Diethyl [3-Cyano-2-Oxo-3-(Triphenylphosphoranylidene)propyl]phosphonate: A Useful Horner-Wadsworth-Emmons Reagent for α-Keto (Cyanomethylene)triphenylphosphoranes from Carbonyl Compounds

Abstract

Recently α-keto amide/ester units have attracted much attention in organic and medicinal chemistry due to the presence of these highly electrophilic units in many bioactive natural products and synthetic peptides. Among a number of synthetic routes reported recently, Wasserman’s approach in which α-keto (cyanomethylene)triphenylphosphoranes 2 are used as the key intermediates according to Scheme 1, stands out for its mild conditions and the outstanding convergence. Although this approach has been widely employed in the synthesis of bioactive compounds including tricarbonyls and human lipase inhibitors by us, there is a shortcoming that the key intermediates 2 can be derived only from carboxylic acids/acid chlorides. In light of that there are abundant carbonyl compounds (aldehydes/ ketones) as natural/synthesized form, it would be highly desirable to use carbonyl compounds for the synthesis of phosphoranes 2. Herein we wish to report the first direct synthesis of phosphoranes 2 from carbonyl compounds utilizing a new Horner-Wadsworth-Emmons (HWE) reagent 4 according to the procedures described in Scheme 2. The requisite HWE reagent 4 was successfully prepared via two-step route: coupling of chloroacetyl chloride with phosphorane 1/BSA (N,O-Bis(trimethylsilyl)acetamide) gave 4-chloro-3-oxo-2-(triphenyl-phosphoranylidene)butanenitrile, which was heated in P(OEt)3 (110 C, 24 h, Ar) to afford a new HWE reagent 4, diethyl [3-cyano-2-oxo-3(triphenylphosphoranylidene)-propyl]phosphonate, in excellent overall yield. A variety of carbonyl compounds were condensed with 4, and the resulting β,γ-unsaturated α-keto cyanophosphoranes 5 were hydrogenated over Pd-C/H2 (1 atm) to α-keto cyanophosphoranes 2 (Table 1). Condensation of simple aryl/aliphatic aldehydes with 4 gave exclusively (E)-olefins in excellent yields (run 1, 2, 4, 5). 2,6-Dimethylbenzaldehyde, however, required longer reaction time obviously owing to the steric hindrance (run 3). The reaction of N-BOC-2-aminoacetaldehyde with 4 is of special interest since it ultimately could afford γ-aminobutyric acid (GABA)-derived α-keto amide/ester units, which have been incorporated into bioactive compounds. Under the standard conditions, N-BOC-2-aminoacetaldehyde furnished 5f in 89% yield (run 6). The condensation of ketones with 4, however, has been found very sluggish and incomplete. In the case of acephenone, 5g was obtained in 69% yield with a ratio (E/Z, 4/6) (run 7). The hydrogenation went straightforward simply by stirring the slurry of 5 and Pd-C (10%, 10-30 wt%) in solvent (THF/MeOH, 1/1) under H2 (1 atm). Simple aryl/aliphatic/ α-aminoacetaldehyde-derived phosphoranes 5 were hydrogenated perfectly in 3 h with Pd-C (10 wt%) (run 1',2',4'-6'). However, sterically hindered aryl aldehyde/ketone-derived phosphoranes 5 required higher loading of Pd-C (30 wt%) and longer reaction time (run 3',7',8'). No detectable byproducts were formed during this step (confirmed by TLC and H NMR), signifying the stability of cyanophosphorane subunit under hydrogenation conditions. Thus, usual workup without chromatography afforded pure phosphoranes 2 in quasi-quantitative yields. In conclusion, we have developed a new synthetic route for β,γ-unsaturated α-keto cyanophosphoranes/α-keto cyanophosphoranes from carbonyl compounds utilizing a new HWE reagent 4. We are currently applying this new synthetic route for heterocyclic/heteroaromatic/chiral αamino aldehydes, and extending the same approach to the synthesis of α-keto alkoxycarbonylphosphoranes from carbonyl compounds utilizing a new HWE reagent having alkoxycarbonylphosphorane subunit.