[COXTRIBUTION FROM THE
DEPARTMENT O F CHEMISTRY,
~ t I R T I i \ \ ' B S T E R N LYSIVERSITY]
Racemization, Isomerization and Isotopic Exchange of ( +)3-Methylhexane on a Silica-Alumina Catalyst BY ROBERT L. BURWELL, JR., HOWARD .I.PORTE AKD LYILILIY 1 3 . HAMILTOX RECEIVED XUGCST 15, 1958 This paper deals mainly with isotopic exchange with heavy water, with racemization and with isornerizatioii of ( +)3methylliesane on silica-alumina catalysts a t - 4 to 2-IOo. At 200 a n d 240", molecules issue from tlie reaction devoid of rotaticin and in isomeric equilibrium with respect t o 2- a n d 3-methylhexane, 2,3- antl 2,.i-diniethylpetitane antl 3-etliylpcnt a m . A11 molecules which have reacted are extensively exchanged a n d all hydrogen atoms save o n e tertiary arc availablc for cxcli:inge. As temperatures are lowered, dimethylpentanes no longer appear and t h e quantities of 3-methylhcxane fall progressively below t h a t corresponding t o equilibrium with product ( i)3-methylhexane. A i t --I ant1 .5G0, racemization greatly exceeds isomerization and also isotopic exchange. A t 120'. migration of ethyl or propyl groups seems as important as t h a t of methyl. At low temperatures two different types of sites are involved and crevice sites seein indicated. Itleutity of rates in the presence a n d absence of hydrogen indicate t h a t chain initiation dors not involve equilibrium foriliation of hytlrogen.
Evidence for the carbonium ion theory of reactions of hydrocarbons on cracking catalyst^^,^ is primarily analogical. Cracking catalysts are strong acids. A number of reactions which occur on cracking catalysts closely resemble those promoted by strong acids such as aluminum chloride or sulfuric acid : alkylation, skeletal isomerization, polymerization, isotopic exchange, etc. In its totality, this evidence for carbonium ion processes is strong, but many features require further elucidation. A study of the behavior of an optically active hydrocarbon on a silica-alumina catalyst offers some prospect of contributing to such elucidation. The reactions of optically active 3-methylhexane in the presence of sulfuric acid, halosulfonic acids5 and aluminum bromide6 have been studied. These reactions, which are clearly carbonium ion in nature present a model with which the reactions on solid acidic catalysts could be compared. The research reported in the present paper followed that with sulfuric and halosulfonic acids as closely as possible. It involves primarily a study of the simultaneous isomerization, isotopic excha,nge and racemization reactions on a silica-alumina catalyst a t temperatures of - 4 to 240".
Experimental Materials.-The silica-alumina catalyst, Houdry Type S-4.5 cracking catalyst, 12..iyi alumina, 87.5y0 silica, was supplied by l l r . .4. G . Oblad. It was crushed and sieved t o 20-40 mesh. The preparation o f inactive m e t h y l h e x a n e ~and ~ ~ of (+)3niethylhex::inc7 !ias been described. Phillips Petroleum Company Pure Gradcs of 2,1-(Iiinethylpentane, 2,3-dimethylhutaiie a r i d heptane xntl the Research Grade of methylcycl(iperit:ine were emplo>-ed. XI1 alkanes were percolated thr(iug11silica gel which h n d been heated in nitrogen t o 350" before usc. All alkanes :igreed wcll with standard samples ( i f the N:Ltion:tl Bureau ( i f Stand:irds in differential infrared [I) Presented in p a r t a t t h e 128th hIeeting of t h e American Chemical Society, RIinneapolis, September 1 4 , 10BS. ( 2 ) C. L,T h o m a s , Z x d . E?zn Chem , 41, 250-1 (1949). (:1) B. S. Greensfelder, H . H. Voge and G h i . Good, ibid.,41, 2573 (1949). ( I ) ( a ) 11. L. B u r w e l l . J r . , R.B. Scott, I, G. h i a u r y a n d A. S. Hus5ey. T H I SJ O U R N A L , 76, 5822 (10.54); ( b ) G. S. Gordon, 111, and I
+
C C
C
C equilibrium mixture OF five heptanes
Since nearly every molecule must have traversed a necessarily optically inactive species such as the carbonium ion from 2-methylhexane, one must ex(17) D. P. Stevenson, C. D. Wagner, 0. Aeeck and J . W. Otvos, THISJOURNAL, 74, 3269 (1952).
pect that the 3-methylhexane re-formed would be racemic. With sulfuric acid-dz, only those hydrogen atoms adjacent to a tertiary position can exchange." However, a t 200" on silica-alumina, the extensive isomerization should expose al! hydrogen atoms in a carbonium ion to the possibility of exchange as is indeed, observed. Exact isotopic exchange patterns would depend in a rather complicated fashion upon relative rate constants in the isomerization processes, rate constants in exchange and hydride ion transfer and upon the rate of exchange between heavy water and those deuterium atoms on the surface which are involved in isotopic exchange. Thus, a t 200", nearly every molecule which reacts would be both optically inactive and isotopically exchanged. As shown in Table 111, this is true to within the available precision. Isomerization and Loss of Rotation at Temperatures below ZOO".-As shown in Table 11,isomerization declines relative to loss in rotation a t teniperatures below 200". With declining temperatures, the dimethylpentanes fall away first, then 2methylhexane. By 56", one deals with a nearly pure racemization. The products formed and the decline of isomerization with temperature resemble the results of the reaction between (+).?-methylhexane and sulfuric acid-dz and chlorosulfonic acidd.4,5 Table V presents k a / k i as a function of temperature for both systems. k R and ki are the rate constants for loss of rotation and isomerization, thus5 k a / k i = ln(1
-
A @ / a o ) / h ( l - X,/X,'q)
where A,8 is the loss in rotation, cyo is the original rotation, Xz is the mole fraction of 2-methylhexane and X 2 e q is its value in an equilibrium mixture with 3-methylhexane. The quantity of water employed with the silica-alumina catalyst a t -4 and 56" is much less than a t the higher temperatures. One could hardly cover such a large range of temperature with a catalyst of constant activity. While the difference in water content might contribute to the decline in the relative rate of isomerization, the trend with temperature is already clearly visible in the results a t higher temperatures. \'ARIATION
Temp., OC.
Catalyst
OF
TABLE V k n j k i WITH T E M I - ~ R A T U R E kajk,
Temp., 'C.
Catalyst
k ~ / k
96% &SO, 1 240 Si02*X120:I 1 1 100% H2SOa 1 . 4 200 H C ~ S O I - H ~ S O ~3 . 1 125 1 .o 15 21 56" HClSO3 12 51 .56b -78 HClSOs -4" 15 Run 22D. Run 33D. Run 35D. 60 30 0 -33
Results at 120".-Except for the decline in isomerization relative to loss in rotation, results a t 120" are very similar to those a t 200-240°, and, in particular: (1) degrees of exchange and loss in rotation are equal; ( 2 ) depths of exchange are large and nearly as large as a t higher temperatures (compare run 32C, Table IV, with run 27C, Fig. 1). As shown in comment 4 of Results a t 200-240°, a t 200" these results are a necessary consequence of
equilibrium migration of methyl groups in the 2and 3-methylhexane system. The equilibrium nature of the products a t 200-240’ is compatible with a number of possible variations in the basic inechanism and the extensive methyl group migration may well obscure the consequences of other processes. The decline in relative degree of methyl group migration a t 120’ (Table 11, run 2lA), in fact, exposes another process since much 3-methylhexane must be re-formed without traversing the 2-methylhexane structure. Complete racemization of product 3-methylhexane cannot entirely result from all reacting species necessarily traversing the optically inactive 2-methylhexane structure, nor can one invoke mere isomerization to 4-methylhexane since this could hardly racemize all 3-methylhexane without leading to equilibrium formation of 2-methylhexane. The observed degree of methyl group migration is also too little to account for the observed depth of exchange if one permits exchange of only those hydrsgen atoms contiguous to the tertiary position. About 60% of the molecules which exchanged in run 32C appear as (*)3-methylhexane yet only 17yo of the total molecules have seven or fewer deuterium atoms. Isomerization to 4-methylhexane would expose two more hydrogen atoms but nearly half of the (.t)3-methylhexane is more exchanged than this. A similar situation seems to exist in run 30C a t 140” (Table IV). This could support the belief of Hindin, Mills and Oblad12 “that all primary and secondary atoms of the structure, however f a r removed f r o m the chain branch, can exchange.” Their arguments for this conclusion, in essence similar to those above, consider migrations of methyl groups only. However, in sulfuric acid, 3-ethylpentane isomerizes unusually rapidly17 and, here, an ethyl group must migrate. Thus, migrations of ethyl and propyl groups could be important and both would lead only to 3-mefhylhexane with a rearranged carbon skeleton. For a propyl group migration with exchange of contiguous hydrogen atoms only + C H ~ C D ~ C C D Z C H Z+ CH~ CDa CD8;
