Synthesis of Aryl Sulfonamides via Palladium-Catalyzed Chlorosulfonylation of Arylboronic Acids

Abstract

A palladium-catalyzed method for the preparation of sulfonamides is described. The process exhibits significant functional group tolerance and allows for the preparation of a number of arylsulfonyl chlorides and sulfonamides under mild conditions.

The medical significance of aryl sulfonamides can be traced to the 1930’s, with the discovery and development of the first commercially available antibiotics, the so-called sulfa drugs.¹ Today, the presence of sulfonamides in medicinal agents is widespread; close to ten percent of the top 100 pharmaceuticals prescribed in 2011 either bear a sulfonamide subunit or are co-administered with a sulfonamide-containing drug.² In general, aryl sulfonamides can be prepared by the straightforward reaction of a sulfonyl chloride with an amine.³ However, the difficulties associated with sulfonamide synthesis stem not from the amination reaction, but from the preparation of sulfonyl chlorides themselves. Two types of processes represent the current state of the art in arylsulfonyl chloride synthesis: (1) electrophilic aromatic substitution (EAS) with chlorosulfonic acid⁴ and (2) oxidative chlorination of organosulfur compounds.^5–9 The use of both approaches suffers from significant limitations. In particular, the acidic conditions required for EAS processes (and most oxidative chlorination methods) impose severe restrictions on substrate scope. Furthermore, desired substitution patterns may be inaccessible via EAS because the regioselectivity is dictated by the intrinsic properties of the parent arene. Traditionally, oxidative chlorination involves the use of hazardous reagents (e.g., aqueous chlorine)⁶ or strong chlorinating agents (e.g., SOCl₂⁷ and SO₂Cl₂).⁸ Although milder conditions have been reported,⁹ oxidative chlorination of thiophenol derivatives ultimately requires prior formation of a carbon–sulfur bond. ^10–11

In principle, a transition metal catalyst could obviate the need for such reagents and permit the convergent synthesis of sulfonamide analogs, allowing variation of both sulfonyl (–SO₂–) substituents. Unfortunately, success in this area of catalysis is limited to a single Pd-catalyzed aminosulfonylation process, initially reported by Willis in 2010, to prepare N-amino-sulfonamides (1) from aryl iodides and hydrazines (Eq. 1).¹² While this example represents an important achievement in sulfonylation chemistry, amine nucleophiles remain incompatible with these types of couplings.^12–13 To address this limitation, we devised an alternative strategy (Eq. 2), which involves oxidative addition of LPd(0) to an electrophile of the type X–SO₂–X′ (A), where X and X′ represent leaving groups of different reactivity. Subsequent coupling with an organometallic nucleophile would afford electrophilic species B, which, in turn, would serve as a direct precursor to the sulfonamide product (C).

Willis and coworkers (Eq. 1)

Our approach (Eq. 2)

Early in the investigation, we selected arylboronic acids as coupling partners due to their compatibility with other functional groups, ready availability, and ease of handling. Upon examining the reaction sequence from a retrosynthetic perspective, we envisioned exploiting arylsulfonate esters (3, Scheme 1)¹⁴ as the immediate precursors to sulfonamides, and thus potential targets for Suzuki-Miyaura cross-coupling. In turn, we proposed that sulfonate esters (3) could be prepared from an aryl chlorosulfate derivative (2)¹⁵ via Pd-catalysis. However, contrary to our expectations, we discovered serendipitously that this coupling process generates arylsulfonyl chlorides (4) in preference to 3, and we describe herein the development and scope of a Pd-catalyzed chlorosulfonylation reaction.

Scheme 1.

Sulfonylation of Arylboronic Acids

Initial experiments focused on identifying conditions to prepare 3a from the simplest aryl chlorosulfate, phenyl chlorosulfate (2a, Table 1), which is an easy to prepare, easily handled liquid.¹⁶ Results from a preliminary survey of bases are outlined in Table 1, utilizing 4-methoxyphenylboronic acid in combination with a catalyst derived from Pd(OAc)₂ and tert-BuBrettPhos (L8) or DavePhos (L1). Unexpectedly, we observed mixtures of 3a and the corresponding sulfonyl chloride (4a) when excess oxygen base was used (entries 1–3). More importantly, we discovered that 4a could be formed exclusively in the absence of base (entries 4 & 5), and that neither product is formed in the absence of Pd or phosphine ligand.¹⁷ Additional control experiments verified that aryl sulfonate esters (3) are not converted to the corresponding sulfonyl chlorides (4) under the reaction conditions;¹⁸ conversely, ester 3a is generated when crude reaction mixtures of 4a are treated with excess K₃PO₄ or K₂CO₃ (following complete consumption of 2a in the absence of base).

Table 1.

Pd-Catalyzed Chlorosulfonylation^a

Entry	Ligand	Base	3a (%)b	4a (%)b
1	L8	K3PO4	31	4
2	L8	K2CO3	15	22
3	L8	Na2CO3	4	36
4	L8	none	0	59
5	L1	none	0	87

^aPerformed on a 0.5 mmol scale with respect to 2a.

^bGC yields.

The above results suggest a catalytic cycle in which Pd(0) inserts into the SO₂–OPh bond of 2a in preference to the SO₂–Cl bond (Figure 1).¹⁹ The phenoxy substituent of the resulting Pd-sulfinate complex (6) would be expected to facilitate transmetallation without the aid of an oxygen base;²⁰ subsequent reductive elimination from 7 would yield 4 and regenerate the active Pd(0) catalyst. Oxygen bases likely promote the release of phenoxide, which in turn would react with the sulfonyl chloride to provide 3. Examination of the chemical literature revealed that Buncel and coworkers have also observed the displacement of aryloxide from aryl chlorosulfate derivatives (2), but in the context of direct nucleophilic additions.^15,21 Nonetheless, this precedent demonstrates the lability of the SO₂–OPh bond, as we have observed for the transfer of –SO₂Cl in the analogous Pd-catalyzed process.

Figure 1.

Proposed Catalytic Cycle

In light of these results, we focused our efforts on the preparation of sulfonyl chlorides as they represent ideal precursors to sulfonamides. However, the system represented in Table 1, entry 5 proved ineffective for both electron-deficient and ortho-substituted substrates. For example, subjecting 2-methoxyphenylboronic acid to these conditions resulted in near complete recovery of phenyl chlorosulfate (2a). Further optimization revealed that yields of aryl-sulfonyl chlorides could be increased by employing anhydrous acetone as solvent in combination with a catalytic amount of Na₂CO₃ (5 mol %). Although Pd(OAc)₂ proved to be a competent palladium source, the use of palladacyclic precatalysts²² (P1–P7) allowed for these experiments to be conducted at lower temperatures. As exemplified in the coupling of 2a with 2-methoxyphenylboronic acid (Figure 2), precatalysts based on diphenyl-(L3 and L5) and di-tert-butylbiaryl phosphine ligands (L2 and L7) proved to be the most effective. In particular, the PhCPhos precatalyst (P5) afforded 4b in the highest yield (82%), albeit with only a slight improvement over that derived from tert-BuDavePhos, L2 (80%). In contrast, the use of XPhos (L6),²³ previously reported to be an excellent ligand for Suzuki-Miyaura reactions, provided little product.²⁴

Figure 2.

Ligand Effects

With this improved protocol, utilizing 2 mol % P5 and 5 mol % of Na₂CO₃ in acetone, we prepared and isolated a number of sulfonyl chlorides in good yields (Table 2). In general, electron-rich, -neutral and -deficient arylboronic acid reagents represent compatible substrates. Couplings with electron-rich substrates can be conducted at 50 °C, while reactions with electron-deficient substrates are typically slower and require higher temperatures.^25–26 While most of the sulfonyl chlorides are stable to chromatography, electron-deficient compounds, such as 4h and 4j, were found to decompose to varying degrees upon attempted purification.²⁷ Thus to circumvent this problem, the sulfonyl chloride intermediates were converted to sulfonamides (5a–m) directly by simply adding a primary or secondary amine to the crude reaction mixtures of 4 (Table 3).²⁸ Weakly nucleophilic aniline derivatives may also be incorporated into the sulfonamide moiety, but pyridine is required to facilitate amination in these cases. As illustrated in Tables 2 and 3, the chlorosulfonylation reaction tolerates chloro-, bromo-, and iodo-substituted arylboronic acids as well as substrates containing TBS-ethers, ester and acetyl functional groups. Heteroaryl substrates, such as 3-thiophene and 2-dibenzofuran boronic acid, also represent suitable coupling partners.²⁹

Table 2.

Chlorosulfonylation of Boronic Acids^a

^aYields represent isolated yields (average of two runs): 2a (1 mmol), ArB(OH)₂ (1.5 mmol), Na₂CO₃ (5 mol %), P5 (2 mol %), degassed, anhydrous acetone (2 mL), 50–70 °C, 12 h.

Table 3.

Preparation of Sulfonamides

^aIsolated yields (average of two runs). Step 1: See conditions in Table 2.

^bP4 (2 mol %) was used. Step 2:

^cR₂NH (2.2 mmol), rt, 1.5 h.

^d1-methylpiperazine (1.2 equiv.) and DIPEA (2.0 equiv.) were used.

^eRNH₂ (3.0 equiv.), rt, 1.5 h.

^fArNH₂ (1.2 equiv.), pyridine (3.0 equiv.), rt, 5 h.

Several additional features of this chemistry are noteworthy. First, Pd(0) reacts with phenyl chlorosulfate (2a) in preference to aryl iodide groups bearing electron-withdrawing sulfonyl groups para to the iodo-substituents (e.g., 4e and 5i). Second, while others have demonstrated that arylsulfonyl chlorides themselves are efficient cross-coupling partners for various Pd-catalyzed processes,^30–32 these intermediates are essentially unreactive under the chlorosulfonylation conditions. Third, palladium typically catalyzes the desulfonylation of arylsulfonyl chlorides; as a result, these substrates (4) are often used as aryl halide equivalents for carbon–carbon bond-forming processes.³² In contrast, the Pd-catalyst derived from L5 promotes carbon– sulfur bond formation, thus enabling the installation of a sulfonyl chloride (–SO₂Cl) functional group.³³ Finally, regarding the oxidative addition step, we surmised that the sp³ oxygen atom of 2a might direct the insertion of Pd(0) into the proximal PhO–SO₂ bond; however, results from a competition experiment between 2b and the bulkier 2c (Eq. 3) revealed that 2c is slightly more reactive than 2b (ratio of recovered 2b/2c = 1.7:1).³⁴ Moreover, this counterintuitive trend in reactivity is not specific to Pd-catalysis. For example, reacting a 1:1 mixture of 2b and 2c with piperidine (Eq. 4) resulted in a comparable ratio of recovered 2b/2c (1.6:1) along with formation of the sulfamoyl chloride (8) derived from piperidine. To explain the displacement of phenoxide from non-catalyzed nucleophilic additions to 2, Buncel has invoked an S_N2 mechanism in which elongation of the S–O bond relieves considerable strain in the bipyramidal transition state; according to the authors, such relief of strain would not accompany partial rupture of the S–Cl bond.²¹ In this regard, increasing the size of the aryloxy substituent of 2 may intensify this stereoelectronic effect and further weaken the S–O bond.

(Eq. 3)

(Eq. 4)

In summary, we demonstrated that phenyl chlorosulfate (2a) represents an excellent [SO₂Cl]⁺ synthon in the context of Pd-catalyzed Suzuki-Miyaura cross-coupling. The chlorosulfonylation reaction exhibits considerable functional group tolerance and the transformation is inherently regioselective; the substitution patterns of many of the products shown (such as 4h–4j) cannot be accessed by EAS processes. Furthermore, the sulfonyl chloride intermediates can be derivatized in situ and isolated as the corresponding sulfonamides. Therefore, both the aryl and amine components of arylsulfonamides can be installed in a single synthetic operation from readily available reagents. Investigations aimed at broadening the scope of this transformation and further delineating the mechanism of sulfur-oxygen bond scission are currently in progress.