Chromic acid oxidation of aromatic alcohols1

Transcription

1 Chromic acid oxidation of aromatic alcohols1 Ross STEWART AND FARIZA BANOO~ Department of Chemistry, University of British Colunzbia, Vat~couver 8, British Colu~nbia Received March 13, 1969 The mechanism of the chromic acid oxidation of di- and tri-aryl carbinols has been studied in 80 wt % acetic acid containing sulfuric acid. The reactions are cleanly second-order and give benzophenones in the case of the secondary alcohols and benzophenones plus phenols in the case of the tertiary alcohols. Electron-donating substituents in the tertiary alcohol appear predominantly in the phenol component and the rate-controlling step is believed to be a 1,2-aryl shift. The reaction constant for the migration is pf = and that for the overall reaction is pf = The presence of manganous ions does not alter these values although it lowers the overall rate of reaction. Although an analogous 1,Zhydride shift mechanism can be written for the oxidation of the secondary alcohols, there are enough points of difference between the oxidations of the secondary and tertiary systems to make this appear unlikely. Canadian Journal of Chemistry, 47, 3207 (1969) Introduction The chromic acid oxidation of primary and secondary alcohols has been very extensively studied and is believed by most workers in the field to proceed via a chromate ester whose decomposition may proceed either by proton loss to the solvent or by an internal hydrogen shift (14). Tertiary alcohols which can be dehydrated to alkenes often suffer oxidation via the alkenes. Thosewhichcannot bedehydrated undergo oxidative cleavage, triphenylcarbinol producing benzophenone. Bartlett and Cotman (5) showed that p-nitrotriphenylcarbinol gives p-nitrobenzophenone and they considered the reaction pathway to involve rearrangement of a phenyl group to electron deficient oxygen. Cleavage products also appear in the case of the secondary alcohol, phenyl t-butyl carbinol (6, 7) which gives a 60-70% yield of benzaldehyde and t-butyl alcohol, phenyl t-butyl ketone being the minor product. The cleavage reaction can be virtually eliminated, however, by the addition of manganous or cerous ions to the system. Theseions scavengechromium- (IV) and chromium(v) intermediates (1,6-8) and it is clear that the cleavage is not caused by chromium(v1). 'Taken in part from the Ph.D. Thesis of Fariza Banoo, University of British Columbia, ZColombo Plan Scholar, on leave of absence from the East Regional Laboratories, P.C.S.I.R., Pakistan. In order to learn more about the mode of the chromic acid cleavage of tertiary aromatic alcohols we have examined the reaction kinetics, the effect of substituents, and the product distribution in a series of triarylcarbinols. A similar study of an analogous series of diarylcarbinols was made for the purposes of comparison. Experimental The preparation of the alcohols is described elsewhere (90). The reaction rates were followed spectrophotometrically, essentially as previouly described (9b). A 3:2 ratio of alcohol to Cr(V1) was used for the secondary alcohols. Because of the lower acidities used with the tertiary alcohols and because of the production of phenol, which is subject to some degradation, a considerable excess of tertiary alcohol was used and the second-order rate constants obtained by correction. The extinction coefficient of Cr(V1) at 350 mp is 3.33 x lo3 whereas that of (C6H5),Cf at the same wavelength is 1.03 x 10'; hence the change of concentration of (C6H5),Cf (and the other triaryl cations) will have only a small effect on the decrease in absorbance at 350 mp. In a typical run, 10 p1 of 0.14 M potassium dichromate was added by syringe to 2 n ~l of a solution made by mixing 80 wt % acetic acid and the required quantity of 99.9% sulfuric acid and containing 2.3 x M benzhydrol. Product Analysis The secondary alcohols gave nearly quantitative yields of ketones. In the case of 4-methylbenzhydrol a yield of 96.4 % of ketone was obtained and the small amount of oily residue showed no absorption bands characteristic of benzaldehyde, the product that would be formed if aryl migration occurred instead of hydrogen loss. With the tertiary alcohols a weighed quantity (0.001 mole) of the carbinol under investigation was dissolved in 85 ml of 80 wt% aqueous acetic acid and 15 ml of 99.9 % sulfuric acid was added slowly and with stirring. This gave a solution containing approximately 23 % sulfuric acid. Next, 0.10 g potassium dichromate (a slight excess) dissolved in 3 ml of 80 wt % acetic acid in water

2 CANADIAN JOURNAL OF CHEMISTRY. VOL. 47, 1969 TABLE I Analysis of chromic acid oxidation products Molar ratio of benzophenones Substituted triphenylcarbinols Benzophenones (a) without Mn+ + (b) with Mn+ + A 4,4'-diMe (i) 4,4'-diMe (ii)/(i) (a) 6.60 (ii) 4-Me (6) 6.55 (i) 4-Me (ii) Unsubstituted (ii)/(i) (a) 1.70 (6) 1.72 C 4-C1 (i) 4-C1 (ii) Unsubstituted (ii)/(i) (a) 0.38 (6) 0.37 D 4-NO2 (i) 4-NO2 (ii)/(i) (a) (ii) Unsubstituted (b) E 3-CF3 (i) 3-CF3 (ii) Unsubstituted (ii)/(i) (a) 0.18 (6) 0.17 F 3-CF3-4'-CH3 (i) 3-CF3 (i):(ii):(iii) (ii) 3-CF3-4'-CH3 (iii) 4-CH3 (a) 3.1:1:0.34 (b) 3.0:1:0.35 containing 0.5 ml of 99.9 % sulfuric acid was added. After standing for several days the mixture was diluted with water and extracted several times with ligroin. The extract was treated with dilute sodium bicarbonate solution to remove acid and then with 10% sodium hydroxide solution to remove the phenols. The ligroin extract was evaporated to dryness and the infrared (i.r.) spectrum of the residue was taken. The residue was then dissolved in a measured quantity of ether and the ether solution was analyzed by vapor-phase chromatography (v.p.c.). The exact quantities of each component of the products were determined by measuring the areas of the peaks and comparing them with those obtained from known quantities of the corresponding compounds. The alkali washings of the ligroin extract were acidified with dilute hydrochloric acid and extracted with ether. The ether extract was analyzed for phenols by preparing the corresponding bromo-derivatives. Acidity Functions The solvent used for the oxidations and for the acidity function determinations was made by adding a known weight of 80 wt % aqueous acetic acid (80 g acetic acid and 20 g water) to a given weight of % sulfuric acid. The concentrations are expressed as wt% sulfuric acid and refer to that % of the solution weight that is contributed by the sulfuric acid. (A consequence of this procedure is that the % of acetic acid in the system varies somewhat as the sulfuric acid concentration changes.) The Ho and HR values were determined by standard techniques using previously established values of pkbh+ and pkr+ of the indicators (10, 11). The results are given in Table 11. Results and Discussion Acidity Dependence As with other alcohol-chromic acid systems that have been examined the kinetics are cleanly second-order, first-order in each of the alcohol TABLE I1 Acidity functions in 80 wt% acetic acid containing various concentrations of sulfuric acid %H2S04 Ho * HR t *Indicators: 4-nitroaniline, 3-rnethoxydiphenylamine, 3,4'-dichlorodiphenylamine, 3-nitrodiphenylamine, 4-nitrodiphenylamine, 2-nitrodiphenylamine. tlndicators: tri(4-tolyl)carbinol, 4,4'- dimethoxytriphenylcarbinol, triphenylcarbinol, tri(4-ch1orophenyl)carbinol. and oxidant. (In the case of the tertiary alcohols deviations were apparent beyond 60% reaction; this is presumably due to some degradation of the phenol produced in the reaction.) The acid-catalysis of the oxidation of benzhydrol was similar to that found previously with isopropyl alcohol (9) i.e. a linear relation of unit slope between log k and H, up to an H, of approximately - 1, fol-

3 STEWART AND BANOO: CHROMIC ACID OXIDATION 3209 FIG. 1. Effect of acidity on the rate of the chromic acid oxidation of diphenylcarbinol (open circles) and triphenylcarbinol (closed circles). lowed by a more gradual increase in rate as the acidity of the solution was increased (Fig. 1). The source of this effect, which is believed to be connected with ligand substitution on chromium, has been discussed previously (9). Curiously enough this effect is not present in the case of the tertiary alcohol, triphenylcarbinol. Indeed, the greater dependence of rate on acidity in this case causes the tertiary alcohol to be oxidized faster than the secondary alcohol at high acidities. (When HR is used instead of H,, a somewhat better straight line of slope is obtained for triphenyl carbinol.) Substituent Effects Analysis of products showed that both the secondary and tertiary alcohols produced high yields of ketones, with phenols also being formed in the latter case. The cleavage which occurs with the tertiary alcohols produces mixtures of products when substituted triarylcarbinols are used and Table I gives the distribution of these products with various substituents. The tendency for elec- tron-withdrawing substituents to be retained in the ketone, which is clearly apparent from the results in the table, suggests that a 1,2-shift to an electron-deficient atom is occurring in agreement with the early suggestion of Bartlett and Cotman (5). (In contrast, homolytic rearrangements of these systems have quite different migration aptitudes (5, 12).) On this basis the migration aptitudes for the various rings were calculated, with appropriate statistical corrections being made, and the results are given in Table I11 and plotted against o+ (13) in Fig. 2. The p+ value for migration is and this was not affected appreciably by the addition of manganous ions. The effect of substituents on the overall rate of oxidation of diarylcarbinols is similar to that found with other secondary alcohols except that the reaction constant is lower, being (Fig. 3). In the case of the tertiary alcohols the effect of substituents was best correlated with of (Fig. 4). The reaction constant for the oxidation, pf, was found to be In the presence of added

4 3210 CANADIAN JOURNAL OF CHEMISTRY. VOL. 47, 19G9 TABLE , 1 Migration aptitude of aryl groups in the chromic acid oxidation of triaryl carbinols Relative migration Aryl group aptitude, y log Y CsHs 1.oo CH3C6H4-3.35* ClC6H CF3CsH4-0.36t N02C6H 'Average of the valucs obtained with compounds A and B in Table I; 3.30 and 3.40, respectively. Cornpound F gave a value of 3.1. tcompound F gave a value of FIG. 2. Effect of substituents on the migration aptitude of aryl rings in the chromic acid oxidation of triarylcarbinols. manganous ions, an almost uniform reduction in rate was observed for all the tertiary alcohols (Table IV), the p+ value, , being almost identical with that found in the absence of manganous ions. The cleavage reaction, which evidently occurs via a 1,Zshift of an aryl group, thus takes place in the same way whether chromium(v1) or chromium ions of intermediate valence are the active oxidants. Mechanism The mechanism of the chromium(v1) oxidation of triarylcarbinols to substituted benzophenones and phenols can be written as follows I FIG. 3. Effect of substituents on the rate of the chromic acid oxidation of diarylcarbinols in 80 wt% aqueous acetic acid at 25', H,, = , correlation coefficient, Substituents are from upper left to lower right: 4,4'-dimethyl, Cmethyl, none, 4-chloro, 4,4'-dichloro, 2,4,6,2',4',6'-hexamethyl. The latter's o was calculated using the pk,, of mesitoic acid. FIG. 4. Effect of substituents on the rate of the chromic acid oxidation of triarylcarbinols. Open circles, manganous ion present; closed circles, manganous ion absent. L transition state _I ~r~c==b-ar + H20 + ArzC=O + ArOH + H+

5 STEWART AND BANOO: CHROMIC ACID OXIDATION The effects of substituents and acidity on the rate of the oxidation are in agreement with the above path. Preferential migration of rings containing electron-donating substituents is expected and is found, although the influence of substituents on migration aptitude3 is somewhat smaller than that found in the analogous acid-catalyzed rearrangement of triarylmethyl hydroperoxides (14) + Ar3C-0-OH + H+ $ Ar3C-0-OHz -> ~r~c==6~r + HZO -> Ar2C==0 + ArOH + H+ The possibility that secondary and primary alcohols are oxidized by a similar mechanism, i.e. a 1,2-hydride shift to give the conjugate acid of benzophenone, should be considered. Such shifts occur with ease in many other systems (15). Over most of the acidity range examined, benzhydro1 is oxidized considerably faster than triphenylcarbinol and if the two pre-equilibrium steps are comparable, this would require a hydride shift to be faster than a phenyl shift. There is evidence that this order prevails in some, though not all, 1,2-shifts (1 5). On the whole, however, the weight of evidence is against a 1,2-hydride shift mechanism. First, a careful examination of the products of the oxidation of 4-methylbenzhydrol revealed no trace of 4-methylphenol. It seems unlikely that a hydride shift would be that much more facile than a 4-methylphenyl shift. (It was not possible to study alkoxybenzl~ydrols under the same conditions as were used for the other carbinols because complex formation between the alkoxy group and chromium(v1) occurs.) Second, the dependence of the oxidation rate on acidity is sufficiently different for secondary and tertiary aromatic alcohols (Fig. 1) as to suggest different reaction paths. (The effect of substituents also shows an important difference-the secondary alcohols follow p and the tertiary alcohols follow 3A referee has made the interesting suggestion that the higher p+ for migration than for reaction may be due to partial equilibration of bridged ions after the ratecontrolling step. TABLE IV Chromic acid oxidation of triphenylcarbinol. Effect of substituents. HR = -3.95, T = 25" 10' x kz MnZ + Substituent 1 mole-is-' Present* '3.045 x lo-+ M manpanous acetate added. pf-but this may simply be due to the participation of the substituent in the migrating ring in the latter case.) Third, the activation parameters are rather different for the chromic acid oxidation of the two kinds of alcohol. At an HR of , the activation parameters for the oxidation of triphenylcarbinol are AH? = 11.8 kcal mole- l and AS? = e.u., whereas for benzhydrol at HR = the values are AH? = 5.9 kcal mole- ' and AS? = e.u. (The weight to be attached to this factor is uncertain since significant variations in these quantities can occur even with secondary alcohols (16).) We conclude that tertiary aromatic alcohols are oxidized by means of a 1,2-aryl shift (5) but that primary and secondary alcohols undergo some other form of carbon-hydrogen bond rupture involving, we would suggest, the 5-membered transition state that has been previously favored (2, 3, 17). The two transition states are shown below 2" alcohol 3" alcohol

6 3212 CANADIAN JOURNAL OF CHEMISTRY. VOL. 47, 1969 Acknowledgments 7. J. HAMPTON, A. LEO, and F. H. WESTHEIMER. J. Amer. Chem. Soc. 78, 306 (1956). The financial support of the National Research 8. J. ROCEK and A. E. RADKOWSKY. J. Amer. Chem. Council of Canada and the External Aid Ofice, SOC. 90, 2986 (1968). 9. (a) F. BANoo and R. STEWART. A~~~~~~~~~~~ Government of Canada, is gratefully acknowl- paper. (b) D. G. LEE and R. STEWART. J. Amer. edged. Chem. Soc. 86, 3051 (1964). 10. D. DOLMAN and R. STEWART. Can. J. Chem. 45, 903 (1967). 1. K. B. WIBERG. In Oxidation in organic chemistry. 11. N. C. DENO, J. J. JARUZELSKI, and A. SCHRIESHEIM. Edited by K. B. Wiberg. Academic Press, Inc., New J. Amer. Chem. Soc. 77, 3044 (1955). York Chap W. H. STARNES, JR. J. her. Chem. Soc. 90, R. STEWART. Oxidation mechanisms. W. A. Ben- (1968). jamin, Inc., New York Chap H. C. BROWN and Y. OKAMOTO. J. Amer. Chem. 3. D. G. LEE and R. STEWART. J. Org. Chem. 32,2868 Soc. 80,4979 (1958). (1967). 14. D. E. BISSING, C. A. MATUSZAK, and W. E. MCEWEN. 4. K. B. WIBERG and H. SCHAFER. J. Amer. Chem. J. her. Chem. Soc. 86, 3824 (1964). SOC. 89, 455 (1967). 15. C. J. COLLINS. Quart. Rev. 14, 357 (1960). 5. P. D. BARTLETT and J. D. COTMAN. J. Amer. Chem. 16. J.-C. RICHER and CLAUDE GILARDEAU. Can. J. SOC. 72, 3095 (1950). Chem. 43, 538 (1965). 6. J. J. CAWLEY and F. H. WESTHEIMER. J. Amer. 17. R. STEWART and D. G. LEE. Can. J. Chem. 42,439 Chem. Soc. 85,1771 (1963). (1964).