Isothiourea‐Catalysed Acylative Kinetic Resolution of Aryl–Alkenyl (sp2 vs. sp2) Substituted Secondary Alcohols

Abstract The non‐enzymatic acylative kinetic resolution of challenging aryl–alkenyl (sp2 vs. sp2) substituted secondary alcohols is described, with effective enantiodiscrimination achieved using the isothiourea organocatalyst HyperBTM (1 mol %) and isobutyric anhydride. The kinetic resolution of a wide range of aryl–alkenyl substituted alcohols has been evaluated, with either electron‐rich or naphthyl aryl substituents in combination with an unsubstituted vinyl substituent providing the highest selectivity (S=2–1980). The use of this protocol for the gram‐scale (2.5 g) kinetic resolution of a model aryl–vinyl (sp2 vs. sp2) substituted secondary alcohol is demonstrated, giving access to >1 g of each of the product enantiomers both in 99:1 e.r.


Introduction
Non-enzymatic, acylative kinetic resolution (KR) is ap owerful methodf or the preparation of enantiomerically enrichedalcohols. [1] In this regard, enantioselective Lewis base-catalysed acylationsa re one of the most widely employed methodologies, and variousc atalyst structures and acyl transfer agents have been developed. In terms of substrate scope, non-enzymatic acylative KRs are mostcommonly trialed on benzylic secondary alcohols for which the catalytic acylating agentm ust differentiate between the enantiomerso fa lcohols bearing ap lanar aryl (sp 2 )a nd at etrahedral alkyl (sp 3 )s ubstituent in order to obtain high selectivity (Figure 1a).
To date there are very few examples of the KR of secondary allylic alcohols bearingb oth planar alkenyl and planar aryl sub-   stituents (sp 2 vs. sp 2 ). [8] This is likely to be due to the challenge of the catalytic acylating agentd ifferentiating between enantiomeric alcohols with two planar sp 2 hybridized substituents during the selectivity-determining acylation step. To this end, Connona nd co-workersh ave studied the KR of ar ange of Morita-Baylis-Hillman (MBH) adducts 8 bearing aryl substituents, obtaining moderate selectivity (S up to 13) using chiral DMAP derivative 3 and isobutyric anhydride (Scheme 1a). [9] Mandai and Suga have also reported as ingle example of the KR of an aryl MBH adduct using ac hiralp hosphoric acid catalyst alongside acetyl chloride and DABCO (1,4-diazabicyclo[2.2.2]octane). [10] Deng and co-workersh ave used amidine 7 as ac atalyst for the acylative KR of aryl-alkenyls ubstituteda lcohols 10,w ithm oderate to good selectivity (S up to 24) obtained for ar ange of aryl substituents and simple 1,1-disubstituted alkenes (Scheme 1b). [11] Herein, the challenge of resolving aryl-alkenyl( sp 2 vs. sp 2 ) substituted secondary alcohols is addressed using an isothiourea-basedo rganocatalyst (Scheme 1c). [12,13] Isothioureas have previously been used as catalysts for the acylativeK Ro fv arious secondary alcohols, [14] as well as the desymmetrization of meso-diols. [15] In this report, we demonstrate that the isothiourea HyperBTM 12 can differentiate between the enantiomers of aryl-alkenyl (sp 2 vs. sp 2 )s ubstituted secondary alcohols. The selectivity of the KR has been assessed across aw ider ange of allylic alcohols, with good to excellente nantiodiscrimination observedf or substrates bearing either electron-rich or naphthyl substituents alongside an unsubstituted vinyl substituent.

Results and Discussion
The reaction of (AE)-1-(4-methoxyphenyl)prop-2-en-1-ol 15 with propanoic anhydride (0.5 equiv) and i-Pr 2 NEt (0.5 equiv) in CHCl 3 was chosen as the starting point to identify suitabler e-action conditions for the acylativeK Ro fa ryl-alkenyl (sp 2 vs. sp 2 )s ubstituted alcohols. The commerciallya vailable and readily prepared isothiourea HyperBTM 12 (1 mol %) was identified as the most promising in an initial screen of readily available catalysts, giving 44 %c onversioni nto ester 16 with S = 8, [16][17][18] whereas both tetramisole 17 and BTM 18 gave poor conversion and lower selectivity ( Table 1, entries 1-3). The absolute configuration of the major enantiomer of recovered alcohol (S)-15 was confirmed by comparison of its specific rotation with literature values. [19] Further optimizationr evealed that using isobutyric anhydride and lowering the reaction temperature to À40 8Cg ave improved selectivity (Table 1, entry 4). A solvents creen showed that both THF (S = 16) and in particular toluene (S = 21) gave improvements in selectivity( Ta ble 1, entries 5a nd 6). Further lowering the reaction temperature to À78 8Cl ed to the efficient KR of (AE)-15 with excellent selectivity (S = 29) considering the challenging aryl-alkenyl (sp 2 vs. sp 2 ) alcohol substitution (Table 1, entry 7). The catalystl oading could also be lowered to 0.25 mol %w ithout an appreciable drop in either conversion or selectivity ( Table 1, entry 8), althoughf or practicality 1mol %H yperBTM 12 was used to assess the reaction scope.
The optimized conditions for the KR of (AE)-15 were then tested for ar ange of vinyl alcohols bearing various aryl substituents (Tables2,3 ,a nd 4). Initiali nvestigations probed the effect of varying the steric and electronic nature of the aryl group bearing as ingle substituent in either the para-, meta-, or ortho-position (Table 2). Unsubstituteda nd aryl rings bearing electron-donatingm ethoxy substituents in either the para-, meta-, or ortho-positions worked well, with excellent selectivity obtained in all cases (Table 2, entries 1, 2, 6a nd 9, S = 29-59). In contrast, the presence of an electron-withdrawing CF 3 substituentinany of the positionsaround the aryl ring led to ano-Scheme1.Lewis base-catalysed acylative KR of aryl-alkenyl alcohols.  (Table 2, entries 3, 7a nd 10, S = 7-11). For example, although 3-methoxy substituted alcohol (AE)-23 gave S = 59, the analogous3 -CF 3 substituted (AE)-24 gave S = 11.V arious halogens ubstituents were tolerated, allowing KR of alcohols 21, 22 and 25 with moderate levels of selectivity (Table 2, entries 4, 5and 8, S = 8-17). This observation is consistent with previous proposals for the acylative KR of aryl-alkyl (sp 2 vs. sp 3 )s ubstituted secondary alcohols using isothioureas, which typicallyg ive highers electivity in the resolution of alcohols bearing electron-rich aryl substitutents. [14] In these processes, the aryl unit is thought to be the key recognition motif for enantiodiscrimination, being involvedi np-stacking with an electron-deficient acyl ammonium intermediate during the acylation step. Subsequents tudies aimed to exploit this observation throught esting the KR of aryl-vinyl alcohols bearing either poly-substituted electron-rich aryl-substitutentsorextended aromaticn aphthyl units (Table 3). Excellent selectivity was observed with electron-rich2 ,6-dimethoxy substituted aryl-alkenyl alcohol (AE)-28 (S = 110), although the presenceo ft wo orthosubstituents resulted in lower,b ut still acceptable,c onversion over an extended 48 hr eactiont ime due to the slower rate of acylation (Table 3, entry 1). The methodology was then applied to the KR of lignin-derived alcohols (AE)-29 and (AE)-30 bearing methoxy-substituted aryl rings (Table 3, entries 2a nd 3). Pleasingly,t he resolutions proceeded with excellent selectivity in both cases (S = 44 and 33, respectively), allowing the recovered alcohols 29 and 30 to be isolated with high e.r. This demonstrates that the methodology can be used to access enantiomericallyp ure syntheticb uildingb locks from renewable monomers derived froml ignin,w hich is important for the continued drive for valorization of such feedstocks. [18] Mesityl-substituted allylic alcohol (AE)-31 also gave lower conversion into the corresponding ester,b ut the KR selectivity was reasonable (Table 3, entry 4, S = 11). The KR of 2-naphthyl substituted vinyl alcohol (AE)-32 gave exceptional selectivity,w ith the remaining   (Table 3, entry 5). The presence of a1 -naphthyl substituent also led to excellent selectivity (S = 108)u nder the standard conditions (Table 3, entry 6). The selectivityo bserved with naphthyl substituents was surprisingly sensitive to further substitution on the naphthylener ing. For example, 6-methoxy substituted naphthyl alcohol( AE)-34 gave dramatically lower selectivity (S = 5) compared with the unsubstituted analogue( Ta ble 3, entry 7). To probe the origin of the high selectivity using unsubstituted naphthyl alcohols,t he KR protocol was tested on aryl substrates (AE)-35 and (AE)-36 containing 4-phenyl and 3vinyl substituents,r espectively (Table 3, entries 8a nd 9). In both cases the KR gave good selectivity (S = 13 and 26), althoughn either match the levels of enantiodiscrimination observed with the extended conjugation within the unsubstituted naphthyl examples. For the resolution of (AE)-32,t he exceptionally high selectivity,c oupledw ith the accuracy of the HPLC analysisu sed to measuret he e.r. values of both alcohola nd ester,m akes the calculation of an exact selectivityf actor difficult. To validate the reported S value,r epeate xperiments were performed and product enantioselectivities measured atv arying reaction conversions.T he data obtained was plotted as shown in Figure 2, allowing the selectivity factor to be determined using linear regression. [19] Good linear correlation of the datao ver ar ange of reactionc onversions suggests that S = 1980 for the KR of (AE)-32.
Finally,a st he catalytic system can effectively discriminate between the two planar sp 2 hybridized substituents within aryl-alkenyl alcohols, the KR of some alternative classes of secondary alcohol were compared under the same reactionconditions (Table6). Interestingly,t he KR of aryl-vinyl substituted alcohol (AE)-32 (sp 2 vs. sp 2 )g ave higherl evels of enantiodiscrimination than the analogous aryl-alkyl substituted alcohol( AE)-47 (sp 2 vs. sp 3 ), althoughi nb oth cases the selectivity is excellent ( Table 6, entries 1a nd 2). However,t he use of aryl-alkynyl alcohol (AE)-48 (sp 2 vs. sp) gave poor selectivity (S = 3) in the KR process( Ta ble 6, entry 3). The catalytic system was also only poorly selectivef or the KR of vinyl-alkyl alcohol (AE)-49 (sp 2 vs. sp 3 )( Ta ble6,e ntry 4, S = 3). This suggests that both aryl (sp 2 ) and alkynyl (sp) groups are effective recognition motifs for enantiodiscrimination and may interact with the proposed acyl ammonium intermediate (vide infra) during the acylation step. Conversely,v inyl (sp 2 )a nd alkyl (sp 3 )s ubstituents are poor recognition units and are unlikely to interactw ith the catalytic intermediate. Consequently,c ombining an effective recognition motif (such as aryl (sp 2 )a nd alkynyl (sp) groups) with ap oor one (such as vinyl (sp 2 )a nd alkyl (sp 3 )u nits) leads to high enantiodiscrimination during KR, whereas alternative combinations result in low selectivity.
To demonstrate the synthetic utility of this KR processt of acilitate the separation of the two enantiomerso faracemic alcohol, the KR was performed on ap reparative laboratory scale using 2.5 g( 13.6 mmol) of (AE)-32 and 1mol %o fH yperBTM (Scheme 2). This highly selectiver eaction proceeded to 50 % conversion,a llowing unreacted (S)-32 to be recovered in 43 % yield (1.08 g) and 99:1 e.r. Isolated ester (R)-37 was readily hydrolyzed under basic conditions to give (R)-32 in 45 %y ield (1.12 g) over the two steps and > 99:1 e.r.
The proposed catalytic cycle starts with areversiblea cylation of HyperBTM 12 with isobutyric anhydride to form acyl ammonium intermediate 50 (Scheme 3a). Turnover-limiting acylation of the favourede nantiomer of the aryl-alkenyl alcohol is thought to occur with concomitant proton transfer to the carboxylate anion. [20,21] The i-Pr 2 NEt may possibly act as as huttle base to regenerate the catalyst and remove isobutyric acid. The sense of enantioselectivity observed can be rationalized by considering the interactions of the incoming alcohol with acyl ammonium 50 during the selectivity-determining step (Scheme 3b). Acyl ammonium 50 is thought to be conformationally locked due to as tabilizingn on-bondingO ÀSi nteraction (n O to s* CÀS ), [22] with the Re faceb locked by the pseudoaxial phenyl group. The fast-reactinge nantiomer of the aryl-alkenyl alcohol can adopt ac onformation that has ap otentially stabilizing aryl p-cation interaction with the isothiourea (52), which is favouredo ver the potentiala lkenyl p-cation interaction in the slow reactinge nantiomer (53). [23] This model is consistent with the higher selectivity observed for substrates bearing electron-rich aryl rings due to the increased strength of the proposed cation-p interaction in the favoured transition state in these cases. [24] Conversely, increasing the substitution on the alkene makes this p-system more electron rich, which decreases the difference in energy between the diastereomerict ransition states and accountsf or the lower selectivity obtainedf or these examples. Apossible explanation for the enhanced selectivity with naphthyl substituents is the presence of an additional stacking interaction with the benzenoid ring of acyl ammonium 50 for the fast reacting enantiomer.S ubstitution of the naphthyl ring with electron-donatings ubstituents may destabiliset hese additional interactions, [25] resulting in the observed loss in enantiodiscrimination.

Conclusion
The isothiourea HyperBTM 12 (1 mol %) can catalyzet he acylative KR of ar ange of aryl-alkenyl (sp 2 vs. sp 2 )s ubstituted secondary alcohols with isobutyric anhydride. The catalytic system achievese ffective enantiodiscrimination between the enantiomers of secondary alcohols bearing two planar sp 2 hybridized substituents. The efficiency of the KR process has been assessed for ar ange of substituted aryl and heteroaryl moieties and various alkenes ubstitution patterns. The highest selectivity is obtained when either electron-rich or naphthyl aryl substituents are presenti nc ombination with av inyl substituent. Conversely,t he presence of either electron-deficient aryl rings or substituted alkenes leads to lower levels of selectivity.T he optimized KR process can be used to separatet he two enantiomerso fs ynthetically usefula ryl-vinyl alcohols with high enantioselectivity (up to > 99:1 e.r.)o napreparative scale at low catalystl oading (1 mol %). Ongoing work within this laboratory is focused upon the development of practical KR processes of challenging substrates and their applications in synthesis.

Experimental Section
General: For general experimental details, full characterisation data, 1 Ha nd 13 C{ 1 H} NMR spectra, and HPLC traces, see the Supporting Information. [26] Representative procedure for the KR of aryl-alkenyl alcohols The appropriate alcohol (1 equiv) was dissolved in PhMe (0.35 m) and the solution cooled to À78 8C. HyperBTM 12 (1 mol %), i-Pr 2 NEt (0.6 equiv) and isobutyric anhydride (0.5 equiv) were added and the solution stirred at À78 8Cf or 16 h. The reaction was quenched with 1 m HCl, the solution diluted with EtOAc and washed successively with 1 m HCl ( 2), NaHCO 3 ( 2) and brine. The organic layer was dried over anhydrous Na 2 SO 4 ,f iltered and concentrated under reduced pressure. The alcohol and ester were purified by column chromatography and analysed by chiral HPLC.