Catalytic Asymmetric Synthesis of Cyclohexanes by Hydrogen Borrowing Annulations

Abstract Hydrogen borrowing catalysis serves as a powerful alternative to enolate alkylation, enabling the direct coupling of ketones with unactivated alcohols. However, to date, methods that enable control over the absolute stereochemical outcome of such a process have remained elusive. Here we report a catalytic asymmetric method for the synthesis of enantioenriched cyclohexanes from 1,5‐diols via hydrogen borrowing catalysis. This reaction is mediated by the addition of a chiral iridium(I) complex, which is able to impart high levels of enantioselectivity upon the process. A series of enantioenriched cyclohexanes have been prepared and the mode of enantioinduction has been probed by a combination of experimental and DFT studies.

Enolate alkylation is af undamental process in organic chemistry and is widely used as as trategy for CÀCb ond formation. [1] In this chemistry acarbonyl substrate is typically deprotonated with astrong base (e.g.,LDA)and the resulting enolate is then trapped with ar eactive electrophile.A lkylation of asubstituted enolate results in the generation of anew a-stereogenic center and an abundance of methods (both stoichiometric and catalytic) have been developed which enable this process to be carried out in an asymmetric manner (Scheme 1A). [2] Whilst this approach is highly effective for alkylation with primary electrophiles,a lkylation with secondary electrophiles is significantly more challenging and often results in sluggish reactivity accompanied by competing elimination processes. [1] Moreover,w hen unsymmetrical secondary electrophiles are employed, anew stereogenic center is formed at the b-position and only ah andful of methods have been reported, which allow control over the stereochemical outcome of such aprocess. [3] Hydrogen borrowing catalysis represents ap owerful alternative strategy to classical enolate alkylation, enabling direct alkylation of enolates with unactivated alcohols. [4] Within this manifold, we recently reported that an achiral iridium(III) catalyst can promote alkylation of pentamethylphenyl (Ph*) ketones with alcohols leading to a-a nd bbranched ketones. [5] This was subsequently extended to a( 5 + +1) annulation process in which racemic cyclohexanes could be accessed from 1,5-diols (Scheme 1B). [6] These reactions proceed by oxidation of the alcohol by the iridium catalyst to generate the corresponding carbonyl compound in situ. After aldol condensation with an enolate and loss of water, the catalyst "returns" the abstracted hydrogen to provide the CÀCcoupled product and complete the catalytic cycle.T he Ph* group plays ak ey role in facilitating this chemistry;the bulky doubly ortho-substituted aromatic group is oriented orthogonal to the carbonyl and shields against competing reduction and homodimerization processes. [5] Moreover,a cyl Ph* derivatives can readily be converted to aw ide range of functional groups via an ipso-substitution process (> 30 examples). [5,6] Remarkably,d espite numerous recent advances in the field of enolate hydrogen borrowing Scheme 1. Previous work and strategy for catalytic asymmetric hydrogen borrowing. LDA = lithium diisopropylamide;THF = tetrahydrofuran.
[ catalysis,n og eneral strategy has been reported allowing the absolute stereochemical outcome of this process to be controlled. [7,8] We rationalized that the enantiodetermining step in these reactions involves the return of iridium hydride to an achiral enone.Since this step bears some resemblance to existing methods for asymmetric hydrogenation we anticipated that achiral transition-metal complex might be able to control the facial selectivity of this process (Scheme 1C). [9] We recognized that the key to success would lie in identifying atransition metal complex that can perform three key roles: (i)efficient oxidation of alcohols;(ii)achallenging reduction of sterically demanding Ph* substituted enones;(iii)c ontrolling facial selectivity within this reduction process resulting in high levels of enantioselectivity. We commenced our study by investigating the reaction between pentamethylacetophenone 1 and commercially available hexane-1,5-diol 2a.I nl ine with our previous studies, [6] in the presence of an achiral Ir III catalyst along with 4equiv of KO t Bu in toluene at 110 8 8Cw eo btained racemic cyclohexane 3a in 75 %y ield and 91:9 d.r. (Table 1, Entry 1). We have previously shown that the high trans-diastereoselectivity in this reaction is ar esult of reversible deprotonation of the product. [6] We were delighted to find that by switching to an Ir I precatalyst along with 5mol %(R)-BINAP (4)w eobtained cyclohexane 3a in 76 %y ield with am odest but promising 68:32 e.r. (Table 1, Entry 2). At this point we embarked upon an extensive program of optimization (for full details,s ee Supporting Information). Changing the ligand to (R)-H 8 -BINAP (5)r esulted in lower enantioselectivity whereas (R)-MeO-BIPHEP (6)a fforded 3a with similar selectivity (Entries 3,4). We next evaluated aseries of MeO-BIPHEP based ligands (6-10)b earing phosphine groups with different steric and electronic properties. Difuryl-substituted phosphine 7 resulted in as ignificant decrease in enantioselectivity,b ut when a3 ,4,5-trimethoxy substituted ligand 8 was employed, 3a was isolated in an improved 73:27 e.r. (Table 1, Entries 5,6). Increasing the steric bulk of the phosphine clearly provided ab eneficial effect-ligands 9 and 10 afforded 3a in improved selectivities of 86:14 and 87:13 e.r. respectively (Table 1, Entries 7,8).
We found that changing the biaryl backbone of the ligand from MeO-BIPHEP to SEGPHOS provided as mall additional increase in enantioselectivity to 88:12 e.r. (Table 1, Entry 9). Conducting the reaction in tert-butanol led to af urther incremental improvement to 80 %y ield and 89:11 e.r. (Table 1, Entry 10). Under these conditions we then screened as eries of Ir,R h, and Ru precatalysts (see Supporting Information for full details) and found that the best result was obtained with Ir(cod)(acac), which afforded 3a in 85 %yield and 90:10 e.r. (Table 1, Entry 11). Finally,we found that with areduced Ir loading (2 mol %) and increased dilution (0.1m)w ew ere able to isolate 3a in 87 %y ield and 92:8 e.r. (Table 1, Entry 12).
With optimal conditions in hand, we set out to evaluate the generality of the process.S ubstitution on the diol backbone was well tolerated with ad iol bearing ag eminal dimethyl group at the d-position cyclizing to afford 3b in 67 %y ield, 90:10 d.r.a nd 94:6 e.r( Table 2, Entry 2). With substitution at the g-position we isolated cyclohexanes 3c-3e in high yields and with excellent levels of diastereo-and enantioselectivity ( Table 2, . Ad iol bearing a nbutyl group reacted to afford 3fin 87 %yield, 89:11 d.r.and 91:9 e.r. (Table 2, Entry 6). Interestingly,i ntroduction of an isobutyl group resulted in poor conversion to cyclohexane 3g which was isolated in 24 %y ield albeit still with good enantioselectivity (Table 2, Entry 7). [10] Aromatic and heteroaromatic groups were well tolerated and cyclohexanes 3h and 3i were isolated in good yields with high levels of enantioselectivity (Table 2, Entries 8,9). Diols bearing ether and thioether groups also cyclized smoothly to afford products 3j and 3k in excellent yields and high levels of stereoselectivity (Table 2, Entries 10,11). Even an acetal was tolerated in the chemistry providing 3lin 80 %yield and 86:14 e.r. with no evidence of any competing side-reactions (Table 2, Entry 12). We also investigated an enantiopure diol derived from b-thujone which we had previously found to undergo annulation with very poor diastereoselectivity (51:7:42 d.r.). [6] We hoped that our optimized conditions might be able to augment this lack of substrate control and were pleased to find that 3m was isolated as a90:10 mixture

Angewandte Chemie
Communications of diastereoisomers. [11] Finally,w ei nvestigated formation of ac yclopentane from 1 and pentane-1,4-diol (Table 2, Entry 14). In this case, 3n was isolated in ar educed yield of 43 %a lbeit still with high levels of diastereo-and enantioselectivity. Afurther benefit of the Ph* group is its highly crystalline nature.A ll of the products 3a-3n described above are crystalline solids and this provides an opportunity to enhance the enantiomeric purity by stereoselective crystallization. As ar epresentative example,w ec arried out the reaction of pentamethylacetophenone with hexane-1, 5-diol (2a)ongram scale,o btaining 3a in 92 %y ield with 93:7 d.r. and 92:8 e.r. (Scheme 2A). After asingle recrystallization (81 %recovery) we were able to significantly enhance this stereochemical purity to > 95:5 d.r. and 98:2 e.r.
To probe the mechanism of the asymmetric hydrogen borrowing annulation, we independently synthesized the proposed key intermediate,c yclic enone 4 and subjected it to the optimized conditions with a n-butyl substituted diol (Scheme 2B). After this reaction we isolated 3a in 77 %yield and 90:10 e.r. Them ajor enantiomer was the same as that obtained in the full hydrogen borrowing sequence and the yield, diastereo-and enantioselectivity were also very similar (c.f., Table 2, Entry 1). Based upon this result, we arrived at the following conclusions:(i) it is likely that cyclic enone 4 is an intermediate in the asymmetric hydrogen borrowing reaction;( ii)t he absence of any crossover products implies that formation of 4 is an irreversible process;(iii)the similar enantioselectivities observed in the resubjection experiment and annulation process implies that the initial C À Cb ond formation between 1 and 2a occurs with complete regioselectivity at the primary end of the diol (i.e., reduction of isomeric enones such as 5 do not account for formation of the minor enantiomer). We have previously shown that Ph* containing products such as racemic 3a-3n can be readily [a] Reaction conditions: 1 (1 equiv), diol (2 equiv), Ir(cod)acac (4 mol %), (R)-DTBM-SEGPHOS (5 mol %), KO t Bu (4 equiv), t BuOH (  cleaved to the corresponding acid bromide in an ipsosubstitution reaction with Br 2 and that the resulting acid bromides can be employed in situ to afford esters,a mides, alcohols,c arboxylic acids,a nd aldehydes without erosion of stereochemical purity. [5,6] This procedure gave us aconvenient opportunity to determine the absolute stereochemistry of the cyclohexane products.T othis end, ketone 3awas treated with Br 2 to generate the corresponding acid bromide.F ollowing addition of LiAlH 4 ,alcohol 6 was isolated in 90 %yield with no stereochemical erosion (Scheme 2C). Correlation of the specific rotation value of 6 with that previously reported in the literature allowed us to determine that the absolute configuration of 6 (and by extension 3a)i s( R,R). [12] The remaining examples in Table 2a re assigned by analogy.
To gain insight into the mechanism of the stereochemical determining step,density functional theory (DFT) modelling studies were conducted, employing ac omputationally tractable [Ir] complex ligated by (R)-BINAP (  Table S1 in the Supporting Information). [13] Si-coordination of 4 (Si-INT0) is computed to be favoured by 4.8 kcal mol À1 over its Re counterpart (Re-INT0). 1,4 hydride insertion then proceeds from the Si-face with af ree energy barrier 0.8 kcal mol À1 lower than that for Reinsertion and accounts for the experimentally observed e.r. (68:32 = 0.6 kcal mol À1 at 383 K, Tables 1a nd S1). This preference results from the steric clash between Ph* and (P)Ph observed in the Re-TS (Figure 1). Structures were optimized and thermodynamic/ solvent effects calculated at the PBE0-D3BJ/def2-SVP,def2-TZVP(Ir) level of theory with the solvent accounted for using the SMD model. Single-point energetics were evaluated on these stationary points at the PBE0-D3BJ/def2-TZVPP level of theory. [14] In conclusion, we have developed ah ighly enantioselective synthesis of multisubstituted cyclohexanes via hydrogen borrowing catalysis.T his process is mediated by two commercially available reagents:Ir(cod)(acac) and DTBM-SEG-PHOS and provides enantioenriched cyclohexanes with control over both diastereo-and enantioselectivity.T he origins of stereoselectivity in this system have been probed by both experimental studies and DFT calculations.T his approach constitutes the first general catalytic asymmetric strategy within the rapidly developing field of enolate hydrogen borrowing catalysis.