Asymmetric Synthesis of Tertiary Alcohols and Thiols via Nonstabilized Tertiary α‐Oxy‐ and α‐Thio‐Substituted Organolithium Species

Abstract Nonstabilized α‐O‐substituted tertiary organolithium species are difficult to generate, and the α‐S‐substituted analogues are configurationally unstable. We now report that they can both be generated easily and trapped with a range of electrophiles with high enantioselectivity, providing ready access to a range of enantioenriched tertiary alcohols and thiols. The configurational stability of the α‐S‐organolithium species was enhanced by using a less coordinating solvent and short reaction times.

lithium compounds are av ersatile class of nucleophiles that are useful in the asymmetric synthesis of chiral alcohols, amines,a nd thiols. [1] Although the use of secondary and mesomerically stabilized (e.g., benzylic and allylic) tertiary a-O-and a-S-substituted organolithium reagents in synthesis is well established, [1c-e] the use of non-mesomerically stabilized (i.e., dialkyl-substituted) tertiary reagents is not. This discontinuity is due to contrasting problematic features governing a-O-and a-S-substituted organolithium species. Nonstabilized tertiary a-O-organolithium compounds are configurationally stable but are difficult to generate owing to their reduced kinetic acidity (Scheme 1B); [1e, 2] tertiary a-Sorganolithium species are easily formed but are not configurationally stable (Scheme 1A). To apply a-O/S-substituted organolithium compounds in asymmetric synthesis,both ease of generation and configurational stability are essential requirements. [1a-d,3] Through variation of the directing group,b ase,s olvent, and additives we discovered reaction conditions enabling the stereospecific deprotonation of secondary dialkyl benzoates (TIB esters) [4] and demonstrated their configurational stability in lithiation-borylation reactions. [5] Herein, we report the broad applicability of these novel enantioenriched nucleophiles in reactions with ab road range of electrophiles.I n addition, we have discovered reaction conditions for the generation of enantioenriched, tertiary,n onstabilized a-Sorganolithium compounds and report their subsequent trapping with electrophiles in high enantioselectivity (Scheme 1C).
Boronic esters represent an iche class of electrophiles, [6] and therefore,w ei nitially embarked on as tudy of the trapping of tertiary a-O-substituted organolithium species, which were generated by lithiation of the enantioenriched TIB esters 1a-1e,w ith ar ange of electrophile classes (Scheme 2). Upon exposure of av ariety of enantioenriched benzoates to sec-BuLi and TMEDAinCPME at À60 8 8C, the corresponding organolithium species Li-1a-1e were generated. Pleasingly,L i-1a-1e were successfully trapped with ar ange of electrophiles,i ncluding methyl chloroformate (2aa), benzoyl chloride (2ab), isocyanates (2ac and 2ad), aldehydes (2ae and 2af), and trialkyl tin chlorides (2ag, 2ah, 2ba-2ea). Reactions with aldehydes gave mixtures of diastereomers (see 2ae and 2af)b ut in the case of PhCHO,h igh Scheme 1. A, B) Previous studies regardingnonstabilized tertiary a-Sand a-O-substituted organolithium compounds.C)This work, reporting their straightforward generation,c onfigurational stability,a nd electrophilic trapping. CPME = cyclopentyl methyl ether,TBME = tertbutyl methyl ether,TMEDA = tetramethylethylenediamine. diastereoselectivity was observed (11:1 d.r.). In all cases,t he desired tertiary alcohol derivatives 2 were obtained in good to high yields and, importantly,w ith universally complete enantioselectivity and retention of configuration, as determined by X-ray crystallographic analysis of 2ad and 2af (see the Supporting Information). In addition, the TIB group could be easily removed upon reduction with LiAlH 4 to give the tertiary alcohol 3.
Having demonstrated the scope of the electrophilic trapping of nonstabilized, tertiary a-O-substituted organolithium intermediates,w ec onducted studies to evaluate whether stannanes 2ag and 2ah could serve as bench-stable organolithium precursors by tin-lithium exchange (Scheme 3). [7] However, treatment of the tributyltin derivative 2ag with n-BuLi with or without TMEDAf ollowed by quenching with CH 3 OD only gave [D]-1a with poor conversion, albeit with excellent stereoselectivity (with retention of configuration, see entries 1and 2). In contrast, treatment of the less hindered trimethyltin derivative 2ah with n-BuLi/ TMEDAw as much more successful and yielded [D]-1a in excellent yield and with complete stereoselectivity and retention of configuration (entries 3a nd 4), indicating that 2ah could indeed serve as au seful precursor to the corresponding organolithium species. [8] An understanding of the configurational stability of chiral organolithium species is crucial for exploiting their use in synthesis. [3,9] Interestingly,t he tertiary a-O-substituted organolithium species Li-1 were found to be chemically and configurationally stable below À40 8 8Cb ut at higher temperatures,d ecomposition rather than racemization occurred. Similar observations have been made with secondary a-Osubstituted organolithium TIB esters whilst the corresponding carbamates are slightly more stable and only decompose above À20 8 8C. [2b, 7c, 10] All of these nonstabilized a-O-organolithium species decompose before they racemize.
Having demonstrated the broad applicability of tertiary nonstabilized chiral a-O-organolithium species,w et hen embarked on as tudy of the more challenging sulfur analogues. [1c, 11] Pioneering work by Beak had revealed that while the a-deprotonation of dialkyl-substituted tertiary thiobenzoates (such as 4)w as facile at low temperature,t hey were configurationally unstable even at À98 8 8Ci nT HF (Scheme 1A). [12] Only tertiary or hindered, branched a-S-substituted organolithium compounds (e.g., 5a-5c)h ave been reported to be configurationally stable; [13] all others are unstable (Scheme 4A). [14] This can be explained based on the mechanism of racemization of a-S-substituted organolithium compounds,w hich involves solvent separation of the ion pair, rate-determining rotation of the hyperconjugated C À S bond, and recombination of the ion pair (Scheme 4B). [15] Thus configurational stability in a-S-organolithium compounds is only observed with hindered substrates where there is ah igh barrier to CÀSb ond rotation. Given this observation, we proposed that nonstabilized, tertiary a-Ssubstituted organolithium species derived from thiobenzoates 4 should be configurationally stable and therefore re-examined the conditions for lithiation.
were lithiated under av ariety of reaction conditions and subsequently reacted with CH 3 OD (Scheme 5). Using THF as the solvent gave the product [D]-4a as ar acemate (entry 1), thus confirming Beakso bservations.G ratifyingly,t he use of Et 2 Oa st he solvent generated [D]-4a in 90:10 e.r. and complete conversion (entry 2). TMEDAw as crucial to facilitate deprotonation as in its absence,[ D]-4a was formed only in low yield (entry 3). [16] TBME was found to be the most suitable solvent (entry 4), and even av ery short deprotonation time (5 min) was sufficient to give the desired product [D]-4a with 100 %c onversion and 97:3 e.r. with retention of configuration (entry 5). These results show that the racemization of the tertiary a-S-substituted organolithium species can be minimized when less coordinating solvents and short reaction times are employed, reaction conditions that maintain atight ion pair.
Having identified the most suitable reaction conditions for the a-deprotonation of STIB ester 4 and trapping with CH 3 OD,weevaluated the scope of electrophiles that could be employed. Thes ame range of electrophiles that were compatible with the lithiated OTIB esters Li-1a-1e were also successful in trapping nonstabilized, tertiary a-S-substituted organolithium species derived from 4a-4e,and gave the tertiary thiol derivatives 6 in good yield and very high enantioselectivity in all cases with the exception of Si-and Snbased electrophiles (6ae-ag,S cheme 6; see below). In the case of ClSnMe 3 (6ag), the predominant enantiomer arose from retentive addition to the organolithium (S E 2ret), whereas for ClSnBu 3 (6af), inversion was observed (S E 2inv). [17] We extended the method further by preparing the STIB cholesterol derivative 8,w hich upon exposure to the optimized reaction conditions gave 9 (ClCO 2 Me quench) in excellent yield and as asingle diastereoisomer (Scheme 7). [18] This example highlights two key features in the generation of nonstabilized, tertiary a-S-substituted organolithium intermediates under our conditions.1 )Although the Li atom adopts an equatorial position in Li-8,a no rientation that according to Beak [12] and Reich [19] will favor epimerization by forming the more stable configuration with an axially positioned lithium (where the bulky TIB group is placed at an equatorial position), epimerization was not observed, underscoring the remarkable configurational stability of nonstabilized tertiary a-S-substituted organolithium species generated under our conditions.2 )The kinetic acidity of the a-S-proton in STIB esters 4 and 8 is remarkably high and outcompetes the lithiation of the allylic position.
[a] Determined on the crude reaction mixture by HPLC analysis on achiral stationary phase.
p-Br-C 6 H 4 -NCO gave carbamate 6ab with different levels of enantioenrichment (see the Supporting Information). The erosion of e.r. observed in these experiments gave rise to an Eyring plot from which the parameters DH°and DS°were determined to be + 13 kcal mol À1 and + 14 cal mol À1 ,r espectively [21] (Scheme 8B). This equates to a DG°r ac(-78 8 8C) value of + 10 kcal mol À1 ,w hich is in line with Hoffmannsr esults. Aside from providing the thermodynamic parameters for racemization, these studies also showed that racemization of Li-4 occurred at À78 8 8Co ver an extended period of time, which is in contrast to the behavior of oxygen analogues Li-1, which do not racemize.
Finally,w ed ecided to investigate the factors that are responsible for the low selectivity observed in the reaction of nonstabilized, tertiary a-S-substituted organolithium species with Si-and Sn-based electrophiles (Scheme 6, 6ae-ag). We speculated that as low electrophilic quench might have resulted in partial racemization of Li-4 and therefore monitored the reaction by in situ IR spectroscopy (Scheme 9). [22,23] With the test electrophile ClCO 2 Me,w here high e.r. values were observed, the in situ IR studies revealed that both lithiation of 4a and subsequent electrophilic quenching were extremely rapid at À78 8 8C.
In the case of Me 3 SiCl, the electrophilic trapping was slow with only 50 %o ft he organolithium compound quenched after 1h,t hus leaving time for competing racemization. [24] However,i nt he case of the Bu 3 SnCl and Me 3 SnCl electrophiles (6af and 6ag), where poor e.r. values were observed, the IR profile revealed that quenching of Li-4a was rapid, as in the case of ClCO 2 Me.This excluded racemization of Li-4a as the sole cause of the poor e.r. observed. These unusual findings can be rationalized on the basis of competing retentive (S E 2ret) and invertive (S E 2inv) electrophilic substitution pathways in the quenching of Li-4a with tin electrophiles, [25] where the more hindered Bu 3 SnCl electrophile favors the invertive pathway.
In conclusion, we have found that difficult-to-generate, nonstabilized, tertiary, a-O-substituted lithiated secondary dialkyl benzoates (OTIB esters) and previously regarded as configurationally unstable a-S-substituted lithiated secondary dialkyl thiobenzoates (STIB esters) can be generated, and are configurationally stable.Key to success in both cases was the use of mildly coordinating solvents together with TMEDAto enable deprotonation and, in the case of the a-S-organolithium species,short reaction times.The subsequent trapping of these rare,nonstabilized tertiary organolithium intermediates with electrophiles proceeded with excellent enantioselectivity,e nabling the synthesis of highly enantioenriched tertiary alcohol and tertiary thiol derivatives.T herefore,w e have established an ew class of organolithium reagents for asymmetric synthesis.