Benzhydrol, from benzophenone

The preparation of benzhydrol from benzophenone is described as an example of reduction of an aromatic ketone 325... 

Hypochlorites are very good oxidizers of alcohols and are frequently selective enough to oxidize secondary alcohols in preference to primary alcohols see equations 288-291). Solutions of sodium hypochlorite in acetic acid react exothermically with secondary alcohols within minutes [693]. Calcium hypochlorite in the presence of an ion exchanger (IRA 900) oxidizes secondary alcohols at room temperature in yields of 60-98% [76 5]. Tetrabutylammonium hypochlorite, prepared in situ from 10% aqueous sodium hypochlorite and a 5% dichloromethane solution of tetrabutylammonium bisulfate, oxidizes 9-fluorenol to fluorenone in 92% yield and benzhydrol to benzophenone in 82% yield at room temperature in 35 and 150 min, respectively [692]. Cyclohexanol is oxidized to cyclohexanone by teit-butyl hypochlorite in carbon tetrachloride in the presence of pyridine. The exothermic reaction must be carried out with due precautions [709]. 

Barium manganate, prepared from potassium manganate and barium chloride [5JJ] or by the reduction of potassium permanganate with potassium iodide in the presence of barium chloride and sodium hydfoxide [5J2], is used for the quantitative oxidation of benzhydrol to benzophenone. The reaction mixture is refluxed in benzene for 0.5-2 h [SJ5]. The result is comparable with and even better than that of oxidation with manganese dioxide [250, 525]. 

Figure 3.24 (a) Photoacoustic waves of references (solid line) and sample (dashed line) containing benzophenone and benzhydrol in benzene (dotted hne), measured by a 2.25 MHz transducer following excitation by a pulse from a nitrogen laser = 337 nm). (b) Comparison between the calculated and experimental waves when two consecutive exponentials are used to describe this process. The dotted line is the difference between the experimental and calculated waves. The first exponential corresponds to the rapid formation of the triplet state of benzophenone, and the second to the abstraction of a hydrogen atom from benzhydrol by benzophenone triplet. Results obtained with the photoacoustic calorimeter in the Coimbra Chemistry Centre. (Courtesy of Monica Barroso.)... 

The conclusions of Hammond and of Backstrom were confirmed by a flash spectroscopic investigation which permitted direct measurement of the rate constants for triplet decay, k i, = 1X10 sec in benzene, hydrogen abstraction from benzhydrol, fcss = 2 X 10 I mole - sec h triplet excitation transfer from benzophenone to suitable acceptors (Bell and Linschitz, 1963). 

Category 5. Hydrogen Atom Abstraction. When benzophenone is irradiated in isopropyl alcohol, the initially formed Si state crosses to the Ti state, which abstracts hydrogen from the solvent to give the radical 7. Radical 7 then abstracts another hydrogen to give benzhydrol (8) or dimerizes to benzpinacol (9) ... 

A number of different mechanisms have been proposed to account for the fact that this product is not observed. Recently, however, a report appeared that described the formation of the mixed pinacol from the photoreduction of benzophenone with isopropyl alcohol and the photoreduction of acetone with benzhydrol. The data from this study are presented in Table 3.10.(73)... 

We saw earlier that when benzophenone is photoreduced in the presence of optically active 2-butanol, the alcohol recovered from the reaction loses no optical activity/541 This was presented as evidence that there could be no appreciable reversibility of the initial hydrogen abstraction since this should lead to racemization of the unreacted alcohol. However, if one uses relabeled benzhydrol and examines the initially produced benzpinacol for the presence of the label, one finds that the product pinacol contains no 14C. This would indicate that there must be some type of rapid transfer of the hydrogen radical from the ketyl radical produced upon abstraction from benzhydrol,... 

An interesting example of hydrogenation with hydrogen in the absence of transition metal catalyst is reduction of benzophenone to benzhydrol with hydrogen in /er/-butyl alcohol containing potassium /er/-butoxide at 150-210° and 96-135 atm. Although the yields range from 47 to 98% the method is not practical because of its drastic conditions, and because of a cornucopia of more suitable reductions. 

Oxidation of diphenylmethane in basic solutions involves a process where rate is limited by and equal to the rate of ionization of diphenylmethane. The diphenylmethide ion is trapped by oxygen more readily than it is protonated in dimethyl sulfoxide-text-butyl alcohol (4 to 1) solutions. Fluorene oxidizes by a process involving rapid and reversible ionization in text-butyl alcohol solutions. However, in the presence of m-trifluoromethylnitrobenzene, which readily accepts one electron from the carbanion, the rate of oxygen absorption can approach the rate of ionization. 9-Fluorenol oxidizes in basic solution by a process that appears to involve dianion or carbanion formation. Benzhydrol under similar conditions oxidizes to benzophenone by a process not involving carbanion or dianion formation. 

The CT complexes are considered to evolve to excited CT complexes. In Scheme 40, an electron is transferred from the amine to the benzophenone (in the triplet state) forming the CT complex. A proton transfer produces an amine radical 100 and a benzhydrol... 

In a 1-litre three-necked flask, equipped with a reflux condenser, a mechanical stirrer and a thermometer dipping into the reaction mixture, place 50 g (0.275mol) of benzophenone (Expt 6.121), 500ml of rectified spirit, 50g of sodium hydroxide and 50 g (0.76 mol) of zinc powder. Stir the mixture the temperature slowly rises to about 70 °C. After 3 hours, when the temperature has commenced to fall, filter the reaction mixture with suction and wash the residue twice with 25 ml portions of hot rectified spirit. Do not allow the residual zinc powder to become dry as it is flammable. Pour the filtrate into 2 litres of ice water acidified with 100 ml of concentrated hydrochloric acid. The benzhydrol separates as a white crystalline mass. Filter at the pump and dry in the air. The yield of crude benzhydrol, m.p. 65 °C, is 49 g. Recrystallise from 50 ml of hot ethanol and cool in a freezing mixture of ice and salt. Collect the colourless crystals and dry in the air 36 g of pure benzhydrol, m.p. 68 °C, are obtained. Dilute the mother-liquor with water to precipitate the residual benzhydrol, and recrystallise this from a small quantity of hot alcohol. 

Equimolar quantities of benzhydrol and the phosphorane (3a) were also reacted in dimethyl sulphoxide solution. Apart from bis(benzhydryl) ether (18%) and catechol monobenzhydryl ether (39%), a small amount of benzophenone (17%) was obtained. It was shown that, in the absence of phosphorane, benzhydrol is not oxidised to benzophenone by dimethyl sulphoxide. Reaction similar to that outlined in Scheme 1 is a likely possibility. Again, the alcohol is activated by reaction with the phosphorane toward nucleophilic attack, in this case by dimethyl sulphoxide. Significantly, oxidation of alcohols by dimethyl sulphoxide is usually carried out using the Pfitzner-Moffatt reagent (dicyclohexyl carbodiimide and anhydrous phosphoric acid in dimethyl sulphoxide) (13) whereas the reaction using the phosphorane (3a) is carried out under neutral conditions. Unfortunately, however, attempts to improve the yield of benzophenone have hitherto failed. 

Marcus treatment does not exclude a radical pathway in lithium dialkyl-amide reduction of benzophenone. It does, however, seem to be excluded (Newcomb and Burchill 1984a,b) by observations on the reductions of benzophenone by N-lithio-N-butyl-5-methyl-l-hex-4-enamine in THF containing HMPA. Benzophenone is reduced to diphenylmethanol in good yield, and the amine yields a mixture of the acyclic imines no cyclic amines, expected from radical cyclization of a putative aminyl radical, were detected. An alternative scheme (17) shown for the lithium diethylamide reduction, accounts for rapid formation of diphenylmethoxide, and for formation of benzophenone ketyl under these conditions. Its key features are retention of the fast hydride transfer, presumably via the six-centre cyclic array, for the formation of diphenylmethoxide (Kowaski et al., 1978) and the slow deprotonation of lithium benzhydrolate to a dianion which disproportion-ates rapidly with benzophenone yielding the ketyl. The mechanism demands that rates for ketyl formation are twice that for deprotonation of the lithium diphenylmethoxide, and, within experimental uncertainty, this is the case. 

Another route to cyclic peroxides from carbonyl compounds is illustrated by the reaction of dichlorodiphenylmethane with hydrogen peroxide to form the dimeric benzophenone peroxide (77),62 which reacts with zinc in acetic acid to form benzopinacolone (Ph3C—CO— Ph), and with aluminum amalgam to form benzhydrol. On fusion (183-225°), 77 decomposes to form benzophenone. 

In order to assess the importance of initiation by semi-pinacol radicals at a lower temperature, Hutchison et al. (52) compared rat of polymerization observed on photolysis of methyl methacrylate/benzene mixtures containing benzophmone ( 5 x 10 M) alone, and benzo-phenone + benzhydrol ( 0.15 M). The aim was to encourage increased semi-pinacol radical formation in the latter mixtures, since it was anticipated that photoexcited benzophenone would hydrc en-abstract from benzhydrol (rather than methyl methacrylate or benzene) thus giving rise to semi-pinacol radicals from both precursors ... 

Sandner et al (21) observed that addition of small amounts of tri-ethylamine (0.02 M) greatly enhanced the jAoto-induced polymerization of methyl acrylate (1.0 M in tert-butanol, nitrogen-flushed) in the presence of benzophenone (0.02 M). Photoinitiation was not effected by triethyl-amine alone, nor by triphenylamine, isopropanol, or benzhydrol (all 0.02 M) in the presence of benzojAenone. Quantum yidds measured for benzophenone disappearance indicated that methyl acrylate itself acted as a quencher of photoexcited benzophenone, effectively suppressing hydrogen abstraction from terf-butanol, and benzhydrol. How-... 

Apart from these SET reactions, solvent effects in reactions of organomagnesium reagents with carbonyl compounds have been studied rather extensively. The reaction of ethylmagnesium bromide with benzophenone (Scheme 15) in diethyl ether yields 94% of the expected addition reaction product, 1,1-diphenyl-1-propanol, and 6% benzhydrol, resulting from a reduction reaction of the Grignard reagent [36]. In tetrahydrofuran this reaction yields 21 and 77%. respectively, of both products. 

Several new products [e.g. the lactone (307)] were obtained as by-products from remote oxidation of steroids by photochemically excited benzophenone substituents. The benzhydrol asymmetric centre is generated with slight stereoselectivity. Deuterium labelling (at 15a) confirmed that the main product of reaction, the 14-ene (308), results from hydrogen abstraction from C-14 followed by transfer of the 15a-hydrogen to the benzhydryl radical. 

There are, however, other examples. In the presence of potassium tertiary butoxide, quinine and benzophenone are equilibrated with quininone and benzhydrol.37 Benzyl,38 methyl,39 ethyl,39 40 propyl,39 butyl,39 and amyl39 alcohols will reduce ketones to the corresponding secondary alcohols on heating at high temperatures with sodium or potassium hydroxide. In all these cases it is probable that the reaction proceeds by the intermolecular transfer of a hydride ion from the alkoxide ion to the carbonyl group. 

Treatment of benzophenone with 153 in THF gave good yields of benzhydrol and imines 156 and 157 but no trace of cyclic products derived from 155 (formed via oxidation of 153 to give 154). The anion 153 does not cyclize under the reaction conditions. Cyclic products would be expected if 153 had been oxidized to radical 154 since 154 cyclizes 111). This and other experiments led to the conclusion that the reduction of benzophenone by dialkyl amides containing P-hydrogen atoms occurs via hydride and not via electron transfer. 

Interestingly, the discovery of imine anions as potentially useful synthetic intermediates was made serendipitously by Wittig approximately 30 years ago. Acting upon an earlier discovery that lithium diethylamide may serve as a hydride donor to benzyne, Wittig examined the reaction of lithium diethylamide with benzophenone to ascertain whether the latter might also serve as a hydride acceptor. Surprisingly, the products observed from this reaction were benzhydrol (7) and the p-hydroxy aldimine (8) rather than the expected oxidation product (9). The sequence of reactions depicted in Scheme 2 was then proposed to account for this unexpected transformation. The initially produced aldimine (9) undergoes deprotonation with a second equivalent of lithium diethylamide to provide the imine anion (10),... 

---