Copper-Catalyzed Enantio- and Diastereoselective Addition of Silicon Nucleophiles to 3,3-Disubstituted Cyclopropenes.

A highly stereocontrolled syn-addition of silicon nucleophiles across cyclopropenes with two different geminal substituents at C3 is reported. Diastereomeric ratios are excellent throughout (d.r.≥98:2) and enantiomeric excesses usually higher than 90 %, even reaching 99 %. This copper-catalyzed C-Si bond formation closes the gap of the direct synthesis of α-chiral cyclopropylsilanes.

Silylboronic acid esters are highly useful silicon pronucleophiles which have had significant impact on synthetic silicon chemistry.1 A broad variety of enantioselective C−Si bond formations can be achieved by using these Si–B reagents,2 and their copper‐catalyzed addition across α,β‐unsaturated acceptors is a prominent example (Scheme 1, top).3 CuI‐NHC4 (NHC=N‐heterocyclic carbene) as well as CuII‐bipyridine5 complexes do promote these reactions with high fidelity. A related enantioselective addition to strained alkenes, such as cyclopropenes, is not known to date (Scheme 1, bottom).6, 7 The resulting silylated cyclopropanes are versatile building blocks in organic synthesis,8 yet is their direct preparation by C−Si bond formation at an existing cyclopropane skeleton rare.9, 10, 11, 12 Gevorgyan and co‐workers developed palladium‐ and platinum‐catalyzed diastereoselective insertion reactions of cyclopropenes into Si−Sn and Si−H bonds, respectively.9 Established methods therefore start with silicon‐containing substrates,13 and a common method is the cyclopropanation of vinylsilanes.14 A fascinating approach by Ito, Sawamura, and co‐workers involving a regioselective copper‐catalyzed borylation of vinylsilanes containing an allylic leaving group by a 3‐exo‐tet ring closure stands out.15 The idea to access silylated cyclopropanes from cyclopropenes was inspired by Marek's16 and, in particular, Tortosa's17 work. Tortosa and co‐workers have accomplished a copper‐catalyzed desymmetrization of cyclopropenes by borylation.17 We report here a highly stereoselective silylation of cyclopropenes without the aid of a directing group (Scheme 1, bottom).18

Scheme 1

Copper‐catalyzed enantioselective addition of Si−B reagents across activated alkenes. EWG=electron‐withdrawing group. R3Si=triorganosilyl.

We started our investigation by reacting 3‐phenyl‐3‐methylcyclopropene (1 a) with Me2PhSiBpin (2 a)19a (1.5 equiv) in the presence of Cu(CH3CN)4PF6 as the copper precatalyst in THF at 0 °C (Table 1). NaOtBu (0.5 equiv) was used as an alkoxide base and MeOH (3.0 equiv) as a proton source (see the Supporting Information for the complete set of optimization data). With no ancillary ligand, almost no conversion of the cyclopropene was seen (<5 %, entry 1). This situation changed completely in the presence of bidentate phosphine ligands. Excellent diastereoselectivity was obtained with binap ligands L1–L3, and the enantioinduction increased with the steric demand of the PAr2 groups (entries 2–4). This high level of stereocontrol could not be further improved by changing the solvent to toluene or by lowering the reaction temperature to −20 °C (entries 5 and 6). A similar outcome was found with segphos ligands L4 and L5 (entries 7 and 8), and we eventually continued with L5, which led to the formation of the silylated cyclopropane 3 aa in good yield with a diastereomeric ratio (d.r.) ≥98:2 and an enantiomeric excess (ee) of 97 %.

Table 1

Selected examples of the optimization reactions.[a]

Entry	Copper/Ligand	Yield [%]	d.r.[b]	ee [%][c]
1	Cu(CH3CN)4PF6	n.d.	–	–
2	Cu(CH3CN)4PF6 /L1	71	≥98:2	80
3	Cu(CH3CN)4PF6 /L2	73	≥98:2	90
4	Cu(CH3CN)4PF6 /L3	73	≥98:2	96
5[d]	Cu(CH3CN)4PF6 /L3	81	96:4	84
6[e]	Cu(CH3CN)4PF6 /L3	74	≥98:2	92
7	Cu(CH3CN)4PF6 /L4	73	97:3	92
8	Cu(CH3CN)4PF6 /L5	74	≥98:2	97

[a] All reactions were performed on a 0.20 mmol scale with the isolated yield determined after flash chromatography on silica gel. [b] Determined by 1H NMR analysis. [c] Determined by HPLC analysis on a chiral stationary phase. [d] Toluene instead of THF. [e] Run at −20 °C. n.d.=not determined. binap=2,2′‐bis(diphenylphosphanyl)‐1,1′‐binaphthyl. sephos=5,5′‐bis(diphenylphosphanyl)‐4,4′‐bi‐1,3‐benzodioxol.

–

–

We then examined the substitution pattern of the cyclopropene (1 a–s, Scheme 2). Yields were generally good, and the level of enantioselection was consistently high. 3‐Arylated cyclopropenes bearing a substituent in the para or/and meta position(s) were tested, and it was found that the X group did not exert any electronic effect on either yield or stereoselectivity (1 a–j→3 aa–ja); the silylated cyclopropanes were all isolated as single diastereomers (d.r.≥98:2). Likewise, a thien‐2‐yl as well as naphthyl groups were tolerated (1 k–m→3 ka–ma). Bulkier alkyl groups instead of the methyl group at C3 of the cyclopropene had no influence on the enantiofacial selectivity, but a little on diastereoselectivity; yields were lower with increasing steric demand (1 n–p→3 na–pa). These results imply that the diastereoselectivity is affected by the steric discrimination of geminal substituents (Aryl/Me vs. Ph/Alkyl). This observation was also made when replacing the phenyl by a benzyl group (Ph/Me versus Bn/Me); the diastereomeric ratio dropped from ≥98:2 to 85:15 (1 q→3 qa). In turn, a spiro derivative reacted with high diastereoselectivity but in low yield (1 r→3 ra). For completion, the 3,3‐diphenyl‐substituted cyclopropene afforded the silylated cyclopropane in good yield and with high ee (1 s→3 sa).

Scheme 2

Scope I: Variation of the cyclopropene.[a–c] [a] All reactions were performed on a 0.20 mmol scale with the isolated yield determined after flash chromatography on silica gel. [b] Diastereomeric ratios determined by 1H NMR analysis. [c] Enantiomeric excesses determined by HPLC analysis on chiral stationary phases.

We next probed the transfer of different silyl groups from silylboronic acid esters R3SiBpin 2 b–g 19 to model compound 1 a (Scheme 3). It became quickly clear at the size of the silyl group substantially influences the yield. MePh2SiBpin (2 b) furnished acceptable 65 % yield (1 a→3 ab). The enantiomeric excess was 97 % ee and was even higher with another substituent in the para position (not shown; additional substrates in the Supporting Information). tBu(Me)PhSiBpin (2 c) did yield trace amounts of 3 ac, and the diastereomeric ratio of 62:38 is due to the stereogenicity at the silicon atom; no formation of 3 ad was seen with Ph3SiBpin (2 d). Trialkylsubstituted Si−B reagents 2 e–g,19b even with a tBu group attached to the silicon atom, reacted in mediocre yields. Enantio‐ and diastereocontrol were excellent though.

Scheme 3

Scope II: Variation of silylboronic acid ester.[a–c] [a] All reactions were performed on a 0.20 mmol scale with the isolated yield determined after flash chromatography on silica gel. [b] Diastereomeric ratios determined by 1H NMR analysis. [c] Enantiomeric excesses determined by HPLC analysis on chiral stationary phases. [d] Diastereomeric ratio determined by GLC and GC‐MS analysis. n.r.=no reaction.

Running the reaction 1 g→3 ga on a tenfold scale was neither detrimental to yield nor stereoselectivity (Scheme 4, top). From this sample, single crystals suitable for X‐ray diffraction were obtained.20 The absolute and relative configuration of 3 ga was found to be R,S. The stereochemistry of the other silylated cyclopropanes was assigned accordingly. Also, oxidative degradation of the C−Si bond in (R,S)‐3 ga employing the Tamao–Fleming protocol was attempted.21 This transformation is usually low yielding due to competing ring opening.22 The corresponding alcohol was obtained in 6 % yield with d.r.≥98:2 and 92 % ee under retention of the configuration (see the Supporting Information for details).

Scheme 4

Determination of the absolute configuration (top) and deuterium‐labeling experiments (bottom). [a] Deuteration grade estimated by NMR analysis.

To learn about the stereochemical course of the copper‐catalyzed addition of the silicon nucleophile across the C−C double bond, we subjected dideuterated cyclopropene 1 a‐d 2 23 (>99 % 2H) to the standard setup (Scheme 4, bottom). Cyclopropane 3 aa‐d 2 did form in 72 % yield with excellent diastereo‐ (d.r.≥98:2) and enantioselectivity (96 % ee). The syn‐addition of the silylcopper intermediate to the cyclopropene was confirmed by 2D NOE measurements between the ring proton in 3 aa‐d 2 and the methyl groups on the ring and the silicon atom (see the Supporting Information for details). To gain further mechanistic insight, an additional deuterium‐labeling experiment was performed (1 a→3 aa‐d 1, Scheme 4, bottom). MeOH was replaced by CD3OD as an exogenous proton source, and 3 aa‐d 1 was isolated in 71 % yield and 82 % deuterium incorporation. This corroborates that the proton originates from the alcohol additives.

Based on these observations and literature precedence,1, 2 we propose the catalytic cycle shown in Scheme 5. The silicon nucleophile (=silylcopper complex) is generated by transmetalation of the Si−B linkage at the Cu−O bond of the in situ formed copper alkoxide. Cyclopropene 1 then coordinates to copper to form a π‐complex followed by syn‐addition of the Cu−Si bond across the strained alkene.24 Diastereofacial selectivity is likely controlled by sterics with the bond formation occurring on the side of smaller R2 (usually methyl) and opposite to larger R1 (usually aryl). Protonation of the Cu−C bond with MeOH releases the cyclopropane 3 and closes the catalytic cycle.

Scheme 5

Proposed mechanism.

In summary, we described here the first example of a highly enantio‐ and diastereoselective addition of silylboronic acid esters across a broad range of prochiral 3,3‐disubstituted cyclopropenes. It is a syn‐addition that does not rely on a coordinating/directing group. The silyl‐substituted cyclopropanes were obtained in good yields and with superb stereoselectivity. Expansion of this methodology is currently underway in our laboratory.

Conflict of interest

The authors declare no conflict of interest.

As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer reviewed and may be re‐organized for online delivery, but are not copy‐edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.

Supplementary

Click here for additional data file.