|Home | About | Journals | Submit | Contact Us | Français|
Two new ligands 1a,b have been reported. Upon treatment with one equivalent of NiCl2·(MeOCH2)2, 1a,b give the corresponding Ni-complexes. X-ray analysis of 1a·NiCl2 has established that the NiCl2 is selectively coordinated to the phenanthroline nitrogens. Ni/Cr-heterobimetallic catalysts 1a,b·CrCl2/NiCl2, prepared from 1a,b·NiCl2, have been shown to behave exceptionally well for the catalytic asymmetric Ni/Cr-mediated couplings, with the highlights including: (1) 1~2 mol % catalysts are sufficient to complete the coupling, (2) only a negligible amount of the dimers, by-products formed through the alkenyl Ni-species, is observed, (3) the coupling completes even with a 1:1 molar ratio of the coupling partners, and (4) the asymmetric induction is practically identical with that obtained in the coupling with the Cr-catalyst prepared from (S)-sulfonamide 2a,b. Using 4 additional aldehydes, a scope of the new Ni/Cr-heterobimetallic catalysts is briefly studied. Applicability of new catalysts to polyfunctional substrates has been demonstrated, with use of two C-C bond-formations chosen from the halichondrin/E7389 synthesis as examples.
CrCl2-mediated Grignard-type addition of an alkenyl halide to an aldehyde was first reported by Takai, Hiyama, Nozaki and coworkers in 1978, cf., eq. 1 in Scheme 1.1 Since then, it was shown that the addition is initiated by a catalytic amount of NiCl2.2 It is now known that this coupling involves: (1) oxidative addition of Ni(0), formed from NiCl2 via reduction with CrCl2 in situ, to an alkenyl halide to form the alkenyl Ni(II)-halide, (2) transmetallation of the resultant Ni(II)-species to the Cr(II)Cl2 to form the alkenyl Cr(III)-halide, and (3) carbonyl addition of the resultant Cr(III)-species to an aldehyde to form the product Cr(III)-alkoxide (Scheme 1).3 One of the by-products commonly found under both stoichiometric and catalytic conditions is iv, the dimer of i formed through the alkenyl Ni-halide.4,5 In order to suppress formation of this by-product, it is critically important to keep a low ratio of Ni- over Cr-salts. For example, the catalytic asymmetric couplings recently reported typically uses 10~20 mol % of the Cr-catalyst and 1~5 mol % of Ni-catalyst.6 Under this condition, iv is observed only in a negligible amount. However, to ensure complete consumption of an aldehyde, the coupling reaction is performed with a slight excess (typically 1.5 equiv) of an alkenyl halide. It is noteworthy that the coupling rate observed is roughly proportional to the amount of Ni(II)-catalyst and, in this respect, the catalyst loading in the Ni/Cr-mediated coupling is already low, i.e., 1~5 mol %. In this communication, we report a solution to improve not only the catalyst loading, but also the molar ratio of coupling partners.
Both problems mentioned above are connected with the efficiency in the alkenyl-group transfer from nickel to chromium, cf., alkenyl Ni(II)-halide ≤ alkenyl Cr(III)-halide vs. ≤ iv. To enhance the transmetallation, we have been curious about the possibility of placing Ni- and Cr-metals in close proximity. Specifically, we have been interested in a ligand bearing two ligation sites, each of which is complexed specifically to Cr- and Ni-metals, respectively.7 Two observations have encouraged us to study ligands represented by 1a,b in which a sulfonamide is tethered with 2,9-dimethylphenanthroline (DMP). First, upon addition of NiCl2·(MeOCH2)2 (1.05 equiv) to a 1:1 mixture of sulfonamide 2a (1.0 equiv) and DMP (1.0 equiv) in CD3CN at rt, DMP is selectively precipitated out as yellow crystalline solid (>95%), leaving the almost colorless solution containing only 2a (1H NMR). Second, NiCl2·DMP 3 is the best Ni-catalyst to achieve the catalytic asymmetric Ni/Cr-mediated couplings.8 Under the coupling condition, 3 is stable and exhibits no ability to facilitate carbonyl addition; in other words, the C-C bond formation takes place through the Cr-sulfonamide catalyst.9
Using the method outlined in Supporting Information, we synthesized ligands 1a,b and studied their behaviors in Ni-complexation. Upon treatment of 1a (150 mg) with 0.95 equiv of NiCl2·(MeOCH2)2 in MeCN (2 mL) at rt, light violet crystalline powder (168 mg, 94% yield based on 1a; 99% yield based on NiCl2·(MeOCH2)2) precipitated out. In a 1H NMR study (CDCl3), this substance caused severe signal-broadenings for the protons on the phenanthroline moiety, but little signal-broadenings for the protons on the sulfonamide moiety. This NMR experiment suggested that the paramagnetic nickel is coordinated to the phenanthroline nitrogens and also that 1a·NiCl2 likely exists in an extended conformation. The structure of 1a·NiCl2 was ultimately established by X-ray analysis (Figure 1). Similarly, upon treatment with 0.95 equiv of NiCl2·(MeOCH2)2 in MeCN at rt, 1b also gave light violet crystalline powder in ~95% yield.
Taking into account the relative rate of ligand exchange on Cr(II)- vs. Cr(III)-species,10 we routinely use the following protocol for preparation of a Cr-sulfonamide complex. A sulfonamide anion, generated from a given sulfonamide (proton sponge), is treated with CrCl2 in MeCN at rt for 1 h. Upon addition of NiCl2·DMP 3, an alkenyl halide, an aldehyde, and additives (Mn, ZrCl2(cp)2), and LiCl), the resultant Cr(II)-complex enters into the Cr(II)Cr(III) catalytic cycle and catalyzes the coupling. Notably, the color of sulfonamide Cr(II)-complex is deep-green, but turns into deep-brown, once the catalytic reaction begins. Based on the X-ray structure of three Cr-complexes, we proposed the octahedral structures for those Cr-complexes (Scheme 2).11
We applied this protocol for preparation of the Cr-catalysts in the tethered series and observed that 1a,b·NiCl2 exhibit behavior identical to that of 2a,b. Primarily based on three reasons, we assume that the MeCN-solution thus prepared contains the heterobimetallic catalysts, with Cr and Ni coordinated to the sulfonamide and phenanthroline moieties, respectively, cf., 1a·CrCl2/NiCl2 in Scheme 2. First, the color of MeCN-solution is deep-green as found for the MeCN-solution of 2a,b·CrCl2. Second, our previous work suggests no metal-exchange between 2a,b·CrCl2 and NiCl2·DMP.8 Third, the degree of asymmetric induction by 1a,b·CrCl2/NiCl2 is virtually identical with that by the corresponding 2a,b·CrCl2 (vide infra). Structurally, we speculate that the Ni-phenanthroline and Cr-sulfonamide adopt tetrahedral and octahedral coordinations, respectively, and that C-C bond-formation takes place at these catalyst-surfaces.
To reveal their catalytic capacity of promoting the Ni/Cr-mediated couplings, we first tested 1a,b·CrCl2/NiCl2 for the standard model system, i.e., 5 + 6 7 (Table 1). To our delight, both 1a,b·CrCl2/NiCl2 were found to behave exactly as we hoped, with the following highlights. First, even with 1 mol % catalyst loading, the coupling progressed to completion in MeCN, to furnish the coupled product in >90% yield (entries 1–3). Second, only a small amount of dimer 8 (≤3%) was observed; thus, the coupling completed even with a 1:1 molar ratio of 5 and 6 (entries 1, 2, 4, 6).12 Third, the asymmetric induction by 1a,b·CrCl2/NiCl2 was practically identical with that by the corresponding previous Cr-catalysts derived from (S)-sulfonamide 2a,b (entries 1–5 vs. entry 9).13 Fourth, the coupling rate was slightly faster at substrate-concentration of 0.8 M than that of 0.4 M, but no significant difference was noticed in the coupling yield (entries 1 vs. 2).
To demonstrate the difference between the new and previous catalysts, we ran the (5+6 7)-coupling side by side in MeCN (entries 5 and 8). In the coupling employing 1a·CrCl2/NiCl2 (2 mol %), both 5 (1.0 equiv) and 6 (1.1 equiv) were consumed within 1.5 h, to furnish 7 (>95%, er = 10:1) and 8 (~3%).12 On the other hand, in the coupling employing the Cr-catalyst from 2a (2 mol %) with NiCl2·DMP 3 (2 mol %), vinyl iodide 6 (1.1 equiv) was consumed in 4 h, to give a mixture of 7 (~45%, er = 8.6:1), 8 (~55%) and recovered 5 (55%). Their difference is further illuminated by the comparison of the result under entry 9 (the optimized condition for the previous catalysts6a) over the result under entry 5.
The Cr-catalyst (2 mol %) derived from the antipode of 1a·NiCl2 was tested for additional aldehydes 9~12, thereby demonstrating that the new catalyst matches well with all of them (Table 2).
Encouraged by these results, we chose two C-C bond-forming reactions from the synthesis of halichondrin/E7389, to show the applicability of new catalysts for polyfunctional substrates.14 The first example is the C26-C27 bond-formation employing aldehyde 13 (1.0 eq) and vinyl iodide 14 (1.2 eq) (Scheme 3).6c Employing the previous Cr-catalyst (20 mol %) derived from 2a, the desired allylic alcohol was obtained in ~90% yield with dr = 19:1.15 The new Cr-catalyst, prepared from 1a·NiCl2, was found to effect this coupling well, to furnish the desired allylic alcohol in 86% yield with dr = 19:1. We should note, however, that 3 mol % catalyst loading was required to complete the coupling with an acceptable rate. As reported previously, reductive cyclization of the allylic alcohol stereoselectively gave 15 in 95% yield.6c
The second example is the C19-C20 bond-formation employing aldehyde 16 (1.0 eq) and vinyl iodide 17 (1.1 eq) (Scheme 3).6c With the previous Cr-catalyst (20 mol %) derived from the antipode of 2a, the desired allylic alcohol 18 was obtained in ~90% yield with dr = 20:1.15 Again, the new Cr-catalyst, prepared from the antipode of 1a·NiCl2, was found to work well for this bond formation, to furnish 18 in 91% yield with dr = 19:1. It is noteworthy that 2 mol % catalyst loading was sufficient to complete the coupling with an acceptable rate. As previously reported, the allylic alcohol was converted to 18 in 88% yield.6c
In conclusion, we have reported two new ligands 1a,b. Upon treatment with one equivalent of NiCl2· (MeOCH2)2, 1a,b give the corresponding Ni-complexes. X-Ray analysis of 1a·NiCl2 has established that the NiCl2 is selectively coordinated to the phenanthroline nitrogens. Ni/Cr-heterobimetallic catalysts 1a,b·CrCl2/NiCl2, prepared from 1a,b·NiCl2, behave exceptionally well for the catalytic asymmetric Ni/Cr-mediated couplings, with the highlights including: (1) 1~2 mol % 1a,b·CrCl2/NiCl2 are sufficient to complete the coupling, (2) only a negligible amount of the dimers, by-products formed through the alkenyl Ni-species, is observed, (3) the coupling completes even with a 1:1 molar ratio of the coupling partners, and (4) the asymmetric induction is practically identical with that obtained in the coupling with the Cr-catalyst prepared from (S)-sulfonamide 2a,b. Applicability of these catalysts to polyfunctional substrates has been demonstrated, with use of two C-C bond formations chosen from the halichondrin/E7389 synthesis as examples. We are currently engaged with further refinements of the tethered ligands as well as preparation of other tethered heterobimetallic catalysts, including tethered Co/Cr-catalysts.
We are grateful to the National Institutes of Health (CA 22215) and to the Eisai Research Institute for generous financial support.