Cu-Catalyzed Enantioselective Protoboration of 2,3-Disubstituted 1,3-Dienes

A Cu-catalyzed regio- and enantioselective protoboration of 2,3-disubstituted 1,3-dienes is described. The protocol operates under mild conditions and is applicable to symmetrically and unsymmetrically substituted dienes, providing access to homoallylic boronates in consistently high yield, regioselectivity, and enantiomeric ratio. Preliminary investigations point to a complex mechanism.

T he growing interest for Cu-catalyzed selective borofunc- tionalizations of 1,3-dienes witnessed in recent years stems from the diversity of polyfunctional (chiral) alkenyl boronates that can be generated and the net increase in molecular complexity enabled by these approaches. 1 At their most sophisticated, the development of these methods requires overcoming challenges associated with chemo-, regio-, and enantioselectivity (Figure 1A).The substitution pattern of the 1,3-dienes has a profound influence on these aspects and affects product distribution.Therefore, much of the work in developing enantioselective borofunctionalization reactions hinges on the design or identification of a chiral ligand capable of exerting high levels of chemo-, regio-, and stereocontrol irrespective of the innate biases imposed by the substrate.Among notable achievements, typical electrophiles include acyl silanes, aldehydes, ketones, imines, aryl halides, and cyanamides. 1 In contrast, examples of enantioselective protoborations remain scarce (E + = H + , Figure 1A). 2 Even though little is known regarding their exact mechanism, it can be surmised that protonation of the putative Cu−allyl intermediate is likely to be enantio-determining in certain cases and, thereby, adds a layer of complexity to the development of these processes, in particular for acyclic substrates. 3In 2010, Ito and co-workers reported the preparation of highly enantioenriched homoallylic boronates by Cu-catalyzed protoboration of cyclic 1,3-dienes using L1 (Figure 1B).2a While 1,2-regioselectivity was predominant at low temperatures, the product of 1,4-addition was generated preferentially at ambient temperature.Of note, 3,4-protoboration of isoprene led to an achiral homoallylic boronate, and a modest er was obtained with a linear diene.Capitalizing on the modularity of a chiral monophosphane ligand (L2), 4 the Mazet group disclosed a 1,2-regioselective Cu-catalyzed enantioselective protoboration of 2-(hetero)aryl branched 1,3-dienes (Figure 1C).2b After in situ oxidation, acyclic homoallylic alcohols featuring a benzylic tertiary stereocenter were obtained with high levels of regio-and enantiocontrol.For alkyl substituted dienes, selectivity was enhanced with increasing steric demand, suggesting that for large substituents formation of a 1,4-σ-allyl copper intermediate is favored and subsequent S E 2′ protonation is enantio-determining.Reduced performances were disclosed for all other substitution patterns surveyed.In 2022, Zhang and co-workers reported the 1,4selective Cu-catalyzed enantioselective protoboration of linear dienes using L3 (Figure 1D).2c Mechanistic investigations pointed to an enantio-determining borocupration followed by a stereoretentive S E 2′-type protonation.In the same year, the Diver group developed a Cu-catalyzed enantioselective protoboration of acyclic 2,4-disubstituted 1,3-dienes accessed by ene-yne metathesis (Figure 1E).2d Using Duphos-type ligands (L1), excellent 3,4-regioselectivity was achieved, and er values ranging from 82:18 to 92.5:7.5 were measured.Mechanistic investigations were consistent with borocupration being both rate-and enantio-determining and with protonation occurring via an S E 2-type mechanism.Of note, a handful of Cu-catalyzed enantioselective protoborations using electronically biased conjugated dienes has also been reported (i.e., 1,2dihydropyridines, [n]dendralenes, trifluoromethylated linear 1,3-dienes). 5o expand the repertoire of borofunctionalizations of nonactivated precursors, we became particularly attracted to the challenge of developing a Cu-catalyzed enantioselective protoboration of acyclic 2,3-disubstituted 1,3-dienes that would be applicable to both symmetrical and unsymmetrical substrates (Figure 1F).We recognized that the successful development of this reaction would provide access to enantioenriched homoallylic boronates that would be difficult to prepare using traditional approaches.We report the results of our progress in this direction, along with preliminary mechanistic insights.We began our study with the survey of standard reaction parameters and the evaluation of several chiral ligands using 2,3-diphenyl-1,3-butadiene 1a as model substrate. 6An overview of these results is presented in Table 1.Additional details are available in the Supporting Information (SI).After an initial round of optimizations, we found that QuinoxP* (L6) afforded the homoallylic boronate 2a in 71% conv.with an excellent regioselectivity (>20:1) and high enantiomeric ratio (er 89:11) using B 2 pin 2 and MeOH as the proton source (entry 1).Ligands L1−L3 which were used by Ito, Mazet, and Zhang for the protoboration of other 1,3dienes did not provide significant improvements, even though regioselectivity was high in all cases (entries 2−4).Noticeably, this was not the case with Binap (L4, 4:1 rr, entry 5).Several other privileged bisphosphine ligands (including L5, entry 6) were surveyed, but none could compare with L6.Introduction of large silyl substituents on the 5 and 8 positions of the quinoxaline core had only marginal impact on the reaction outcome (L7, entry 7). 7Much reduced catalytic activity and enantiomeric ratio were obtained with L8, a C 1 -symmetric analogue of L6 (entry 8).Representative results of the influence of the solvent, the temperature, the copper precursor, and the proton source are disclosed in entries 9−14.Finally, using BenzP* (L9), 8 we found that both the reactivity and er could be slightly improved without impairing the regioselectivity (entry 15).This last set of conditions was employed to delineate the scope of symmetrically 2,3-disubstituted 1,3-dienes.For ease of purification and analysis, the boronic esters obtained were converted to the corresponding alcohols by oxidation (Figure 2).Variations of the electronic nature of the aryl groups using ether, alkyl, various halides, or trifluoromethyl substituents did not affect the catalytic performances in terms of activity, regiocontrol, or enantiocontrol (2′a−2′i).A diminished yield was obtained for ortho-substituted derivatives (2′h).While a primary alkyl substituent was found to be compatible, no reaction was observed for a cyclohexyl-derived substrate (2′j− 2′k).
When the robustness of the reaction was evaluated by performing the Cu-catalyzed protoboration using 1.10 g of diene 1d, homoallylic boronate 2d was isolated as the sole regioisomer in 91% yield (1.36 g) and 92:8 er.We showed that the enantiopurity of the protoboration product could be further increased to >99:1 er by a single recrystallization using a 10:1 MeOH/CHCl 3 solvent mixture (Figure 4).A series of complementary experiments was conducted to gather preliminary insights into the mechanism of the catalytic reaction (Figure 5).We first measured the initial rates of protium and deuterium incorporation in parallel experiments run in separate vessels using MeOH and MeOD (Figure 5A). 10 Under the standard reaction conditions, a primary kinetic isotope effect was measured (KIE = 1.9), a result consistent with protonation being the rate-determining step (RDS).Quite noticeably, the er values were identical in both experiments.In line with this observation, we did not observe any significant variation in er using sterically more demanding alcohols (Figure 5B).Taken together, these data suggest that borocupration rather than protonation is likely to be the enantio-determining step.The model reaction was conducted using scalemic mixtures with different proportions of the two enantiomers of ligand L9. 11 The homogeneity of the reaction was checked to avoid a false positive.11e The plot of the ee of product 2a as a function of the ee of chiral ligand L9 showed both positive and negative nonlinear effects (NLE), as observed in some rare occasions for other Cu-catalyzed processes (see SI). 11,12 Using the method developed by the Bureś group, variable time normalization analyses (VTNA) were performed by monitoring formation of 2a by 1 H NMR (Figure 5D). 13The graphical overlay obtained for the "same excess" experiment indicates that there is neither catalyst deactivation nor product inhibition over the course of the reaction.An unusual partial order of 0.2 in [Cu] was probed by means of a "different excess" experiment.Coupled with the observation of NLE, this finding reveals catalyst speciation and that oligomeric copper complexes probably directly participate in the prevailing productive catalytic cycle.These observations are reminiscent of those reported by Blackmond and co-    workers in the context of Pd-catalyzed enantioselective C(sp 3 )−H bond arylations. 14 tentative catalytic cycle consistent with the results of our mechanistic investigations is disclosed in Figure 6.Activation of B 2 pin 2 by a copper-alkoxo complex followed by diene coordination (I → II → III) is well documented and is expected to proceed smoothly. 1,7,15Collectively, our experimental data are in agreement with the subsequent borocupration (III → IV) being the enantio-determining step.The tertiary Cu-alkyl intermediate generated undergoes rate-determining S E 2 protonation to liberate the product and regenerate the starting copper-alkoxo complex (IV → I).The tertiary allylcopper complex (IV) is potentially in equilibrium with the primary allylcopper (V) via a σ−π−σ mechanism.Because protonation is not enantio-determining, formation of 2 from V is unlikely to occur.The product of formal 1,4protoboration (3) could be accessed either by S E 2′ protonation of IV or by S E 2 protonation of V.The high levels of regiocontrol observed under the optimized conditions suggest either that protonation is slow or that V is kinetically not accessible.Finally, the results of the NLE study and our kinetic measurements indicate that a composite catalytic system arising from catalyst speciation is likely operating.Even though the exact nature of the copper complex remains unclear, formation of dimeric μ-boryl-bridged copper complexes from a mononuclear intermediate similar to II is thermodynamically favorable (using L6 instead of L9).7a,16 Overall, the exact composition of the mixture of catalytically active intermediates is unclear and whether these numerous species lie on or off-cycle is difficult to assess. 11,14Finally, our data also point to the fact that product coordination is not inhibitory, and displacement by binding of an incoming molecule of substrate to regenerate III should be favorable.
In conclusion, we have developed a regio-and enantioselective Cu-catalyzed protoboration of 2,3-disubstituted 1,3dienes, which provides access to chiral homoallylic alcohols with high levels of selectivity.The method operates under mild conditions and is applicable to symmetrical and unsymmetrical derivatives.Preliminary mechanistic investigations have shed light on unusual features indicative of catalyst speciation and the formation of catalytically competent oligomeric copper species.Further studies are underway in our laboratories to understand the intricacies of this mechanism, which may well extend to other Cu-catalyzed selective borofunctionalization reactions.

Data Availability Statement
The data underlying this study are available in the published article and its Supporting Information.
Experimental procedures, characterization of all new compounds, and NMR spectra.(PDF) ■

Figure 5 .
Figure 5. (A) Kinetic isotope effect.(B) Influence of the structure of the protonation agent.(C) Variable time normalization analyses.All data are the average of at least two experiments.Er was measured by HPLC.

AUTHOR INFORMATION Corresponding Author
AuthorsSensheng Liu − Department of Organic Chemistry, University of Geneva, 1211 Geneva, Switzerland Yangbin Liu − Department of Organic Chemistry, University of Geneva, 1211 Geneva, Switzerland; Present Address: Shenzhen Bay Laboratory, Shenzhen, 518055 (China) Arthur Flaget − Department of Organic Chemistry, University of Geneva, 1211 Geneva, Switzerland