A concise and stereoselective synthesis of (+/−)-erythro-methylphenidate

A concise and stereoselective synthesis of racemic erythro-methylphenidate (1) is described. The coupling reaction between piperidine-2-thione (3) and 2-bromo-2-phenylmethylacetate (4) afforded the -enaminocarbonyl compound 2 in 60% yield by a modified Eschenmoser sulfide contraction reaction. In most cases the bicyclic thiazolidinone 5 was produced. Diastereoselective reduction of 2 in the presence of borohydrydes furnished the (+/−)-erythro-methylphenidate in good yields with dr >95%. © 2003 Elsevier Science Ltd. All rights reserved. The (+/−)-threo-methylphenidate hydrochloride (Ritalin), is an indirect catecholamine agonist, and is the drug treatment of choice for the attention deficit/ hyperactivity disorder (AD/HD), one of the most common behavioral disorders of childhood which affects 5–10% of the general population. The activity of Ritalin promotes an enhancement of the cognitive performance in both adults and children diagnosed with AD/HD. Although Ritalin is available in the market as the racemate, the (2R,2 R)-(+)-threo-methylphenidate has been reported to be up to 38 times more active than (2S,2 S)-(−)-threo-methylphenidate and several racemic and chiral synthesis of threo-methylphenidate were reported previously. Although the erythro-methylphenidate exhibits very few therapeutic properties and toxic hypertensive effects, the stereoselective synthesis of the erythro-isomer can be helpful due to the possibility of resolution and epimerization of the erythro-isomer. This approach was used early in the original synthesis of (+/−)-threomethylphenidate. Close to this, an enantioselective synthesis of (2S,2 S)-erythro-methylphenidate was recently published by Prashad and co-workers. Due to this fact and relating to our interest in the synthesis of -aminocarbonyl compounds we would like to communicate in this paper our approach to the synthesis of (+/−)-erythro-methylphenidate 1 applying a new concise stereoselective synthetic strategy. We envisaged obtaining 1 based on a stereoselective reduction of a -enaminocarbonyl compound 2 which is readily accessible by the Eschenmoser sulfide contraction reaction (Scheme 1). The needed piperidine-2-thione (3) was prepared in 91% yield after reaction of piperidine-2-one with Lawesson reagent. The bromoester 4 was obtained by sterification and subsequent -bromination of 2-phenylacetic acid in 71% overall yield after purufication by horizontal destilation at reduced pressure. Next, the condensation reaction of compounds 3 and 4 was carried out under the Eschenmoser sulfide contraction reaction Scheme 1. The retrosynthetic analysis of 1 based on the eschemoser sulfide contraction reaction. * Corresponding author. Tel.: +55-51-33166284; fax: +55-51-33167304; e-mail: [email protected] 0040-4039/03/$ see front matter © 2003 Elsevier Science Ltd. All rights reserved. doi:10.1016/S0040-4039(03)00420-9 D. Russowsky, B. A. da Sil eira Neto / Tetrahedron Letters 44 (2003) 2923–2926 2924 protocol to provide the key intermediate -enaminocarbonyl compound 2 (Table 1). The classical conditions to carry out this reaction generally employ CH2Cl2 as the solvent, Et3N as the base and excess of Ph3P as the sulfur scavenger. Other tertiary amines and phosphorous thiophiles are reported in the literature as being effective, as well as reaction performed in the absence of a thiophile. While trisubstituted -enaminocarbonyl compounds are readily prepared starting from thiolactams and -bromoesters under the usual protocol, the tetrasubstituted derivatives, such as 2, require carefully controlled conditions to avoid undesired side reactions. The main examples in the literature to prepare the tetrasubstituted -enaminocarbonyl compounds are limited to the employment of a secondary -triflates or -bromocarbonyl reagent with a second electron withdrawing group (CN, COR, CO2R, NO2) attached to the -carbonyl carbon. In our first attempt to carry out the condensation of compounds 3 and 4 by the usual Eschenmoser protocol we were able to isolate the -enaminocarbonyl compound 2 in 42% yield along with 20% yield of a second product which was identified as the bicyclic thiazolidinone 5 (Scheme 2). The products 2 and 5 were separated and purified by column chromatography and their H and C NMR and IR spectra are in accordance with the proposed structures. The formation of a thiazolidinones was recently reported in the literature by Padwa and Michael for analogous reactions and the factors controlling its formation are not yet clear. The geometry of the double bond in 2 was shown to be Z by the analysis of the NOE experiment: irradiation of allylic protons of the piperidine ring at 2.10 showed a 5.30% increment in the orthohydrogens of the phenyl ring. Attempting to achieve the best yields of compound 2 and to avoid the formation of thiazolidinone 5, we carried out the reaction under various different conditions employing three different thiofiles and a set of bases. The results are summarized in Table 1. We observed that different amounts of Ph3P as the thiophile in the reaction using Et3N as the base did not produce a significative change in the proportion of 2 and 5 and the yields were essentially the same (Table 1, entries 1–3). The use of n-Bu3P caused a small decrease in the yield of thiazolidinone 5 (Table 1, entry 4) and no products were observed employing (EtO)3P as the thiophile (Table 1, entry 5). Changing from Et3N to DBU (Table 1, entry 6) caused a decrease in the formation of product 2 and an increase in the yield of 5. With these initial results, we decided to monitor the reaction course by GC analysis. In the first hour of reaction, it was possible to detect the formation of compounds 2 and 5 and partial consumption of piperidine-2-thione, even in the absence of base and thiophile. We supposed that a competitive side reaction would form the thiazolidinone 5 from the initially formed thioalkylated salt 6 (pathway B) and the addition of a base could enhance the yield of compound 2 (pathway A) by the current accepted mechanism of the Eschenmoser sulfide contraction reaction (Scheme 3). Table 1. Reaction conditions and yields of compounds 2 and 5 via Scheme 2 Thiophile (no. equiv.) Entry Base (2.0 Yield (%)