A Novel, Nonaqueous Method for Regeneration of Aldehydes from Bisulfite Adducts

Abstract

Additionally, storage of an aldehyde as the bisulfite adduct gives a significant stability advantage. Yet, the technique appears to be seldom applied. In a search of the Chemical Abstract Services database from 1967 to present we uncovered only 26 papers which describe use of bisulfite adducts in this manner. It is probable that the rarity of the method’s use results from two liabilities: (1) bisulfite adducts will not form with hindered aldehydes, (2) the pH extremes required to regenerate the aldehyde are not tolerated by many aldehydes and other functional groups. We report conditions which broaden the utility of the method by addressing the second concern. The aldehyde 1 is an intermediate in the synthesis of LY231514‚2Na,2 an investigational oncolytic agent.3 Synthesis of 1 in one step from 4-iodobenzoic acid methyl ester and 3-buten-1-ol has been reported by LaRock (Scheme 1).4 This method appeared very attractive to us, if the chemoselectivity could be improved. The LaRock coupling method produces a mixture of ca. 90% 1, 5% 2, and 5% olefins 3 and 4. Clearly, the presence of the undesired side products, particularly the isomeric aldehyde 2, limits the utility of the LaRock method. We were unable to improve the product ratio through optimization of the reaction conditions. Therefore, we sought to develop a commercially viable purification. Purification via an intermediate bisulfite adduct 5 appeared attractive. In fact, 5 crystallizes readily when the reaction mixture is treated with sodium bisulfite in a 2:1 (v:v) ethyl acetate:ethanol mixture containing traces of water. Excess sodium bisulfite must be avoided or competitive precipitation of 6 is unavoidable. The desired purity can then be obtained (at the expense of yield) by increasing the water content. At ca. 10% water content (by Karl Fisher titration) approximately 90% of the available 1 can be reproducibly isolated as 5 with less than 0.5% of the isomeric bisulfite adduct 6. The overall yield for the condensation and bisulfite adduct formation is 75-80%. The difficulty with this approach became apparent when we began to attempt to regenerate 1 from 5. The typical approaches found in the literature involve dissolving the bisulfite adduct in water and treating with either acid5 or base.6 These conditions increase the equilibrium concentration of aldehyde 1, which is then either reacted in water or extracted into an organic solvent. We required extraction into an organic solvent for subsequent chemistry. Therefore, we studied extraction of 1 into methylene chloride as a function of pH.7 These data are presented in Figure 1. Clearly, high pH is needed for efficient formation and extraction of 1. However, these conditions carry a liability which is illustrated by Figure 2. If the layers are separated after 3 min, a 97% yield of 1 in the organic phase can be obtained. However, prolonged stirring rapidly consumes 1 and 5 as the ester moiety is hydrolyzed. On a commercial scale at least 30 min would be required which would be expected to result in less than a 50% recovery. We discovered conditions which avoid ester saponification by elimination of water from the reaction mixture. Treatment of the bisulfite adduct 5 with in excess of 2 equiv of chlorotrimethylsilane (TMS-Cl) in acetonitrile at between 40 and 60 °C rapidly regenerates the aldehyde 1. The balanced chemical equation for this reaction is shown below (eq 2). Components in the equation below have been confirmed by 1H NMR and GC analysis of the liquid phase, FT-IR analysis of the gas phase, and ion chromatography of the solid phase.