Diastereo- and Enantioselective CuH-Catalyzed Hydroamination of Strained Trisubstituted Alkenes

Abstract

Amine-substituted cyclobutanes and cyclopropanes are important substructures in biologically active compounds. Moreover, many of the cycloalkane units bear multiple substituents and stereocenters. Therefore, synthetic methods that produce polysubstituted aminocyclobutanes and aminocyclopropanes in a highly diastereo- and enantioselective manner are of importance. Herein, we describe the diastereo- and enantioselective synthesis of various types of polysubstituted aminocyclobutanes and aminocyclopropanes through CuH-catalyzed hydroamination of 1-substituted cyclobutenes and cyclopropenes. These strained trisubstituted alkenes exhibit much higher reactivity compared to their unstrained analogues in the initial hydrocupration step of the reaction. Moreover, an interesting reversal of regioselectivity was observed in the hydroamination of 1-aryl-substituted cyclobutenes compared to the cyclopropene analogues. The origins of the enhanced reactivity of strained trisubstituted alkenes as well as the differences in the regio- and enantioselectivity between reactions with cyclobutenes and cyclopropenes were investigated computationally.

Graphical Abstract

INTRODUCTION

Cyclobutylamine and cyclopropylamine substructures are found in a variety of bioactive molecules and pharmaceutical compounds (Figure 1a).¹ Moreover, the stereoisomers of these compounds can exhibit remarkable differences in bioactivity.² Thus, considerable effort has been expended to developing methods for the stereoselective construction of these structural units. Currently, most synthetic approaches to enantioenriched cyclobutylamines are based on [2+2] cycloadditions³ and the Amadori–Heyns rearrangement.⁴ Among these methods, few are catalytic and effective in an intermolecular context.^3f–h,4 In contrast, for the catalytic, enantioselective synthesis of cyclopropylamines, a number of elegant methods have been developed.⁵ The most well-known processes include intramolecular C–H activation of prochiral aminocyclopropanes,⁶ cyclopropanation of vinylcarbamates,⁷ carboamination of cyclopropenes,⁸ and rare-earth-metal-catalyzed hydroamination of cyclopropenes.⁹ Despite these advances, the development of a unified synthetic strategy that allows for the stereoselective formation of multiple types of polysubstituted cyclobutyl- and cyclopropylamines would be desirable. We proposed that a general, enantioselective hydroamination of cyclic alkenes could address this challenge.

Figure 1.

Proposed CuH-catalyzed hydroamination of 1-substituted cyclobutenes and cyclopropenes. (a) Representative biologically active cyclobutylamines and cyclopropylamines. (b) Proposed catalytic strategy. (c) DFT-computed activation barriers of the hydrocupration of 1a-1c. Geometries were optimized at the B3LYP/SDD–6–31G(d) level of theory. Single point energies were calculated at the M06/SDD–6-311+G(d,p)/SMD(THF) level.

Our group and others have reported copper-hydride(CuH)-catalyzed enantioselective hydrofunctionalization reactions of various unsaturated substrates.^10,11 In particular, CuH-catalyzed hydroamination has been applied on a broad range of olefins,¹² such as styrene derivatives,^12a–e 1,1-disubstituted alkenes,^12d–f and unactivated trans-1,2-disubstituted alkenes^12g. These reactions proceed through enantioselective hydrocupration of the alkene to form a chiral alkylcopper species, which is then trapped by an electrophile such as a hydroxylamine ester.¹³ We postulated that the CuH-catalyzed stereoselective hydroamination of 1-substituted cyclobutenes and cyclopropenes could potentially furnish cyclobutyl- and cyclopropylamines bearing various types of substituents and with adjacent stereocenters (Figure 1b).

Despite the generality and broad utility of published CuH-catalyzed approaches, a common obstacle has been the unwanted CuH-catalyzed reduction of the electrophile, a side reaction that depletes the electrophilic reagent.^11b,12c,14 For unactivated substrates with high barriers for reaction with CuH (hydrocupration), the reduction of the electrophile is a significant side reaction, resulting in greatly diminished yields or even no product formation. For instance, CuH-catalyzed hydrofunctionalization of unactivated trisubstituted alkenes has been extremely challenging in general. However, the hydroamination of certain activated trisubstituted alkenes such as allylic alcohol derivatives^12a,15 and β,β-disubstituted styrenes^12a have been achieved. Activation of olefins through ring strain is a strategy that has long been utilized by organic chemists to achieve new strain-enabled reactivities.¹⁶ Thus, we wondered whether the partial release of the ring strain energy of 1-substituted cyclobutenes and cyclopropenes in the hydrocupration step would help lower their hydrocupration barriers, and thereby allow the hydroamination of these strained trisubstituted alkenes to proceed faster than competing reduction of the electrophile.

At the outset, it was unclear to us whether these strained trisubstituted alkenes¹⁷ would be reactive enough to undergo CuH-catalyzed hydroamination reactions. Thus, we performed DFT calculations on the hydrocupration barriers of strained and unstrained trisubstituted alkenes (Figure 1c). The DTBM-SEGPHOS(L1)-based CuH catalyst system has been shown to exhibit high levels of enantioselectivity as well as enhanced reactivity in hydroamination reactions^12,18 and was therefore chosen in our calculations. Our results indicated that the hydrocupration barriers of 1-methylcyclobutene (1b) and 1-methylcyclopropene (1a) are 3.5 and 11.0 kcal/mol lower, respectively, than that of the acyclic trisubstituted olefin 2-methylbut-2-ene (1c). Based on these promising results, we proceeded to experimentally investigate the CuH-catalyzed hydroamination of 1-substituted cyclobutenes and cyclopropenes.

RESULTS AND DISCUSSION

Hydroamination of 1-Substituted Cyclobutenes.

We initiated our study by examining the CuH-catalyzed hydroamination reaction of 1,3-diphenyl-3-methylcyclobut-1-ene (2a) using Bn₂NOBz (3a). In previous CuH-catalyzed hydroamination reactions, simple styrene derivatives were generally converted selectively to the Markovnikov isomers of products^12a–b (Figure 2a). However, when α-methyl-styrene^12f was employed, the anti-Markovnikov isomer was instead favored by a 7:1 ratio (Figure 2b), indicating that tertiary alkyl copper species are perhaps hard to form by hydrocupration or react slowly with the electrophile. In contrast to this observation, we found that, using Cu(OAc)₂ as the precatalyst and DTBM-SEGPHOS (L1) as the ligand, the reaction of phenylcyclobutene (2a) with Bn₂NOBz (3a) afforded the Markovnikov product exclusively in 85% yield (Figure 2c). The reaction not only generated a fully substituted carbon center on the cyclobutane ring, but also formed the 1,1,3,3-tetrasubstituted aminocyclobutane product with a 13:1 cis/trans ratio. 3-Substituted 1-arylcyclobutylamine subunits, though found in a number of interesting molecules,^1b,19 have been difficult to prepare through the diastereoselective installation of 3-substituents,²⁰ and highly stereoselective approaches to directly access these structures are rare.²¹ Our CuH-catalyzed hydroamination approach on 3,3-disubstituted 1-arylcyclobutenes can provide a method to rapidly construct a diverse range of 3,3-disubstituted 1-aryl-1-aminocyclobutanes in good stereoselectivities.

Figure 2.

Regioselectivity of the CuH-catalyzed hydroamination using different phenyl- substituted alkenes.

After identifying the optimal reaction conditions (shown in Figure 2c),²² we investigated the substrate scope for the hydroamination of 1-arylcyclobutenes (Table 1). In all cases, the Markovnikov products were formed exclusively. We found that 1-arylcyclobutenes bearing para- (4b, 4h), meta- (4c), and ortho- (4d) substituents were all suitable for the hydroamination reaction. An electron-withdrawing group on the arene (4c) greatly improved the stereoselectivity of the reaction, while electron-donating substituent (4b) led to a slightly diminished yield and diastereomer ratio. A 1-pyridyl cyclobutene was also well tolerated (4e). Moreover, arylcyclobutenes without any substituents at the 3-position were also able to undergo the hydroamination reaction to selectively form the Markovnikov products (4d, 4e) in good yields, suggesting that the regioselectivity was not a result of the steric repulsion from the 3-substituents. A 1-phenylcyclobutene with a spiro-fused cyclohexyl group at the 3-position also reacted efficiently (4f).

Table 1.

Substrate Scope for the CuH-Catalyzed Hydroamination of 1-Arylcyclobutenes^a

^aIsolated yields on 0.5 mmol scale (average of two runs).

^bRatio refers to the ratio of major and minor stereoisomers.

^cReaction was carried out with 1.2 equiv of Bn₂NOPiv, THF (0.5 M) at 40 °C.

^dReaction was carried out with (R)-DTBM-SEGPHOS instead.

^e4 equiv of (MeO)₂MeSiH was used.

We next evaluated a range of hydroxylamine esters in this reaction. It was found that a number of functional groups such as an alcohol (4h), an ester (4i), and a phenol (4i) were tolerated under the hydroamination reaction conditions. Moreover, heterocycles such as pyrimidine (4g) and furan (4h) were shown to be compatible in the reaction.

We also examined the hydroamination of 1-alkylcyclobutenes, which lack the activating influence of an aryl substituent on the alkene. We chose (3-(cyclobut-1-en-1-yl)propyl)benzene (5a) as our model substrate, for which a series of amination reagents were evaluated.²² We found that the use of the mesitoyl hydroxylamine ester 6a gave the highest yield. Under the optimal reaction conditions, hydroamination of 5a provided the 1,2-disubstituted aminocyclobutane product 7a in 79% yield with >99.5:0.5 er and >20:1 dr (Table 2). That the regioselectivity is totally opposite of that observed with the aryl-substituted cyclobutene substrates is consistent with what we observed in, e.g., a comparison of the hydroamination of styrene and 1-dodecene.^12a

Table 2.

Substrate Scope for the CuH-Catalyzed Hydroamination of 1-Alkylcyclobutenes^a

^aIsolated yields on 0.5 mmol scale (average of two runs).

^bReaction was carried out with 1.2 equiv of Bn₂NOC(O)Mes.

A range of functional groups and heterocycles were found to be compatible with the reaction conditions (Table 2). For example, amination reagents containing a thiophene (7b) and an acetal (7c) as well as 1-alkylcyclobutenes bearing a silyl ether (7d) and a pyridine (7e) were all suitable coupling partners in this hydroamination reaction, each providing the corresponding product in good yield with >99.5:0.5 er and >20:1 dr (Table 2).

To demonstrate that our hydroamination method is also applicable to 1-silyl substituted four-membered cycloalkenes, we carried out the hydroamination reaction of 1-silyl-4-azacyclobutene (5d) with the amination reagent (6a), which resulted in the formation of an aminoazetidine product (7f) in excellent yield, enantio- and diastereoselectivity (Table 2).

Hydroamination of 1-Substituted Cyclopropenes.

We were also interested in applying the hydroamination chemistry to other strained trisubstituted olefins, and thus we turned our attention to the hydroamination of 1-substituted cyclopropenes. While exploring different types of cyclopropenes, we had two interesting observations regarding the selectivity of these reactions. First, in contrast to the regioselectivity observed with 1-arylcyclobutenes (2) (Scheme 1b), the formation of the anti-Markovnikov hydroamination product was found to be preferred when using 1-phenylcyclopropene derivative 8 as the substrate (Scheme 1a). Second, while the hydroamination of 1-alkylcyclobutenes were able to proceed with excellent enantioselectivity (Scheme 2b), the reaction with 1-alkylcyclopropene 11 provided the hydroamination product 12 with only 55.5:44.5 er (Scheme 2a).

Scheme 1.

Comparison of the regioselectivity in CuH-catalyzed hydroamination of 1-phenylcyclopropene and 1-arylcyclobutenes

Scheme 2.

Comparison of the enantioselectivity in the CuH-catalyzed hydroamination of 1-alkylcyclopropene and 1-alkylcyclobutene

We reasoned that installation of a bulky group at the 1-position of the cyclopropene may help restore the enantioselectivity due to increased ligand-substrate repulsion in the disfavored hydrocupration transition state. Thus we investigated the hydroamination of 1-silyl cyclopropenes. First, we examined the hydroamination reaction with 1-silyl-3,3-dimethylcyclopropene (13a) and Bn₂NOPiv (9) (Table 3), finding that the reaction proceeded smoothly to give the 1,2-disubstituted aminocyclopropane product (15a) in 70% yield, with 98.5:1.5 er and >20:1 dr. As previously shown by our DFT calculations, cyclopropenes hydrocuprate much faster than cyclobutenes. As a result, we discovered that 1-silylcyclopropene (13a) was even able to react with 1,2-benzisoxazole (14), an electrophile that is highly susceptible to competing Kemp elimination in the presence of CuH and therefore couples only with the most activated olefins.^{12e, 30} The protected primary amine product 15b was obtained in 63% yield with excellent enantio- and diastereoselectivity. Moreover, 1-silylcyclopropenes bearing 3-spirocycloalkyl substituents were also capable of reacting with 1,2-benzisoxazole (14) to give the corresponding hydroamination products (15c, 15d) in moderate or good yields and with high stereoselectivities. The latter is related to a key intermediate for the synthesis of a T-type Ca_V3 channel inhibitor (Table 3).^1e

Table 3.

Substrate Scope for the CuH-Catalyzed Hydroamination of 1-Silylcyclopropenes^a

^aIsolated yields on 0.5 mmol scale (average of two runs).

^bCu(OAc)₂ (5 mol%), (R)-DTBM-SEGPHOS (5.5 mol%), 1,4-dioxane (0.5 M).

^cCu(OAc)₂ (2 mol%), (R)-DTBM-SEGPHOS (2.2 mol%), THF (1.0 M).

^d14 was added via syringe pump over 2 h.

^e14 was added via syringe pump over 2.5 h.

COMPUTATIONAL STUDIES

Our experimental results not only demonstrated the generality of the CuH-catalyzed hydroamination of cyclopropenes and cyclobutenes, but also led to several interesting mechanistic questions regarding reactivity and selectivity. First, what is the origin of the enhanced reactivities of 1-substituted cyclobutenes and cyclopropenes as compared to acyclic trisubstituted alkenes? Second, why do the hydroamination reactions with 1-phenylcyclobutene (2a) and 1-phenylcyclopropene (8) form opposite regioisomers? Lastly, why is the hydroamination with 1-alkylcyclobutene (5a) highly enantioselective, while the reaction with 1-alkylcyclopropene (11) occurs with low enantioselectivity? To address these questions, we performed density-functional theory (DFT) calculations to reveal factors that control reactivity, regio-, and enantioselectivity in the CuH-catalyzed hydroamination of strained cyclic alkenes. We surmised that the angular strain,²³ the ease to distort the alkenyl carbon to a pyramidalized transition state geometry,²⁴ and the diminished steric repulsions with the DTBM-SEGPHOS ligand may all affect the reactivity and selectivity of cyclopropenes and cyclobutenes. Therefore, we employed the distortion/interaction model²⁵ to analyze the effects of catalyst/substrate distortion and the interaction energies between the CuH catalyst and the alkene in the hydrocupration transition state.

Computational Details.

Geometries were optimized in the gas phase using the B3LYP²⁶ functional and a mixed basis set of SDD for Cu and 6-31G(d) for other atoms. Single point energies were calculated with the M06²⁷ functional and a mixed basis set of SDD for Cu and 6-311+G(d,p) for other atoms. Solvation energy corrections were considered in tetrahydrofuran (THF) solvent using the SMD²⁸ solvation model. All geometry optimizations and single-point energy calculations were performed using Gaussian 09.²⁹

A modified version¹⁸ of the distortion/interaction model (or activation strain model),²⁵ namely the ligand-substrate interaction model, was employed to decompose the activation energy (ΔE^‡) using Eq. 1.

$$\Delta E^{‡}=\Delta E_{\text{sub-dist}}+\Delta E_{\text{cat-dist}}+\Delta E_{\text{int-space}}+\Delta E_{\text{int-bond}}$$ (Eq. 1)

where ΔE^‡ is the gas-phase electronic energy of the hydrocupration transition state with respect to the separated alkene substrate and the L*CuH catalyst; ΔE_sub-dist and ΔE_cat-dist are the energies to distort the alkene substrate and the catalyst into the transition state geometries, respectively; ΔE_int-space is the through-space interaction energy between the (R)-DTBM-SEGPHOS ligand and the substrate calculated from the interaction energy of a supramolecular complex of the ligand and the substrate at the transition state geometry in the absence of the CuH moiety; ΔE_int-bond is the through-bond interaction energy between the catalyst and the substrate calculated from ΔE_int-bond = ΔE^‡ − ΔE_sub-dist – ΔE_cat-dist − ΔE_int-space. The overall distortion energy of the catalyst and the substrate (ΔE_dist) is calculated from ΔE_dist = ΔE_sub-dist + ΔE_cat-dist.

Origin of the Enhanced Reactivity of Strained Trisubstituted Alkenes.

We performed the ligand-substrate interaction model analysis to investigate the origin of the enhanced reactivities of 1-methylcyclopropene (1a) and 1-methylcyclobutene (1b) in the hydrocupration as compared to the acyclic trisubstituted alkene, 2-methylbut-2-ene (1c) (Figure 3). Using this approach, the computed activation energy (ΔE^‡) is dissected to distortion energies of the substrate and the catalyst (ΔE_sub-dist and ΔE_cat-dist), and the through-space and through-bond interaction energies between the L*CuH catalyst and the substrate (ΔE_int-space and ΔE_int-bond) (Figure 3b). Among the four different energy components, the main factor that promotes the hydrocupration of 1-methylcyclopropene (1a) is the highly favorable through-bond interaction energy (∆E_int-bond = −33.9 kcal/mol). The strong catalyst-substrate interaction in TS-1a is due to the prominent pyramidalization of both sp² carbons of 1a as evidenced by the out-of-plane dihedral angles of the C1-Me and C2-H groups (α_Me and α_H, Figure 3a). Frontier molecular orbital (FMO) theory analysis indicates the pyramidalization of 1-methylcyclopropene decreases its LUMO energy, and thus promotes the FMO interactions between the alkene π* orbital and the HOMO of CuH (σ_Cu-H, see SI for details). Interestingly, although the pyramidalization of 1a in TS-1a is much more significant than that of 1b and 1c in TS-1b and TS-1c, the energies to distort these substrates are comparable (∆E_sub-dist = 22.4, 22.2, and 23.0 kcal/mol, respectively). This observation is consistent with previous reports that indicated easier distortion of cyclopropene as compared to cyclobutene and acyclic alkenes.²⁴ Because the sp² carbons of cyclopropene have significant angular strain,²³ pyramidalization of cyclopropene is promoted by strain release. The propensity of out-of-plane distortion of 1-methylcyclopropene 1a is further demonstrated in Figure 3c, where the distortion energies of three alkenes (1a, 1b, and 1c) are plotted against the out-of-plane dihedral angle of the alkenyl Me and H groups. In the transition state region (α = 120~140°), 1-methylcyclopropene (1a) requires much smaller distortion energy than 1-methylcyclobutene (1b) and 2-methylbut-2-ene (1c). The ligand-substrate interaction model analysis also revealed the impact of catalyst distortion energy (∆E_cat-dist) on the reactivity. TS-1a and TS-1b both have smaller catalyst distortion energies than TS-1c. This indicates the smaller sizes of the strained cyclic alkenes as compared to 2-methylbut-2-ene (1c) also contribute to the reactivities of these substrates through decreasing steric repulsions with the L*CuH catalyst.

Figure 3.

Origin of enhanced hydrocupration reactivity of strained cyclic alkenes 1a and 1b. All energies are in kcal/mol.

Taken together, the above analysis indicates the greater reactivities of 1-methylcyclopropene (1a) and 1-methylcyclobutene (1b) in hydrocupration are due to the combination of two effects. First, the ease to distortion of strained cyclic alkenes leads to greater pyramidalization of the alkenyl carbons in 1-methylcyclopropene, which in turn promotes the bonding interactions with the CuH catalyst. Second, the smaller sizes of cyclopropene and cyclobutene than the acyclic analogues decrease the catalyst-substrate steric repulsions in the hydrocupration transition state.

Origin of the Regioselectivity Reversal in the Hydroamination Reactions with 1-Phenylcyclobutene and 1-Phenylcyclopropene Derivatives.

Next, we computed the regioisomeric hydrocupration transition states with 1-phenylcyclobutene and 1-phenylcyclopropene derivatives 2a and 8 (Figure 4). These substrates were chosen in the computational study because their hydroamination reactions lead to opposite regioisomers (Table 1 and Scheme 1). Our DFT calculations indicated that the hydrocupration of 2a favors the formation of the tertiary benzylic copper intermediate 16a by 1.9 kcal/mol. By contrast, in the reaction with 8, hydrocupration to form the secondary alkylcopper intermediate 17b is favored by 0.8 kcal/mol. These computed regioselectivities of hydrocupration are consistent with the experimentally observed hydroamination regioselectivities with these substrates. Although we have not yet computationally confirmed the regioselectivity-determining step in the catalytic cycle, the exergonicity of the hydrocupration of strained cyclic alkenes 2a and 8 (−13.9 and −29.5 kcal/mol, respectively, see SI for details) suggests that the hydrocupration is most likely irreversible and thus regioselectivity-determining.

Figure 4.

Origin of the reversed regioselectivities of the hydroamination of 1-phenylcyclobutene derivative (2a) and 1-phenylcyclopropene derivative (5). All energies are in kcal/mol.

The origin of the regioselectivity reversal was analyzed using the ligand-substrate interaction model, as shown in Figure 4b. In the hydrocupration of 1-phenylcyclopropene derivative 8, regioisomer TS8-b is more favorable because of the strongly stabilizing through-bond interactions between the L*CuH catalyst and the substrate (∆E_int-bond = −34.5 kcal/mol). At first glance, these results are counter-intuitive because TS8-b forms a secondary alkyl-copper bond which is expected to be less electronically favorable than the formation of the benzylic copper bond via TS8-a. Closer examination of the four-membered cyclic hydrocupration transition states (TS8-a and TS8-b) revealed an unusual rhombus-shaped geometry, in which the diagonal Cu–C bond is shorter than the forming Cu-C bond. Therefore, TS8-b is stabilized by the favorable bonding interaction between the Cu center and the benzylic carbon due to the short Cu–Cα distance (2.08 Å). By contrast, TS8-a has a much longer distance between Cu and the benzylic carbon (2.20 Å) that leads to a less favorable through-bond interaction energy. In the hydrocupration with 1-phenylcyclobutene derivative 2a, the through-bond interactions in both regioisomeric transition states are weaker than those in the transition states with 1-phenylcyclopropene derivative 5 because of smaller degrees of pyramidalization of 2a (vide supra). Nonetheless, TS2a-b still has more favorable through-bond interactions than TS2a-a (∆∆E_int-bond = −3.2 kcal/mol) because of the shorter Cu–Cα distance (2.18 and 2.24 Å in TS2a-b and TS2a-a, respectively). However, TS2a-b requires a much higher energy (∆E_dist = 34.8 kcal/mol) to distort the cyclobutene derivative 2a to facilitate the through-bond interactions with CuH. Therefore, the regioselectivity in the reaction with 2a is distortion-energy controlled (∆∆E_dist = 5.2 kcal/mol) and favors the formation of the benzylic copper intermediate (16a) via TS2a-a.

Enantioselectivity of the Hydroamination Reactions with 1-Alkylcyclobutene and 1-Alkylcyclopropene.

Finally, we investigated the origin of the notably different enantioselectivities in the hydroamination of 1-alkylcyclobutene and 1-alkylcyclopropene (Table 2 and Scheme 2). We computed the hydrocupration transition states with the two different π faces of 1-methylcyclobutene 1b and 1-methylcyclopropene 1a (Figure 5). In the reaction with 1b, DFT calculations predicted strong preference for TS-1b that leads to the experimentally observed (1R,2R)-aminocyclobutane. The computed enantioselectivity (∆∆G^‡ = 2.7 kcal/mol) is comparable to the difference between the distortion energies of the hydrocupration transition states (∆∆E_dist = 2.6 kcal/mol), indicating the enantioselectivity is controlled by steric effects that lead to distortions of the catalyst and the substrate. Indeed, the less favorable transition state TS-1b’ is destabilized due to steric repulsions between the cyclobutene moiety and the P-aryl group in the more-occupied quadrant (quadrant II). The ligand-substrate steric repulsions in TS-1b’ are evidenced by the short C…H distance of 2.52 Å between the P-aryl group and the methylene group on cyclobutene. The C…H distance between the P-aryl group and the 1-methyl substituent is much longer (2.79 Å in TS-1b’), indicating that the steric repulsions with the cyclobutene moiety, rather than the 1-substituent, dictate the enantioselectivity.

Figure 5.

Origin of enantioselectivities in the hydroamination of 1-alkylcyclobutene and 1-alkylcyclopropene. All energies are in kcal/mol.

In the hydrocupration of 1-methylcyclopropene (1a), the two enantiomeric transition states TS-1a and TS-1a’ have comparable activation energies. This is consistent with the low e.r. observed in the hydroamination of 1-alkylcyclopropene 11. Their similar distortion energies (30.1 and 30.9 kcal/mol for TS-1a and TS-1a’, respectively) indicate that the ineffective stereoinduction is due to comparable ligand-substrate steric repulsions in both enantiomeric transition states. The transition state quadrant diagrams in Figure 5b show that due to the smaller size of the cyclopropene moiety compared to the cyclobutene, steric repulsions with the P-aryl group in quadrant II of TS-1a’ are diminished. This is evidenced by the much longer C…H distance (2.82 Å) between the P-aryl group and the methylene on the cyclopropene in TS-1a’.

CONCLUSION

In summary, we have developed the diastereo- and enantioselective CuH-catalyzed hydroamination reactions of 1-substituted cyclobutenes and cyclopropenes. DFT studies showed that strained trisubstituted olefins exhibit enhanced rates of hydrocupration compared to unstrained trisubstituted analogues, which allows for the effective hydroamination reactions of these substrates. For 1-arylcyclobutenes, Markovnikov products were selectively formed in the hydroamination reactions and a tetrasubstituted carbon center was generated in the cyclobutane product. By contrast, the opposite regioselectivity was observed for the hydroamination of 1-phenylcyclopropene derivatives. DFT studies revealed the Markovnikov-selectivity with 1-arylcyclobutenes is due to a smaller distortion energy in the hydrocupration transition state to form the benzylic copper intermediate, while the anti-Markovnikov-selectivity with 1-arylcyclopropenes is controlled by catalyst-substrate through-bond interactions. Moreover, the hydroamination reactions of 1-alkylcyclobutenes as well as 1-silyl substituted three- and four-membered cycloalkenes were shown to produce a variety of aminocyclobutanes and aminocyclopropanes bearing contiguous stereocenters in excellent enantio- and diastereoselectivity. We also showed that the small size of the cyclopropene moiety in 1-alkylcyclopropenes leads to insufficient ligand-substrate steric interactions for the chiral induction in hydrocupration. Accordingly, the hydroamination of 1-alkylcyclobutenes proceeds with considerably higher levels of enantioselectivity compared to 1-alkylcyclopropenes. We anticipate that our studies on the scope, regio-, and enantioselectivity of CuH-catalyzed hydroamination using various types of strained trisubstituted alkenes can facilitate further developments in asymmetric hydrofuctionalization of strained alkenes.