Sulfamyl Radicals Direct Photoredox-Mediated Giese Reactions at Unactivated C(3)–H Bonds

Abstract

Alcohol-anchored sulfamate esters guide the alkylation of tertiary and secondary aliphatic C(3)–H bonds. The transformation proceeds directly from N–H bonds with a catalytic oxidant, a contrast to prior methods which have required pre-oxidation of the reactive nitrogen center, or employed stoichiometric amounts of strong oxidants to obtain the sulfamyl radical. These sulfamyl radicals template otherwise rare 1,6-hydrogen-atom transfer (HAT) processes via seven-membered ring transition states to enable C(3)–H functionalization during Giese reactions.

Graphical Abstract

INTRODUCTION

Nitrogen-centered radicals are an important and versatile class of chemical intermediates.¹ Yet, nitrogen-centered radicals remain underutilized, as most methods for their generation rely on harsh conditions to oxidize the nitrogen center. Recently, photocatalytic strategies have been developed as mild processes to form nitrogen-centered radicals,^2–6 however, only amides, carbamates, and sulfonamides have served as precursors to neutral nitrogen-centered radicals.^4–6 These groups template position-selective C–H functionalization technologies, transforming C(4)–H bonds through 1,5-HAT processes (Scheme 1A).⁷ In-contrast, sulfamate esters guide functionalization to C(3)–H centers through otherwise rare 1,6-HAT processes (Scheme 1B), providing complementary positional selectivity to established processes.^8–16 As a complement to known guided methods, sulfamate esters are attractive directing groups because they derive from alcohols, which are ubiquitous in biologically active small molecules. To date, protocols templated by sulfamate ester substrates require pre-oxidation of the reactive nitrogen center, or the use of strong stoichiometric oxidants to access nitrogen-centered sulfamyl radicals.^{14, 15} As such, a strategy to facilitate sulfamyl radical formation directly from N–H bonds under mild conditions would enhance the substrate tolerance of these directing motifs, and could also enable previously unrealized synthetic disconnections.^{17, 18} Herein disclosed is the first catalytic process to access sulfamyl free radicals directly from N–H bonds.¹⁹ These sulfamyl radicals have been engaged in Giese reactions, in the only examples of C(3)–H alkylation reactions guided by alcohol surrogates (Scheme 1C, 1 → 2).

Scheme 1.

Sulfamate esters offer complementary site selectivity to developed templates for C–H alkylation reactions

RESULTS AND DISCUSSION

At the outset of these investigations, we sought conditions that would facilitate oxidation of sulfamate ester 1a to sulfamyl radical 3a. We envisioned that this could occur through an initial deprotonation to provide sulfamate ester anion 4a. Anion 4a could undergo single electron oxidation to generate sulfamyl radical 3a, with concurrent reduction of the excited iridium catalyst 6a (Scheme 2). Consistent with this proposal, in acetonitrile, sodiated 5-methylhexyl N-tert-butyl sulfamate ester anion 4a has a half-peak potential (E_p/2) of +0.753 V versus saturated calomel electrode (SCE), indicating that oxidation of the anion by a strongly oxidizing and broadly used iridium(III) photocatalyst [Ir(dF(CF₃)ppy)₂(dtbbpy)]PF₆ (6a, E_1/2[Ir^III*/Ir^II]= +1.21 V versus SCE, dF(CF₃)ppy = 2-(2,4-difluorophenyl)-3-trifluoromethylpyridine, dtbbpy = 4,4’-di-tert-butyl-2,2’- bipyridine) is thermodynamically feasible.²⁰ Stern-Volmer analysis of the related sulfamate ester anion 4b quenches 6a. Based on this data, reductive quenching of [Ir^III*] 6a to [Ir^II] 7a by an in situ generated sulfamate ester anion, such as 4a, can furnish sulfamyl radical 3a. Alternatively, oxidation could also occur through a concerted proton-coupled electron transfer (PCET) event. Presently, we cannot exclude the possibility that electron transfer could proceed through a concerted PCET event.

Scheme 2.

Proposed mechanism for sulfamate ester directed Giese reaction

By analogy to previously reported sulfamyl radical-mediated C–H functionalization reactions,^{14, 15} 3a is poised to guide an intramolecular 1,6-HAT process to furnish carbon-centered radical 8a. Nucleophilic radical 8a engages electron-deficient olefin 9a in a Giese reaction. At this point, reduction of an ester stabilized radical (E⁰[R•/R–] ≈ –0.63 V versus SCE),²¹ such as 10a, by strong reductant 7a (E_1/2[Ir^III/Ir^II]= –1.37 V versus SCE)²⁰ to furnish carbon anion 11a should be thermodynamically favorable. Finally, anion protonation would provide C(3)-alkylated product 2a.

To develop this guided Giese reaction, we employed tert-butyl acrylate (9a) as a radical trapping agent with 5-methylhexyl N-tert-butylsulfamate ester (1a, Table 1). With this system, we could probe the ability of this process to overcome the innate position-selectivity that dictates the site of unguided C–H functionalization reactions. In undirected processes, functionalization would be expected to proceed preferentially at the weakest and the most electron-rich C–H bond in the molecule, which would be the distal tertiary C–H center (bond dissociation energy (BDE) ≈ 96 kcal/mol).²² The targeted C(3)–H bond is expected to be stronger (BDE ≈ 98 kcal/mol), and thus less likely to engage in an unguided functionalization. Nevertheless, the directed Giese reaction proceeds in the presence of light, photocatalyst 5a, and K₂CO₃ to afford desired 2a in 77% yield (entry 1). Using a range of strongly oxidizing excited state photocatalysts 5a–5e, those that become stronger reductants prove more efficient in this transformation (entries 1, 2, 5). Consistent with mechanistic hypotheses, control experiments confirm that the photocatalyst, light, base, and inert atmosphere are required for product formation (See supplementary information).

Table 1.

Photocatalysts that are stronger reductants upon initial quenching are more efficient

entrya	photo-catalyst	E1/2(PC*/PC−) [V]b	E1/2(PC/PC−) [V]b	1a [%]c	yield 2a [%]d
1	5a	+1.21	−1.37	17	77
2	5b	+0.97	−1.23	46	44
3	5c	+1.68	−0.69	98	nde
4	5d	+1.65	−0.79	96	nde
5	5e	+1.37	−1.21	12	34

^aGeneral reaction conditions: 1.0 equiv sulfamate ester, 2.2 equiv tert-butyl acrylate, two 34 W blue lamps, photocatalyst (1.0 mol %), 1.0 equiv K₂CO₃, MeCN (0.2 M), 900 mot/min stirring, 21–28 °C, 48 h.

^bV versus SCE. ^{5, 20, 23}

^cRecovered 1.

^dIsolated yields.

^eNot detected in ¹H NMR of the crude reaction mixture.

Under the optimized reaction conditions, the methylene center in sulfamate ester 1a engages selectively in a single alkylation event to furnish Giese product 2a, which incorporates a tertiary C(5)–H bond. This tertiary C–H bond is weaker than the secondary C(3)–H bond in substrate 1a, and might be expected to be more susceptible to further alkylation to furnish fully-substituted 12. To explain this selectivity, we hypothesize that the newly generated tertiary C(3)–H bond may prove too sterically hindered to engage in further functionalization. Consistent with this hypothesis, we found that less sterically encumbered 1b–1d undergo initial monoalkylation with tert-butyl acrylate, followed by a second Giese reaction to generate fully-substituted 12b–d (Scheme 3). Furthermore, compounds with distal branching patterns (i.e. 1e–h) react to furnish exclusive monoalkylated products 2e–h.

Scheme 3. Secondary centers engage in productive Giese reactions^a,b.

^a General reaction conditions: 1.0 equiv sulfamate ester, 2.2 equiv electrophilic olefin, two 34 W blue lamps, photocatalyst (1.0 mol %), 1.0 equiv K₂CO₃, MeCN (0.2 M), 900 mot/min stirring, 21–28 °C, 48h. ^b Isolated yields. ^c 11 days, additional 2.2 equiv acrylate added after 142 h. ^d 144 h, 4.4 equiv ethyl acrylate, and additional 4.4 equiv methyl acrylate added after 48 and 96 h.

In more complex substrates, C(3)-methylene centers react with substrate-induced site- and diastereoselectivity (Scheme 3). To our delight, the reaction of isosteviol derivative 1i gives product 2i as a 2:1 mixture of diastereomers in 58% isolated yield. Furthermore, while many electron-rich aryl groups are oxidatively labile, these conditions are sufficiently mild to affect functionalization of dehydroabietyl-derived 1j. Indeed, 1j provides alkylated products 2j as a 1.6:1 mixture of diastereomers in a remarkable 81% isolated yield. These are the first documented cases of sulfamate ester-guided process where the radical-containing intermediate unambiguously induces a diastereoselective reaction.²⁴

In spite of sensitivity to steric hindrance, sulfamate esters guide Giese reactions at tertiary centers with predictable selectivity (Scheme 4). While the electron densities of the C(5)–H and C(8)–H bonds of menthol-derived 13a are predicted to be similar,²⁵ this Giese reaction functionalizes solely the C(8)–H bond. Analogous site-selectivity is displayed in sulfamate ester-guided chlorination¹⁴ and bromination¹⁵ reactions, as well as in iron- and manganese-mediated intramolecular amination reactions.^{10d, 10g} Presumably, the C(8)–H bond is geometrically disposed to interact with the directing groups that guide these reactions. By comparison, unguided intermolecular oxidation²⁵ and amination²⁶ processes engage the more sterically accessible C(5)–H bond selectively.

Scheme 4. Reactions can form fully-substituted centers and tolerate N-substituent variations^a,b.

^a General reaction conditions: 1.0 equiv sulfamate ester (0.2 mmol), 2.2 equiv electrophilic olefin, two 34 W blue lamps, photocatalyst (1.0 mol %), 1.0 equiv K₂CO₃, MeCN (0.2 M), 900 mot/min stirring, 21–28 °C, 48h. ^b Isolated yields. ^c 16 h. ^d 24 h. ^e Reaction performed with 13f (1 mmol). ^f Yield as average of two trials. ^g Product not detected by ¹H NMR of the crude reaction mixture in a reaction with tert-butyl acrylate.

Furthermore, this directed reaction surmounts electronically-induced preferences for functionalization. In undirected processes, inductively electron-withdrawing groups can deactivate proximal C–H bonds to oxidation. When 3,7-dimethyloctanol is masked with an electron withdrawing group, it undergoes preferential oxidation at C(7) in fluorination,²⁷ oxygenation,²⁸ amination,²⁹ azidation,³⁰ and trifluoromethylthiolation³¹ reactions. By contrast, this sulfamate ester guided reaction installs new carbon–carbon bonds predictably and selectively at C(3).

Additional fully-substituted C(3) centers can be generated from unactivated acyclic or cyclic tertiary C(sp³)–H bonds, or ethereal centers (i.e. 13c–g). Remarkably, while ethereal 13e features more sterically accessible and electronically activated C–H bonds, the reaction occurs selectively at the C(3)–H site.

Notably, optically active 13h is converted to racemic 14h, such that the generated radical must have a sufficient lifetime to epimerize. Nevertheless, remaining substrate is recovered without detectable epimerization, which suggests that the C–H abstraction is not reversible on a time scale that would allow for epimerization.

While tertiary and secondary C–H bonds engage in productive Giese reactions, primary centers, and secondary benzylic centers do not. Additionally, while racemic substrate 13g reacts productively in this transformation, analogue 13k is ineffective, providing further evidence that the reaction is sensitive to steric encumbrance. To date, the efficient transformation is highly dependent on nitrogen substitution. Unfortunately, N-iso-propyl, N-(1,1,1-trifluoro-2-methylpropyl), and N-2-(2-phenylpropyl) sulfamate esters (13l– n) guide the Giese reaction less effectively than N-tert-butyl sulfamate esters.

A wide range of electron-deficient olefins engage in this directed Giese reaction (Table 2). Alkyl, aryl, allyl, and alkynyl acrylates react in good to excellent yields (entries 1–4). This is remarkable, as many aryl, allyl, and alkynyl groups would undergo deleterious processes under more oxidizing conditions. Furthermore, as alkynes are well-developed as handles to facilitate structure-activity-relationship or mode-of-action studies, incorporation of an alkyne-bearing trapping agent suggests that this Giese reaction could be applied to complex small molecules to generate chemical probes.³²

Table 2.

Reaction traps various electron deficient olefins^a

entry	olefin	product		R2		yield [%]b	yield [%]c
1			15a	Me		22	74
2			15b	Ph		07	72
3			15c			38	58
4			15d			09	66
				R2	R3
5			16a	Et	Me	ndd	58
6			16b	Me	Ph	–e	80
7f			16c	Me	NBoc2	ndd	89
8f			16d			21	78
9			16e			11	87
				R2
10			15e	Ph		47	50
11			15f	Me		24	72
12			15g	H		58	43
13			15h	NMe2		04	66
14			15i	NHMe		04	55
15			15j			52	44
16			16f			ndd	≥98 >20:1 dr

^aGeneral reaction conditions: 1.0 equiv sulfamate ester, 2.2 equiv olefin, two 34 W blue lamps, photocatalyst (1.0 mol %), 1.0 equiv K₂CO₃, M eCN (0.2 M), 900 mot/min stirring, 21–28 °C.

^bRecovered 13b or 13f

^cIsolated yields.

^dNot detected in crude ¹H NMR of the crude reaction mixture.

^eNot isolated.

^f1.2 equiv olefin.

Despite the documented challenges with substrate steric encumbrance, bulky electrophilic olefin engage effectively, including 2-substituted acrylates, a dehydroalanine derivative, and dimethyl fumarate, which contains an internal olefin (entries 5–8). Moreover electrophilic olefins also react effectively, including α,β-unsaturated ketones that are cyclic, acyclic, or aryl-substituted, and acrolein (entries 9–12). Finally, nitrogen-containing olefins are competent in the reaction (entries 13–15).

Radical conjugate addition reactions benefit from a rich library of enantioenriched electrophilic olefins, designed to induce asymmetry in the course of radical addition reactions, even with acyclic nucleophilic free radicals.³³ In principle, this rich array of enantioenriched electrophilic olefins could be used to intercept this radical translocation protocol and induce asymmetry in the course of these Giese reactions. Gratifyingly, sulfamate ester 13f reacts with a known enantioenriched methyleneoxazolidinone^{34, 35} to furnish amino acid derivative 16f in quantitative yield with complete diastereocontrol (>20:1 d.r., entry 16). This approach could provide access to a broad array of enantioenriched non-natural alkylated amino acids.

Ultimately, to maximize the utility of this process, it is necessary to cleave or displace the sulfamate ester template for this alkylation process. To make sulfamate esters more liable to displacement, typically they are N-acylated.^10a To our disappointment, the N-tert-butyl sulfamate ester products are not efficiently carbamoylated under standard conditions. Fortunately, N-vinylation of N-tert-butyl sulfamate esters can be achieved with diethyl acetylene dicarboxylate (Scheme 5). This renders vinylated 17 susceptible to displacement to furnish azide,¹⁵ iodide,¹⁵ acetate,^10a thioacetate,^10a or free alcohol^10a analogues.

Scheme 5.

Strategies for sulfamate ester displacement

CONCLUSION

This sulfamate ester-guided photoredox-mediated reaction provides a powerful and general platform for directed C–H functionalization. This is the first research to establish that sulfamate ester anions can be photochemically oxidized. Moreover, the resultant nitrogen-centered radicals guide 1,6-HAT to furnish tertiary or secondary carbon-centered radicals, which can be trapped to generate racemic and diastereoselective products. Thereby, this process affords complementary position-selectivity to that achieved using known photoredox-mediated carbon–carbon bond-forming reactions. Finally, a new tactic for sulfamate ester vinylation and displacement introduces further diversity to the array of accessed small molecules.