Intermolecular Anti-Markovnikov Hydroamination of Alkenes with Sulfonamides, Sulfamides, and Sulfamates

Abstract

A general method for the light-driven intermolecular anti-Markovnikov hydroamination of alkenes with primary sulfonamides, sulfamides, and sulfamates is presented. The reaction is mediated by a ternary catalyst system composed of an iridium(III) chromophore, a fluorinated alkoxide base, and a thiol H-atom donor. We hypothesize that the reactions proceed through a proton coupled electron-transfer (PCET) mechanism wherein implementation of the alkoxide base imparts additional thermochemical driving force for the homolytic activation of strong N–H bonds that were previously inaccessible using this methodology. This furnishes electrophilic N-centered radicals that subsequently interface with a wide range of unactivated alkenes for C–N bond formation. This protocol exhibits a broad substrate scope and great functional group tolerance, further highlighting the advantages of excited-state PCET as a platform for catalytic radical generation from common organic functional groups.

Graphical Abstract

INTRODUCTION

Sulfur(VI) compounds, such as sulfonamides, sulfamides, and sulfamates, are valuable motifs in synthetic and medicinal chemistry.¹ Within the context of drug discovery, sulfonamides are often employed as bioisosteres for carboxylic acids and amides, and they have also been shown to confer chemical and metabolic stability, three-dimensionality, and crystallinity to target compounds.^2,3 As such, these functionalities are found in a variety of approved pharmaceuticals exhibiting a broad spectrum of biological activities (Figure 1A).⁴ More recently, secondary N-alkyl sulfonamides have garnered interest as potential anticancer agents due to their selective inhibition of carbonic anhydrase enzymes.⁵ These compounds are typically synthesized via nucleophilic addition of alkyl amines to sulfonyl chlorides, which are hydrolytically unstable and can be challenging to prepare from functionalized substrates.⁶ Alternatively, they can be accessed through N-alkylation of primary sulfonamides, which relies on the use of undesirable alkyl halide electrophiles and often results in overalkylation.⁷ In contrast, catalytic intermolecular olefin hydroamination represents a complementary and atom-economical approach to the synthesis of secondary N-alkyl sulfonamides by directly coupling accessible primary sulfonamides with readily available olefin starting materials.

Figure 1.

Motivation and prior work in intermolecular anti-Markovnikov alkene hydroamination with sulfur(VI) compounds.

Numerous intermolecular olefin hydroamination protocols with sulfonamides have been developed,^8–11 including notable methods from Tilley,¹² He,¹³ Hartwig,¹⁴ Cheng,¹⁵ and Cook.¹⁶ These methods utilize transition metal or Brønsted acid catalysts and all proceed with Markovnikov regioselectivity. By contrast, catalytic methods for sulfonamide-based anti-Markovnikov hydroaminations generally rely on electronically-activated olefin substrates.^17,18 For unactivated alkenes, the state of the art is much less developed, with the only reported methods being photoredox-based radical transformations reported by Nicewicz,¹⁹ Doyle,²⁰ and our own laboratory (Figure 1B).²¹ While notable, all three protocols are restricted in scope with respect to either the alkene or sulfonamide component, and there remains no general solution for an intermolecular anti-Markovnikov hydroamination reaction that can accommodate a wide variety of both olefins and sulfur(VI)-based amine derivatives.

Our laboratory’s prior work in sulfonamide-based hydroamination relied on PCET, wherein homolytic activation of the strong substrate N–H bonds by an excited-state oxidant and a weak Brønsted base provides access to N-centered radical intermediates.^22,23 These electrophilic radicals subsequently undergo rapid addition to nucleophilic olefins with high levels of anti-Markovnikov regioselectivity (see Supporting Information, Figure S1 for the catalytic cycle).²⁴ We previously leveraged this strategy to generate sulfonamidyl radicals under the joint action of [Ir(dF(CF₃)ppy)₂(5,5’-d(CF₃)bpy)]PF₆ ([Ir-A]PF₆) photocatalyst and tetrabutylammonium dibutyl phosphate base co-catalyst for the intra- and intermolecular hydroamination of a broad array of unactivated olefins.²¹ However, sulfonamides other than electron-rich para-methoxybenzenesulfonamide (PMP sulfonamide) were found to exhibit poor intermolecular reactivity, which we hypothesized may be due to a lack of thermochemical driving force for the PCET activation of less activated primary sulfonamide N–H bonds (Figure 1C). More recently, our group has disclosed the use of nonafluoro-tert-butyl alkoxide base for the PCET activation of electron-rich sulfonamides in the context of an asymmetric hydroamination protocol.²⁵ During those investigations, we observed that the perfluorinated alkoxide base was not only compatible with our chosen excited-state oxidant, but also provided product conversions that were superior to those provided by the phosphate base. Thus, we sought to utilize this base to expand the reactivity of our PCET-enabled intermolecular hydroamination protocol, thereby developing a general fragment coupling reaction for expedient access to complex secondary N-alkyl sulfonamides.

METHODS

We began our optimization efforts by studying the hydroamination of 2-methyl-1-hexene with benzyl sulfonamide (Table 1). We observed that blue light irradiation (456 nm) of these starting materials in the presence of 2 mol% [Ir-A]PF₆ photocatalyst, 10 mol% tetrabutylammonium nonafluoro-tert-butyl alkoxide, and 30 mol% 2,4,6-triisopropylthiophenol (TRIP-SH) as a hydrogen atom transfer (HAT) co-catalyst in toluene at 35 °C afforded desired hydroamination product 1 in 85% yield (entry 1). Exchanging nonafluoro-tert-butyl alkoxide (pK_a((CF₃)₃COH) = 21 in MeCN) for weaker bases gave significantly decreased yields of 1 (entries 2–5).²⁶ Replacing the tetrabutylammonium cation of the alkoxide with tetrabutylphosphonium improved the reaction yield to 97% (entry 6), likely due to the improved solubility of the phosphonium salt in toluene.²⁷ Using the more basic benzoate (pK_a(BzOH) = 22 in MeCN) (entry 7) or n-butoxide (pK_a(BuOH) ~39 in MeCN) (entry 8) salts afforded 1 in 41% and 6% yield, respectively, possibly due to protonation of the alkoxide by TRIP-SH or competitive oxidation of the base (E_p/2 (NBu₄[BzO]) ~1.00 V vs Fc⁺/Fc in MeCN) by the excited-state photocatalyst.^26,28–30 Evaluation of less oxidizing iridium(III) photocatalysts resulted in lower yields (entries 9–11). Aromatic solvents such as benzene and trifluorotoluene (entries 12–13) led to moderate levels of product formation, whereas other solvents provided dramatically reduced yields (entries 14–16). Control experiments run in the absence of light, photocatalyst, or base furnished no product (entries 17–19), while poor conversion was observed in the absence of the thiol HAT co-catalyst (entry 20).

Table 1.

Reaction Optimization

Entry	Base	Photocatalyst	Solvent	Yield (%)a
1	NBu4[(CF3)3CO]	[Ir-A]PF6	PhMe	85
2	NBu4[(BuO)2OPO]	[Ir-A]PF6	PhMe	2
3	2,6-lutidine	[Ir-A]PF6	PhMe	0
4	2,4,6-collidine	[Ir-A]PF6	PhMe	3
5	DMAP	[Ir-A]PF6	PhMe	8
6	PBu4[(CF3)3CO]	[Ir-A]PF6	PhMe	97
7	PBu4[BzO]	[Ir-A]PF6	PhMe	41
8	PBu4[BuO]	[Ir-A]PF6	PhMe	6
9	PBu4[(CF3)3CO]	[Ir-B]PF6	PhMe	92
10	PBu4[(CF3)3CO]	[Ir-C]PF6	PhMe	51
11	PBu4[(CF3)3CO]	[Ir-D]PF6	PhMe	2
12	PBu4[(CF3)3CO|	[Ir-A]PF6	PhCF3	74
13	PBu4[(CF3)3CO]	[Ir-A]PF6	C6H6	75
14	PBu4[(CF3)3CO]	[Ir-A]PF6	CH2Cl2	19
15	PBu4[(CF3)3CO]	[Ir-A]PF6	MeCN	10
16	PBu4[(CF3)3CO]	[Ir-A]PF6	THF	4
change from optimal conditions (entry 6)
17	no light			0
18	no photocatalyst			0
19	no base			0
20	no TRIP-SH			14

^aReactions were performed on a 0.05 mmol scale and the yields were determined via ¹H NMR spectroscopic analysis of the crude reaction mixture relative to 1,3,5-trimethoxybenzene as an external standard.

^bAll photocatalyst potentials are reported relative to Fc⁺/Fc in MeCN.

Having established optimized reaction conditions, we proceeded to examine the scope of sulfonamide derivatives using two equivalents of 2-methyl-1-hexene as the model olefin (Table 2). In addition to benzyl sulfonamide, various other alkyl sulfonamides were converted to their corresponding anti-Markovnikov hydroamination products in good to moderate yields (1–5). We were pleased to find that a wide range of aryl sulfonamides were also successful substrates for this transformation. Subjecting simple phenyl sulfonamide to the optimized conditions resulted in nearly quantitative formation of hydroamination product 6. Ortho- and meta-substitution on the phenyl ring was tolerated (7, 8), as were a variety of electronically diverse functional groups at the para-position (9–13). Functional handles for further product diversification, such as an alkyne (14), bromide (15), and boronic ester (16), were also accommodated. Heterocyclic sulfonamides bearing pyridine (17), thiophene (18), isoxazole (19), and pyrazole (20) functionalities smoothly underwent hydroamination in good to excellent yields, as did more complex sulfonamide examples. Celecoxib, a non-steroidal anti-inflammatory drug, was converted to its secondary N-alkyl sulfonamide analogue 21 in excellent yield, as was a derivative of the ACE inhibitor captopril (22). Furthermore, we observed that saccharin, an electron-deficient secondary sulfimide, could be alkylated to give 23 in 64% yield.

Table 2.

Scope of sulfonamide, sulfamide, and sulfamate substrates

Yields are for isolated material and are the average of two experiments. Reactions were conducted on 0.5 mmol scale with 2 equiv of olefin unless otherwise noted.

^a[Ir(dF(CF₃)ppy)₂(dtbbpy)]PF₆ was used.

^b3 equiv of olefin were used.

This methodology was subsequently extended to sulfamide (N–H BDFE ~105 kcal/mol) and sulfamate (N–H BDFE ~108 kcal/mol) substrates, which are medicinally relevant sulfur(VI)-based functional groups that are underexplored relative to sulfonamides.^31–34 We were pleased to find that using [Ir(dF(CF₃)ppy)₂(dtbbpy)]PF₆ ([Ir-B]PF₆) (E_1/2(*Ir^III/Ir^II) = 0.81 V vs Fc⁺/Fc in MeCN), a more mildy oxidizing photocatalyst, enabled acyclic N,N-dialkyl sulfamides to undergo hydroamination with 2-methyl-1-hexene in high yields (24 25). Cyclic voltammetry experiments (Supporting Information, Figure S3) suggest this is potentially due to competitive quenching of the more oxidizing excited-state [Ir-A]PF₆ via single electron transfer at the sulfamide’s tertiary nitrogen site, rather than productive PCET at the substrate’s primary amine site. Thus, using the less oxidizing [Ir-B]PF₆ photocatalyst results in more efficient reactions and higher yields. Sulfamides bearing saturated heterocyclic moieties such as morpholine and N-Boc piperazine were also tolerated, delivering 26 and 27 in 89% and 88% yield, respectively. Notably, selective alkylation of the primary site over the secondary site was observed in an unsymmetrical sulfamide with multiple N–H bonds (28). This selectivity, as well as the lack of overalkylation products, is attributed to the reduced electrophilicity and reactivity of the secondary sulfonamidyl radical. [Ir-A]PF₆ remained the optimal photocatalyst for the activation of highly electron-deficient sulfamate N–H bonds. The sulfamates reacted to furnish desired hydroamination adducts 29–31 in synthetically useful yields. Notably, 31 can be deprotected under mild reductive conditions to furnish the formal anti-Markovnikov hydroamination product of ammonia.³⁵ Finally, anticonvulsant drug topiramate could also be successfully monoalkylated in near quantitative yield (32).

We next evaluated the scope of the olefin partner in this reaction with benzyl sulfonamide as the model amine derivative (Table 3). A wide range of aliphatic alkenes bearing di-verse substitution patterns gave moderate to excellent yields, including terminal (33), 1,1-disubstituted (1 and 34), 1,2-disubstituted (35 and 36), trisubstituted (37), and tetrasubstituted (38) olefins. The reaction was not sensitive to the geometry of the alkene, as both cis- and trans-3-hexene afforded 36 with similar efficiency. Al-kenes bearing trialkylsilane (39), silyl ether (40), ester (41 and 42), phthalimide (43), carbamate (44 and 45), and epoxide (46) functional groups were all tolerated, demonstrating the mild nature of this hydroamination protocol.

Table 3.

Scope of alkene substrates

Yields are for isolated material and are the average of two experiments. Reactions were conducted on a 0.5 mmol scale with 2 equiv of olefin unless otherwise noted.

^a5 equiv of olefin were used.

^b10 mol% TRIP-SH was used.

^c1.5 equiv of olefin were used.

^d10 mol% (TRIP-S)₂ was added.

Although styrenes were a challenging class of substrates in our original PCET-based method, we discovered that by lowering the loading of TRIP-SH to 10 mol% and adding 10 mol % TRIP disulfide, styrene and α-methylstyrene could now be hydroaminated to give 47 and 48 in 48% and 65% yield, respectively. We believe that the disulfide increases the quantum yield of the reaction by serving as an oxidant to help turn over the reduced-state iridium(II) photocatalyst.^36,37 Formation of thiol ene products under the optimized conditions was not observed. Electron-rich, heteroatom-substituted olefins were also amenable substrates, with an enamide affording 49 in 64% yield and a monosaccharide-derived vinyl ether affording 50 in 76% yield. An enol phosphinate resulted in 60% of 51, from which the corresponding N-protected aziridine may be accessed via a base-mediated intramolecular nucleophilic substitution (Supporting Information, Section 7). We further showed that this methodology can be used to functionalize more complex olefins. An estrone derivative bearing a 1,1-disubsituted olefin reacted to furnish 52 as a single diastereomer in 80% yield. Similarly, dehydroepiandrosterone acetate, which bears a sterically encumbered, endocyclic trisubstituted alkene, underwent hydroamination to give 53 in 50% yield and 4.6:1 d.r.

The ability of an oxidant/base pair to serve as a formal H-atom acceptor in a PCET reaction can be quantified using a thermodynamic cycle popularized by Mayer and coworkers as a joint function of the reduction potential of the oxidant and the pK_a of the conjugate acid of the base.³⁸ The resulting effective bond dissociation free energy (BDFE) provides a measure of the strength of bonds that can be activated by a given oxidant/base pair in a thermoneutral reaction. Using this formalism, the combination of [Ir-A]PF₆ (E_1/2(*Ir^III/Ir^II) = 1.30 V vs Fc⁺/Fc in MeCN) and tetrabutylammonium dibutyl phosphate (pK_a((BuO)₂P(O)OH) ~13 in MeCN) provides an effective BDFE of ~100 kcal/mol.^21,39,40 This value rationalizes why various secondary alkyl and aryl sulfonamides (N–H BDFE ~97 kcal/mol) could be activated for intramolecular hydroamination in our prior work, but not primary alkyl sulfonamides, electron-poor aryl sulfonamides, and sulfamates, which were computationally assessed to have higher N–H BDFEs ranging from approximately 106 to 108 kcal/mol (Supporting information, Section 8). In contrast, the combination of [Ir-A]PF₆ and the more basic nonafluoro-tert-butyl alkoxide provides an increased effective BDFE of ~111 kcal/mol.²⁶ The results of our base screen support our initial hypothesis that the higher effective BDFE afforded by the [Ir-A]PF₆/alkoxide system is necessary for the activation of these stronger primary sulfonamide N–H bonds. However, we later observed that systems with lower effective BDFEs utilizing the fluorinated alkoxide base and less oxidizing photocatalysts also provided modest levels of conversion. We hypothesize that stepwise PCET, arising from sulfonamide deprotonation by the alkoxide and subsequent single-electron oxidation of the resulting anion (E_p/2(NBu₄[BnSO₂NH]) ~0.41 V vs. Fc⁺/Fc in MeCN), could also be a competing pathway for generation of the sulfonamidyl radical, though this proton exchange equilibria is considerably unfavorable (pK_a(MeSO₂NH₂) ~29 in MeCN).^29,41

Efforts to evaluate the mechanism of sulfonamide activation using cyclic voltammetry and Stern-Volmer luminescence quenching experiments were confounded by competitive single electron oxidation of the tetrabutylphosphonium nonafluoro-tert-butyl alkoxide base (E_p/2(PBu₄[(CF₃)₃CO]) = 1.17 V vs Fc⁺/Fc in MeCN) (Supporting Information, Figures S5 and S9). While direct H-atom abstraction of strong and acidic sulfonamide N-H bonds by the resulting electrophilic alkoxy radical is thermodynamically feasible, it has no literature precedent and is expected to be kinetically disfavored relative to the abstraction of weaker substrate allylic C–H bonds on the basis of polar effects.^42,43 Mechanisms proceeding through direct oxidation of the sulfonamide (E_p/2(BnSO₂NH₂) ~2.00 V vs Fc⁺/Fc in MeCN) or the alkene substrate (E_p/2(2-methyl-1-pentene) ~2.10 V vs Fc⁺/Fc in MeCN) are significantly endergonic,³⁰ and Stern-Volmer studies show that neither benzyl sulfonamide nor 2-methyl-1-hexene alone quench the luminescence of [Ir-A]PF₆ (Supporting Information, Figures S7 and S8). Further mechanistic work aimed at elucidating the precise nature of the electron and proton transfer steps involved in N-radical formation will require additional investigation. Nevertheless, we believe these reactions are valuable additions to the hydroamination literature, as they provide new opportunities for the synthesis of valuable secondary N-alkyl sulfur(VI) derivatives under mild conditions without leading to overalkylation products.

CONCLUSION

In summary, we have developed a catalytic protocol for the intermolecular anti-Markovnikov hydroamination of alkenes with a diverse range of sulfonamides, sulfamides, and sulfamates. Key to this method is the use of nonafluoro-tert-butyl alkoxide, a stronger Brønsted base, to enable direct radical generation from a variety of strong N–H bonds. This work further demonstrates excited-state PCET as a general catalytic platform for the generation of open-shell intermediates and their subsequent applications in challenging alkene functionalizations.