknight, and L. Teng, C h e m . Rev., 73, 407 (1973). M. J. Gibian and R. C. Corley, C h e m . Rev., 7 3 , 441 (1973). C. H. Bamford and C. F. H. Tipper, “Comprehensive Chemical Kinetics”, Vol. 5 , Elsevier, New York, 1972. k , = 2k61i2k5a1iZ, ‘ L. Teng and W. E. Jones, J. Chem. SOC.,Faraday Trans. 1, 68 (1972). J. V. Michael, D. T. Osborne, and D. G. Keil, J. Phys. Chem., 58, 2800 (1973). G. B. Skinner, R. C. Sweet, and S. K. Danis, ibid., 75, 1 (1971). P. M . Holt and H. A. Kerr, Int. J . Chem. Kinet., 9 , 185 (1977). A. F. TrotmanDickenson and E. W. R. Steacie, J. C h e m . Phys., 19, 169 (1951). C.-J. Chen, M. H. Back, and R. A. Back, Can. J. Chem., D. G. Keil. K. P. Lvnch. 54. 3175 11976). Esti” , J. A. Cowfer. and J. V. Michael, Int. J. Chem. Kinet., 8, 825 (1976). mated f r o h k,,’by assuming k i / k C H , = 10. -+

-+

-+

‘

‘

’

of reaction conditions, no attempt can be found in the literature to correlate the obtained product distributions of shock tube studies with any modified mechanism. In 1973 Rurcat et al. reported product distributions of four shock tube experiments, performed a t 1206 f 8 K, on cracking a 6.3 atm mixture of 0.1% CzHs in argon.12 The reproducibility of these four experiments is remarkable. Since computer simulation work has been successfully applied to propane pyrolysis in shock tubes,16 a similar computer study was tried in the present work to correlate literature rate constants of elementary processes to experimental product distributions as reported by Burcat et al. for ethane pyrolysis at 1206 K, a much simpler reaction than propane pyrolysis. The initial concentration of ethane was 6.1 X 10-5M-l. A constant step size of 1 X lo-@s was used for this integration. The pyrolysis time of the simulation was 700 ~ s .

Prototype Mechanism Table I lists a prototype mechanism and published Arrhenius parameters for the reactions involved. Since hydrogen atoms, methyl radicals, and ethyl radicals have been established as the main intermediates of the pyrolysis, all the possible radical combination reactions (processes 5-10) were therefore added to the reaction scheme as candidates for significant termination steps. (As can be seen in Table 11, the derived concentrations of hydrogen atoms and methyl radicals become comparable to that of the ethyl radical at high temperature and low pressure according to the mechanism of Rice and Herzfeld).’ H + CH, CHI (9) +

2H+M +Hz+M

(10)

Reactions 11 and 1 2 were also listed in this prototype H C C?H, CZHj H + CZH, Hz + CzH, CH, + CZH, CBH7 CHB + CzH4 CH, + CzH, +

+

+

+

(1W (1lb) (124 (lab)

TABLE 11: Variation of Steady-State Concentrations of Radicals in Ethane Pyrolysis with Cracking Temperature and Ethane Concentration’ T, K 800

1000 1300 1000 1000 io00

steady-state radical concn, M [c,H,I, M [HIb [CH31b [CAIb io-, 6.2 x 10-14 1.1 x 10-13 2.2 x 10-10 io-z 7.4 x io-io 1.7 x 10-9 5.7 x 1 0 - 8 10.: 4.3 x 1.3 x 9.5 x lo-, 10-3 2.3 x 10-9 1.7 x 1 0 - 9 1.8x 1 0 - 8 10-4 7.4 x 10-9 1.7 x 10-9 5.7 x 10-9 2.3 x io-8 1.7 x 10-9 1.8 x 1 0 - 3

a Assuming a Rice-Herzfeld mechanism. [CH,] = 2k1/kz,[C,H,] = (k,”’/h,”’)[C2H,]’’2. [HI = k 3 k 1 1 ” / k,k I ’ [ C2H,]

mechanism to cover the possible interactions among reaction products and reaction intermediates. Finally, reactions 13-15 were introduced to account for the acetylene observed in the pyrolysis product. C2H3 4 H + C2H2 H CzH3 HZ + C2Hz CHB + C2H3 + CH4 + C2H2

+

4

(13) (14)

(15)

This reaction scheme was put into the computer for the present simulation study. One of the most interesting reactions in this kinetic scheme is the abstraction of a hydrogen atom from ethane by methyl radicals. This reaction has been investigated in the temperature range 300-600 K by Trotman-Dickenson and Steacie, who found log (k,/M-’s-l) = 9.0 - 11.2/2.3RT.17 This relationship leads to k, = 1 X lo7 M-’s-l at 1206 K. However, recent studies on this rate constant showed an upward curvature at temperatures over 800 K.ls-zOPacey and Purnell found log k 2 = 11.7 - 21500/2.3RT over temperature ranges 920-1040 K (kz = 7 X l o 7 at 1206 K).6 Clark and Dove, on the other hand, preferred a nonlinear expression log k2 = -3.26 4 log T - 82800/2.3RT for 300-1800 K (4.4 X 10: at 1206 K).19 In the present simulation study, literature Arrhenius parameters in Table I were accepted for all reactions involved with the exception of k2 and h9. These

+

The Journal of Physical Chemistry, Vol. 83, No. 7, 1979 773

Computer Simulation Studies of Ethane Pyrolysis

TABLE 111: Product Distribution for t h e Decomposition of Ethane' typeof expt

temp, K

CH,d

C,Hdd

shocktubeb

1198 1202 1208 1216 1205

0.65 0.63 0.54 0.33 0.45

7.7 10.2 9.7 8.1 8.7

computerC

C,HSd 0.06 0.10 0.04 0.04 0.06

0.1% ethane in 6.3 atm of argon, dwell time is 700 ps. Reference 12. Computer simulated result using the mechanism and rate constants listed in Table I. Moles of product/100 moles of original C,H,.

mechanism is obtained from Table I by eliminatiing reM s-l: actions with rates less than 1 X chain initiation k

C2H6

52CH3 k6

2

4

6

8

Tirne/l O-4s Flguro 1. Calculated concentration-time profile of the computer simulated experiment described in Table I.

two rate constants were varied to seek optimized values that would generate the product distribution 'observed in the experiment of Burcat et al. At room temperature, hg follows third-order kinetics at pressure below 20 torr and approaches its high pressure limiting value (2 X lo1' M-l s-l) when the system pressure is higher than 1atm.21*22However, kg deviates f3ooner from h9with increasing temperature. At 1000 torr, hg decreases from 3.2 x 1O1O at 500 K to 2.5 X 1O1O a t 700 K, and to 2.0 X I O L oa t 1000 K.23 Under the present condition (6.3 atm of argon and 1206 K), hg is expected to be comparable to the rate constant of the combination of methyl radicals (reaction 6). Much controversy has also been caused by reaction 5a, the combination of ethyl radicals. Early sector experiments proposed = 2 X 1O1O M-I s-1.24This rake constant was revised to 2 X IOs by Benson et al. in experiments using radical buffer technique^,^^ and by Hughes e t al. through data on n-butane pyrolysis.26 Recently, Golden et al. increased h5ato 1 X 1O1O again using a modern VLPP technique.27 This newly obtained value was arbitrarily assigned to h5ain the present study. Fortunaitely, interactions between ethyl radicals will be shown as a negligible process in the Results and Discussion. The same product distribution will therefore be obtained no matter which of these reported values is accepted. R e s u l t s arid Discussion A simulated product distribution, when rate constants listed in Table I were used, may be found in Table 111. This distribution is seen to be in good agreement with the experimental distributions of Burcat et al. Figure 1 illustrates the calculated concentration-time profiles. CzH5,the major radical in the low temperature reactions, turns out to be a minor radical, due to a high k3 under this pyrolysis condition. Process 5 as well as other combination reactions involving ethyl radicals therefore no longer plays a major role in chain termination. Table I furnishes the calculated rate of each process a t the end of this pyrolysis (700 ps). It is evident from this list that only two processes, i.e., reactions 6 and 9, in the mechanism are significant in terminating radicals in the system. The overall mechanism for the thermal decomposition of ethane in shock tubes may be considered in the light of the present findings. The following summarized

CH3 + C2HG

k2

CH4 + C2H5

+

chain propagation k3

klla H + C2H4 H + C2Hs Hz + C2H5 C2H5

k4

+

H

+ CzH4 CZH,

kllb

k 13

H2

H

+

+ C2H3

+ C2HZ

chain termination H

+ CH3-

k9

CH4

Only two reverse reactions (reactions 6 and l l a ) are shown in this shortened mechanism. The remainiing reverse reactions are excluded because they are involving either an inactive product (CH4or H,) or a product (C2H2) of very low yield. One of the interesting consequences of this mechanism is that reactions 6 and 9 join reaction 2 in the consuming of methyl radicals generated from reaction 1. Therefore, eq I is no longer valid for ethane pyrolysis at temperatures higher than 1000 K. Equation 11, therefore, leads to hl values that are smaller than those derived from eq 111, as observed in shock tube studies.'@12 In this revised mechanism, reaction 2 and reaction 9 are responsible for the generation and the termination of the C2H5and H. Accordingly, these two reactions will have an equal rate if radicals reach steady state. Since both reactions 2 and 9 generate methane molecule, then d[CH,]/dt = Rz + Rg 2R2 = 2hz[CH3][C,H,]

+-

where R, denotes the rate of reaction n. The reliability of the simulated hz therefore relies heavily on the accuracy of the concentration of methyl radicals, or on the h l , h6, and k g values which decide the steady state concentration of CH3 hl and k6 increase with the system pressure.2s At 1206 K, these rate constants are fairly close to their high pressure limiting condition under 6.3 atm of argon.2s-30 The acceptance of hl and k6 in this study thus would not be expected to cause too much deviation from the measured value of k z of 6.5 X lo7 M-l s-l. However, this determined value is substantially higher than 1.0 X lo7, a value obtained from extrapolation of low temperature data to 1206 K.I7 The non-Arrhenius behavior of reaction 2 is thus supported by this study. Process 4 is another reaction that has been reported to have non-Arrhenius behavior. Clark and Dove relcently proposed a relation log k4 = -4.27 + 3.5 log T - 520012.3RT

774

The Journal of Physical Chemistry, Vol. 83, No. 7, 1979

for this nonlinear rate ~ 0 n s t a n t . lThis ~ relation suggests k4 = 4 X lo9 M-l s-l a t 1206 K. This is around a factor of 2 larger than that expected from the Arrhenius equation used in Table I. According to our calculation, acceptance of this large k4 value will not significantly affect the simulated k2 of this study, but it does increase k9 to 5 X 1010. In conclusion, due to the dramatic increase of the rate constants of reaction 1 (E, = 88 kcal/mol) and reaction 3 ( E , = 40 kcal/mol) with temperature, methyl radicals and hydrogen atoms take the place of ethyl radicals as the dominant radicals in ethane pyrolysis in shock tubes. Accordingly, the mechanism for the thermal decomposition of ethane varies with conditions. Reactions 1-5 describe the cracking in conventional reactors when the temperature is less than 1000 K. At high temperatures, however, observed product distribution may be explained with a different mechanism. This mechanism is characterized by a fast equilibrium between the ethane molecule and methyl radicals, and by the termination of propagating carriers, hydrogen atoms and ethyl radicals, through the combination of hydrogen atoms and methyl radicals. Ethane from natural gas has been used, in tremendous amounts, to generate ethylene, the most important feedstock in modern industry, through cracking at temperatures over 1100 K. The present mechanism may, hopefully, help give decent insight to this important industrial process. Acknowledgment. The authors gratefully acknowledge the support of the National Science Council of the Republic of China. Assistance of the 1979 class of the chemistry department of National Tsing Hua University on the content of Table I1 is also appreciated. References and Notes (1) F. 0. Rice and K. F. Herzfeld, J . Am. Chem. Soc., 56, 284 (1934). (2) H. G. Davis and K. D. Williamson, World Petr. Congr., Prac., 5fh,

S. K. Tokach and R. D. Koob

(3) (4) (5)

(6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16) (17) (18) (19) (20) (21) (22) (23) (24) (25) (26) (27) (28) (29) (30)

1959, 4 (1960); K. D. Williamson and H. G. Davies, Presented at the 169th National Meeting of the American Chemical Society, Phibdelphh, Pa., 1975, abstract no. 195. G. D. Towell and J. J. Martin, AIChE, J., 7, 693 (1961). A. B. Trenwith, Trans. Faraday Soc., 62, 1538 (1966). M. C. Lin and M. H. Back, Can. J. Chem., 44,505, 2357, 2369 (1966). P. D. Pacey and J. H. Purnell, J. Chem. Soc., Faraday Trans. 7 , 68, 1462 (1972). G. B. Skinner and W. E. Ball, J . Phys. Chem., 64, 1025 (1960). G. I. Kozlov and V. G. Knorre, Combust. Flame, 6, 253 (1962). I. F. Miller and S. W. Churchill, AIChE. J., 8, 201 (1962). J. N. Bradley and M. A. Frend, J . Phys. Chem., 75, 1492 (1971). T. C. Clark, T. P. J. Izod, and G. B. K'stiakowsky, J . Chem. Phys:, 54, 1295 (1971). A. Burcat, G. B. Skinner, R. W. Crossley, and K. Scheller, Int. J . Chem. Kinet., 5, 345 (1973). C. P. Quinn, Proc. R. Soc., London, Ser. A , 275, 190 (1963). D. W. Dexter and A. B. Trenwith, Proc. Chem. Soc., 392 (1964). S.W. Benson and H. E. O'Neil, Natl. Stand. Ref. Data Ser., NaN. Bur. Stand., No 21 (1970). A. Lifshitz and M. Frenklach, J . Phys. Chem., 79, 686 (1975); W.-H. Kao and C.-T. Yeh, !bid., 81, 2304 (1977). A. F. Trotman-Dickenson and E. W. R. Steacie, J . Chem. Phys., 19, 329 (1951); J. A. Kerr and A. F. Trotman-Dickenson, Prq. React. Kinet., 1, 105 (1961); R. H. Snow, J. Phys. Chem., 70, 2780 (1966). D. A. Leathard and J. H. Purnell, Annu. Rev. Phys. Chem., 21, 197 (1970). T. C. Clark and J. E. Dove, Can. J. Chem., 51, 2147 (1973). C. J. Chen, M. H. Back, and R. A. Back, Can. J . Chem., 54, 3175 (1976). J.-T. Chen, Y.4. Lee, and C.-T. Yeh, J. Phys. Chem., 81, 667 (1977). J.-T. Chen and C.-T. Yeh, J . Phys. Chem., 81, 1981 (1977). C.-J. Chen, M. H. Back, and R. A. Back, Can. J. Chem., 53, 3580 (1975). K. J. Laidler, "Chemical Kinetics", McGraw-Hill, New York 1965. R. Hiatt and S. W. Benson, J . Am. Chem. Soc., 94, 6886 (1972). D. G. Hughes and R. M. Marshall, J . Chem. Soc., Faraday Trans. 1 , 71, 413 (1975). D. M. Golden, K. Y. Choo, M. J. Perona, and L. W. Piszkiewica, Int. J . Chem. Kinet., 8, 381 (1976). J. Troe, J . Chem. Kinet., 8, 381 (1976). H. E. Van Den Bergh, Chem. Phys. Lett., 43, 201 (1976). K. G l h e r , M. Quack, and J. Troe, Chem. Phys. Lett., 39,304 (1976).

Photolysis of Tetramethylsilane at 147 nm. Reactivity of (CH&Si and (CH3)2SiCH2 S. K. Tokach and R. D. Koob" Department of Chemistry, North Dakota State University, Fargo, North Dakota 58 105 (Received August 25, 1978; Revised Manuscript Received January 8, 1979)

Results of the photolysis of tetramethylsilane at 147 nm are reported. The observed products, Hz, CH4, C2Hs, (CH3)3SiH,and (CH&Si2, along with accumulated evidence from additive experiments, suggest the primary processes (CH3)4Si CH4+ (CH&3CH2, (CHJ4Si CH3+ H + (CH&3iCH2,and (CH3)*Si CH3+ (CH3)3Si followed by combination, disproportionation, and abstraction reactions of the radical intermediates. The disproportionation/combination ratio for trimethylsilyl radicals is about 0.48. (CH3)$i abstracts hydrogen from a variety of donors approximately 20 times faster than CH3. Both observations were unexpected based on previous work involving (CH3)$i. (CH,)2SiCH2behaves as anticipated based on the previously reported photolysis of 1,l-dimethylsilacyclobutane.

-

Introduction Results of studies of the pyrolysis1,2and radiolysis3 of tetramethylsilane and the photolysis of methyl-4 and dimethyl~ilane~ may be found in the literature. However, the gas or liquid phase photolysis of tetramethylsilane has not been reported. The results of such work are of interest in that a number of intermediates whose reactivity is only partially understood are found as contributors to the total reaction mechanism. In the experiments reported below, 0022-3654/79/2083-0774$01 .OO/O

-

-

special attention is paid to trimethylsilyl free radical and dimethylsilylethylene. Experimental Section Tetramethylsilane (Me,Si) obtained from a commercial supplier (Norell Chemical Co., Inc., 99.5%) was purifed to 99.99% using preparative gas chromatography. DZS, CDBODand CD4 were obtained from Merck Sharpe and Dohme and were checked for isotopic purity by mass 0 1979 American Chemical

Society