Modular Synthesis of Alkenyl Sulfamates and β-Ketosulfonamides via SuFEx Click Chemistry and Photomediated 1,3-Rearrangement

Abstract

Herein, we report a synthesis of medicinally-relevant β-ketosulfonamides via a photomediated 1,3-rearrangement of alkenyl sulfamates. This protocol tolerates a wide array of sensitive functional groups including alkenes, alkynes, and nitrogen-based heterocycles. Additionally, this work provides a general approach towards alkenyl sulfamates via a two-step Sulfur(VI) Fluoride Exchange (SuFEx) sequence capitalizing on SO₂F₂ as a linchpin to efficiently couple readily available ketones and amines without a large excess of either partner.

Graphical abstract

Sulfur(VI)-containing functional groups such as sulfonamides, sulfamides, and sulfamates are becoming increasingly prevalent in pharmacophores. Sulfonamides alone were present in 25% of all sulfur-containing drugs approved by the US FDA through 2016¹ and as a result, are one of the most widespread functional groups in active pharmaceutical ingredients.¹ Sulfonamides are also ubiquitous in agrochemicals.² While the introduction of aryl sulfonamides can be efficiently achieved via the intermediacy of aryl sulfonyl chlorides prepared by electrophilic aromatic substitution,^3a Suzuki-Miyaura cross-coupling,^3b or C–H thianthrenation,^3c their alkyl counterparts are typically more challenging to access.⁴ Among these, β-ketosulfonamides are particularly valuable as precursors for a variety of bioactive molecules.⁵ Interestingly, alkenyl sulfamates, which are constitutional isomers of β-ketosulfonamides, are a virtually unexplored class of compounds; the properties and potential bioactivity of these molecules, which are structurally related to common alkenyl (vinyl) triflates, remain unknown. Herein, we report the efficient and modular synthesis of β-ketosulfonamides through a photomediated 1,3-rearrangement of alkenyl sulfamates readily obtained via Sulfur(VI) Fluoride Exchange (SuFEx) click chemistry.⁶

Typical syntheses of β-ketosulfonamides rely on a polar disconnection between a silyl enol ether,^7,8 or an enamine,⁹ and a sulfamoyl chloride reagent (Figure 1a). This approach requires the multistep preparation of highly reactive sulfamoyl chloride reagents that are generally sensitive to moisture and challenging to purify.¹⁰ Additionally, only monoalkylated sulfamoyl chloride derivatives were shown to be compatible⁸ and the propensity of the S–Cl bond to undergo homolytic cleavage^4,5b might lead to side reactions.^6a The reverse strategy, which relies on the Claisen condensation of deprotonated sulfonamide and an ester, remains rare.¹¹ In 2019, Wang and coworkers reported the transformation of enol silyl ethers into β-ketosulfonamides via a radical process catalyzed by a Ru photocatalyst (Figure 1a).^5b While an elegant departure from the traditional polar disconnection, this approach still suffers from the use of sulfamoyl chloride derivatives and requires a large excess of the silyl enol ether (5 equivalents), thereby hampering large-scale applications with advanced coupling partners. Following our investigation of the synthesis of sulfamides via SuFEx,¹² we hypothesized that sulfuryl fluoride (SO₂F₂) could serve as a linchpin to access a broad array of β-ketosulfonamides from simple ketones and amines. Engaging an alkenyl fluorosulfonate (obtained from the reaction of an enolate with SO₂F₂) with a variety of amines should allow the rapid formation of alkenyl sulfamates. Subsequent 1,3-rearrangement should deliver the desired β-ketosulfonamides (Figure 1b). This sequence merges polar and radical approaches, and would provide a modular and efficient route to these desirable intermediates while presenting the opportunity to explore uncharted chemical space through the isolation of alkenyl sulfamates.

Figure 1.

(a) Prior syntheses of β-ketosulfonamides. (b) Synthesis of β-ketosulfonamides from alkenyl sulfamates through photomediated 1,3-rearrangement.

Aryl fluorosulfonates are known alternatives to aryl triflates in Pd-catalyzed cross-coupling reactions,¹³ but only a handful of alkenyl fluorosulfonates have been reported thus far,¹⁴ thereby precluding exploration of their reactivity. α-Tetralone (1) was selected as a test substrate to develop the desired synthetic sequence. Following deprotonation of 1 with lithium diisopropylamide (LDA) in THF at –78 °C and trapping of the resulting enolate with SO₂F₂, alkenyl fluorosulfonate 2 was isolated in 84% yield. However, these conditions led to lower yields with more sensitive substrates (Table S2), which prompted us to screen alternative bases. Potassium tert-butoxide was found to afford 2 in 90% yield (Figure 2), which compares favorably to the 75% yield reported by Sharpless and coworkers using LiHMDS.¹⁵ These optimized conditions allowed the efficient preparation of a broad array of alkenyl fluorosulfonates while circumventing the use of corrosive reagents, such as fluorosulfonic acid.^14a The next challenge was the preparation of alkenyl sulfamates through the second step of the SuFEx coupling. These compounds are exceedingly rare in the literature and only a few alkenyl sulfamates have been previously isolated.¹⁶ Ball, am Ende, and coworkers recently demonstrated that Lewis acidic calcium triflimide (Ca(NTf₂)₂) selectively activates a variety of sulfonyl fluoride derivatives towards clean fluoride exchange with nucleophilic amines.^17,18 Ca(NTf₂)₂¹⁹ presumably coordinates to both Lewis basic oxygens of the sulfonyl group, thereby allowing for efficient SuFEx reaction. These conditions delivered alkenyl sulfamate 3 in 98% yield (Figure 2). With optimized SuFEx conditions in hand, diverse ketones and amines were coupled to form a variety of alkenyl sulfamates (Figure 3). High-to-excellent yields over two steps were obtained with a range of aliphatic and aromatic amines, including primary (4, 6, and 9) and secondary amines (3, 5, 7, and 8). The mild conditions of the second step allowed the incorporation of 5-methylhex-4-en-1-amine, allylamine, and propargyl amine to sulfamates 10–12. Halogenated benzyl amines were successfully coupled to form 13 and 14 in 69% and 56% yield, respectively. Finally, L-proline derivative S5 proved a competent amine in this reaction, which proceeded with minimal racemization (<3%). The scope of the ketone partner was subsequently explored. α-Tetralone and acetophenone derivatives adorned with electron-rich and electron-poor substituents were coupled in moderate yields (16–20). A variety of heterocycle-containing sulfamates were synthesized including drug-like pyridine (21–24) and pyrazine 25 congeners. Impressively, aliphatic ketones such as tert-butylcyclohexanone and the natural product (S)-carvone were successfully transformed into sulfamates 26 and 27.

Figure 2.

Design of a synthesis of alkenyl sulfamates via SuFEx chemistry.

Figure 3.

Scope of alkenyl sulfamates. The sequence was performed with 0.15 mmol of ketone substrate. All yields are isolated. Deviations from general conditions: ^aMeCN (step 2); ^bLDA (step 1); ^cEt₂O (step 1); ^dDCM at 0 °C (step 1).

Alkenyl sulfamates were purified through standard column chromatography. No decomposition was observed under ambient conditions for several months and sulfamate 3 did not hydrolyze in acidic or basic aqueous media at room temperature. Investigation of the final step en route to β-ketosulfonamides was initiated with N-benzyl alkenyl sulfamate 4. Attempts to trigger the 1,3-rearrangement thermally or in the presence of Brønsted or Lewis acids led to poor yields (Table S3). Delightfully, irradiation with visible light led to the formation of β-ketosulfonamide 29 (Figure 4). Optimization of the photochemical 1,3-rearrangement (Table S4) allowed the isolation of 29 in 99% yield using blue LEDs (456 nm) and a 1:1 mixture of DMSO:H₂O (Figure 4). Unfortunately, N-dialkylated alkenyl sulfamates such as 3 did not undergo the 1,3-rearrangement under these conditions (Table 1, entry 1). We hypothesized that using a photosensitizer might encourage the photochemical activation of recalcitrant substrates. While the addition of eosin B to the otherwise identical reaction conditions did not provide any desired product (Table 1, entry 2), adding 9-fluorenone led to the formation of 28 in 50% yield (Table 1, entry 3). Ir(ppy)₃, another commonly used photosensitizer,²⁰ afforded the desired product in 70% yield (Table 1, entry 4). Increasing the reaction time to 48 h resulted in an improvement of the yield to 95% (Table 1, entry 5). Surprisingly, when no effort was made to remove O₂ via “freeze-pump-thaw” degassing, the reaction reached completion in only 6 h with 97% yield (Table 1, entry 6).²¹ In the presence of air but without photocatalyst, a 28% yield was observed (Table 1, entry 7), which supports the non-spectator role played by O₂ in the reaction process. Full recovery of 3 was observed when the reaction was carried out in the absence of light (Table 1, entry 8).

Figure 4.

Scope of β-Ketosulfonamides. The reaction was performed with 0.15 mmol of substrate. All yields are isolated. ^aNo photocatalyst was used. ^b(Ir[dF(CF₃)ppy]₂(dtbpy))PF₆ instead of Ir(ppy)₃.

Table 1.

Optimization of the Photomediated 1,3-Rearrangement

entry	deviation from the initial conditions	yielda
1	None	N.R.
2	Addition of eosin B (1 mol%)	N.R.
3	9-Fluorenone (1 mol%) instead of eosin B	50%
4	Ir(ppy)3 (1 mol%) instead of 9-fluorenone	70%
5	Ir(ppy)3 (1 mol%) and reaction time = 48 h	95%
6	Ir(ppy)3 (1 mol%), no degassing, and reaction time = 6 h	97%
7	No Ir(ppy)3, no degassing	28%
8	same as entry 6 but in the dark	N.R.

Reactions were performed with 0.036 mmol of 3. N.R. = No reaction.

^aYields were calculated by NMR using 1,4-dimethoxybenzene as internal standard.

The scope of this intriguing photochemical 1,3-rearrangement was explored using the optimized conditions and a variety of alkenyl sulfamates (Figure 4). All alkenyl sulfamates derived from tetralone (29 to 41) underwent rearrangement in good to quantitative yields. Importantly, alkene (35 and 36) and alkyne (37) functional groups were tolerated in this photochemical reaction. Alkenyl sulfamate 15 synthesized from L-proline afforded 40 as a 1:2 mixture of diastereomers after rearrangement. Similarly, a modest diastereoselectivity was observed in the synthesis of 31 (1:1.4 dr). Investigation of the reactivity of alkenyl sulfamates prepared from a variety of ketones was subsequently conducted. High yields were obtained for acetophenone derivatives including fluorinated compounds 43 and 44 and dimethyl analog 45. Pyridine-containing sulfamates (46–49) and pyrazine 50 were all isolated in great yields as well, which bodes well for applications in drug discovery. Interestingly, aliphatic sulfamates 26 and 27 were impervious to the 1,3-rearrangement. The high stability of these compounds suggests that aliphatic alkenyl sulfamates may prove a useful motif in medicinal chemistry. Additionally, this sequence can be coupled to the venerable CuAAC click reaction to rapidly build structural complexity from readily available building blocks.²² For example, S13 was synthesized in 94% yield from 37 under typical CuAAC conditions (See SI).

A few experiments were devised to investigate the 1,3-rearrangement. A cross-over experiment with sulfamates 3 and 22 led to the formation of four products (Figure S6), which supports a non-concerted transfer of the sulfamoyl group. Addition of TEMPO to the reaction mixture inhibited the rearrangement (See SI),⁴ which suggests a radical pathway as postulated for the 1,3-rearrangement of alkenyl tosylates,²³ rather than an anionic process.²⁴ Since the reaction takes place without photocatalyst, Ir(ppy)₃ is believed to act as a photosensitizer that facilitates the homolytic cleavage of the S–O bond via energy transfer, which is in line with the report of Li and coworkers on the photochemical rearrangement of alkenyl tosylates.^23b However, further studies are necessary to rule out a mechanism that relies on single electron transfer. Finally, a radical clock experiment was performed using cyclopropyl derivative 53.²⁵ Irradiation of 53 in the presence of (Ir[dF(CF₃)ppy]₂(dtbpy))PF₆ led to the formation of rearranged product 54, as well as partially isomerized starting material as determined by NOESY experiments (Figure 5a). The cyclopropyl ring was left intact during the rearrangement, which suggests that a chain propagation reaction involving attack of a sulfur-centered radical into the double bond is more likely than a radical–radical recombination. A similar chain mechanism has been put forward for the rearrangement of stilbenyl tosylate derivatives at high temperatures in the presence of a radical initiator.^23a However, attempts to induce the 1,3-rearrangement of 7 with AIBN only led to low yields (see SI). While the exact role of O₂ remains puzzling, reactive oxygen species have been shown to serve as radical initiators in other photochemical transformations.²⁶ A postulated chain mechanism consistent with experimental data and literature precedent is summarized in Figure 5b. Homolytic fragmentation of alkenyl sulfamate I in the reaction conditions likely leads to the formation of a small amount of α-ketonyl radical II and sulfamoyl radical III. Addition of III into the double bond of IV is subsequently proposed to take place, followed by regeneration of sulfamoyl radical III and concomitant formation of β-ketosulfonamide V.

Figure 5.

(a) Radical clock-experiment. (b) Postulated mechanism.

In summary, we have developed a modular and efficient synthesis of β-ketosulfonamides through a photomediated 1,3-rearrangement of alkenyl sulfamates accessed via SuFEx click chemistry. The SuFEx process allows the preparation of a broad array of alkenyl sulfamates in good-to-excellent yields by coupling various amines and ketones. Both SuFEx steps and the subsequent 1,3-rearrangement were shown to tolerate a range of functional groups, including alkenes and alkynes, and are therefore promising for the design of medicinally relevant molecules with sulfur(VI) linkages. The mild conditions and non-excess amounts of coupling partners are clear advantages over prior methods. Initial experiments suggest that the photochemical rearrangement follows a radical chain mechanism triggered by initial generation of a sulfamoyl radical through homolytic cleavage. The presence of oxygen was found to be beneficial to the rearrangement, which should facilitate the large-scale implementation of this reaction.