New Chiral Sulfur and Phosphorous Chemistry from Boehringer Ingelheim

Chiral heteroatom-containing compounds have become highly useful reagents for the enantioselective synthesis of important building blocks and intermediates used for the preparation of pharmaceutical compounds. Stephen Han, Chris Senanayake, and other coworkers from the Department of Chemical Research and Development at Boehringer Ingelheim Pharmaceuticals in Ridgefield, Connecticut, recently reported on the development of new and more efficient routes to two types of important sulfur- (1) and phosphorous-based reagents (2).

Need for a new chiral sulfinyl transfer agent

Chiral sulfinamides are widely used for the synthesis of chiral amine-containing compounds, and they also serve as useful chiral ligands for catalytic asymmetric transformations. The number of routes to these useful compounds is limited, and few methods offer the accessibility for structurally diverse sulfonamides (1). The researchers reported that their goal was to develop a route for the synthesis of hindered sulfinamides on a large scale in a practical manner (1). Previously, the researchers at Boehringer Ingelheim had developed a versatile method for the preparation of sulfinamides by means of cyclic-oxathiozolidinone-based chiral sulfinyl transfer agents. Even though this method was used to prepare many hindered substrates, it suffered from the need to use excess NH₂Li/NH₃ (Li/NH₃), which is prepared in situ using a large excess of solid Li metal and anhydrous NH₃ and presents both safety and waste generation issues (2-4) Therefore, the group set out to develop an efficient, cost-effective, and practical method for the synthesis of hindered sulfinamides (1).

They reported that the first step was to replace NH₂Li/NH₃ with a safer and greener reagent, such as lithium bis(trimethylsilyl)amide (LHMDS), for the final S-O bond cleavage. In order to do so, they needed to increase the reactivity of the S-O bond, which meant developing a new template (1). To investigate the reactivity of the S-O bond, the researchers reported that various tert-butanesulfinates were reacted with LHMDS. Not surprisingly, the reactivity correlated well with the electronegativity of the oxygen, which was related to the basicity of the leaving group (RO-). Thus, sulfinates derived from phenol, particularly those bearing electron-withdrawing substituents, reacted very quickly and quantitatively. Importantly, it was also noted that the reaction proceeded in an enantioselective manner (1).

Based on these preliminary studies, the researchers reported that optically pure p-chlorobenzothiazine 2-oxide was prepared as a new chiral sulfinyl transfer agent on a large scale (2 kg) in 87% yield and 99.6:0.4 diastereomeric ratio (d.r.) (S configuration at sulfur) from the corresponding commercially available chiral aminophenol, SOCI₂, and pyridine at 10 °C. (1).

Preparation of hindered, enantiopure sulfinamides

The researchers reported that the reactivity of the transfer agent was first tested by treating it with tert-butylmagnesium chloride in tetrahydrofuran (THF) to cleave the S–N bond to generate a diastereomerically pure, stable, crystalline sulfinate ester, to which was subsequently added LHMDS to cleave the S-O bond to provide enantiopure tert-butylsulfinamide ((R) or (S)-tBSA) in high yield and high enantiomeric excess, even when the two steps were run at 10 °C and 0 °C, respectively (1). This reaction was also run on a 100-g scale (15 °C and 5 °C). After the two steps, the reaction mixture was quenched with phosphoric acid to provide a mixture of the starting chiral aminophenol and (R)-tBSA. The template was crystallized from water and recovered in 99% yield while the product was extracted using methylene chloride and obtained in 85% yield and 99.5:0.5 enantiomeric ratio (e.r.) (1).

The researchers reported that the reaction is applicable to the synthesis of a wide variety of structurally diverse and sterically hindered alkyl and aryl sulfonamides. In some cases, a lower temperature was required to maintain the very high enantioselectivty. In addition, they found that the ester intermediate does not need to be isolated, and a one-pot process is effective (1).

Extension to hindered, chiral sufinyl ketimines

Furthermore, in the same study, the researchers applied the methodology to the preparation of related sterically hindered chiral sufinyl ketimines, which are key intermediates for chiral amines and typically prepared from ketones and sulfinamides in the presence of a Lewis acid. Yields for sterically hindered ketones with this reaction are, not surprisingly, low. With the new chiral sulfinyl transfer agent, however, the researchers reported that imine nucleophiles were added to the sulfinate ester intermediate, making it possible to access hindered products (both alkyl and aryl) in good yields and high enantiomeric purities (1).

The researchers then investigated the value of the chiral sufinyl ketimines as substrates for the preparation of chiral amines. They reported that reduction of the ketimines with NaBH₄ in THF proceeded in very high yields with high selectivity (diastereomeric ratios > 90:10). The sulfinyl group was also found to clearly influence the selectivity, and the researchers are investigating further the preparation of different chiral amines under different reaction conditions (1).

Wanted: structurally diverse P-chiral phosphines

Chiral phosphines are widely recognized as important ligands for asymmetric transition-metal catalysts and as effective organocatalysts for enantioselective transformations. The researchers, however, noted that there are few efficient methods for preparing optically active P-chiral phosphorous compounds, and those that are used are generally limited to nonsterically hindered substrates (5).

Recently, Han and his colleagues at Boehringer-Ingelheim reported on the successful development and use of MeO–BIBOP, MeO–BOP, and P-chiral biaryl ligands in a number of important asymmetric reactions, such as propargylations (6), hydrogenations (7), and Suzuki-Miyaura couplings (8). The ligands were all prepared from a bulky P-chiral phosphine oxide that, at the time of their research could only be obtained using a resolution-based synthetic route. Han, Senanayake, and coworkers, therefore, set out to develop a general method for the efficient and stereoselective synthesis of P-chiral phosphine oxides with diverse structures and functionalities (5).

Another chiral transfer agent

Taking a page from their research notebook on the chiral sulfinyl transfer agent, the researchers focused their efforts on developing a chiral phosphinyl transfer agent based on the 1,3,2-benzoxazaphosphinine-2-oxide structure (5).

The researchers reported that the challenge was to overcome the low reactivity of P–N and P–O bonds. This goal was achieved by activating the P–N bond (increasing its eletrophilicity) with an arylsulfonyl group (specifically a tosyl group) on the N atom, and activating the P–O bond with a phenol derivative that could act as an effective leaving group (as was the case with the sulfinyl transfer reagent). The researchers noted that the design also ensured that the P−N and P−O bonds have differentiated reactivity toward nucleophilic substitutions, thereby enabling their sequential cleavage with two different organometallic reagents. They then described how the desired (R)-1,3,2-benzoxazaphosphinine-2-oxide derivative was prepared from an intermediate aminophenol-based auxiliary that was easily synthesized on a large scale from the readily available compound (R)-2-(1-aminoethyl)-4-chlorophenol.  The transfer agent was obtained in high yield via addition of 1-methylimidazole (1-MeIm) to the chlorophenol in dichloromethane with a diastereomeric ratio of 98:2, and was then recrystallized to give the pure isomer. The yield on a 100-g scale was 85% (5).

Synthesis of bulky P-chiral phosphines

Han et al. reported that the chemoselective cleavage of the P–N bond proceeded upon treatment with Grignard reagents or t-BuLi in THF at low temperature (-20 °C) (5). This reaction occurred with inversion of configuration at phosphorous and generated diasteromerically pure (S)-phosphinates, including t-butyl and mesityl derivatives, in moderate to good yields. When 4 equivalents of the Grignard reagent 2-MeO-PhMgBr were used, the reaction was complete in 15 minutes at -10 °C (5). The researchers reported that cleavage of the P–O bond proceeded using a variety of reagents to generate a wide variety of (R)-chiral phosphine oxides. While bulky Grignard reagents (i-PrMgBr and t-BuMgCl) reacted slowly, addition of the corresponding lithium compounds proceeded in good yields with good enantioselectivity. Primary alkyl and vinyl Grignards were also effective, as were alkynyl and ferrocenyl lithium reagents. Chiral versions of several Buchwald-type ligands were also effectively prepared. It should also be noted that the free phosphine ligands could be obtained in excellent yield without erosion of the enantiopurity via reduction with polymethylhydrosiloxane (PMHS) in the presence of Ti(OiPr)₄ in THF (5).

Demonstration of utility

The researchers asserted that numerous types of functionality can be incorporated into the chiral phosphine oxides prepared using this new method, making it possible to synthesize a wide variety of interesting chiral phosphines for use as ligands and/or organocatalsyts (5). As an example, the researchers reacted a vinyl-substituted chiral phosphine oxide with dialkyl- or diarylphosphine−borane adducts to generate enantiomerically pure bisphosphine compounds in high yield (5).The new approach was also used to prepare the bulky P-chiral phosphine oxide from which Boehringer-Ingelheim’s new P-chiral ligands were derived as earlier described. In this case, a different (R)-1,3,2-benzoxazaphosphinine-2-oxide derivative was first prepared by reacting the chiral template with dichloro(2,6-dimethoxyphenyl)phosphine followed by H₂O₂ oxidation. Addition of t-BuLi to this compound followed by reaction with MeLi/MgBr₂-OEt provided the desired chiral phosphine oxide with a 98.3:1.7 er. The researchers asserted that this new method is a practical, stereoselective, and high-yield route to P-chiral phosphine oxides, which are of interest as intermediates for preparing P-chiral phosphine ligands as well as applicability for novel chiral ligand structures for asymmetric synthesis (5).

References

1. Z. S. Han et al., Angew. Chem. Int. Ed. 52 (26), 6713-6717 (2013).
2. Z. Han et al., J. Am. Chem. Soc.124 (27), 7880 – 7881 (2002).
3. Z. Han et al., Tetrahedron 61 (26), 6386 – 6408 (2005).
4. Z. S. Han, et al., J. Org. Chem. 76 (13), 5480 – 5484 (2011).
5. Z. S. Han et al., J. Am. Chem. Soc. 135 (7), 2474−2477 (2013).
6. Fandrick, D. R. et al., J. Am. Chem. Soc. 132 (22), 7600 (2010).
7. . (b) Tang, W. et al., Org. Lett. 12 (5), 1104 (2010).
8. Tang, W. et al., Org. Lett. 14 (9), 2258 (2012).