Ring‐Opening Regio‐, Diastereo‐, and Enantioselective 1,3‐Chlorochalcogenation of Cyclopropyl Carbaldehydes

Abstract meso‐Cyclopropyl carbaldehydes are treated in the presence of an organocatalyst with sulfenyl and selenyl chlorides to afford 1,3‐chlorochalcogenated products. The transformation is achieved by a merged iminium–enamine activation. The enantioselective desymmetrization reaction, leading to three adjacent stereocenters, furnished the target products in complete regioselectivity and moderate to high diastereo‐ and enantioselectivities (d.r. up to 15:1 and e.r. up to 93:7).

On the basis of these resultsw ew ere keen to investigate whethers uch asymmetric regioselective 1,3-bisfunctionalizations are also feasible using different substituents.W ec hose sulfenylc hlorides as reactants to add acrosst he CÀCb ond in the cyclopropane; [10,11] because of the highly polarized SÀCl bond we anticipated that sulfur would act as the electrophilic component, with chlorine as the nucleophilic counterpart (Scheme 1, right). [12] To test our notion, achiral meso-cyclopropyl carbaldehyde 1a and p-tolylsulfenyl chloride 2a were chosen as substrates for the optimization. [13] As starting point we chose the first generation MacMillan catalyst [14] using different counterions, whichp rovided the desired product, but without any enantioinduction (Table 1, Entries 1and 2).
With the optimized results in hand, av ariety of sulfenyl chlorides and phenylselenyl chloride weres ubjected to the reaction( Ta ble 2). Various aryl sulfenyl chlorides were utilized ( Table 2, Entries1-4);e lectron-withdrawing and also electrondonating substituents were tolerated, affording the desired products in moderatet og ood yields( 61-84 %). The diastereo-selectivity was improved by using electron-rich sulfenyl chlorides (3b), while electron-poor aryl units led to lower selectivities (3d, 3e). In all cases as light decreaseo ft he enantiomeric ratio was observed compared to the optimized model system.
The use of sterically demandinga lkyl sulfenyl chlorides ( Table 2, Entries 5a nd 6) was possible. In the case of the cyclohexyl substituent, as ignificantly higherd iastereomeric ratio was observed, while yield and enantiomeric ratio decreased. In the case of tert-butyls ulfenyl chloride, the system avoided the steric hindrance between the bulky tert-butyl moiety and the aliphatic chain by formingt he disulfide 3g,a nd thus two equivalents of the sulfenyl chloride had to be used;l ower amountsr esulted in the same product and selectivity,b ut with lower yield. With primary alkyl sulfenyl chlorides the desired products weren ot observed. Additionally,c ommercially available methoxycarbonyl sulfenyl chloride and phenylselenyl chloridew eret ested as reagents. The products 3h and 3i were obtained,r espectively,i ng ood yield with acceptable diastereo-ande nantioselectivity.T he use of analogouss ulfenyl and selenyl bromides showed no conversion of 1a,p robably because the less polarizedS ÀBr and SeÀBr bonds are not sufficiently reactive.
After evaluating the scope with respect to 1a,w ev aried the substituents of the cyclopropylc arbaldehyde to determine limitationso ft he reaction ( Table 3). Exchanging the cyclohexyl ring for two ethyl moieties resulted in ap rolonged reaction time, probablyb ecause of the loss of ring strain and increased steric hindrance. The diastereoselectivity was completely lost, while the e.r.s howedo nly as light decrease for the analogous diastereomer 4b.I nt his case the undesiredd iastereomer 4b' Table 2. Scope using cyclopropyl carbaldehyde (1a). [a] Entry Substrate 2 Product Yield [b] [%] d.r. [c] e.r. [ Yield [b] [%] d.r. [c] e.r. [ was also isolated, still showing ane .r.o f7 4:26. Determination of the configuration by NOESY revealed that the stereocenter next to the hydroxymethyl group is inverted. The other two stereocenters are built up in ac atalyst-controlled fashion and are defined during the cyclopropane ring-opening step after iminium activation. Replacement of the ethyl groups by phenyl groups (Table3,E ntry 2) yieldedn or eactivity of the substrate at the optimized conditions. The reaction temperature had to be increased to ambient temperature and the catalyst was changed to III with 5-methylfuryl as substituent R 1 .T hese changes were necessary to overcome the larger steric hindrance of the phenyl rings compared to the ethyl moieties, providing the major diastereomer 4c in reasonable d.r.a nd e.r. Entry 3a nd especially Entry 4( Ta ble 3) reveal the limitations of the transformation.B ecause of the increased polarity of the substrates 1d and 1e,D ME had to be used as solvent to avoid precipitation of intermediates. The reaction time was increased to 72 ha ta mbient temperature. In the case of the substrate 1d,amoderate selectivity was achieved using the catalyst V·DCA,w hile utilization of the Boc-protected pyrrolidine cyclopropylc arbaldehyde 1e as substrate gave the desired product 4e with almostn os electivity regarding d.r.a nd e.r.E xchange of the Boc-for benzyl-or tosyl-protected pyrrolidine led to no conversion of the substrates. The preliminary studies using these polar substrates demonstratet hat they can, indeed,b e converted to the resulting 1,3-chlorochalcogenated products, but to obtain highers electivities differentc atalytic systems need to be examined.
The reaction mechanism consists of am erged iminium-enamine activation (Scheme2). The initial step is the formation of the iminium ion VIb,w hichi st hen attacked at the 3-position by the chloride from the sulfenyl chloride, leading to the enaminec omplex VIc,w hile releasing the positively charged sulfenylium ion.
The benzyl moiety that shields the top face seems to be crucial for enantioselectivity,w hich is determined in this step. The subsequenta ttack of the emerging enamine at the cationic RS forms the iminium complex VId.T his last step is responsible for the diastereomeric ratio of the transformation. As demonstrated in Table 1( Entries 3a nd 17) and Table 2( Entry 5), the larger the substituent, the greater the stericr epulsion and the better the diastereomeric ratio. Final hydrolysis releases the product and regenerates the active catalyst VIa.