was added in 1-g. portions over a 30-min. period below 30". Isolation of the product by the method described above afforded 6.6 g. (66.0%) of relatively pure XIV, n1.p. 110-1 12". The analytical sample was obtained as white prisms from ethyl acetate-hexane, m.p. 11 1.5112". Anal. Calcd. for C10H18N20: C, 65.89; H , 9.96; N, 15.37. F o u n d : C, 66.01; H , 10.09; N , 15.18. I I,I I - Dimethyl-4-oxo-3-aza-1 I-azoniabicyclo[4.4.I]undecane Iodide ( X V ) . A solution of 10.6 g. (0.059 mole) of XIV and 19.0 g. (0.135 mole) of methyl iodide in 250 ml. of absolute ethanol was refluxed for 2 hr., cooled, and filtered to give 8.6 g. of product, m.p. 246-252 ". By concentrating the mother liquor, there was obtained an additional 8.6 g. (total yield of 92.0%) of methiodide, m.p. 245-250". Recrystallization of the combined fractions from aqueous ethanol-ether gave pure white crystals, m.p. 262-263'; vKUj0l 3225 ( N H ) and 1660 cm.-' (amide carbonyl). Anal. Calcd. for Cl1Hz1IN2O: C, 40.69; H, 6.67; N, 8.63. F o u n d : C, 40.65; H , 6.61; N, 8.68. Hofmann Elimination of X V . A solution of 18.65 g. (0.058 mole) of XV in water was passed through a column of Amberlite IRA-400 (hydroxide form). One liter of eluate was collected and concentrated. The
residue was pyrolyzed to give 11.28 g. (100%) of a crystalline distillate. Three recrystallizations of this material from ethyl acetate-hexane gave an analytical sample, m.p. 128-129'; vCHC13 3400 (NH) and 1640 cm.-' (amide carbonyl). The ultraviolet spectrum showed only end absorption. Anal. Calcd. for CllH20N20:C, 67.30; H , 10.27; N , 14.27. F o u n d : C, 67.28; H, 10.40; N , 13.96. Reaction of X V I with Perchloric Acid. A solution of 0.50 g. (2.55 mmoles) of XVI in 5 ml. of absolute ethanol was treated with a n ethanolic solution of perchloric acid (1 : 1). The mixture was heated for 5 min. on a steam bath and cooled, and the precipitate was filtered. There was obtained 0.69 g. (91.5%) of white crystals, m.p. 158-159"; vNulol 3350 ( N H ) and 1610 cm.-' (amide carbonyl). The ultraviolet spectrum showed only end absorption. Anal. Calcd. for CllH21C1N20j:C, 44.52; H , 7.1 1 : N , 9.44. F o u n d : C, 44.59; H , 7.14; N,9.33. Acknowledgment. We are indebted to the National Science Foundation for partial support of this research. L. A. P. gratefully acknowledges permission granted by The Upjohn Company, Kalamazoo, Michigan, to publish the conversion of I1 to Ib which was studied while he was in their employ.
Friedel-Crafts Oxygenation of Toluene with Diisopropyl Peroxydicarbonate' Peter Kovacic and Samuel T. Morneweck2
Contribution f r o m the Department of Chemistry, Case Institute of Technology, Cleveland, Ohio. Received November 12, 1964 Cresols were prepared in about 50 % yield f r o m toluene b y direct oxygenation with diisopropyl peroxydicarbonate in the presence of aluminum chloride. Several reaction variables were studied including time, temperature, ratio of reactants, and nature of the catalyst. Electrophilic oxygenation is proposed on the basis of the cresol isomer distribution (ortho, 34 %; meta, 1I %; para, 55 %), the relative reactivity value of 9.6 f o r kroluenr/kbeneenp, the necessity of a catalyst, and the absence of products which would be derived f r o m free-radical reactions. Evidence indicates that the oxygenated product is present in the reaction mixture mainly in the f o r m ArOC02AIC12. This procedure provides the Jirst case of direct oxygenation of toluene in reasonably good yield with essentially no undesirable side reactions. Introduction The literature contains numerous examples of aromatic oxygenation3 with peroxides, e.g., hydrogen peroxide, peracids, and diaroyl peroxides. In many (1) From the Ph.D. thesis (1965) of S . T. M.; presented at the symposiuni i n honor of Sir Christopher K . Ingold, Vanderbilt University, Nashville, Tenii., Aug. 1964. (2) Union Carbide Co. Felloa, 1963-1964. ( 3 ) This word has been chosen to designate direct introduction of R O ( R = hydrogen, alkyl, acyl, etc.) into the aromatic nucleus.
1566
Journal of the American Chemical Society J 87.7
of these cases the evidence points to participation of free-radical intermediates. However, analogous reactions which proceed by a n electrophilic pathway have received relatively little attention. The historical treatment will be limited largely to those investigations which apparently fall in the latter category. Hydrogen peroxide in the presence of concentrated sulfuric acid5 or boron trifluoride etherate6 has been used t o effect oxygenation of mesitylene, toluene, and m-xylene in low yields. Derbyshire and Waters postulated the involvement of hydroxonium ion. A similar mechanism was proposed in the case of peroxy acids for both the uncatalyzed7 and Lewis acid catalyzed reactions. Chambers, Goggin, and M ~ s g r a v e who , ~ reviewed earlier work on the reaction of peracetic acid with aromatics, examined the behavior of trifluoroperacetic acid toward m-xylene, mesitylene, and pseudocumene. The concept of electrophilic attack is consistent with the isomer distribution observed with m-xylene. Phenol ethers have also been (4) E.g., see J . R. L. Smith and R . 0. C. Norman, J . Chem. Soc., 2897 (1963); D . I . Davies, D . H. Hey, and G . H . Williams, ibid., 1878 (1958). (5) D. H . Derbyshire and W. A . Waters, Nature, 165, 401 (1950). (6) J. D. McClure and P. H. Williams, J . Org. Chem., 27, 24 (1962). (7) R . D. Chambers, P. Goggin, and W. K . R . Musgrave, J . Chem. Soc., 1804 (1959).
April 5 , 1965
used as substrates with this reagent.8 These reactions are generally characterized by modest yields of monooxygenated products since further oxidation was prone to occur. Benzene and toluene, in contrast, gave only tars. Hart and c o - w o r - k e r ~ ,in~ recent investigations of the boron trifluoride catalyzed reaction of trifluoroperacetic acid with aromatics, reported fair to good yields of hydroxylated products from mesitylene, isodurene, and prehnitene. The formation of cyclohexadienones, isolated from hexaalkylbenzenes and prehnitene, indicated the intermediacy of positively charged entities. In the presence of a Friedel-Crafts catalyst, diaroyl peroxides are known to form esters with aromatics, particularly of the activated type. Orientation pointed to an electrophilic mechanism. lo The objectives of this work were to effect FriedelCrafts oxygenation of toluene with diisopropyl peroxydicarbonate and to study the mechanistic aspects.
Table 11. Effect of Temperature"
----Crude
Temp., "C.
phenolic,
7-
g.
g.
--7%
- 30 0 25
3.9 4.6 4.4
3.3 4.0 3.8
43 53 50
Effect of Catalyst : Peroxide Ratioa
Table I.
7---
AICI, : peroxide, mole ratio 3 2 1
Product yield
-
g.
g.
z
4.6 4.2 1.7
4.0 3.7 1.5
53 49 18
Cresols --.
a At 0" for 3 hr. with 0.07 mole of peroxide and 1.4 moles of toluene.
relation to the temperature variable, the -30 to 25" region was investigated. Best yields were observed at 0 to 25" (Table 11). T h e amount of cresol increased (8) J . D. McClure and P. H. Williams, J . Org. Chem., 27, 627 (1962). (9) C . A . Buehler and H. Hart, J . Am. Chem. SOC.,85, 2177 (1963); A . J . Waring and H. Hart, ibid., 86, 1454 (1964). ( I O ) J . T. Edward, H. S . Chang, and S . A . Sarnad, C a n . J . Chem., 40, 804 (1962). (11) D. 2. Denney, T.M. Valega, and D . B. Denney, J . A m . Chem. SOC.,86, 46 (1964).
=
1 :3:20
slightly on increasing the reaction time from 1.5 t o 3 hr., and then remained essentially constant for longer times (Table 111). Short periods could not be examined Table 111. Effect of Reaction Timea Product yield ------.
Crude phenolic,
Time, hr. 1.5 3 6
---Cresols---.
g.
g.
3.9 4.2 4.3
3.3 3.7 4.1
z 43 49 51
a At 0" with per0xide:aluminum ch1oride:toluene (molar) (ca. 0.07 mole of peroxide).
=
1 :2:20
conveniently since the exothermic reaction precluded rapid addition of the peroxide. Doubling the amount of toluene served only to decrease the over-all yield, presumably because of additional losses during work-up, with little alteration in the per cent of high-boiling phenol and residue (Table IV). Table IV.
Effect of Toluene Concentration"
----
Product yield ___-
Toluene : peroxide, molar ratio 20 40
Crude phenolie, 8.
4 . I5 3.45
Cresols
7-
7
---
€5
%
3.7 3.1
49 41
a At 0" for 3 hr. with aluminum ch1oride:peroxide (cn. 0.07 mole of peroxide).
-----
Crude, phenolic,
Cresols
a For 3 hr. with per0xide:aluminum ch1oride:toluene (molar) (ca. 0.07 mole of peroxide).
y -
Results and Discussion The standard conditions (diisopropyl peroxydicarbonate : aluminum chloride :toluene = 1 : 2 : 20 (molar), O", 3 hr.) gave a mixture of crude phenols of which 96% were cresols, a 49% yield based on the peroxide (53% when corrected for losses incurred during work-up). The cresols were identified by the infrared spectrum, g.1.c. retention time, elemental analysis, boiling point, and odor. Infrared spectroscopy provided the isomer distribution: 34% ortho, 11 % meta, 5 5 % para. A higher boiling material (2%) was obtained as a by-product which had a retention time and infrared spectrum similar t o those of the cymenols. The distillation residue constituted 2 % of the phenols. The neutral portion contained mainly cymenes (primarily m e t a ) characterized by the infrared spectrum, g.1.c: retention time, and elemental analysis. I n addition, some 3,5-diisopropyltoluene was present which was identified by the infrared spectrum. A number of variables were examined in order to determine their effect on the reaction. It was found that the yield of cresols was very dependent on the ratio of aluminum chloride to peroxide, at least a 2 : 1 molar ratio being required for good results (Table I). In
-----
Product yield
=
1 :2 (molar)
F r o m a study of various candidate catalysts, the indicated order of reactivity was obtained: AIC13 > BF3, AIBr3 > SbClj > FeCI,, SnCI,, H 3 P 0 4(Table V). Introduction of coordinating solvents, namely ethyl ether, nitromethane, or tetramethylene sulfone, markedly decreased the product yield with aluminum chloride or boron trifluoride. It is well established that reactions of the Friedel-Crafts type are affected by the catalyst strength. For example, acylation of toluene with acetyl chloride revealed the following decrease in catalyst potency12: AICI, > SbCl, > FeCI, > SnCI,. Hence, the order is similar to that obtained in the present study. However, the catalyst rating is known t o depend on conditions, particularly the nature of the (12) 0. C. Dermer, D . M. Wilson, F. M . Johnson, and V . H. Derrner, ibid., 63, 2881 (1941).
Kovacic, Morneweck
Friedel-Crafts Oxygenation of Toluene
1567
Table V.
Catalyst Variationa
-
~~
7
Catalyst
g.
g.
--z
AICI, BF3b AIBr3 AIC13rCHoNO~C SbCIS AIC13-(CHz),SOzc BF3 Et20 FeC& SnCL H3P04 (80 %)
4.15 3.8 1.95 1.6 0.8 0.5 0.2 secondary, > primary, and (b) lower temperatures. Furthermore, the aluminum derivative I1 brings to mind the stoichiometry proposed for acylation with acid anhydrides. l 9 The predominant ortho,para orientation with toluene suggests an electrophilic substitution mechanism. Subjecting a cresol mixture to simulated reaction conditions established the validity of the isomer distribution. No isomerization was found. The relatively large amount of meta isomer indicates that the attacking species possesses high activity and therefore low selectivity. The selectivity factor, Sf, which can be calculated from orientation data, serves as a measure of activity. * l . 2 2 Brown and c o - w o r k e r s 2 1 ~ 2 3postulated ~24 a concerted displacement mechanism for the Friedel-Crafts methylation of toluene. Since the selectivity factor (1.0) for the diisopropyl peroxydicarbonate-toluene-alumh u m chloride reaction is close to the value for methylation (0.84), a similar mechanism for oxygenation appears reasonable.
*”
AIH
--+-
ROCOOCOR I
O...AICI, 11 ROCO-OCOR 6+ 60
..
ArH
+ArOCOzAIC1z+ ArR + HC1 Ha0
ArOCO2AICl2--f ArOH HC1
+ COz + A1Cb
(3) (4)
Alternatively, oxygenation might precede alkylation. 0
O...AICla
I1
11
ROCOOCOR
hrH
+ ArOC02R + ROCOZAICI2+ HCI
(5)
It is likely that a peroxide-catalyst complex plays a crucial role since no reaction occurred in the absence of aluminum chloride. In addition, when diisopropyl peroxydicarbonate was added to aluminum chloride in cyclohexane, a brown flocculent mass was formed. However, no evidence concerning the structure of the brown material could be obtained because of the rapid reaction with atmospheric moisture during attempted identification. Complexes similar to I have been (13) G. A. Olah in “Friedel-Crafts and Related Reactions,” Vol. I, G . A. Olah, Ed., Interscience Publishers, Inc., New York, N . Y., 1963, Chapter XI. (14) G . A. Russell, J . A m . Chem. SOC.,81, 4834(1959). (15) G. F. Hennion and R . A. Kurtz, ibid., 65, 1001 (1943); R . L. Burhell, Jr., and L. M. Elkin, ibid.,73, 502(1951).
1568
Journal of the American Chemical Society J 87.7
(16) N. N . Greenwood and I
