Introduction

Skeletal reorganization is a type of process involving multiple bond cleavage and formation for molecule framework reassembly (Fig. 1a)1,2. Owing to the intriguing reaction mechanism, high atom-economy property and capability of accessing complex and synthetically challenging molecules, the skeletal reorganization process has attracted considerable attention and also been widely applied in organic synthesis toward diverse carbocyclic and heterocyclic compounds3,4,5. For example, a variety of transition-metal-catalysed skeletal reorganizations of enynes have been explored for efficient synthesis of polycyclic compounds6,7,8,9,10. Recently, Sun and co-workers discovered an unusual skeletal reorganization of oxetane for the synthesis of 1,2-dihydroquinolines (Fig. 1b)11. Meanwhile, Liu and co-workers established a skeletal reorganization protocol of olefine via a radical-initiated cyclization/1,n (n = 3, 4, 5) vinyl migration cascade for accessing medium- and large-sized cycles, which are ubiquitous structural motifs in natural products and pharmaceutical agents (Fig. 1c)12. Very recently, Zhu and co-workers reported a novel skeletal reorganization of kojic acid- or maltol-derived alkynes under Indium catalysis, which provided an expeditious access to valuable hydroxylated benzofurans (Fig. 1d)13. Despite these advances, the investigation of skeletal reorganizations toward synthetically challenging and biologically interesting molecules remains continuously interesting and important.

Fig. 1: Skeletal reorganization.
figure 1

a Scheme of a skeletal reorganization in molecules. b In-catalysed skeletal reorganization of oxetanes. c Radical-initiated skeletal reorganization of olefines. d In-catalysed skeletal reorganization of kojic acid- or maltol-derived alkynes.

Full size image

The four-membered sulfur-containing heterocycles, such as thiete sulfones, thietane sulfones and thietanes, are strained small ring compounds that have found wide applications in the discovery of dye, drug and pesticide (Fig. 2)14,15,16,17,18,19. Typically, the unsaturated thiete sulfones showed valuable synthetic utility in organic synthesis. For example, thiete sulfones could be used as dienophiles in the Diels–Alder reaction with tetraphenyl cyclopentadienones or isobenzofurans for accessing bridged and fused-ring compounds20,21,22. Moreover, they could participate in [3 + 2] cycloadditions with diazo compounds or nitrile oxides for the synthesis of heterocycles23,24. Recently, thiete sulfones have been investigated in C–H functionalization to establish axially chiral molecules and macrocyclic compounds25,26. However, the conventional synthesis of thiete sulfones relies on multi-step routes and also suffers from narrow scope27,28,29,30,31,32. Therefore, it would be interesting to explore a distinct and efficient approach to functionalized thiete sulfones.

Fig. 2: Sulfur-containing four-membered ring in useful molecules.
figure 2

Some representative cases are listed.

Full size image

Ynamides are a type of N-substituted electron-rich alkynes that exhibit unique chemical properties and serve as versatile synthons in organic synthesis33,34,35,36,37,38,39,40,41,42. For example, ynamides could act as flexible cyclization partners in heterocycle synthesis43,44,45, carbene precursors46,47,48,49 and enamide precursors50,51, racemization-free coupling reagents for peptide and macrolide synthesis52,53,54 and C2 building blocks of multicomponent reactions)55,56,57. In recent years, the intramolecular cyclizations of ynamides, including transition-metal-catalysed and Brønsted acid-catalysed nucleophilic cyclizations, anionic cyclizations and radical cyclizations, have been extensively investigated for the synthesis of N-heterocycles (Fig. 3a)58,59,60,61. However, the skeletal reorganization of ynamides is rarely reported. In 2012, Evano and co-workers reported an s-BuLi-mediated skeletal reorganization of N-Boc ynamides for de novo synthesis of 1,4-dihydropyridines and pyridines, which invoked a process of carbonyl-directed deprotonation and anionic 6-endo-dig cyclization (Fig. 3b)62. Encouraged by these, we hypothesized that the deprotonation at the α-position of sulfonyl moiety of N-sulfonyl ynamides might initiate an anionic 4-exo-dig or 5-endo-dig cyclization to deliver cyclic sulfonamides, which are privileged structures in medicinal chemistry (Fig. 3c)63,64. In continuation of our interests in ynamide chemistry55,56,57,65, herein we would like to report a skeletal reorganization divergence of N-sulfonyl ynamides for selective entry to thiete sulfones and propargyl sulfonamides (Fig. 3d).

Fig. 3: Skeletal reorganization of ynamides.
figure 3

a Intramolecular cyclizations of ynamides. b Skeletal reorganization of N-Boc ynamides. c Our initial hypothesis. d This work: Skeletal reorganization divergence of N-sulfonyl ynamides.

Full size image

Results

Reaction optimization

We commenced our study by using N-sulfonyl ynamide 1a and lithium base to investigate this reaction. As shown in Table 1, we initially treated 1a with n-BuLi at −40 °C for 1 h and then quenched with MeOH; a new major product 2a and a minor product 3a were isolated (entry 1). The standard analysis, including 1H nuclear magnetic resonance (NMR), 13C NMR and mass spectroscopy, was not able to identify these compounds. Gratifyingly, the X-ray analysis showed that 2a was a thiete sulfone and 3a was a propargyl sulfonamide (for more details, see Supplementary Figs. 8 and 9), indicating an unusual occurrence of skeletal reorganization. This unexpected outcome prompted us to optimize the reaction. The next survey of lithium bases showed that lithium diisopropylamine (LDA) was superior to give 2a in 72% yield (entry 2), while the utilization of lithium bis(trimethylsilyl)amide would decrease the yield to 31% (entry 3). The use of sodium bis(trimethylsilyl)amide and potassium bis(trimethylsilyl)amide as base could only give 3a in very low yield (entries 4–5). Using NaH, 1,8-diazabicyclo[5.4.0]undec-7-ene or trimethylamine as base showed inferior to shut down the reactivity even at room temperature, and 1a was recovered (entries 6–8). When LDA was used as base and the temperature was varied to 0 °C or −78 °C, the yield of 2a would slightly decrease (entries 9–10). Considering that lithiation was involved in this process, we tried to add ligands to increase the yield. Next, using of N,N,N’,N’-tetramethylethylenediamine as ligand showed no improvement (entry 11). Interestingly, the utilization of 1,3-dimethyl-tetrahydropyrimidin-2(1H)-one (DMPU) as ligand would exclusively give 3a in good yield (entry 12, 82% yield). Therefore, the skeletal reorganization is divergent and could be controlled by ligand.

Table 1 Reaction optimizationa.
Full size table

Reaction scope study

With the optimized reaction conditions in hand, we next tested the substrate scope of this skeletal reorganization divergence. The starting material N-sulfonyl ynamides could be easily prepared by coupling of sulfonamides and alkynyl bromides. As shown in Fig. 4, various ynamides 1 could participate well in this skeletal reorganization, leading to the corresponding thiete sulfones 2 in moderate-to-good yields. Diverse substitutions on the amino group of ynamides, including n-butyl, benzyl, i-propyl, cyclopentanyl and thiophene-2-ethyl, were found tolerable in this process (2b–2d). The (S)-1-phenylethyl amine-derived ynamide 1g could also engage in this reaction to give the chiral moiety tethered thiete sulfone 2g in 77% yield. Other functional groups, such as alkene and protected alcohol, were also compatible in this reorganization process to give the corresponding products (2h and 2i), which might offer ample opportunities for the further derivatization. In addition, the aniline-derived ynamide (1j) was also applicable to deliver the desired product 2j in an excellent 92% yield. The structure of 2j was further confirmed by X-ray diffraction (for more details, see Supplementary Fig. 10). For the variation of sulfonyl groups, a variety of substitutions were also tested and found amenable in this process to access functionalized thiete sulfones (2k–2q). In detail, N-methylsulfonyl and N-benzylsulfonyl ynamides could reorganize to corresponding thiete sulfones in moderate yield (2k–2m), and the structure of 2k was confirmed by X-ray diffraction (for more details, see Supplementary Fig. 11). When N-cycloalkylsulfonyl ynamides like N-cyclopentanylsulfonyl ynamides and N-cyclohexanylsulfonyl ynamides were used, the skeletal reorganization could deliver spiro-fused thiete sulfones in moderate-to-good yields (2n–2q), and the structure of 2q was also verified by X-ray diffraction (for more details, see Supplementary Fig. 12). With respect of the substitution at the β-position of ynamides, a set of aryl groups functionalized with 2-chloro, 3-methyl, 4-methyl, 4-pentyl, 4-methoxy, 4-fluoro or 4-chloro were tolerable to furnish the corresponding products in good yields (2r–2x), and these substituents did not show any significant electronic and steric effects with the yields. Other aryl groups, including 2-naphthyl, 3-pyridinyl and thiophene-2-yl, were suitable in this process to produce the products successfully (2y–2a’), and the structure of 2a’ was confirmed by X-ray analysis (for more details, see Supplementary Fig. 13). Meanwhile, the cyclohexenyl-substituted ynamide 1d’ could smoothly transform to the corresponding product 2b’ in 62% yield. Notably, silyl substitution was compatible with the process as well. For example, the TIPS-substituted ynamide 1e’ could reorganize to product 2c’ in 58% yield with retention of the TIPS group, while the TMS-substituted ynamide 1f’ could transform to product 2d’ in a remarkable 94% yield accompanied with TMS desilylation. Unfortunately, β-alkyl-substituted and terminal ynamides were not applicable in this process (2e’–2f’). Since the amino-containing full-substituted thiete sulfones are synthetically challenging66, this skeletal reorganization of ynamides provides a robust and efficient approach toward these molecules with the achievement of structural diversity and molecule complexity.

Fig. 4: Scope of thiete sulfones.
figure 4

Standard condition: 1 (0.2 mmol), LDA (0.3 mmol), THF (2 mL), −40 °C, 1 h, then treated with MeOH (0.1 mL). Yields refer to isolated products. a0.6 mmol LDA was used. bProduct of desilylation. cFrom the terminal ynamide.

Full size image

Next, the scope for another skeletal reorganization toward propargyl sulfonamide was also investigated. As shown in Fig. 5, a variety of ynamides were treated with LDA and DMPU toward the formation of propargyl sulfonamides, and the amino substitutions like n-butyl, cyclopentanyl, thiophene-3-ethyl and protected alcohol were compatible with this process (3b–3e). With respect to the sulfonyl substitutions, the cyclopentanyl and cyclohexanyl were applicable in the process to deliver the corresponding products in good yields (3f–3i). Meanwhile, alkynes with substitution on the aryl ring, including 2-fluoro, 2-chloro, 3-methyl, 4-methyl, 4-pentyl and 4-fluoro, could well engage in this skeletal reorganization process to give the corresponding propargyl sulfonamides in good yields (3j–3o). The naphthyl and pyridinyl groups were also applicable to furnish the products smoothly (3p–3q). In terms of the structure of products, this skeletal reorganization involves an interesting 1,3-alkyne migration from N-atom to C-atom. Propargyl sulfonamide is a versatile synthon and their synthesis is challenging by conventional methods. Thus this protocol provides a simple and efficient method to synthesize these molecules from readily available materials along with the fascinating process.

Fig. 5: Scope of propargyl sulfonamides.
figure 5

Standard condition: 1 (0.2 mmol), LDA (0.3 mmol), DMPU (1 mmol), THF (2 mL), −40 °C, 1 h, then treated with MeOH (0.1 mL). Yields refer to isolated products.

Full size image

Synthetic applications and mechanism study

To demonstrate the synthetic utility of this skeletal reorganization divergence, an 8 mmol scale reaction was conducted (Fig. 6a). Under the standard reaction conditions, 1a could be selectively transformed to 2a and 3a in gram-scale, indicating that these skeletal reorganization processes were practical. Considering that these processes may involve lithium intermediates that could offer opportunities for divergent functionalization of products, the derivatization was carried out. As shown in Fig. 6b, when various electrophilic reagents instead of MeOH were used to quench the reaction after treating ynamide 1 with LDA, corresponding electrophilic groups, such as methyl, allyl, propargyl, protected ethanol-2-yl, acetate and acetyl, were successfully installed in the amino group of products, leading to a series of functionalized thiete sulfones (4a–4f). This result not only confirmed that a lithium amino intermediate is involved in the reorganization process but also offered a vast potential for further derivatization of thiete sulfone skeletons. Similarly, when the LDA/DMPU-mediated skeletal reorganization process of ynamide 1a was quenched with MeI, an N-methyl propargyl sulfonamide product 5 was obtained in good yield (Fig. 6c). This experiment suggested the existence of sulfonamide anion intermediate after this skeletal reorganization process.

Fig. 6: Synthetic applications and mechanistic experiments.
figure 6

a Gram-scale reaction of ynamide 1a. b Electrophilic reagents-quenched reorganization toward N-functionalized thiete sulfones 4. c MeI-quenched reorganization toward N-methyl propargyl sulfonamide 5. d Crossover reaction. e α-Methylation experiment. f 13C-Labelled experiment.

Full size image

In order to view more insights into the reaction, the mechanism exploration was conducted. First, the crossover reaction was carried out to further explore the molecular reaction mode of this skeleton reorganization. As shown in Fig. 6d, when ynamide 1r and 1x were subjected in one pot under the standard reaction condition, 2q and 2v were exclusively formed in 79 and 73% yields, respectively. This result excluded the possibility of intermolecular group exchange and verified this reaction is an intramolecular process. Besides, treating the mixture of 1c and MeI with LDA could exclusively deliver the α-methylation product 6 in an excellent 91% yield (Fig. 6e). This outcome disclosed that the α-lithiation was the initial step of the skeleton reorganization. Meanwhile, isotope-labelling experiment was conducted to probe the skeleton reorganization processes. As shown in Fig. 6f, (β-13C)-1a was prepared and subjected to the standard reactions. 13C-2a and 13C-3a could be formed smoothly under standard conditions. The position of 13C-labelled carbon atom in the products was identified, which indicated that phenyl migration did not occur in these skeleton reorganization processes.

Moreover, we conducted density functional theory (DFT) calculations at the level of M06 using the Gaussian 09 suite of computational programs. The 6-31G(d,p) basis set was applied for the C, H, O, N, S and Li atoms. In the LDA-mediated process, there are three possible cyclizations after α-lithiation (Fig. 7). In paths 1 and 2, Li species Int-A may undergo a 4-exo-dig cyclization to deliver isometric Int-B or Int-C via transition state TS1 or TS2 with a barrier of 17.6 or 18.4 kcal/mol, respectively. In path 3, Int-A undergoes a 5-endo-dig cyclization to give Int-D by TS3 with a higher barrier of 20.4 kcal/mol. Similarly, in the presence of DMPU, the formed Li species Int-E may undergo three types of cyclizations as well (Fig. 8). In paths 4 and 5, Int-E may transform to Int-F or Int-G via TS4 or TS5 with a barrier of 22.5 or 23.7 kcal/mol. However, in path 6, Int-E needs a higher barrier of 25.2 kcal/mol to initiate a 5-endo-cyclization to deliver Int-H. The relative lower barrier of paths 1 and 4 suggests that 4-exo-dig cyclization/cis addition is more favoured in these two skeleton reorganization processes.

Fig. 7: DFT calculations for the LDA-mediated cyclization in the skeleton reorganization.
figure 7

Free energy profiles (kcal/mol) of three possible cyclizations from Int-A. path 1 (red): 4-exo-dig cyclization/cis addition leading to Int-B (via TS1), path 2 (blue): 4-exo-dig cyclization/trans addition leading to Int-C (via TS2), path 3 (green): 5-endo-dig cyclization leading to Int-D (via TS3).

Full size image
Fig. 8: DFT calculations for the LDA/DMPU-mediated cyclization in the skeleton reorganization.
figure 8

Free energy profiles (kcal/mol) of three possible cyclizations from Int-E. path 4 (red): 4-exo-dig cyclization/cis addition leading to Int-F (via TS4), path 5 (blue): 4-exo-dig cyclization/trans addition leading to Int-G (via TS5), path 6 (green): 5-endo-dig cyclization leading to Int-H (via TS6).

Full size image

Based on experimental and computational studies, a plausible reaction mechanism for this skeletal reorganization divergence is proposed (Fig. 9). Initially, the α-position of sulfonyl group would undergo lithiation upon treatment with LDA to deliver A and then a 4-exo-dig cyclization would occur to generate four-membered β-sultam intermediate B63,64,67,68. Subsequently, B presumably undergoes a 1,3-sulfonyl migration to form lithium thiete sulfone intermediate C69,70, which could be protonated by MeOH to deliver product 2. When MeOH was replaced by other electrophilic reagents, intermediate C could be functionalized directly to obtain N-substituted thiete sulfone products. In the presence of DMPU, α-lithiation of 1 could deliver Li species D, which undergoes a 4-exo-dig cyclization to form intermediate E. Unlike intermediate B, intermediate E may undergo a β-elimination to generate lithium propargyl sulfonamide F, probably due to the addition of DMPU that dissociates the intermolecular or intramolecular interaction of Li species and change the reactivity. The following protonation would furnish product 3.

Fig. 9: Plausible mechanism.
figure 9

Possible reaction mechanism of the skeletal reorganization divergence.

Full size image

Discussion

In summary, we have discovered a skeletal reorganization divergence of N-sulfonyl ynamides. Upon treatment with lithium base, the N-sulfonyl ynamides could undergo lithiation/cyclization and the sequential ligand-determining 1,3-sulfonyl migration or β-elimination to deliver thiete sulfones or propargyl sulfonamides. This skeletal reorganization divergence features broad substrate scope and scalability. Mechanistic experiments and DFT calculations are conducted to verify the rationality of the proposed mechanism. Therefore, this protocol not only represents a new skeletal reorganization mode but also provides facile and selective access to privileged molecules from the easily accessible ynamides.

Methods

General procedure for the synthesis of thiete sulfones 2

An oven-dried Schlenk tube equipped with a magnetic stirrer bar was purged with argon three times. Ynamide 1 (0.2 mmol) was dissolved in 2 mL anhydrous tetrahydrofuran (THF) and added by a syringe. The mixture was cooled to −40 °C and LDA (2 mol/L in THF, 0.15 mL, 0.3 mmol) was added dropwise. The reaction was stirred at −40 °C for another 1 h. MeOH (0.1 mL) was added to quench the reaction and then the mixture was concentrated under vacuum to obtain the residue, which was further purified by silica gel column chromatography using ethyl acetate/petroleum ether as eluent to give thiete sulfones 2.

General procedure for the synthesis of propargyl sulfonamides 3

An oven-dried Schlenk tube equipped with a magnetic stirrer bar was purged with argon three times. Ynamide 1 (0.2 mmol) was dissolved in 2 mL anhydrous THF and added by a syringe. DMPU (62 μL, 1 mmol) was added and the mixture was cooled to −40 °C. Subsequently, LDA (2 mol/L in THF, 0.15 mL, 0.3 mmol) was added dropwise. The reaction was stirred at −40 °C for another 1 h, and then MeOH (0.1 mL) was added to quench the reaction. The mixture was concentrated under vacuum to obtain the residue, which was further purified by silica gel column chromatography using ethyl acetate/petroleum ether as eluent to give propargyl sulfonamides 3.