One-Pot Synthesis of Sulfonamides from Unactivated Acids and Amines via Aromatic Decarboxylative Halosulfonylation

Abstract

The coupling of carboxylic acids and amines to form amide linkages is the most commonly performed reaction in the pharmaceutical industry. Herein, we report a new strategy that merges these traditional amide coupling partners to generate sulfonamides, important amide bioisosteres. This method leverages copper ligand-to-metal charge transfer (LMCT) to convert aromatic acids to sulfonyl chlorides, followed by one-pot amination to form the corresponding sulfonamide. This process requires no prefunctionalization of the native acid or amine and extends to a diverse set of aryl, heteroaryl, and s-rich aliphatic substrates. Further, we extend this strategy to the synthesis of (hetero)aryl sulfonyl fluorides, which have found utility as “click” handles in chemical probes and programmable bifunctional reagents. Finally, we demonstrate the utility of these protocols in pharmaceutical analogue synthesis.

Carboxylic acids and amines have high structural diversity, synthetic utility, and broad availability from both natural and commercial sources.^1,2 These two functionalities are the traditional partners of the amide coupling, the most commonly performed reaction by medicinal chemists.^3,4 Amides, however, can be metabolically labile toward hydrolysis and may not possess the ideal binding properties for a given target.⁵ To address these shortcomings, several amide bioisosteres have been identified that retain similar geometric properties to amides while displaying modified metabolic stability or binding affinity.⁶

One notable amide isostere is the sulfonamide, which features similar geometry, an additional hydrogen bond acceptor (HBA), improved hydrolytic stability, and increased polar surface area (PSA).⁷ These features can lead to dramatic improvements in physiochemical properties or binding affinity, with examples showing nearly 200-fold increases in binding with no other structural modification.⁸ Beyond medicinal chemistry, sulfonamides have been used in place of amides to tune the gelation properties of physical gels.⁹ Despite high interest, the preparation of sulfonamide analogues typically necessitates de novo synthesis of a sulfonyl chloride partner or a multistep sequence proceeding through high-energy diazonium intermediates.^8,10 Therefore, a method for the expedient synthesis of sulfonamides from the same partners used in conventional amide coupling would accelerate the preparation of amide analogues in drug discovery campaigns.

We envisioned the synthesis of sulfonamides from aryl carboxylic acids and amines could be achieved through an initial decarboxylative chlorosulfonylation, followed by a one-pot sulfonamide formation via the addition of an amine (Figure 1). This strategy confers several benefits. First, the reaction between amines and sulfonyl chlorides occurs rapidly and is robust in scope.¹¹ Second, sulfonyl chlorides themselves are valuable synthetic intermediates and can be used as a lynchpin functionality toward the synthesis of diverse sulfur containing functionalities, such as sulfonates,¹² sulfones,¹³ sulfinates,¹⁴ and thiophenols.¹⁵ Moreover, we anticipated that a decarboxylative halosulfonylation strategy would extend to aryl sulfonyl fluorides, valuable “click” handles for sulfur(VI)−fluoride exchange^16,17 with broad utility as chemical probes,¹⁸ programmable bifunctional reagents,¹⁹ and versatile functional handles.²⁰ While several methodologies have been developed in recent years for accessing aryl sulfonyl halides from aryl boronic esters,²¹ halides,²² diazoniums,²³ and several sulfur-containing functionalities,²⁴ a direct approach beginning from aryl carboxylic acids has not been reported to date.¹⁷

Figure 1.

Decarboxylative halosulfonylation and subsequent amination of (hetero)aryl acids via Cu-LMCT catalysis.

Despite its appeal, the development of a versatile decarboxylative halosulfonylation protocol presents several challenges. While the decarboxylative halosulfonylation of related aliphatic systems has been reported,²⁵ aromatic decarboxylation is significantly more challenging.¹ Recently, both Ritter and our group have demonstrated that copper ligand-to-metal charge transfer (LMCT) is a mild and general approach for traditionally challenging aromatic decarboxylative functionalizations including halogenation,^26,27 borylation,²⁸ hydroxylation,²⁹ and sulfoximination.^30,31 Notably, our lab rendered this strategy catalytic in copper by including a single electron oxidant.^27,28 Given the broad precedent for engaging aryl radicals with SO₂ to forge C−S bonds,³² we anticipated that a Cu-LMCT strategy could be successful for catalytic decarboxylative halosulfonylation. Additionally, we expected that the mild nature of LMCT decarboxylation could allow a highly efficient one-pot conversion of sulfonyl chlorides to sulfonamides, circumventing the isolation of highly reactive electrophiles. Herein, we describe the successful development of a copper-catalyzed aromatic decarboxylative halosulfonylation, expediently accessing sulfonamides from aryl carboxylic acids and amines.

We propose the general design plan for decarboxylative halosulfonylation, as detailed in Figure 2. First, a Cu(II) catalyst combines with an in situ generated aryl carboxylate, providing a photoactive Cu(II) carboxylate. Upon irradiation, this complex can undergo LMCT, inducing Cu−O bond homolysis and furnishing an aroyloxy radical as well as a reduced Cu(I) species. We hypothesize these two species can recombine, regenerating the ground state Cu(II) carboxylate complex and suppressing competitive, deleterious bimolecular reaction pathways.³³ The aroyloxy radical can also undergo decarboxylation, furnishing the desired aryl radical. Radical capture by SO₂ would then forge the desired C(sp²)−S bond and generate an aryl sulfonyl radical.³² We propose that this sulfur-centered radical could react with various electrophilic halogenation reagents to afford the desired aryl sulfonyl halide products. Finally, a suitable single electron oxidant can oxidize Cu(I) to Cu(II), closing the catalytic cycle.

Figure 2.

General design plan for the decarboxylative halosulfonylation of (hetero)aryl acids.

We began optimization studies by attempting the decarboxylative chlorosulfonylation of 4-fluorobenzoic acid. By employing Cu(MeCN)₄BF₄ as the most generally successful copper catalyst, 1-fluoro-2,4,6-trimethylpyridinium tetrafluoroborate (NFTPT) as the single electron oxidant, 1,3-dichloro-5,5-dimethylhydantoin (DCDMH) as a chlorine atom source, and the integrated photoreactor (IPR) as a high intensity source of 365 nm light,³⁴ we observed a 2% yield of the desired sulfonyl chloride product with sodium metabisulfite as the SO₂ source (Table 1, entry 8). While other SO₂ surrogates from the literature were similarly unsuccessful,³⁵ super stoichiometric loadings of copper did lead to improved yield in some cases (see SI for experimental details). Recognizing the sensitivity of Cu-LMCT decarboxylation to the Lewis basic functionality present in each of these surrogates, we sought to employ SO₂ directly by evaluating stock solutions of SO₂. We were delighted to find that by using a solution of SO₂ in acetonitrile,^22c,35a,36 we observed formation of the desired product in 62% yield (Table 1, entry 2). Though we prepared our solution in house, similar stock solutions of SO₂ in acetonitrile are commercially available. This solution could be stored over extended periods of time in a sealed vessel at −20 °C. Despite frequent use, no decrease in reaction performance was observed when using a single 4.0 M stock solution for >6 months. Control reactions are consistent with the proposed LMCT pathway, as copper, oxidant, and light are all essential for reactivity (Table 1, entries 3−5).

Table 1.

Control Experiments and Optimization for LMCT Decarboxylative Chlorosulfonylation and One-Pot Sulfonamide Formation

entrya	deviations	yieldb
1	none	69%
2	no LiBF4	62%
3	ambient light only	0%
4	no Cu	0%
5	no NFTPT	0%
6	no DCDMH	0%
7	no SO2	0%
8	with Na2S2O5 (2 equiv) instead of SO2	2%
9	with DABSO (1 equiv) instead of SO2	0%
10	entry 2 then morpholine (2 equiv) and DIPEA (2 equiv)	50%c
11	entry 1 then morpholine (2 equiv) and DIPEA (2 equiv)	68%c

^a0.1 mmol scale.

^bYields determined by ¹⁹F NMR analysis.

^cYield of the corresponding morpholine sulfonamide. See SI for experimental details. NFTPT, 1-fluoro-2,4,6-trimethylpyridinium tetrafluoroborate. DCDMH, 1,3-dichloro-5,5-dimethylhydantoin.

We next attempted to convert the sulfonyl chlorides to sulfonamides in the same reaction vessel. We found that this functionalization could be achieved in near quantitative yield from the intermediate sulfonyl chloride by adding morpholine and DIPEA to the crude reaction mixture following irradiation and removal of unreacted SO₂.³⁷ Across both steps, we found that LiBF₄ was a beneficial additive for avoiding the formation of an undesired sulfonyl fluoride byproduct, ultimately delivering the desired morpholine sulfonamide in 68% yield (Table 1, entry 11).

Excitingly, this strategy was extended to efficient decarboxylative fluorosulfonylation using Selectfluor as both the single electron oxidant and fluorine atom source, slightly elevated copper loadings of 50 mol %, and an increased concentration of 0.5 M (see SI for experimental details).

With the optimized conditions in hand, we evaluated the scope of the decarboxylative chlorosulfonylation reaction (Figure 3). To demonstrate that electronically and sterically unbiased substrates are capable of this reaction, benzoic acid was converted to benzenesulfonamide (1, 63% yield). Gratifyingly, electron deficient acids (2−7, 61−85% yield) proceeded with good to excellent efficiencies. Notably, these examples were successfully functionalized with diverse amines, providing efficient access to primary, secondary, tertiary, and anilinic sulfonamides, all in near quantitative yield from the intermediate sulfonyl chlorides. Ortho-substitution was also found to be well tolerated (8 and 9, 58 and 75% yield, respectively). Additionally, O−heteroatom substituents could be effectively introduced around the arene periphery (10 and 11, 71 and 61% yield, respectively), as well as electron donating alkyl substituents (12 and 13, 59 and 61% yield, respectively). Finally, other sulfur-containing functionalities such as trifluoromethyl thioethers (14, 75% yield) and sulfonyl fluorides (15, 69% yield) were successfully tolerated, high-lighting an opportunity to introduce distinct sulfur-containing functionalities of variable oxidation states on the same arene.

Figure 3.

Decarboxylative chlorosulfonylation and one-pot sulfonamide formation. Reactions were performed on a 0.5 mmol scale with (hetero)aryl acid (1 equiv), [Cu(MeCN)₄]BF₄ (20 mol %), 1,3-dichloro-5,5-dimethylhydantoin (1 equiv), 1-fluoro-2,4,6-trimethylpyridinium tetrafluoroborate (1 equiv), LiBF₄ (1.2 equiv), SO₂ (2 equiv), MeCN (0.1 M), 365 nm LEDs, 12 h. Amination conducted in MeCN or THF (0.1 M) with amine or amine·HCl (2 equiv) and DIPEA or pyridine (2−4 equiv). See SI for experimental details. All yields isolated. ^aAmination performed with 25 equiv of NH₄OH. ^bWith N-chlorosuccinimide (2.5 equiv) as the chlorination reagent. ^cWith N-chlorophthalimide (2 equiv) as chlorination reagent. ^dWith 10 mol % [Cu(MeCN)₄]BF_4.

We were eager to evaluate the scope of heteroaryl sulfonyl chlorides accessible via this sequence, as these scaffolds are desirable in medicinal chemistry programs despite the challenges presented by their high electrophilicity and susceptibility to hydrolysis. We were delighted to find that our method proceeded in good efficiency on a range of differentially substituted, electron-deficient pyridines, including picolinic (16−18, 45−64% yield), nicotinic (19−22, 45−55% yield) and isonicotinic (23, 63% yield) acids across a diverse set of amines.

We next attempted to extend our method beyond (hetero)arenes and toward the functionalization of other medicinally relevant s-rich acids, such as cyclopropanes and bicyclo[1.1.1]pentanes (BCPs), which have found widespread adoption as arene bioisosteres.³⁸ An inexpensive and commercially available BCP substrate, 3-(methoxycarbonyl)-BCP-1-carboxylic acid, was efficiently converted to the corresponding sulfonamide (24, 63% yield). While BCPs bearing a sulfonamide exit vector have been reported in the literature, they previously have been accessed via a 5-step synthetic sequence.³⁹ Beyond this BCP scaffold, cyclopropanes were also readily functionalized in high yield (25, 64% yield, see SI for additional examples and limitations).

While each of the acid starting materials that appear in Figure 3 are inexpensive and commercially available,⁴⁰ the corresponding sulfonyl chlorides are in many cases either not commercial (designated by NC) or are prohibitively expensive (>$100 per gram and >18× more expensive per gram than the acid starting material, designated by PE).⁴¹ As such, this procedure to convert bench-stable acids to desirable sulfonamides via a one-pot protocol alleviates a significant limitation associated with sulfonamide preparation.

Next, we evaluated the scope of the decarboxylative fluorosulfonylation (Figure 4). This reaction tolerates both electron-deficient (26−28, 50−82% yield) and electron-rich (29 and 30, 50 and 53% yield, respectively) acids. Additionally, highly activated aryl bromides (31, 56% yield) are well tolerated. This method also extends to a similarly desirable selection of electron deficient isonicotinic (32−34, 51−61%) and nicotinic acids (35, 59% yield). While picolinic acids did not proceed in the standard protocol with high efficiencies, a one-pot, two step procedure for their synthesis from the intermediate sulfonyl chloride was developed (36, 53% yield). Though more challenging, other heteroarenes, such as pyridazines (37, 40%), were also successful.

Figure 4.

Decarboxylative fluorosulfonylation of (hetero)aryl carboxylic acids. Reactions were performed on a 0.5 mmol scale with (hetero)aryl acid (1 equiv), [Cu(MeCN)₄]BF₄ (50 mol %), Selectfluor (1.5 equiv), SO₂ (2 equiv), and MeCN (0.5 M), 365 nm LEDs, 12 h. All yields isolated. ^aIsolated as the S_NAr product with N-Boc piperazine. ^bVia halogen exchange from the corresponding sulfonyl chloride. See SI for experimental details.

To demonstrate the differentiated reactivity of sulfonyl chlorides and fluorides, we subjected 2-chloro-6-(trifluoromethyl)nicotinic acid to both sets of halosulfonylation conditions, followed by an identical one-pot amination. These two sequences gave different products: the sulfonyl chloride was converted exclusively to the expected sulfonamide product (38, 65% yield) while the sulfonyl fluoride gave exclusively the S_NAr product (39, 71% yield). This selectivity is due to the attenuated electrophilicity of sulfonyl fluorides and can allow for programmable S_NAr/sulfonamide formation sequences.

To exemplify the utility of these methods in drug discovery programs, we undertook the synthesis of sulfonamide analogues for amide-containing biologically active molecules (Figure 5). Sulfonamide analogues of (±)-Bitopertin and Vismodegib could be synthesized in high yields directly from their amide coupling partners (40 and 41, 53 and 72% yield, respectively). Additionally, native carboxylate pharmacophores could be directly converted into sulfonyl fluoride-containing chemical probes potentially capable of forming covalent linkages in the binding sites of their target proteins.¹⁸ Specifically, Ataluren and Lumacaftor were converted into their sulfonyl fluoride analogues in good yield (42 and 43, 62 and 41% yield, respectively). Finally, the tolyl group of Celebrex was converted to an aryl sulfonyl fluoride in a two-step sequence involving benzylic oxidation and subsequent decarboxylative fluorosulfonylation (44, 54% yield).

Figure 5.

Late stage functionalization via Cu-LMCT halosulfonylation. All yields isolated. See SI for experimental details. ^aAfter two-step benzylic oxidation and decarboxylative fluorosulfonylation. Yield given is for the fluorosulfonylation step.

In summary, we have developed an approach for aromatic decarboxylative halosulfonylation that is amenable to a one-pot synthesis of sulfonamides from traditional amide coupling partners. We anticipate that these reactions will be of value to the medicinal chemistry community by assisting in the rapid synthesis of diverse organosulfur products.