Lewis base catalyzed, enantioselective, intramolecular sulfenoamination of olefins.

A method for the enantioselective, intramolecular sulfenoamination of various olefins has been developed using a chiral BINAM-based selenophosphoramide, Lewis base catalyst. Terminal and trans disubstituted alkenes afforded pyrrolidines, piperidines, and azepanes in high yields and high enantiomeric ratios via enantioselective formation and subsequent stereospecific capture of the thiiranium intermediate with the pendant tosyl-protected amine.

Nitrogen-containing, saturated heterocycles such as n class="Chemical">piperidines, pyrrolidines, and azepanes are commonly found in many biologically active natural products and pharmaceutical compounds.[1,2] The chemical and biological properties of these molecules can be greatly influenced by the location and configuration of carbon and heteroatom substituents.[2d] Accordingly, numerous strategies have been developed to generate stereodefined, saturated, nitrogen heterocycles with various types of substitution.[3]

As a part of an ongoing program in these laboratories on the concept of Lewis base activation of Lewis acids,[4] we have focused on the activation of group 16 electrophiles with chiral Lewis bases.[5] Recent reports have described the catalytic, enantioselective sulfenoetherification[6] and carbosulfenylation[7] of n class="Chemical">olefins (Scheme 1). These reactions proceed via the enantioselective generation of thiiranium ions[8] that are constitutionally and configurationally stable at low temperatures which allows them to be captured by nucleophiles without racemization.[9]

Scheme 1

The research reported herein expands upon the previous sulfenofunctionalization reactions to develop a catalytic method for the enantioselective, intramolecular sulfenoamination of olefins.[10] Using an achiral n class="Chemical">sulfenylating reagent in concert with a chiral Lewis base catalyst, an intermediate thiiranium ion is generated which is subsequently captured with a pendant amine-based nucleophile (Scheme 2).

Scheme 2

In the studies of sulfenoetherification and carbosulfenylation reactions, it was shown that a Brønsted acid coactivator was required. As a result, for the sulfenoamination reactions, the n class="Chemical">amine-based nucleophile needed to be not only sufficiently nucleophilic to form the C–N bond but also sufficiently non-basic to avoid protonation under the acidic conditions. To satisfy those criteria, a series of amines were selected as candidates for the nucleophile: sulfonamides, benzamides, carbamates, and phosphinic amides (Chart 1).

Chart 1

To evaluate the reactivity of the nucleophiles, each substrate was subjected to the reaction conditions developed for the sulfenofunctionalization reactions (Table 1).[6,7] Preliminary evaluation of the protected amine substrates 5–11 with n class="Chemical">sulfenylating agent 2 (phenylthiophthalimide, PhthSPh) in the presence of an achiral Lewis base catalyst (tetrahydrothiophene, THT) and a Brønsted acid (MsOH) at room temperature showed that sulfonamides 5–7 rapidly formed piperidines in good yields (entries 1, 3, and 5). Although both tosylamide 5 and nosylamide 6 displayed excellent reactivity in the presence of THT, tosylamide 5 possessed a lower background rate when the Lewis base was omitted (entries 2 and 4). Cyclizations of benzamide 8 (entry 7) and benzyl carbamate 9 (entry 8) were slow under the reaction conditions such that 48 h were required to reach high conversions. Unsurprisingly, tert-butylcarbamate 10 (entry 9) and diphenylphosphinic amide 11 (entry 10) decomposed under the acidic reaction conditions.

Table 1

Survey of Amine Protecting Groups

entry	substrate (R)	Lewis base	timea	yield, %d
1	5 (Ts)	THT	5 minb	93
2	5 (Ts)	none	48 hc	4f
3	6 (Ns)	THT	5 minb	95
4	6 (Ns)	none	48 hc	11f
5	7 (Tris)	THT	5 minb	84
6	7 (Tris)	none	48 hc	2f
7	8 (Bz)	THT	48 hb	86
8	9 (Cbz)	THT	48 hb	81
9	10 (Boc)	THT	–e	–e
10	11 (DPP)	THT	–e	–e

Conversion monitored by TLC.

The time full conversion observed.

The time reaction was quenched.

Isolated yields.

Decomposed under the reaction conditions.

Determined by integration of the 1H NMR spectra of crude reaction mixtures.

Identification of the optimal catalyst involved a survey of various BINAM-derived selenophosphoramides bearing different n class="Chemical">dialkylamine substituents (Table 2). Azepane-substituted selenophosphoramide (R)-3a, which was employed in the analogous sulfenoetherification reaction,[6] afforded 12 with an 11:89 e.r. (entry 1). The reaction with diisobutylamine-substituted selenophosphoramide (S)-3b, which was employed in the carbosulfenylation reaction, afforded the same yield and selectivity as (R)-3a (entry 2). Selenophosphoramides with less bulky substituents, (S)-3c, resulted in a slightly eroded enantiomeric ratio (entry 3). Azocane-substituted catalyst (S)-3d and diisopentylamine-substituted selenophosphoramide (S)-3e afforded the product in lower yields but improved the enantioselectivities of 91:9 and 93:7 e.r., respectively (entries 4 and 5). Interestingly, diisopropylamine-substituted selenophosphoramide (S)-3f, bearing the most sterically encumbered substituent adjacent to the nitrogen, afforded the best enantiomeric ratio of 95:5 e.r. but at a lower rate. The absolute configuration of 12 was determined by reductive removal of the sulfide group and comparison of the optical rotation of the resulting piperidine to literature values.[11]

Table 2

Survey of Chiral Lewis Base Catalysts

entry	catalyst, R2	yield, %a	e.r.b
1	(R)-3a, (CH2)6	90	11.5:88.5
2	(S)-3b, (i-Bu)2	88	89.4:10.6
3	(S)-3c, n-Bu, Et	79	88.1:11.9
4	(S)-3d, (CH2)7	67	91.4:8.6
5	(S)-3e, (i-amyl)2	82	92.8:7.2
6	(S)-3f, (i-Pr)2	75	94.6:5.4

Isolated yields.

The enantiomeric ratio was determined by CSP-SFC analysis.

During the reaction optimization studies, isomerization of piperidine 12 to a pyrrolidine was observed. The combination of THT and 1.0 equiv of MsOH afforded piperidine 12 quantitatively in 5 min at rt (Scheme 3). However, piperidine 12 isomerized into a 1:2.8 mixture of 12 and pyrrolidine 13 by allowing the mixture to stir for 12 h. Independent treatment of either 12 or 13 with 1.0 equiv of MsOH at rt resulted in the establishment of an equilibrium mixture of 12/13 (1:2.8) after 12 h.[15] The isomerization of 12 to 13 alleviates the steric interactions between the N-tosyl group and the 2-phenyl group in piperidine 12.[12]

Scheme 3

In view of the MsOH-induced isomerization of the product, additional studies were performed to evaluate the optimal acid loading for the sulfenoamination reaction (Table 3). With 1.0 equiv of n class="Chemical">MsOH at 0 °C, isomerization of 12 to 13 was observed. However, using less than 1.0 equiv of MsOH greatly reduced the amount of product isomerization. Whereas reactions with loadings of 0.5 and 0.75 equiv of MsOH afforded comparable results (entries 2 and 3), 0.5 equiv led to slightly higher enantioselectivity. The reaction with 0.25 equiv of MsOH displayed a slightly slower reaction rate, reaching full conversion at 24 h (entry 4). Although the reaction with 0.10 equiv gave high e.r., the reaction rate was unacceptably slow (entry 5).

Table 3

Survey of Acid Loadings

entry	MsOH, equiv	conv, %a,b	endo:exob	e.r.c
1	1.00	100	85.7:14.3	91.6:8.4
2	0.75	100	98.9:1.1	92.9:7.1
3	0.50	100	99.2:0.8	93.5:6.5
4	0.25	98	99.4:0.6	93.6:6.4
5	0.10	69	99.5:0.5	93.9:6.1

The conversion was monitored by 1H NMR spectroscopy (6, 12, and 24 h).

Determined by 1H NMR spectroscopy of the crude mixture.

The enantiomeric ratio was determined by CSP-SFC analysis.

The scope of the reaction with various olefins was investigated next (Table 4). The influence of the electronic properties of the n class="Chemical">alkene on reaction rate and stereoselectivity was examined first. Substrate 14, with a 4-anisyl-substituted double bond possessing greater electron density than 5, showed comparable reactivity with a slight drop in enantioselectivity (entry 2). In contrast, substrate 16, bearing a strongly electron-withdrawing 4-trifluoromethylphenyl substituent, afforded only a 39% yield after 48 h (54% conv., entry 3). Interestingly, the observed e.r. (91.9:8.1) for 17 was comparable to that for 15. It is important to note that substrates 5, 14, and 16 (entries 1, 2, and 3) afforded piperidines, as established by 1H NMR spectroscopy. Substrate 18 bearing a non-conjugated olefin afforded a mixture of endo and exo products in a 1:3 ratio in good yields and excellent e.r.’s (entry 4). The reduced endo to exo ratio is likely due to the less-biased electron density of the alkene. In contrast, isopropyl-substituted olefin 20 showed a much greater exo selectivity (along with a high yield and e.r.), thus implicating an important role for the steric bulk of the substituent (entry 5). Interestingly, the reaction of 22, containing 2,2-dimethyl substitution on the tether, afforded a good yield and retained the excellent enantioselectivity (entry 6). However, substrate 24, with 1,1-dimethyl substitution in the tether, resulted in a lower enantioselectivity (entry 7). Other olefins with different substitution patterns were also investigated. Olefin (Z)-5 reacted slowly (75% conv. in 48 h) with poor enantioselectivity (62.8:37.2) (entry 8). The reaction of terminal olefin 27 gave a good yield and high enantioselectivity of 92.5:7.5 with exclusive exo cyclization (entry 9). Carboxamide 29 afforded the product in good yield with constitutional selectivity, but showed diminished e.r. (entry 10). Presumably, protonation of the carbonyl group attenuates the nucleophilicity of the nitrogen and prevents rapid capture of the intermediate thiiranium ion, thus allowing racemization.

Table 4

Scope of the Enantioselective Intramolecular Sulfenoamination Reaction

Isolated yields of analytically pure material.

Constitutional selectivity determined by 1H NMR spectroscopy of the crude mixture.

The enantiomeric ratio of the major constitutional isomer was determined by CSP-SFC analysis, and the absolute configurations of the products were assigned by analogy to 12.

Incomplete conversion on quenching at 48 h.

The influence of tether length on cyclization was also investigated. Two-carbon-tethered substrate 31 cyclized to n class="Chemical">pyrrolidine 32 in 86% yield and 91.3:8.7 e.r. with complete endo selectivity (entry 11). Interestingly, four-carbon-tethered substrate 33 showed the impact of conjugation on biasing the two olefinic carbons by affording exclusively azepane 34 (entry 12). The structure and the absolute configuration of 34 were established by X-ray crystallography.[13] In contrast, the non-conjugated substrates 35 and 37 afforded only piperidine products via exo cyclization, indicating the preference to form the six-membered rings for dialkyl-substituted olefins. Additionally, reactions with both 35 and 37 gave the products in good yields and excellent enantioselectivities.

The proposed catalytic cycle for the sulfenoamination reaction is shown in Figure 1.[14] Sulfenylating agent 2 is activated with n class="Chemical">MsOH and then transfers the sulfenyl moiety to the Lewis base (S)-3f, forming the chiral sulfenylating complex .[7c] Subsequent transfer of the sulfenium ion from to the alkene furnishes the enantioenriched chiral thiiranium ion intermediate . Finally, capture of the thiiranium ion with the pendant tosylamide and subsequent deprotonation affords the enantioenriched product.

Figure 1

Proposed catalytic cycle for the sulfenoamination.

In conclusion, a Lewis base catalyzed, enantioselective, intramolecular sulfenoamination of unactivated olefins has been developed. The reaction produces saturated n class="Disease">N-heterocyclic rings with high enantioselectivities for a wide range of trans olefins. Extensions to intermolecular sulfenoamination reactions are under investigation.