Table IV.
Dehydroisomerization
Catalyst % Chromia-aluminaa Molybdena-alumina b Chromia gel0 a 538’ C. 150 lb,/scj, ratio, 3.0. b 4 8 2 O C. 300 lb./sq. ratio, 2.7. c 4 8 2 O C. 50 lb./sq. ratio, 1.0.
Table V.
of Methylcyclopentane
Conversion/Pass Ultimate Yield of Benzene, % 27.0 88.1 53.7 70.0 15.0 94.0 inch gage. 1.0 LHSV. Hz/hydrocarbon mole inch gage.
1.0 LHSV.
Hz/hydrocarbon mole
inch gage.
1.0 LHSV.
Hz/hydrocarbon mole
Hydrogenation of isobutylene Dimer over Chromium Oxide Gel Catalyst
Temp. 390’ C Total pressure 700 lb./sq. inch gage. LHSV, 1.0. Hdhydrocarbon mdle ratio, 2) % Saturation 70 of Charge Catalyst of Cs Product Cracked Chromia gel 99.5 12-15 Chroinia gel NaOHa 9s.5 1.9 a 100 ml. of catalyst soaked in 300 mi. of 2.6% sodium hydroxide solution for 3 hours. (Conditions.
+
Table VI.
Acknowledgment
The data presented in this paper are the results of the work of many individuals in the Research Division of the Phillips Petroleum Co. Special acknowledgment should be made to F. E. Frey, R. IT-. Blue, V. C. F. Holm, W. C. Lanning, J. P. Hogan, R. I,. Banks, and J. R. Owen. Acknowledgment is also made to the Phillips Petroleum Co. for permission to publish these data. Literature Cited
Effect of Alkali on Running Cycle for Butane Dehydrogenation over Chromia Gel
4 t m . pressure. Temp. 500’ C. av.) Running Cycle a t 17% Conversion, Catalyst Hours Chromia gel 12-14 Chromia gel S a O H “ 65-70 100 ml. of catalyst soaked in 300 ml. of 2.5% NaOH solution for 3 hours. (Conditions.
+
a
The effect of alkali in butane dehydrogenation to butenes OVPT chromia gel catalyst is also pronounced. I n this case, the rate of deposition of coke resulting from cracking reactions is important in determining the length of running cycles between regenerations. Table VI shows the effect of alkali in increasing the running cycle at a given per pass conversion. Other alkaline agents which also had an appreciable effect are: potassium hydroxide, sodium metaborate, strontium hydroxide, and sodium aluminate.
Bevan, D. J. &I.,and Anderson, J. S.,Discussions F a i a d u y Soc., NO. 8,238-46 (1950).
Blue, R. W., and Engle, C . J., IND. ENG.CHEM.,43, 494 (1951). Clark, Alfred, unpublished data on iron catalysts. Carter, N. C., and Cromeans, Clark, Alfred, Matuszak, 1LI. P., J. s., I N D . ENG.CHEhf., 45, 803 (1953). Couper, A., and Eley, D. D., Discussions Faraday Soc., No. 8, 172-84 (1950).
Dowden, D . 9 . , J . Chem. SOC.,1 9 5 0 , 2 4 2 4 5 . Dowden, D. A., and Reynolds, P.W., Discussions Faraday Soc., NO.8,184-90 (1950).
pentane to benzene, but rnolybdena or chromia supported on alumina and chromia gel all show appreciable activity, as the iesults in Table I V indicate. Because chromia gel and alumina both contain residual 15 ater, and the nongel unsupported oxides do not, one might presume that an acid-type mechanism is coming into play. On the other hand, the addition of alkali to the active catalysts causes some decrease, but not the sharp drop to negligible activity which usuallj- occurs in true acid-type reactions. Cracking Reactions. Two examples of the inhibition of acidtype catalytic cracking reactions are given using a chromium oxide gel catalyat. The first example given in Table V relates to the hydrogenation of isobutylene dimer. There is a slight reduction in hydrogenation activity of the order of 1.0% attributable to the presence of alkali, but a very marked reduction in cracking activity of the order of 85%. The residual crarking observed in the presence of alkali-treated chromia gel may be accounted for almost entirely by thermal cracking.
Fowle, M .J., Bent, R. D., Ciapetta, F. G., Pit&, P. ll,,and Leum, L. N.,Advances in C h e m Sei-.,S o . 5 , 7 6 (1951). Garner, W.E., Discussions Faraday Soc., KO. 8 , 211 (1950). Greensfelder, B. S., Archibald, R. C., and Fuller, D. L., Chem. E n g . Progr., 4 3 , 5 6 1 (1947).
Griffith, R. H., Chapman, P. R., and Lindars, P. R., Discussions Faraday Soc., S o . 8,258-64 (1950).
Haensel, V., and Donaldson, G. R., 1x0. ENG.CHEX.,43,
2102
(1951).
Holm, V. C. F., and Blue, R. W., Ibid.,4 3 , 5 0 1 (1951). Ibid., 44, 107 (1952). Hubbard, H. D., W ,iLI. Welch Manufacturing Co., “Periodic Table,” revised by W.F. Meggers, 1950. Mills. G. A , , Heinemann, Heinz, Milliken, T. H., and Oblad, A. G. “Houdriforming Reactions,” Division of Petroleum Chemistry, 121st Meeting, A x , CHEM.Soc., Milwaukee, Wis. Pauling, Linus, Proc. R o y . SOC.(London), 196A. 343 (1949). Rideal, E. K., Ibid.,184A, 434 (1945). Steiner, H., Discussions Faraduu SOC.,N o . 8, 264-70 (1950). Tamele, ill. W., Ibid.,No. 8,270-8 (1950). Thomas, C . L., IND. ENG.CHEM.,41,2564-73 (1949). Winter, E. R., Discussions Faradail Soc., No, 8, 231-7 (1950). RECEIYED for review December 8, 1952.
ACCEPTED April 27, 1953.
Hydrogenation of Olefins over Metals J. N. WILSON, J. W. OTVOS, D. P. STEVENSON,
AND
C. D. WAGNER
Shell Developmenf Co., Emeryvilfe, Calif.
T
HE mechanism of thr catalytic hydrogenation of olefins over metals has been the subject of intensive investigation by very able investigators for several decades. This formally simple reaction appears to involve a considerable number of elementary reactions, some of which may not even yet be identified. The recent understanding of the situation is admirably summarized by the Faraday Society ( 4 ) . A detailed understanding of these elementary reactions would be useful not only in the study of hydrogenation, but also in the study of the related reactions of catalytic dehydrogenation and
1480
hydrogenolysis, in which many of the same or related elementary reactions doubtless participate. I t is interesting to consider how the present incomplete knowledge of these elementary rractions ha3 been gained. -4surprisingly small part has come from the study of reaction kinetics alone, though this part is indeed an indispensable one. Part has come from the study of the phyaics and chemistry of adsorption, and part from the use of deuterium as a tracer. The early applications of deuterium (3, 6, 6, I S ) afforded considerable insight into the mechanism but were limited by the
INDUSTRIAL AND ENGINEERING
CHEMISTRY
Vol. 45, No. ‘2
L
In a study of the mechanism of catalytic hydrogenation olefins were treated over a commercial nickel catalyst with a large excess of pure deuterium and the isotopic composition of the products was determined by mass spectrometry. The product of the reaction between cis-2-butene C. contained all isotopic species and deuterium at -78' from C4H10 to C4D10 in an essentially random distribution of the deuterium atoms among the hydrogen positions, even though the average composition of the product was close This suggests that at low temperatures both to C4H&. alkane formation and the redistribution of hydrogen atoms among chemisorbed hydrocarbon species occur largely by elementary reactions in which a hydrogen atom is transferred from one chemisorbed hydrocarbon fragment to
another, rather than by reactions involving chemisorbed hydrogen atoms directly. At higher temperatures exchange reactions with chemisorbed hydrqgen become more prominent. On the basis of these observations and the results of other investigators, a mechanism is proposed for the catalytic hydrogenation of olefins over metals, which provides for competition among a number of elementary reactions. The detailed course of reaction under a particular set of conditions will depend on the effects of temperature, pressure, ratio of olefin to hydrogen, and catalyst on this competition. Catalytic hydrogenation is of limited utility for the preparation of specific isomers of deuterium-la beled paraffins, even when only the stoichiometric quantity of deuterium reacts with an olefin,
nature of the available analytical tools, which permitted only a determination of the average concentration of deuterium in the products. The results showed the rather general concurrence of an exchange reaction with the addition reaction, but the detailed nature of the processes involved has remained controversial. Additional information has been obtained by the use of infrared spectrometry for analysis of the products of the reaction ( l a ) ,but even this technique is limited in its ability to detect minor products and is applicable only with extreme difficulty t o the quantitative analysis of mixtures of polydeuterated hydrocarbons heavier than ethane. The use of the analytical mass spectrometer in combination with essentially pure deuterium provides information on the distribution of deuterium among the product molecules. This combination is an extremely powerful tool for the elucidation of hydrocarbon reaction mechanisms. This technique was first applied to the study of the catalytic hydrogenation of olefins by Turkevich, Bonner, Schissler, and Irsa (11). The present authors have employed a similar procedure independently in a study of the hydrogenation of cis-2butene, isobutylene, and ethylene with a tenfold excess of deuterium over nickel (14). The butylenes were studied initially because the distribution of isotopic species among the deuterobutanes can be determined with comparative ease by mass spectrometry with ionizing electrons of low energy.
used to measure gas ressures. On evacuation of the system, a U-tube between it a n f t h e system was cooled to 19Goto prevent access of mercury vapor to the catalyst.
Materials
.
The cis-2-butene, isobutylene, and ethylene used were Phillips research grade hydrocarbons. Deuterium gas was obtained from the Stuart Oxygen Co., under allocation from the Atomic Energy Commission, and contained 0.45% H D and