ISOMERIZATION OF ETHYLTOLUENE
Sept. 20, 1960
4853
ORGANIC A N D BIOLOGICAL CHEMISTRY [COFTRIBUTION FROM
THE
POLYMER RESEARCHLABORATORY, MIDLANDDIVISIO?;,THE Dom CHEMICAL Co , MIDLASD, MICH.]
Kinetics of Three-compound Equilibrations. 111. The Isomerization of Ethyltoluene BY ROBERTH. ALLEN,LARRYD. PATS AND DUNCAN S. ERLEY RECEIVED DECEMBER 2, 1959 The isomerization of ethyltoluene was studied as a three-compound equilibration involving six rate constants Dilute solutions of five ethyltoluene isomer distributions in toluene were isomerized at room temperature with aluminum chloride. In each case samples were taken periodically and analyzed by differential infrared spectroscopy and vapor phase chromatography. The isomer distributions obtained are consistent with the following relative rate constants and equilibrium percentages: k,, = 1.0, k,, = 2.9, k,, = 5.3, k,, = 38.2, k,, = 60.8, k,, = 24 G, O* = 9.0, li* = 64.8, P* = 26.2. This work shows t h a t the mechanism of isomerization of ethyltoluenes under these conditions is not exclusively an intramolecular l,Z-shift, as was the case with xylenes. From the magnitude of k,, and k,, it appears likely t h a t most of the isomerization proceeds by an intramolecular 1,2-shift.
Introduction It has been shown' that for certain three-compound equilibrations involving six rate constants, as oitho
I----
kopJ.
rkpn
1 c1
Lo1 i k o m
k Pm
para
r+nzeta k w
the integrated rate expressions are O ( t ) = O* P(t) = P*
+ Ae-nr + Be-Br + Ce-ar + De-87
where U ( t ) is the concentration of the ortho isomer a t time t, O* is the equilibrium concentration of the ortho isomer, A and B are determined from the starting concentrations, a and @ from the rate constant set, and r is some function of time. The use of this scheme for the study of isomerization of cymenes' and xylenes2 has been reported. The use of this scheme for the study of the isomerization of ethyltoluenes is now being reported. Much literature has appeared on the ethylation of aromatics3 and the disproportionation of ethylaromatic^,^ but very little has been reported on the isomerization of ethylaromatics. The fact that oethyltoluene isomerizes is an important feature of the Dow vinyltoluene p r o c e ~ s . ~ The thermodynamic equilibrium of ethyltoluene isomers in the vapor phase has been calculated to contain 8% 0-,48% m- and 44% P-ethyltoluene.6 The isomer distributions obtained by the continuous ethylation of toluene with various catalyst have been re(1) R . H. Allen, T. Alfrey, Jr., and L. D. Yats, THISJ O U R N A L , 61, 42 (1959). (2) R . H. Allen and L. D. Y a t s , ibid., 81, 5289 (1950). (3) E. M. Hodnett and C. F. Feldman, ibid., 81, 1638 (1959); C. C. Lee, M . C. Hamblin and h-. James, Co?:. J Chem.. 36, 1507 (19.58) ; C . R. Smoot and H. C. Brown, THISJOURNAL, 78, 6245. 6249 (1056). (4) R. M Roberts. G. A. Ropp and 0. K. Seville, i b i d . , 7 7 , 1764 (1955); A I . C. Hoff, ibid., 80, 6046 (19.3); D. A hIcCaulay and A. P. Lien, i b i d , 75, 2411 (1953); H. C. Brown and C. R. Smoot, ibid, 78, 2176 (1956); W. J. Roberts and b1. B. Price, First Regional Meeting, Delaware Valley Section, A.C.S., Feb., 1956. ( 5 ) R €1. Roundy and R. F. Boyer, "Styrene," A.C.S. Monozraph N o . 115, Reinhold Publishing Corp., New York, PZ. Y., 1952, p. 1232; J. L. Amos and K. E. Coulter, U. S. Patent 2,763,702 (Sept. 18, 1956). (6) W. J. Taylor, D . D. Wagman, M . G . Williams, K. S. Pitzer, and F. D. Rossini, J . Res. Nal2. Bur. Slandavds, 87, 95 (1940).
ported.' These distributions probably differ only in the extent of isomerization that accompanies the alkylation, since the distributions appear to be consistent with the results of the present work. One of the distributions, 11% 0-,64% m- and 25% e-, is quite close to the distribution taken to be the equilibrium distribution in the present work. One experiment has been reported on the isomerization of ethyltoluene per S E . Brown has shown that the relative rates of Friedel-Crafts isomerization of alkyltoluenes decrease in the order isopropyltoluene > ethyltoluene > xylene.*
Experimental Materials.-Redistilled commercial grade toluene was used. The o-ethyltoluene was obtained from the Dow vinyltoluene plant. Comparison with an A.P.I. sample by infrared analysis showed no significant difference. Clemmensen reduction of recrystallizedQEastman Kodak Co. p-methylacetophenune produced p-ethyltoluene, which when treated with concentrated sulfuric acid, washed and distilled showed no sigiiificaut iiripurities. T h e preparation of m-ethyltoluene was patterned after the process of Schlatter for separation of m- and p-xylenes.10 A miiture of m- and p-ethyltoluenes obtained from the Dow vinyltoluene plant was exhaustively t-butylated at 35" with isobutylene using BFa-HzO as the catalyst. The 3ethyl-5-t-but>-ltoluerie obtained by distillation of the alkylate was then cracked a t 300' over a Houdry S 6 5 silica-alumina catalyst. Distillation of this product yielded nz-ethyltoluene that was analyzed by infrared to contain 1yo0- and 1yop-ethyltoluenes. Baker and .%dams reagent anhydrous aluminum chloride was used with Matheson anhydrous hydrogen chloride. Isomerization Procedure .-To a one-liter 3-necked creased flask, equipped with a gas inlet tube, thermometer, stirrer and a water trap connected to the flask through a calcium chloride drying tube, was added 450 g . (4.9 moles) of toluene and 5.4 g. (0.041 mole) of aluniinum chloride. Through the stirred reactants at room temperature was passed anhydrous hydrogen chloride for 5 minutes. T o this solution a t room temperature was added 50.0 g. (0.41 mole) of ethyltoluene. At specific times, 25-ml. p o r t i o ~ ~ofs the reaction mixture were withdrawn, quenched, washed and dried. The p-ethyltoluene isomerization was run on a 3 / r scale in a oneliter flask. The 8553 p-, 1570 o-ethyltoluene isomerization was run on a a i l 0 scale in a 500-nil. flask. Analytical Procedure.-The samples, untreated and undiluted, were scanned on a double beam infrared spectroni(7) W. M. K u t z , J. E. Iiickels, J. J. McGovern and 13. B. Cnrpon, J . Org. Chrm., 16, 699 (1951). (8) H . C. Brown and H. Jungk, THISJ O U R N A L , 7 7 , 5579 [1935). (9) W. hl. Schubert. J. Robins and J. L. Haun, i b i d . , 79, 910 (1957). (10) M . J. Schlatter. U. S. Patent 2,734,930 (Feb. 14, 1956).
sorb W. T h e instrument12 was connected through i t Brown Recorder t o a n Instroil two counter automatic integrator. One peak for m- and p-ethyltoluenes arid one peak for o-ethyltulueiie was obtained. T h e infrared aiialyses indicate a i equilibriuIIi value of i . 5 7 0 - , while t lie vapor phase chromatograph>- m d y s e s indicate : i i i rquilibriuni value o f 9 .O(i'; o-ethb-ltolueiie. Selective cornplesiiig of ethyltoluene b ~ -the cirtalyst evidentally did not significantly alter t h e equilibriuiii coiiiIxlsition of isomers, since the analysis for the last sample was identiha1 t o the analysis for t h e residue in each isomeriz;itioii.
I:ig. l.---EthyltidUene isonier t1istril)utions obtained ti>AlCl:>isomerization: 0, iiifrared d a t a ; 0 , x7.p.c. data.
PARA
MET4
1:ig. 2.-~~llth\.ltolueiieisomer tiistributir)ns ohtainetl b y -41ClJ iminerizaticiti: 0, itifrared d a t a ; 0 , i-p c . c1at;t.
eterll mith tlie toluene :ibsorptioiis of the saiiiple iiiutched o u t . The instruinent autijinatic gain featurell was used to compensate for the low iiptical energy in regions of high tiiluene absorption. Stantlard absorbances for each isomer a t 12.24, 12.78 and 13% p were determiiieti using 3 straight line between the absorption miniiiia at 11.8 and 14.1 p for the base line. Three absorbances a t the analytical wave lengths were determiiietl f x each sample a n d introduced into a set o f three simultaneous linear equations itivol\-ing the uine standard :ibsiirbatices. T h e equations were then solved for ccincetitrations using a Royal McBec LGP:113 cmnputer. This c;ilculation assumes :istraight line rclation hetweeu absrirbmce ant1 ciiiice~itr:itioii. T I Icheck tlie accur~tc!-( i f tlie i i i c t h r ~ l, t syiitlietic inistw-e \Y:LS :ciinlyzctl with tile result Tc>lliene, ,'
IC11 0 \",I
Foilrid
) J / - l < t l l 1>
l i i iI .i!l 1 )
!I 3 .TU, il
/5-1:t11!.i
:;I
:i1 7
I!
~
~~~~
1111hI a n d S .n ' r i r h t , J . Or,/ .Ycic.
k,,,, = 1 i . 3 k,,,= 7.0
0
=
k,,, = 7.2 k,,, = 1 0
il (1
=
64 8
N
=
1 0
8 =
3 . 19 P" = 2 6 . 2
This set generated the curved lines on Fig. 1. The correlation with the experimental isoilier distributions is rather poor, which suggests that the isomerization of ethyltoluenes under the conditions used here does not proceed exclusively by an intramolecular 1,?-shift. By varying both parameters the best relative rate constant set found is k , , , = 1. O kT2,,= 2 . 9 0" = 9 . 0
k,,,,, = 5 3 k , , : &= 38 2 11- = -'16 2
k,,,, = t ; O . M k,, = 24 G If" = G-l 8
This set produces the following integrated equatiolis for the percentage of ovth,o and puvn isomers i11 terms of the parameter 7 , where
.The m- wid p-ethyltoluene determinations are withiii the estimated error of i 3 c b of the amount present. Since the broclti vi-ethyltciluene absorbtiiin overlaps the o-ethyltoluene ahsorbtion, several samples low iii o-isomer were se chromatography. e carried out at 113" usiiig :L 10aig polyester adipate o n Chri i i i i o 111) I.. 'A-. Hersclicr, H.
Results and Discussion The experimental isomer compositions are presented in Table I and plotted as the points on Figs. 1 and 2. The determination of a relative rate constant set for a three-compound equilibration must provide five rate constants, since one will be set equal to 1.0. The three equilibrium constants involved provide three parameters, so that only two further parameters are required. In the cymene case,' the fact that the ortho trajectory was a straight line indicated that kpo = k,,,, so that only one parameter was varied in order to fit the data to the theoretical equations. In the xylene case,? it was evident that kflo = kOi, = 0, so that only one parameter was varied. In the ethyltoluene case presented here, unfortunately, it was necessary to vary two parameters, so that the rate constants are more arbitrary than those for the previous two cases. Haag and Pines have pointed out that for a n xppropri ate three-compound equilibration, tangents a t the 100% point for each of the three isomerization composition trajectories must intersect a t a common point.13 *Whough this observation is helpful in obtaining a rough approximation of the rate constants, it didn't provide a sufficiently accurate additional parameter. The best relative rate constant set found based on the assumption that kbo = knp = 0 is
iliii
.
T =
f(t)df
some unknown function of time. il) R . S . Gohike, A n d ( : i f ? ? . , 23, 1723 (1957). (13) \T. 0 . Haag and H. Pines, Isomerization and I
3
composition. mole %--
-Isomer
Time, min
no
0 1
0
1
27.8 45 4 53.5 80.5 64.8 67. 1 66 9 1 4 15 .5 38.2 54.1 60.3 64 7 65.8 66.9 0 3 13 fi 22 9 33 2 40 1 47 2 -53.1 59 6 63.6 65 8 63 2 0.0 4 1
10 9 I -5 1 $1 17.1 15 27.6 15 $91) 8 0 36 1 12 I1 1x0 13.8 31 X 17 1 360 21 ti :I, 2 720 24 4 61 5 1 MI1 65.1 27 R WOO 26 7 6fi.3 Res > 0 1 1 97.8 1 ethyl > methyl, correlates inversely with basicity, the basicity differences alone are probably not large enough to account for the rate deviations. It seems more likely that the rate deviations are due to alkyl hydrogen transfer and cracking reactions.l5 If the above values for r are introduced into the equations for three of the four remaining trajectories, the curved lines shown on Fig. 3 are obtained. The points repre2ent the experimental data. The agreement indicates that the relation of r to time is fairly consistent from one reaction mixture (14) W. von E. Doering, h l . Saunders, H. G Boyton, H. W. Earhart, E. F. Wadley, W. R. Edwards and G. Laber, Tetrahedron, 4, 178 (1958). (15) We wish t o thank Dr. D. A. XcCaulay for this explanation.
[CONTRIBUTION FROM
THE POLYMER
to another. That is, although the reactions are not pseudo first order, their deviation from a pseudo-first-order rate are somewhat reproducible. If the above treatment is applied to the equations from the rate constant set in which kpo = kO* = 0, Fig. 4 is obtained. The 100% p-trajectory was not included in either figure because its deviation from the experimental data obscured the figures without adding any useful information. If any conclusion may be drawn from the deviation of the 100% o-trajectory in Fig. 4, the most reasonable conclusion would be that the ratio of ,B to a: is more accurate from the rate constants set for Fig. 3 than the rate constant set for Fig. 4. Thus, the absolute rate data tends to support the thesis that the isomerization of ethyltoluenes is not exclusively an intramolecular 1,2-shift.
RESEARCH LABORATORY, THEDOW CHEMICAL CO., MIDLAND,M I C H . ]
IV. The Isomerization of
Kinetics of Three-compound Equilibrations. Alkylaromatics BY ROBERTH. ALLEN
RECEIVEDDECEMBER 2, 1959 The isomerization of xylene in toluene solution by the action of AlCla.HC1 has been shown t o proceed by an intramolecular 1&shift. The present work provides evidence t h a t under the above conditions, the isomerization of t-butyltoluenes proceeds by alkylation of the toluene, and the isomerization of ethyltoluenes and isopropyltoluenes proceed by both mechanisms.
-
Introduction Data presented in this paper indicate that the Friedel-Crafts positional isomerization of alkyltoluenes in toluene proceeds by either an intramolecular 1,2-shift mechanism, an alkylationdealkylation mechanism, or both. The alkyl shift mechanism has been offered for the isomerization of xylenes.’ The alkyl transfer mechanism is here offered for the isomerization of t-butyltoluenes. and a combination of both mechanisms for ethq’ltoluenes ane isopropyltoluenes. The intramolecular 1,2-shift mechanism may be pictured as
7 -
CH3
1’
i y LH
..RJ
+
CH3
A
d
-
plex displaces the first aromatic ring. Loss of proton from the second aromatic ring yields the product. This mechanism is quite similar to that proposed for the disproportionation of alkylaro,‘-‘. matics. Isomerization of t-Butylto1uene.-The postulate that t-butylaromatics isomerize by alkylationH R R dealkylation is supported by the ease with which +, I f t-butyl groups split off of molecules under acidic CH8 -: CH3 conditions; for example, from t-butylbenzeneZc I and methyl pentamethylethyl k e t ~ n e . ~ More c--, , direct evidence for the present argument has been H ‘.’ RS R 2 reported by S ~ h l a t t e r . ~ The hydrogen fluorideI1 rIr catalyzed reaction of 3,5-di-t-butylethylbenzene The alkyl group of the aromatic u-complex bridges with toluene produced m-t-butylethylbenlene t.hat two adjacent aromatic carbons and then swings was better than 90% pure. Thus t-butyl groups over to its final position. Loss of proton yields (2) (a) K. L. Nelson and H. C. Brown in “The Chemistry of Pethe product. The intermolecular mechanism may troleum Hydrocarbons,” Vol. 111, Reinhold Publishing Corp., New be pictured as is shown in the next diagram. Y o r k , N. Y.,1955,p. 528; (b) 8. P. Lien and D. A. McCaulay, THIS Nucleophilic attack by a second aromatic on the JOURNAL, 75, 2407 (1953); (c) D.A. hIcCaulay and A. P. Lien, i b i d . . 5 , 2411 (1953); (d) 11. E. Kinney and L. A. Hamilton, ibid., 76,786 electronically deficient alkyl group of the u-corn- 7(1954).
5
a.
‘Q
(1) (a) G. Baddeley, G . Holt and D. Voss, J . C h r m Soc., 100 (1952); b) D. A. McCaulay and A. P. Lien, THISJ O U R N A L , 74, 6246 (1952); ( c ) H.C. Brown and H. Jungk, i b i d . , 77, 5579 (1958). f
( 3 ) H. D. Zook, W. E. Smith and J. I.. Greene. i b i d . , 79, 4436
(1957). (4)
M.J. Schlatter, U.S. Patent 2,768,985(Oct. 30, 195C).
