HERMAN PINESAND
4953
lized from benzene, it formed almost black needles with bluish lustre (1n.p. 2113). The red benzene solution slruwed a green fluorescence. rlnal. Calcd. for C18H15S3: C, 79.09; H, 5.53; S,15.38. Found: C, 79.20; H, 5.28; iY,15.33. 2 . Condensation with p-Diethylaminobenzaldehyde .The dinitrile (1.42 g., 0.01 mole) was heated with a n excess of aldehyde and 1 5 drops of piperidine for 4 hr. a t 1%'. After preliminary purilrlcatioii with etlianol and one crystallization from benzene-etlianol, 1. l l g. (37yq) of 4-(p-diethylaminostyryl~-isophthalonitrile was obtained. Further purification over alumina and recrystallization from benzene-ethanol brought it t o the form of orange prisms melting at 190".
Anal. Calcd. for CmH19N3: C, 79.70; H, 6.35; Found: C, 79.80; H, 6.33; X, 14.12.
[CONTRIBUTION FROM
N,13.94.
JOSEPH
Vol. 79
T. ARRIGO
VI. 2,4-Bis-(benzenesulfonyl)-toluene. Condensation sulfone (3.7 with p-Dimethylaminobenzaldehyde .-The g., 0.01 mole) was heated 4 hr. a t 145' ~ ~ i t 2l i g. of aldehyde and 10 drops of piperidine. Even a t this temperature a solid crystalline mass formed, which after a washing with ethanol and crystallization from dioxane formed ;L brick-red powder (2.4 g., 48y0). Recrystallized twice from pyridine, i t formed wine-red prisms of 2,4-bis-(heiizenesulfonyl)-4'-dimethylaminostilbene (m.p. 265'). Anal. Calcd. for C Z B H ~ ~ O ~ NC, S Z66.77; : H, 5.00; S, 12.73. Found: C,66.68; H, 5.30; S,12.52. Absorption spectra were determined on a Beckn~arimodel DU spectrophotometer with 10-mm. quartz crlls. A11 solutions were made up of 1 nig. of sillutc iii 100 nil. of methanol. FRIBOURG, SWITZERLAND COLLEGEVILLE, INDIANA
HIGHPRESSURE AUD CATALYTIC LABORATORY, DEPARTMEYT O F CHEMISTRY, NORTHWESTERN UNIVERSITY ]
THE IPATIEFF
High Pressure Thermal Alkylation of Monoalkylbenzenes by Simple Olefins' BY HERMAN PINESAND
JOSEPH
T. ARRIGO~
RECEIVEDMARCH14, 1957 The thermal reactions of monoalkylbenzenes with simple olefins were studied in a flow-type system a t 420 atm. pressure and in the temperature range of 400-485'. With respect to the formation of 1: 1 adducts with the arenes, it was observed t h a t t h e olefin reactivity decreased in the order propylene > isobutylene > 2-methyl-2-butene. Arene reactivity decreased in the sequence toluene > ethylbenzene > cumene. Selectivity in the addition of aralkyl radicals to unsymmetrical olefins was found t o parallel the order of stability of free radicals, I11 > I1 > I. Intermediate radicals were found t o undergo cyclization or 1,2-pheny1migration in certain cases. Free radical mechanisms for the various reactions are discussed.
The side-chain alkylation of alkylbenzenes by simple olefins a t elevated temperatures and presTABLE I TOLUESE-PROPYLENE EXPERIMENTS Experiment Temperature, 'C. Reactants, moles Toluene Propylene
1 409 4.31 0.85
Results: Propylene reacted, mole yo 34 Yields based on propylene reacted, mole % Propylene dimers 24 Propylene trimers 10 Monoadducts 28 Toluene reacted, mole 70 3.4 Yields based on toluene reacted, mole %" Benzene Ethylbenzene Cumene n-Propylbenzene Cymenes( 0 - , m-, and
P-1 Isobutylbenzene n-Butylbenzene Diadduct5 Diar, lalkanes Unsatn. of monoadducts, mole 5%
2 430 4.31 0.85 51
21 11 34 6.4
3 456 4.14 0.82 54
19 7.4 44
10.4
4 475 4.11 0.85 42
19 3.8 37 10.4
5 485 2.46 0.50
35
16 4.1 13 13.7
....
....
0.8 8 0.3 1.8
....
....
....
11 47 36
9.4 47.6 27.2 12.5
9.8 40.3 12 27
1 9.3 24 8 38.4
5 9c
5.3
4.6
6.8
14
0.7 5
0 3
.... 2
0.2 3 0.1
1.3 13.2 0.6
4.2
4
22.5 0.5 3 0.8 2.0 4.6
3.6
Arenes boiling above t h e reactant and through the monoadduct range were analyzed after selective hydrogenation (see Experimental). Ir 1: 2 arene-olefin products. There was also isolated 0.04 g. of anthracene from expt. 5 . ______ a
(1) Taken i n part from a dissertation submitted b y J. T. Arrigo t o the Graduate School in partial fulfillment of the requirements for the I'h 3 d,.rrer. October. 1956. ( 2 : Cnii %.rralOil Frodiicts Co. Predoctoral Fellow, 1953-1956.
TABLE I1 TOLUESE-ISGBUTYLEXE EXPERIMENTS S I Experiment 6 456 Temperature, OC. 411 432
-
Reactants, moles Toluene Isohutylene Results Isobutl-lene reacted, mole
4.40
4.88
4 72
(1.92
1.0.7
1 00
70
28
41
S1
Yields based on isobutylene reacted, mole % Isobutylene dimers Isobutylene trimers Monoadducts Toluene reacted, mole 70
40 10
48
47
9.7 31 3 6
24 7 5
31 i'T,
Yields based on toluene reacted, mole yo Benzene 1.8 E thylbenzene 3 Unident. butylbenzene ( ? ) ,... Neopentylbenzene 1.6 Isopentylbenzene 71.8 Diadduct 19.8 Diarylalkanes .... Vnsatn. of monoadducts, mole 70 5 . 5
0 4 1
,...
1.6 56.3 9.4 27.3 5.7
73
1 1 81 3 7 1 7 35 4 11 39 10
sures has been reported r e ~ e n t l y . ~Results of the reaction of toluene with propylene indicated that there was about a sevenfold predominance of n- to isobutylbenzene formed. I n view of these data, the purpose of the present work was to make a (3) V . N. IDatieff. H. Pines and E.Kvetinskas, U. S. Patent 2,755-, 140 (1956).
49.59
THERMAL ALKYLATION OF MONOALKYLBENZENES
Sept. 20, 14.57 TABLE I11 Experiment Temperature, "C. Reactants, moles Toluene 2-Methyl-2-butene
5.40 1.08
Results 2-Methyl-%butene reacted, mole % Ratio of 2-methyl-2-butene/2methyl-1-butene recovd. 3-Methyl-1-butene
....
Yields based on 2-methyl-2-butene reacted, mole % Methylbutene dimers Monoadducts Toluene reacted, mole %
66 18 0.25
Yieldsb based on toluene reacted, mole % Benzene Unident. pentylbenzene ( ? ) 2,3-Dimethyl-l-phenylbutane 3-Methyl-1-phenylpentane Diadduct Diar ylal kanes
systematic study of the direction and extent of addition of representative monoalkylbenzenes to utisymmetrical olefins a t various temperatures. Toluene, ethylbenzene and cumene were chosen to illustrate a phenyl-activated primary, secondary and tertiary radical, respectively. The competi4.92 tion between primary vs. secondary, primary V S . 0.99 tertiary and secondary vs. tertiary intermediate monoadduct (1:1 arene-olefin adduct) radical formation was represented by propylene, isobutylene and 2-methyl-Z-butene, respectively. 24 The reactions were carried out in a flow-type system having a copper-lined reaction tube with a 2.0 zone of copper punchings as contacting medium for D the arene-olefin mixtures. Extensive studies have shown that copper has no catalytic effect upon the thermal reactions of hydrocarbons (ref. 3 and unpublished results). The conditions and results 71 are summarized in Tables I-VIJ and the skeletal 21 1 . 9 4 structures of the monoadducts isolated (after selective hydrogenation) from experiments 1-19 are given in Table VIII.
EXPERIMENTS 9 11 10 456 403 432
TOLUENE-2-METHYL-2-BUTEXE
4.1
4.3
7
5.37 1.08
12
2.3
71 20 0.67
13 36 25 19
6 13 38 34 9
3 7 25 28 6
....
....
31
Unsatn. of monoadducts, mole % 38 37 34 a Only a negligible and small amount of 3-methyl-l-b~The tene was detected in expts. 10 and 11, respectively. presence of 2,2-dimethyl-l-phenylbutane(VI) was excluded by comparison of its infrared spectrum with those of the hexylbenzene cuts from expts. 9-11. TABLE IV ETHYLBENZEXE-PROPYLENE EXPERIMENTS Experiment 12 13 14 Temperature, "C. 402 434 456 Reactants, moles Ethylbenzene 4.69 4.22 4.46 Propylene 0.93 0.85 0.89 Results Propylene reacted, mole % Yields based on propylene reacted, mole % Propylene dimers Propylene trimers Monoadducts Ethylbenzene reacted, mole % Yields based on ethylbenzene reacted, mole % Benzene Toluene Cumene n-Propylbenzene Unident. butylbenzene ( ? ) 2-Methyl-3-phenylbutane 2-Phenylpentane Diadduct 2,3-Diphenylbutanea Diarylalkanes
23
35
41
25 6.5 34 2.8
23 4.2 45 5.5
18
.... 36 8.1
TABLE V ETHYLBEYZENE-~SOBUTPLENE EXPERIMENT
Experiment Temperature, "C.
15 4!5.5
Reactants, moles Ethylbenzene Isobutylene
4.53 1 00
Results Isobutylene reacted, mole % ' Yields based on isobutylene reacted, mole Isobutylene dimers Isobutylene trimers Monoadducts Ethylbenzene reacted, mole %
37
%
Yields" based on ethylbenzene reacted, mole % ' Benzene Toluene Cumene n-Propplbenzene Unident. pentylbenzene ( ? ) 2-Methyl-4-phenylpentane Isohex ylbenzene Diadduct 2,3-Diphenylb~tane~ Diarylalkanes
54 6 8 24 5.2
1 8
9 3 2 18 20 5 2 32
Unsatn. of monoadducts, mole % 31 2,2-Dimethyl-3-phenyIbutane, the expected "alternate" monoadduct, was not detected in the product cuts. The infrared spectrum of a n authentic sample (b.p. 205", nZoD 1.4954) had a characteristic band at 12.95 p . The presence of this arene's possible rearrangement product, 3,3-dimethyl-l-phenylbutane(lit.48 b.p. 211.6-212' a t 750 mm., n M D 1.4826) was not indicated by the properties of the monoadduct cuts from expt. 15. 6This represents only the material which crystallized out of the higher boiling cuts. a
1.3 5.1 14 0.3
1 1 6 1
1 1 12
....
....
7.6 49 21.3 1.4
8 51 15 2 1.5
2 7 29 13 1 29
....
5
Unsatn. of monoadducts, mole yo 21 13 13 This represeiits only the material which crystallized out of the higher boiling cuts.
The experiments were carried out a t 420 + 15 atm. and a t an hourly liquid space velocity (H.L.S.V.) of 2.0 except in experiment 5 which was performed a t H.L.S.V. 1.0. The amount of noncondensable (-78") gases was negligible, in most cases being less than 0.5 liter.
1ro1. 79
HERMAN PINESAND JOSEPH T. ARRIGO
4960
TABLE VI EXPERIMENTS CUMENE-PROPYLENE Experiment 18 17 16 454 404 433 Temperature, “C.
TABLE VI1 CUIMENE-ISOBUTYLENE EXPERIMENT Experiment Temperature, “C.
K e x t a n t s , moles Cumene Propylene
Reactants, moles Cuxnene Isobutylene
Results Propylene rcacted, mole yo Yields based on propylene reacted, mole % Propylene ciimcrs Propylene trimers hlonoadducts Cumene reacted, mole % Yields” based on cumene reacted, mole 70 Benzene Toluene n-Propylbenzene” Isobutylbenzene Unident. pentylbenzene ( 7 ) 2-hlethyl-2-pheiiylpentane 2-hIethyl-1-phenylpentanrC 1,1,3-Trimethylindan 1,1,2-Tritnethylindari Diadduct Diarylalkanes Unsatn. of monoadducts, mole
4.17 0.83 SO
37 37 14 0 63
1 3 ;i 0.02 .... 10 20
4 14 3 13 ....
4.02 0.80 16
50 16 22 1.34
1 2 0.21 2 4 26 9 19 8 14 15
4 31 0 86
39
42 20 20 5 92
1 7
2.8 3 8 1!5 15 14 5 17 17
70
41 29 37 The expected ‘‘alternate” monoadduct, 2,3-dimethyl-2pheiiylbutane, could not be detected in the product cuts by comparison of their infrared spectra with that of an authentic sample (b.p. 210”, X ~ O D 1.4995). *Yields based on cumene charged. c This compound could not be isolated in a pure form; its infrared spectrum contained two bands not present in that of an authentic sample. a
Results Isobutylene reacted, mole
%
Yields based on isobutylene reacted, mole % Isobutylene dimers Isobutylene trimers I\.Ionoadducts Cumene reacted, mole % Yields“ based on ciiinene reacted, mole Betizene Toluene n-Prop ylbenzenc” I w b u t ) Ibenzene Unident lieptylbenzcne ( 7 ) 2,4-Dimethyl-l-plienylpentane 1,1,3,3-Tetrninethyliridan Diadduct Diarylalkanes
%
Unsatn. of monoadducts, mole % The expected monoadducts could not be detected in the product cuts of expt. 19. 2,3,3-Trixnethy1-2-phenylbutane53 (b.p. 219-233” cor. t o 760 mm., ~ Z O D 1.5055) exhibited a sharp infrared band at 12.91 p not present in the spectra of the product cuts. 2,4-Diniethyl-2-phenglDen(b.p. 215” cor. t o 760 mm., T Z ~ O D 1.4932) has a characteristic infrared band a t 13.04 p which was not detected in the spectra of tile product cuts. * Yields based o i i cumene charged.
I t is assumed that the main source of benzylic free radicals was the direct pyrolytic homolysis of benzylic C-H bonds. The relevant bond dissociaDiscussion tion energies for toluene, ethylbenzene and cumene The products of the thermal reactions are clearly are 77.5, 74 and 71 kcal.,/mole, respecti~ely.~In the result of free radical processes. The following the latter two cases, although C-C cleavage of the mechanism of monoadduct formation is suggested, side chain has a lower energy requirement, under similar to that proposed for the thermal alkylation the present conditions only negligible amounts of of alkane^.^ products which could have been formed by such (1) PhCH3 PhCH2 + H . cleavage were found. Benzylic radicals were I I1 probably formed in part by hydrogen abstraction I” from arenes by allylic radicals (generated by olefin ( 2 ) I1 JJ PhCHZCHzCHCH3 Ph(CHz)sCH3 pyrolysis), which have comparable stabilities.6 I11 + 11 (“normal”)b Toluene is known to serve as an effective hydrogenCH3 CH? donating solvent to allyl radicals formed by the I“ I H&=CHCH3 PhCH2CHCH2. r’PhCH2&HCH3 pyrolysis of allyl chloride.6 The formation of alkenylbenzenes by hydrogen IVC + 11 (‘‘alternate”)h atom elimination (equation 3 and 4) appears to be (3) IIId H . + PhCH*CH2CH=CH* and isomeric of only minor importance under the conditions olefin used, judging from the monoadduct unsaturation CH3 data (Tables I-VII), which, although approximate, I give an upper limit of alkenylbenzene concentra(4) IVc H . + PhCH2C=CH2 and isomeric olefins tion. The olefin may also serve as hydrogen donor, yielding The effect of structure on reactivity is evident allyl radicals in this case. For convenience, addition leading to the more stable intermediate radical is termed “norfrom the data in Table IX on the conversions of mal” and t o the least stable radical “alternate.” c IV may arenes to monoadducts (based on arenes charged). also eliminate a methyl radical t o yield allylbenzene. Al-
+
though the concentrations of radicals I11 and IV are probably low, the formation of alkenylbenzenes by the disproportionation of these radicals should not be excluded. (4) F. E . Frey and H.3. Hepp, I n d . Eng. Chcm., 28, 1439 (1936).
( 5 ) (a) M. Szwarc, C h e m . Reus., 4’7, 75 (1950); (b) for a recent tabulation of bond dissociation energy values, see E. W. R. Steacie, “Atomic and Free Radical Reactions,” Vol. I, 2nd ed., Reinhold Publishing Corp., New York, N. Y . , 1954. (6) L. M. Porter and F. F. Rust, THISJOURNAL, 18, 5571 (1966).
THERMAL ALKYLATION OF MONOALKYLBENZENES
Sept. 20, 1957
49G1
TABLE VI11 MONOADDUCT PRODUCTS OF EXPERIMENTS 1-19” Expts.
C
1-5: PhC
+ C=C-C
C
C 6-8: PhC
I
+ C=C-C
+PhC-C-C-C
c c
9-11: PhC
I
C
I
+ PhC-C-CI I
C
c c
/
I + C=C-C +PhC-C-C-C C
PhC
+ PhC-C-CI
+PhC-C-C-C
I
C
+ C=C!!-C-C
+PhC-C-C-C-CI
c c
C 12-14: PhC-C
+ C=C-C C
15: PhC-C
I
+ C=C-C
I 1 + PhC-C-C
I
+PhC-C-C-C c c I I +PhC-C-C-C
C
+ PhC-C-C-C-C
I
I
C
c C 19: PhJ--C
‘ Ph
c
C
+ C=C!!-C
--f
C
c
PhC-A-C-C-C
+ cAc
=
CsHs.
The observed olefin reactivity toward a given arene is seen to decrease in the order propylene > isobutylene > 2-methyl-2-butene. This sequence concurs with that reported in the thermal alkylation of paraffins with olefin^.^!^ As with the olefin reactants, i t was found that increasing substitution decreased the reactivity of the alkylbenzenes toward a given olefin in the sequence toluene > ethylbenzene > cumene. It is of interest t o note that in both series there is a small decrease going to the second member but a sharper drop-off in reactivity with the most substituted compound. The latter decrease is less marked as the temperature increases, presumably due to increased activation of the reactant molecules. A similar effect of increasing substitution was encountered by Huisgen and co-workersg in the thermal additions of arenes to azodicarboxylic ester. As the temperature increased, the conversions of arenes charged also rose, falling off, however, a t (7) G . G. Oberfell and F. E. Frey, Oil and Gas J . , 88, 70 (1939). (8) A. A. O’Kelly and A. N. Sachanen, Ind. Eng. Chrm., S8, 462 (1946). (9) R. Huisgen, F. Jekob, W. 8iegel end A. Cadus, ann., 690, 1 (1954).
475 and 485” where secondary reactions became intensified. The yields of monoadducts based on arenes reacted (in parentheses in Table IX) decreased a t the higher temperatures, further indicating the increased importance of side reactions. Selectivity.-The selectivity of benzylic radical addition to unsymmetrical olefins may be expressed in terms of the selectivity ratio, defined as the ratio of “normal” t o “alternate” addition, e.g., n-lisobutylbenzene in the toluene-propylene reactions. The results are given in Table X . The selectivity data support the following generalizations : 1. The order of relative ease of formation of the intermediate monoadduct radicals decreases in the series tertiary > secondary > primary. There is a tenfold increase in the selectivity of addition of toluene to isobutylene as compared to propylene a t 405”, while only tertiary radical products were detected in the methylbutene reaction. Comparable selectivity has been reported in the additions of methyl radicals to propylene and isobutylene. lo 2. The selectivity among the aralkyl radicals toward a given olefin de(10) F. F. Rust, F. H. Seubold and w. E. Vaughan, THISJOURNAL, 70, 95 (1948).
4962
HERMAN PINESAND JOSEPH T. ARRIGO
Vol. 79
TABLE IX CONVERSIONS OF ALKYLBENZENES TO MONOADDUCTS~ Reactants
+
I
PhC-C PhC
Approximate temperature 4550
430'
405O
+
2.0 (58) 1 . 6 (57)
3.7 (57)
3 . 2 (59)
5 . 2 (50) 2 . 9 (36)
+ C=C-C
0 . 2 5 (41)
0 . 7 0 (62)
1 . 6 (49)
1 . 8 (73)
2 . 7 (58)
2.8 (37)
PhC C=C-C C=C-C PhC-C C
+ C=C-CI
4750
485'b
3 . 5 (33)
0 . 9 1 (6.6)
C PhC-C C
I
PhC-C
+ C=&-C
2 . 0 (38)
C
+ C=C-CI
0 . 6 2 (40)
C
+
I
PhC C-C=C-C 0.15 (61) 0.48 (72) 1 . 0 (53) a Mole ?& yields based on arene charged; values in parentheses are yields based on arene reacted. others H.L.S.V. 2.0.
H.L.S.V. 1.0, all
TABLE X SELECTIVITY RATIOS Competing intermediate radical types
Secondary vs. primary
405'
Reactants
+
PhC C=C-C PhC-C C=C-C " c F'hA-C
+
+ C=C-C
430'
Approximate temperature 4550 4750
5.1 6.6
4.1 4.2
13
7
9
45
35
20
4.4
6.5
2.6
4850a
2.3
C
Tertiary vs. primary
PhC
+ C=C-CI C
PhC-C C
I
PhC-C
+ C=A-C
All tertiary
C
+ C=C-CI
All tertiary
C
Tertiary vs. secondary PhC H.L.S.V. 1.0, all others H.L.S.V. 2.0.
+ C-&=C-C
All tertiary
* Partial isomerization of the olefin to 2-methyl-1-butene occurred. a,a-dimethylbenzyl > a- have been found to yield products of both modes of
creases in the order phenylethyl > benzyl, as is evident from their additions to propylene a t 405". It is of interest to note that a selectivity of similar magnitude has been reported in the addition of t-butyl radicals to propylene.8 The decrease in selectivity in the higher temperature reactions may be due to the enhanced activation of the reacting species; however, due to certain inherent complexities, l1 further data are necessary before any clear-cut relationship can be established. Orientation in free radical additions to olefins has received much study in recent years. It is well established that moderate temperature peroxide- or light-induced additions are unidirectional, attack leading exclusively to the most stable intermediate radical. l 2 I n contrast, the higher temperature additions of alkyl radicals to olefins (11) Any preferential secondary reactions of either of the monoadduct radical species would affect the apparent selectivity t o some extent, as would coupling of allylic and benzylic radicals which is known t o proceed t o a small extent in a similar reaction (ref. 6). Monoadduct unsaturation d a t a indicate t h a t the latter factor is only of minor importance. (12) For a general review, see J. I. G. Cadogan and D. H. Hey, Quart. Rcrr., 8 , 308 (1954).
a d d i t i ~ n . * * ~In - ' ~the present study, the observed selectivity of benzylic radical addition was found to depend on the structure of the reactants and the reaction temperature, The order of increasing reactivity is in accord with bond dissociation energy data5 and hydrogen abstraction result^.'^^'^ It has been reported that selectivity in hydrogen atom abstraction reactions paralleled the stability of the attacking species.15 The selectivity-activity concept16 used in correlating isomer distribution in electrophilic aromatic substitution has been applied recently in explaining the thermal polymerization of 0lefins.1~ I n accordance with these views, the present results indicate that the most stable benzylic radicals are the most selective in the addition to unsymmetrical olefins. Although the relative importance of polar ZIS. steric effects in determining the point of radical (13) G. A. Russell, TRISJOURNAL, 78, 1047 (1956). (14) A. L. Williams, E. 9. Oberright and J. W. Brooks, ibid., 78, 1190 (1956). (15) G. A. Russell and H. C. Brown, ibid., 77, 4578 (1955). (16) H. C. Brown and K. L. Nelson, ibid., 1 6 , 6292 (1953). (17) V. Mark and H. Pines, ibid., 78, 5946 (1956).
Sept. 20, 1957
THERMAL ALKYLATION OF MONOALKYLBENZENES
attack on an unsymmetrical olefin is difficult to assess, the results of many studies appear to favor the predominance of the former. l 2 The decreasing susceptibility of olefins to attack as methyl groups are substituted on the unsaturated carbons indicates that steric factors are quite important in the present study. Maximum hindrance was found in the 2-methyl-2-butene reactions. Isomerization of the olefin led to approximately an equilibrium mixture of methylbutenes.18 If the relative ease of addition of benzyl radicals t o 2-methyl-2butene and 2-methyl-1-butene were comparable, it might be expected that roughly proportional amounts of 2,3-dimethyl-l-phenylbutaneand 3methyl-1-phenylpentane would have resultedlg (Table VIII). By dividing the ratio of the two recovered olefins by that of the corresponding monoadducts, i t was found that addition to the terminal double bond of 2-methyl-1-butene was favored by a t least a factor of about 2:l a t the upper temperatures. Nuclear Alkylation.-In the cumene-propylene and isobutylene reaction, alkylindans were identified and probably arose as
C
I PhC-C-C-C I
4963
C
I
+PhC-C-C-C-C (b)
C
c I
c I
[PhC-C-C-CIZ3 I
c ---f
c
PhC-&-C-&-C
(c)
C
Cumene also was found to rearrange to n-propylbenzene as previously reported.z4 The mechanism may be formulated as
PhC-C-C-C-C
I
The initial primary radicals could have been formed by direct hydrogen abstraction or hydrogen atom isomerization of related species. It is noteworthy that the rearranged products found all appeared to be due to phenyl migration t o a terminal carbon. The 1,2-phenyl migration in liquid phase free radical reactions has been investigated in detail in c c recent years.25 Evidence has been reported which favors the discrete primary radical rather than a cyclic intermediate in the case of neophyl radic a l ~ .More ~ ~ recently, ~ ~ ~ the unique rearrangement of the 4-methyl-4-phenyl-1-pentyl radical t o C R yield ultimately isohexylbenzene has been reC C ported.20 It was suggested that aryl participation led to a spiro free radical intermediate capable of forming the product. Although such a process cannot be excluded in (c), it fails to account for the products in (a) and (b). C R C R Side-Reaction Products.-The hydrogenated expts. 16-18 (R = H ) ; expt. 19 (R = C) propylene dimer cuts were found to contain nhexane with the presence of methylcyclopentane Recently, similar cyclizations of 4-phenyl-1-butyl also indicated, while isobutylene yielded mainly free radicals to form tetralin have been reported.20t21 1,1,3-trimethylcyclopentane. These findings conIt is possible that the presence in the intermediate cur with results observed in reactions a t lower radicals of gem-dimethyl groups, which are known pressures. 17,28 Evidence also was found for the to exert a marked enhancement on ring closure re- formation of polyalkylcyclopentanes as well as actions,2zfavored cyclization in experiments 16-19. linear dimers from 2-methyl-2-butene. The only other occurrence of nuclear alkylation Side-chain decomposition products of arene reacin the present study, vzz., the formation of isomeric tants and products were found in all reactions in cymenes in minute amounts in experiments 4 and small amounts increasing with the severity of con5, is not explainable readily by known free radical ditions, in accordance with the reportz4 that mechanisms. cumene and s-butylbenzene yielded all possible Monoadduct Rearrangement.-The following ex- lower monoalkylbenzenes under similar condiamples of 1,2-phenyl migrations were noted in the tions. present work I n most of the experiments, the arene reactants’ next higher homolog was found, e.g., toluene yielded c c C I I I ethylbenzene. Similar “methylation” by the comPhC-C-C-C +PhC-C-C-C-C (a) bination of methyl and benzylic radicals has been shown t o O C C U T . ~ ~ ,The ~ ~ detection of methane in (18) R. H. Ewe11 and P. E. Hardy, Abstracts of the 102nd Meeting of all analyzed non-condensable gases suggests such a the American Chemical Society, Atlantic City, N . J., September, 1941, p. 231. (19) Only the two arenes formed from tertiary intermediate radicals were found; secondary radicals would have yielded 2,Z-dimethyl-lphenylbutane in each case. (20) S.Winstein, R. Heck, S. Lapporte and R. Baird, Ezpcrientia, 12, 138 (1956). (21) D. F. DeTar and C. Weis, THISJOURNAL, 1 8 , 4296 (1956). (22) For a discussion of the gem-dimethyl effect, see G. S. Hammond, “Steric Effects in Organic Chemistry,” edited by M. S. Newman, John Wiley and Sons, Inc., New York, N. Y.,1956,pp. 462-469.
(23) Tbis expected monoadduct was not detected, only its rearrangement and cyclization products were. (24) V. N. Ipatieff, B. Kvetinskas, E. E. Meisinger and H. Pines, THIS JOURNAL, 1 6 , 3323 (1953). (25) For literature references, see D . Y. Curtin and M . J. Hurwitz, i b i d . , 14,5381 (1952). (26) W.H. Urry and N. Nicolaides, ibid., 14, 5163 (1952). (27) F. H. Seubold, Jr., ibid., 16, 2532 (1953). (28) J. B. McKinley, D. R . Stevens and W. E. Baldwin, ibrd., 61, 1455 (1945).
HERMAN PINESAND JOSEPH T. ARRIGO
4964
reaction was occurring. The data also indicate that some decomposition of the normal adducts took place a t higher temperatures, e.g., the formation of ethylbenzene and ethylene by cleavage of n-butylbenzene. The complexity of secondary reactions makes any detailed conclusion beyond the scope of this work. The diadduct products were only investigated in the simplest case, from toluene and propylene. It was reasoned a priori that the main component might be either n-heptylbenzene (from benzyl radical addition to 1-hexene, an expected olefin dimer) or 3-methyl-l-phenylhexane, formed as PhC-C-C-C
+ C-C-C
+
H.
Hydrogenation.-The selective hydrogenations were performed in a 114-1111. rotating autoclave a t temperatures of 150-225' and initial hydrogen pressure of 100-130 atm., using 10 wt. yo or the charge of copper chromite36 catalyst and pentane solvent. The approximate mole yo of unsaturation of the monoadduct products was calculated from the pressure drop, after correction was made for the hydrogen uptake of the catalyst36 as determined in a blank run. Chromatography.-The separation of arenes from saturates was done by displacement chromatography on silica ge13'j38 using absolute ethanol as the desorbing agent. Separations of arene mixtures39 were similarly carried out using a n-pentane pre-pass. Infrared Spectral Analysis.-The analyses were performed on a Baird double beam recording infrared spectrophotometer40 by the base-line technique.41 The wave lengths of the absorption bands used in the calculations are given in Table XI."
TABLE XI ABSORPTION BANDS
c Ph-C-C-C-C-C-C
I
The purified heptylbenzene did not correspond to either of the above compounds and was not further investigated. The infrared spectrum established that the arene had two benzylic hydrogens, thus excluding its formation by dialkylation of the methyl group of toluene. Diarylalkane fractions probably consisted of two types of compounds: one being composed of two arene molecules, e.g., anthracene from toluene and 2,3-diphenylbutane from ethylbenzene. I t has been shown that the decomposition of benzoyl peroxide in ethylbenzene leads to 2J-diphenylbutane as well as ethyl biphenyl^.^^ The fact that toluene with propylene yielded three times the amount of diarylalkanes as i t did with 2-methyl-2butene a t 456" suggests that such fractions arose in part by the combination of two arenes and one olefin molecule.
Experimental Apparatus and Procedure .-The experiments were carried out in a previously described high-pressure flow-type app a r a t u ~ ,using ~ ~ ? 45 ~ ml. of '/a X '/a'' copper punchings as contacting agent. T h e vertical furnace was automatically controlled. T h e arene-olefin charge was accurately delivered by means of a Ruska proportioning pump31 into the system which was initially pressurized t o about 408 a t m . with nitrogen. The liquid product was collected in a calibrated receiver and the effluent gases were passed through a Dry Ice-acetone-cooled trap, metered by a wet test incter and collected in a gas collectiiig bottle. The liquid product was distilled initially through a LeckyEwell column32t o remove the unrcactcd arene and lower boiling components; then the residue was distilled through a Vigreux column. The fractions boiling up t o the lower diadduct range were combined, selectively hydrogenated and chromatographed t o remove saturutes. The aromatic portion was then carefully fractionated through a Podbiclniak whirling band column33 a t atmospheric pressure and the composition of the individual cuts determined by infrared spectral analysis.34 Non-condensable gases wcre analyzed by means of a mass spectrograph. (29) R. L. Dannley and B. Zaremsky, Txns J O U R N A L , 77, 1588 (1955). (80) V. N. Ipatieff, G. S. Monroe and L. E. Fischer, I u d . Eizg. Cheiib., 40, 2054 (19-IS). (31) Kuska Instrument Corp., Houston, Texas. ( 3 2 ) H. S. Lecky a n d R. H. Ewell, I d . Ewg. Chdnr., A i d . E d . . 1 2 , 544 (1940). ( 3 3 ) Podbielniak. Inc., Chicago, Ill. (:34) I n expts 1 - 1 9 , charged areneb were U3-98';/, &iccounted f u r , \\ hile t h e ulelins were accounted f u r tu t h e following extent (average) : C3H', 5 9 % ; CIH3, 7 2 % ; C,HlO 035;.
Vol. 79
AXALYTICAL I N F R A R E D
Compound
Wave length,=
fi
Benzene Toluene Ethylbenzene Cumene n-Propglbenzene 12-Butylbenzene Isobutylbenzene p-Cymeneb nt-Cymeneb o-Cy*nene~ Isopeiitylbenzene Neopentylbenzene 2-Phenylpentane 2-bIetli~l-3-plietiylbutaiie 2-?Lletliyl--l-phenylpentane Isohcsyll~enzenc 3-hIetliy1-l-plien~l~e1it~1ie
14.86 13.74 13.30 13.15 13.47 8.99 7.28,8.58 12.28 12.78 13.21 9.33 13.87 7 .G3, IO,07 7.80,8.83,0.81,10.28 13.07 13.35 9.63 2,3-r)iinetiiyl-l-plieii~-lbutane 7 , 2 S "-RIethyl-"-pheiiylpeiitane 7.28,9 .F1 ~-hIetl~~-l-l-pIieriylpetltalle 8 . 6 3 , 9 .64 I , l,:;-Triniethylitidaii 7.58,Q.69 1,1,2-Tri1nethylindan 13.66 2,1-1)i1rieth~.l-1-phen~-lpentane 8.51 1,1,3,3-Tetratnetii~linda~i 7.60 2-Metliyl-2-butene 8.98,12.44 2-hZetliyl-1-butene 6.04,11.24 TVave leiigths are uncorrected. b Approximate calculations were made based 011 the corrected absorptivities reported by I
