We herein report the successful synthesis of beta-hydroxy nitriles in very good to excellent yields from aldehydes and ketones in a simple reaction that is promoted by strong nonionic bases of the title type. The reaction occurs in the presence of magnesium salts which activate the carbonyl group and stabilizes the enolate thus produced.

Herein we report a very effective and mild procedure for the silyl protection of a wide variety of substrate alcohols, including primary, secondary, allylic, propargylic, benzylic, hindered secondary, tertiary, acid-sensitive, and base-sensitive alcohols and also hindered phenols. The silylation reagent used is tert-butyldimethylsilyl chloride (TBDMSCl) and the catalyst is P(MeNCH2CH2)3N, 1b, both of which are commercially available. The reactions are carried out in acetonitrile from 24 to 40 °C and on rare occasions in DMF from 24 to 80 °C. The effect of solvent, catalyst concentration, and temperature and reaction time on the silylation of alcohols and the excellent compatibility of our method with a variety of functional groups is discussed. An efficient method for recycling the catalyst is also presented. Although representative primary alcohols, secondary alcohols, and phenols were silylated using the more sterically hindered reagent tert-butyldiphenylsilyl chloride (TBDPSCl) in the presence of 1b as a catalyst, tertiary alcohols were recovered unchanged.

The symmetric active-methylene compounds CH2(CO2Et)2 and CH2[C(O)Me]2 are selectively monoalkylated in the presence of 1.1 equiv of a variety of alkyl halides and 1 equiv of the nonionic superbase P(MeNCH2CH2)3N in 85−98% yields in 30 min at room temperature. The unsymmetrical active-methylene compound EtO2CCH2C(O)Me is selectively monoalkylated under the same conditions, except for the temperature, which is 0 °C, in 59−88% yields. The observation of selective C- rather than O-alkylation is rationalized in terms of the formation of an enolate whose negatively charged oxygen is sterically protected by a nearby HP(MeNCH2CH2)3N+ counterion in a tight ion pair.

The dimerization of methyl acrylate to the head-to-tail 2-methylene-pentanedioic acid dimethyl ester product was realized in 82 and 85% yield in only 4 h at room temperature in THF in the presence of catalytic amounts of P(RNCH2CH2)3N (R = i-Bu and Bn, respectively). These phosphines are to our knowledge the best nonmetallic catalysts so far reported for this reaction. In contrast, less sterically hindered P(MeNCH2CH2)3N failed to catalyze this dimerization, giving oligomer or polymer instead.

Thermal cyclization of arylhydrazones in the absence of acid catalyst has been investigated. Indole was not obtained from acetaldehyde phenylhydrazone; aniline, N-ethylaniline, and a n i~nidentihed conlpound were isolated. Phenylhydrazones of cyclohcsanone ant1 cyclopentanone and their 2-substituted derivatives were cyciized, often in good yield, to the indole and indolenine; the nature of the substituent inarlcedly influences the ratio of the two products. Gaslicluid chronlatographic analysis of the reaction mixture from indolization of two arylhydrazones revealed no crossed products. Nitrophe~lylhydrazo~ies do not cyclize in good yield, but thernlal indolization of some pyridylhydrazones is a good niethod for the preparation of pyrrolopyridines.One of the most useful methods available for the preparation of indoles is the acidcatalyzed cyclization of phenylhydrazones discovered by Fischer (1). RIIany Brdnsted and Lewis acids have been used as catalysts. T h e mechanisnl of the reaction was investigated by Robinson and the proposed intralllolecular cyclization process has been accepted by later worlrers, with some slight modification. A review of the mechanism and some aspects of the synthetic utility of the reaction has appeared recently (2).Early worlr led t o the idea that a n acid catalyst and a high temperature were essential for a successful Fischer indole synthesis. However, later n-orlr (3) has shown that the acid may assist in the first stage of the reaction, i.e. the tautomerization of the hydrazone t o the enehydrazine, but is not generally essential for this or later stages of the cyclization. T h e early ideas may explain why an observation (4) that distillation of acetophenone phenylhydrazone gave some 2-phenylindole was ignored, until Fitzpatriclr and Hiser ( 5 ) reported several phenylhydrazones were converted into t h e corresponding indoles on boiling a solution of the phenylhydrazone in a suitable high-boiling solvent. T h e reaction occurred in the absence of any acid catalyst and, indeed, in one case, in the presence of sodiuin hydroxide. T h e exainples quoted by Fitzpatriclr and Hiser were somewhat linlited in variety. Since the reaction has obvious interest and application due to the experilnental simplicity, we have investigated this thermal indolization reaction.Indole cannot be prepared by the acid-catalyzed cyclization of acetaldehyde phenylhydrazone, though propionaldehyde and higher aldehyde phenylhydrazones do give the 3-allcylindoles (2). Fitzpatriclc and Hiser (5) report t h e thermal cyclization of propionaldehyde @-tolylhydrazone but malre no mention of attempts t o cyclize acetaldehyde phenylhydrazone. Our efforts to obtain indole in this way were unsuccessful, though ammonia was evolved from the reaction mixture. T h e product consisted of a steam-volatile fraction and a nonvolatile black residue. Fractional distillation of the volatile nlaterial yielded N-ethylaniline, aniline, and a n unidentified component. Since our experiments were completed, Robinson (2) has r...