J. Org. Chem. 1986,51, 5390-5393
5390
Free-Radical Reagents. 2. Oxidation and Addition Products from the Reaction of Di-tert - butyliminoxyl with Phenols M. Ngo,+K. R. Larson, and G. D. Mendenhall* Department of Chemistry and Chemical Engineering, Michigan Technological University, Houghton, Michigan 49931 Received April 7, 1986 The blue, persistent free-radical di-tert-butyliminoxyl, t-Bu&=NO' (l),oxidizes p-hydroquinones and catechol in organic solvents to the corresponding quinones in good yield. With simple phenols, the reaction takes three pathways. 2-Naphthol and 9-phenanthrol are oxidized to the corresponding o-quinones, while 1-naphthol, phenol, and 2,6-di-tert-butylphenol give 4,4-disubstituted oxime ethers. Para-substituted phenols afford cyclohexadienones with one di-tert-butyliminoxyl a t the original para position. Some of these adducts can be prepared by cooxidation of 1-H and the phenol with ceric ion. Rate constants for some of the reactions of 1 with phenols have been measured by kinetic EPR spectroscopy.
Phenolic oxidations are very widespread in natural systems,l and they are commercially important because of the extensive use of phenols as antioxidants2 and as precursors for the synthesis of polymer^.^ We have studied the oxidation reactions of the blue free-radical di-tert-butylimino~yl~?~ (I) because of a possible analogy to the well-known, inorganic Fremy's radical (2), which characteristically oxidizes phenols to quinones in high yielda6 t-Bu&=NO* 1
(-03S)2NO' 2
Results Stock solutions of 1 in pentane, prepared as de~cribed,~ readily oxidized stoichiometric hydroquinonesand catechol to the corresponding quinones in good yield (Table I). The byproduct 1-H, unlike the case with oxidations with 2, is soluble in organic solvents. The Experimental Section gives exemplary procedures for its separation from the phenol-derived products by fractional crystallization, sublimation of 1-H, preferential extraction of one product into a solvent, extraction of 1-H into methanolic base, or chromatography. p-Cresol and its 2,6-di-tert-butyl derivative (BHT) gave colorless 1:l adducts 3a and 3b, respectively, in the yields indicated in Table I. These compounds could also be prepared more conveniently by cooxidation of equimolar phenol and 1-H with 2 equiv of ceric ion.
o:*
Table I. Products from 1 and Phenols phenol product, % 1-H, % hydroquinone quinhydrone, 99 76 2,5-di-tert-butylhydro- t-BuzCBHz02, 69 68 quinone o-catechol o-benzoquinone, 80 63 2-methyl-l,4-naphtho- vitamin KS,92 63 hydroquinone phenol 3c, 58 61 p-cresol 3a, 43 66 BHT 3b, 78 89 2,6-di-tert-butylphenol 3d, 81 36 1-naphthol 4, 87 68 1,2-naphthoquinone, [82In 2-naphthol 9-phenanthrol 9,10-phenanthrenequinone, 51 66 From visible absorption spectrum.
chloroform in the presence of aqueous NaOH, a 23% yield of benzoquinone could be detected by chromatography, but numerous other products were present. 2-Naphthol with 1, on the other hand, led to a succession of vivid colors, culminating with the appearance of an orange precipitate identified as o-naphthoquinone. Spectroscopic experiments subsequently showed that 2naphthol consumed about 4.1 equiv of 1 and generated 81% of the o-quinone product. 9-Phenanthrol and 1 similarly gave 9,lOphenanthrenequinone. When a 4.5-fold excess of 1 reacted with 1-naphthol under the same conditions, however, no 1,Cnaphthoquinone could be detected, but a cream solid, mp 108.5 "C, to which we assigned the novel bis-oxime structure' 4, was isolated in 87% yield.
X
3a. X = H, Y =Me, 2 = t-BupCNO b. X = t - B u s Y = Me. Z :t-BupCNO C, X
ti, Y
2
t-BupCNO
d . X = ~ - B u Y, = 2 = t-BupCNO e . X = Y = H , 2 = t-BupCNO
4
Initial triah with phenol and 2 equiv of 1 in pentane gave black reaction mixtures from which only 1-H could be isolated. We then made a number of attempts to divert PhOH
+
21
-
C1-H
+
3e3
-
0
0
0
+
Ct-B~2C=Ntll
the presumed intermediate para adduct of 1 and phenol (3e) with base. When the reaction was conducted in
'MTU Dean's Fellow. summer 1985.
(1)Recent Advances in Phytochemistry; Swain, T., Harborne, J. B., Eds.; Plenum: New York, 1979; Vol. 12. (2) (a) Schnabel, W. Polymer Degradation; Hanser International: Vienna, 1981; pp 49-50. (b) Howard, J. A. In Free Radicals; Kochi, J., Ed.; Wiley-Interscience: New York, 1972; Vol. 2, Chapter 12. (3) Finkbeiner, H. L.; Hay, A. S.; White, D. M. In Polymerization Processes; Schildknecht, C. E., Skeist, I., Eds.;Wiley-Interscience: New York, 1977; Chapter 15. (4) Mendenhall. G. D.: Inaold, K. U. J. Am. Chem. SOC. 1973, 95, 2963-2971. (5) Part 1in this series: Corneio, - . J. J.:. Larson. K. R.: Mendenhall. G. D. J..Org. Chem. 1985, 50, 5382. (6) Zimmer, H.; Lankin, D. C.; Horgan, S. W. Chem. Rev. 1971, 71, 229-216.
0022-3263/86/ 1951-5390$01.50/0 0 1986 American Chemical Society
J. Org. Chem., Vol. 51, No. 26, 1986 5391
Reaction of Di-tert-butyliminoxyl with Phenols Table 11. Rate Constants for Reactions of 1 with Phenols (Acetonitrile, 22 "C) compound
PhOH JJ-CIC~H~OH p-BrC6H40H p-02NCeH4OH p-CH&H,OH BHT p-MeOC,H,OH
[compoundlo, M 0.0063 0.00905
0.0226 0.0070 0.00578 0.00394 0.00357
lo4 [l]",M 6.75 6.87
7.39 6.63 6.87
6.87 6.87
104kOMa
299 394 218 c4.9 1940 >5000b >5000b
The mean of three measurements (average SD = 9%) in M-' s-' from fit of ESR decay to the equation In (hO/h,) = kobad[AH]Ot, where h is the peak-to-peak height of one of the three multiplets of 1 in the partially resolved derivative spectrum, recorded with a Varian E9 spectrometer. Reaction complete within preparation time. (I
The acid-sensitive acetal group in 4 immediately suggested an explanation for the unsatisfactory results with the reaction of phenol with only 2 equiv of 1. Indeed, when phenol was reacted with excess 1, a pale yellow, lightsensitive 2:1 adduct 3c was isolated by chromatography. 2,6-Di-tert-butylphenol afforded an analogous, yellow product 3d in good yield, and the latter compound could also be prepared by a cooxidation of 1-H and the phenol with ceric ion. All of these adducts were rather labile, and poor resolution of 'H NMR spectra obtained for 3c in CDCl, was traced to apparent dissociation to give 1, identified by its characteristic ESR spectrum. Satisfactory, well-resolved lH NMR spectra were obtained in acetone-d, and benzene-d,. The spectra of the adducts showed two kinds of tert-butyl groups (syn and anti) derived from 1, and other resonances in the numbers and positions consistent with the assigned structures. A low-temperature 'H NMR spectrum of the freshly prepared reaction products from p-cresol and 1 in CDC13 showed only resonances ascribed to 1-H, 3a, excess pcresol, and solvent-derived protons. Kinetic ESR studies of the reactions of phenol and substituted phenols under pseudo-first-order conditions showed a trend toward faster reaction rates with electron donation into the aromatic ring (Table 11). In this respect, the radical 1 resembles tert-butylperoxy12band aroyl tert-butyl nitroxidesas
a number of reactions of 2 with phenols, unstable intermediates have been isolated and ascribed variously to C-N and C-0 addition (5a, 5b).13 RON(SO3-)2 RN+(O-)(S03-)2 RN+(O-)=CBu2-t 5a 5b 5c We assume that our addition compounds are oxime ethers and not nitrones (5c) because the latter would engender substantial steric strain from both tert-butyl groups, and because the adducts displayed solubility characteristics that were not at all consistent with the presence of two polar nitrone groups in the molecules. Barton and co-w~rkers'~ found that the Fremy's analogue (PhS)2N'gave a higher yield of product arising from addition to the ortho position of phenol, and they ascribed this result to the higher free spin density at the 2- and 6-positions of the intermediate phenoxyl radical. This explanation is clearly not a general one, since 3c appears to be the only product from 1 and phenol. The reaction of phenoxyl with 1 and 2 appears to respond more to steric factors in giving the less hindered para adduct.I5 Where the reagent 1 distinguishes itself from Fremy's radical and even from organic analogues such as acyl alkyl nitroxidess is in the formation of the unusual geminal dioximes 3c, 3d, and 4. 2,5-Cyclohexadienonessimilar to these are valuable synthetic intermediated, and have been the subject of numerous photochemical studies.17 The preference for these structures over other possible products may reflect a relatively slower rate of 0-N scission in the presumed intermediate 6 vs. coupling with 1 compared \=J-oN=cB"2-t
e
with the analogous reactions of intermediates from phenols and other radicals. The enhanced thermodynamic stability of the 0-C-0 sequence18 may explain why the geminal bis-oxime ethers are preferred over adducts to two sites in the aromatic ring. We have noted that the selectivity of 1 toward H abstraction in related substrates does not follow the order of published C-H bond strengths. These studies are still in progress and will be reported separately.
Discussion Experimental Section The oxidation of hydroquinones with 1 deserves little References to 1 in the following procedures refer to stock comment because this transformation can be accomplished solutions 0.04-0.3M in pentane6 that were stored
