Introduction

Since the first successful isolation in 19911 and the first use in catalysis in 19952, N-heterocyclic carbenes (NHCs) have emerged as a powerful class of ligands in transition metal catalysis3,4,5,6,7,8,9,10,11,12,13,14. The tremendous utility of NHCs hinges on strong σ-donation15,16 in combination with tunable steric environment17,18, supercharging the catalytic activity of transition metals beyond other ligands. The most remarkable impact is in the development of Ru-catalyzed olefin metathesis19,20 and Pd-catalyzed cross-couplings21,22,23, where the strong σ-donation and high stability of M–C(NHC) bond render NHCs superior to the more ubiquitous phosphine ligands. Thus far, NHC ligand development in transition metal catalysis has been almost exclusively limited to N-aryl-imidazolylidenes A5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23, such as IPr24,25,26, prepared by deprotonation of symmetrical imidazolium salts (Fig. 1). This is presumably due to enhanced electronic and steric stabilization of the carbene center by two nitrogen atoms as well as two N-Ar wingtip substituents, which render N-aryl-imidazolylidenes more stable and easier to handle27. The pioneering studies by Bertrand and co-workers established that cyclic carbene systems with a marked decrease of heteroatom stabilization, such as CAACs B (cyclic (alkyl)amino)carbenes)28,29,30,31,32,33,34,35, are readily available, showing unique reactivity as supporting ligands in transition metal catalysis. More reactive and less stabilized systems, such as diamidocarbenes C36,37,38, mesoionic carbenes D39,40,41 and remote carbenes E42,43,44,45, have been developed, each class showing varying degrees of heteroatom stabilization and distinctive promise in transition metal catalysis46,47,48.

Fig. 1: Structures of N-heterocyclic carbenes with different degrees of stabilization.
figure 1

Energies calculated at B3LYP 6-311 + +g(d,p) level, R = Dipp.

In this context, N-aryl thiazol-2-ylidenes F are an intriguing class of N-heterocyclic carbenes (Fig. 1). Following the isolation of a stable thiazol-2-ylidene by Arduengo in 199749, this class of ligands stayed dormant until 2008, when Grubbs demonstrated the unique reactivity of Ru-based thiazol-2-ylidene olefin metathesis catalysts50. To our knowledge, this is the only application of N-aryl thiazol-2-ylidene ligands in transition metal catalysis reported to date51,52,53,54. More recently, there has been a resurgence of organocatalyzed radical relays and decarboxylative couplings made possible through the persistent radical stabilization by thiazol-2-ylidenes55,56,57,58,59,60. In the meantime, studies by Boydston demonstrated organocatalyzed anodic oxidation of aldehydes through in situ generation of electroauxiliaries of thiazol-2-ylidenes61, while the first characterization of elusive Breslow intermediates from thiazol-2-ylidenes by spectroscopic and crystallographic methods has been reported62,63. Thiazol-2-ylidenes are key intermediates in biochemical transformations of vitamin B164,65.

Geometrically, replacement of one of the nitrogen atoms in imidazol-2-ylidene systems with sulfur in thiazol-2-ylidenes results in disrupting the ring geometry of imidazolylidenes66,67,68. At the same time, there is a strong electronic effect in decreasing stabilization of the carbene center through diminished π donation from sulfur69. Finally, the sulfur atom does not bear any wingtip substituents that in imidazolylidene systems often provide a significant contribution to the stabilization of the carbene center5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,66,67,68. These geometrical and electronic factors might explain why, with exception of the report by Grubbs50, N-aryl thiazol-2-ylidenes have been unexplored as NHC ligands in transition metal catalysis.

In terms of electronics, the diminished π donation from sulfur due to ring geometry and large sulfur radius is expected to result in more electrophilic carbenes than traditional imidazol-2-ylidene systems, while maintaining strong donor ability (Fig. 1)69.

In terms of geometry, the effect of typical NHC ligands on M–C(NHC) bond is defined as “umbrella” shaped, in contrast to cone shaped phosphines (Fig. 1)17,18. The combination of a nitrogen atom with a quaternary carbon in CAACs renders these ligands as “wall-shaped” in some cases with regard to the M–C(NHC) bond28,29,30,31,32,33,34,35. The steric properties of N-aryl thiazol-2-ylidenes render these ligands “half umbrella” shaped with the nitrogen N-wingtip oriented toward the M–C(NHC) bond and lack of substitution on the sulfur atom.

As a part of our interest in NHC catalysis70,71,72,73,74,75,76, herein, we report the study on steric and electronic properties of thiazol-2-ylidenes. Most importantly, we demonstrate that the thiazole heterocycle and enhanced π-electrophilicity result in a class of highly active carbene ligands that supersede imidazol-2-ylidenes. We present the evaluation of steric, electron-donating and π-accepting properties as well as structural characterization and coordination chemistry. Considering the key role of N-heterocyclic ligands, we envision that N-aryl thiazol-2-ylidenes will be of broad interest as ligands in chemical synthesis.

Results

Synthesis of Thiazol-2-ylidene precursors

N-Aryl thiazol-2-ylidenes carbene precursors are readily available on multigram scale following protocols for organocatalytic transformations (see Supplementary Method 2)55,56,57,58,59,60. Four N-aryl thiazol-2-ylidenes carbene precursors were selected as a starting point (Fig. 2). For the study, we chose structural sulfur analogues of IPr and IMes on 3-aryl-4,5-dimethylthiazol-2-ylidene framework; MeIPrS and MeIMesS. IPr ligand (IPr = 1,3-bis(2,6-diisopropylphenyl)imidazol-2-ylidene; N-Dipp, Dipp = 2,6-diisopropylphenyl) is by far the most common NHC ligand used in transition metal catalysis, while its smaller IMes (IMes = 1,3-bis(2,4,6-trimethylphenyl)imidazol-2-ylidene, N-Mes, Mes = 2,4,6-trimethylphenyl) counterpart is often used for transformations requiring lower steric demand of the N-wingtip substituents. Furthermore, 3-aryl-4,5-cyclohexylthiazol-2-ylidene and 3-aryl-4,5-cycloheptylthiazol-2-ylidene, 6IPrS and 7IPrS were selected on the basis of the recent reports in organocatalytic transformations55,56,57,58,59,60, where the fused cyclic ring on the backbone of N-aryl thiazol-2-ylidenes often provided advantageous stability of the system.

Fig. 2: Synthesis of thiazol-2-ylidene precursors.
figure 2

See Supplementary Method 2 for details.

Synthesis of Thiazol-2-ylidene Complexes

With access to N-aryl thiazol-2-ylidene precursors, we next prepared Ag(I) complexes [Ag(NHC)2](ClO4) 4a-4d by the reaction with Ag2O in CH2Cl2 (Fig. 3). Interestingly, complexes 4a-4c ([Ag(MeIPrS)2](ClO4) (4a), [Ag(6IPrS)2](ClO4) (4b) and [Ag(7IPrS)2](ClO4) (4c)) were found to be stable to air and moisture and could be fully characterized by X-ray crystallography (Fig. 4, for more details, see Supplementary Note 1 and Supplementary Data 13). In contrast, the less sterically-hindered Ag(I) complex 4d [Ag(MeIMesS)2](ClO4) was found to be significantly less stable. Arduengo reported that small N-wingtip substituents in thiazol-2-ylidenes result in unstable carbenes49. Unsurprisingly, the “half-umbrella” shape of N-aryl thiazol-2-ylidenes requires larger groups at the nitrogen atom for easy handling and isolation. Likewise, we found that the formation of bis-NHC–Ag(I) is needed to prevent decomposition of monomeric Ag(I)–NHCs in these thiazol-2-ylidene systems. From the outset, we were interested in Ag(I)–NHC complexes because of the untapped potential of Ag(I)–NHC complexes in catalysis as compared to other coinage metals77,78.

Fig. 3: Synthesis of Ag(I) complexes.
figure 3

Conditions: Ag2O (0.5 equiv), NaCl (2.0 equiv), CH2Cl2, 25 °C, 16 h, 4a: 96%; 4b: 95%; 4c: 97%; 4d: 90%.

Fig. 4: X-ray crystal structure of Ag(I) complexes 4a-4c.
figure 4

Two views: front (A. top); side (B. bottom). Hydrogen atoms and counterion have been omitted for clarity. Selected bond lengths [Å] and angles [°]: 4a: Ag–C13, 2.082(3); Ag–C30, 2.086(3); N1–C13, 1.335(4); N2–C30, 1.337(4); S1–C13, 1.703(3); S2–C30, 1.697(3); C13–Ag–C30, 175.0(1). 4b: Ag–C1, 2.075(4); Ag–C4, 2.078(4); N1–C1, 1.342(6); N2–C4, 1.331(6); S1–C1, 1.691(5); S2–C4, 1.693(6); C1–Ag–C4, 172.2(2). 4c: Ag–C40, 2.087(2); Ag–C39, 2.081(3); N1–C40, 1.339(3); N2–C39, 1.336(3); S1–C40, 1.699(4); S2–C39, 1.704(3); C40–Ag–C39, 175.5(1). 4a: CCDC 2117719; 4b: CCDC 2117722; 4c: CCDC 2117721.

We next comprehensively evaluated steric and electronic properties of these N-aryl thiazol-2-ylidene ligands. As shown in Fig. 5, the linear copper(I) complex [Cu(MeIPrS)Cl] (5a) was prepared after deprotonation with an excess of KOt-Bu (2 equiv), while Rh(I) complexes, [Rh(6IPrS)(CO)2Cl] (6b) and [Rh(7IPrS)(CO)2Cl] (6c) were prepared by a two-step procedure via [Rh(NHC)(cod)Cl] and the reaction with carbon monoxide. We have also prepared the selenium adducts [Se(NHC)] (5a5d) by adding the free carbene generated in situ to excess of selenium (The respective experimental conditions could be found in the Supplementary Method 46.).

Fig. 5: Synthesis of thiazol-2-ylidene complexes.
figure 5

The respective experimental conditions can be found in the Supplementary Method 46.

Cu(I)–NHC complex [Cu(MeIPrS)Cl] (5a) was fully characterized by X-ray crystallography (Fig. 6, for more details, see Supplementary Note 4 and Supplementary Data 4). Studies by Cavallo and co-workers demonstrated that catalytic pockets of M–NHCs are best described by the % buried volume (%Vbur) of model linear [M(NHC)Cl] complexes79. Complex [Cu(MeIPrS)Cl] (5a) is linear (C(NHC)–Cu–Cl, 177.0°; C–Cu, 1.871 Å), making it a good model for evaluating %Vbur of N-aryl thiazol-2-ylidene ligands. Thus, the (%Vbur) of (5a) is 37.0%, which can be compared with the (%Vbur) of 47.6% determined for [Cu(IPr)Cl] (C–Cu–Cl, 176.7°; C–Cu, 1.881 Å)80. A graphical representation of the steric mapping is shown in Fig. 3b. Importantly, the X-ray crystallographic analysis revealed the (%Vbur) of 50.2%, 52.0%, 22.6%, 23.1% for each quadrant (Fig. 6). The values can be compared with the (%Vbur) of 55.5%, 39.6%, 39.6%, 55.5% for each quadrant of [Cu(IPr)Cl], revealing a “half-umbrella” steric arrangement of N-aryl thiazol-2-ylidene ligands.

Fig. 6: X-ray crystal structure and Topographical steric map.
figure 6

A X-ray crystal structure of complex (5a). Hydrogen atoms have been omitted for clarity. Selected bond lengths [Å] and angles [°]: Cu–C1, 1.871(2); Cu–Cl, 2.0947(7); C1–N1, 1.338(2); C1–S1, 1.700(2); C6–N1, 1.460(3); N1–C3, 1.402(3); S1–C2, 1.723(2); C1–Cu–Cl, 176.98(7); N1–C1–S1, 107.3(1); C6–N1–C1, 121.3(2); C3–N1–C1, 117.0(2); C2–S1–C1, 94.5(1). B Topographical steric map of [Cu(MeIPrS)Cl] (5a) showing % Vbur per quadrant. CCDC 2117739. Note dissymmetry of the ring.

The Tolman electronic parameter (TEP) has been determined from the CO stretching frequencies of [Rh(7IPrS)(CO)2Cl] of νsym = 2078.0 cm−1 and νasym = 2001.4 cm−1 (CH2Cl2, 0.20 M), respectively. This corresponds to a TEP of 2051.9 cm−1 as a combined measure of the electronic properties of N-aryl thiazol-2-ylidene ligands81. These values can be compared with the classical imidazol-2-ylidene IPr (TEP of 2051.5 cm−1) and a model cyclic (alkyl)amino)carbene CAACCy (TEP of 2048.6 cm−1)28, indicating strong donor ability of N-aryl thiazol-2-ylidenes.

In the same vein, selenourea adducts allow to determine π-backbonding of NHC ligands from the 77Se NMR spectra82,83,84,85. As such, the δSe values of 375.99 ppm for [Se(MeIPrS)] and 374.88 ppm, 366.70 ppm, 329.96 for [Se(6IPrS)], [Se(7IPrS)], [Se(MeIMesS)] (CDCl3), respectively, indicate significantly better π-acceptance of N-aryl thiazol-2-ylidenes than imidazol-2-ylidene, IPrSe = 90 ppm), as expected from the sulfur substitution.

Moreover, one-bond CH J coupling constants from 13C satellites of the 1H NMR spectrum give a good prediction of σ-donating properties of NHC ligands86,87. The values of 218.70 Hz for MeIPrS and 218.34 Hz, 218.82 Hz, 219.36 Hz for 6IPrS, 7IPrS, and MeIMesS (HClO4 salts, CDCl3), respectively, are consistent with N-aryl thiazol-2-ylidenes as strongly σ-donating NHC ligands, which can be compared with imidazol-2-ylidene IPr (1JCH = 223.70 Hz). However, at the same time, N-aryl thiazol-2-ylidenes are significantly more π-accepting and feature a distinct “half-umbrella” steric impact.

Ag–NHC-catalyzed cyclization

With structural and electronic characterization of N-aryl thiazol-2-ylidenes, we next evaluated the activity of Ag(I)–thiazol-2-ylidene complexes in catalysis (Table 1, and Figs. 78). As stated above, we selected Ag(I)–NHC complexes because Ag(I) complexes have been much less explored in catalysis than other group 11 metals77,78 as well as to probe electrophilic π-activation of the ligands. Electrophilic O-cyclization of N-propargylic amides was selected as a model reaction due to the importance of the product oxazoline heterocycles in medicinal chemistry research88. As shown, the reaction proceeds under very mild conditions using bis-NHC–Ag(I) salts 4a-d (5-10 mol%) in the presence of AcOH in CH2Cl2 at room temperature (Table 1, entries 1-8, see Supplementary Method 8 for details). AcOH is required as an additive (vide infra)89,90. Likewise, no reaction takes place in the absence of N-aryl thiazol-2-ylidene Ag(I) complexes (Table 1, entries 9-10). Out of the complexes 4a-d, the cycloheptyl complex [Ag(7IPrS)2](ClO4) showed the highest reactivity and was selected for scope studies. The loading could be further decreased to 1 mol% with excellent efficiency (>95%) (Table 1, entries 11-14). Most importantly, the use of classical imidazol-2-ylidene complexes [Ag(IPr)Cl] and [(Ag(IMes)Cl] resulted in negligible reactivity (7-16%) (Table 1, entries 15-16), indicating superior reactivity of N-aryl thiazol-2-ylidenes. Further comparative studies between thiazol-2-ylidene and IMes/IPr to eliminate the effect of counterion were conducted. Specifically, we also prepared and tested thiazol-2-ylidene [NHC–Ag]2PF6 (NHC = 3c, 90% yield), imidazol-2-ylidene [NHC–Ag]2ClO4 (NHC = IPr, <5% yield, IMes, 30% yield) and imidazol-2-ylidene [NHC–Ag]2PF6 (NHC = IPr, <5% yield, IMes, 21% yield). These results are consistent with the superior reactivity of N-aryl thiazol-2-ylidene Ag(I) complexes (See Supplementary Method 10). As expected, marginal reactivity was observed in the presence of soluble silver salts (AgOTf, 11%; AgSbF6, 5%).

Table 1 Optimization of Ag–NHC-Catalyzed Cyclization of N-Propargylic Amidesa.
Fig. 7: Scope of Ag–NHC-catalyzed cyclization of N-propargylic amides.
figure 7

Conditions: amide (1.0 equiv), catalyst 4c ([Ag], 1 mol%), CH3CO2H (1.0 equiv), CH2Cl2 (1.0 M), 25 °C, 8 h. See Supplementary Method 9 for details, 9ad: CCDC: 2125052.

Fig. 8: Late-stage functionalization in Ag–NHC-catalyzed cyclization of N-propargylic amides.
figure 8

Conditions: amide (1.0 equiv), catalyst 4c ([Ag], 1 mol%), CH3CO2H (1.0 equiv), CH2Cl2 (1.0 M), 25 °C, 8 h. See Supplementary Method 9 for details.

Having established the optimal conditions for electrophilic cyclization using N-aryl-thiazol-2-ylidene–Ag(I) complexes, next the scope was examined (Fig. 7).

As shown, the scope of the reaction is very broad and encompasses a variety of N-propargylic amides (Fig. 7). As such, aromatic amides with neutral (9a), electron-donating (9b-9f), electron-withdrawing (9g-9n) substituents at the para, ortho and meta positions could be successfully reacted to give diverse 2-aryl-2-oxazolines. Importantly, medicinally-relevant substituents, such as 3,4,5-trimethoxyphenyl (9f), nitro (9g), cyano (9h), trifluoromethyl (9i-9j) as well as halide functional handles that enable further derivatization (9k-9n) were well compatible with the reaction. Furthermore, π-conjugated substituents, such as naphthyl (9o) and cinnamyl (9p) furnished the products in high yields. Heterocyclic amides, such as 2-thienyl (9q) were well tolerated. Interestingly, highly selective mono-cyclization is possible using meta-substituted propargylic diamide (9r). Pleasingly, aliphatic amides with α-branching (9s) as well as 1° (9t, 9w), 2° (9v), and 3° sterically-hindered aliphatic amides (9u) are tolerated in this process despite the lack of Ar→π*C=O conjugation with the amide oxygen atom. Furthermore, substitution at the methylene carbon adjacent to the nitrogen to deliver 4-substitued oxazolines is also compatible (9x-9aa), increasing the overall synthetic utility of the process. Further, substitution of the alkyne is tolerated without loss in reaction efficiency (9ab-9ad), furnishing fully substituted oxazolines. The product 9ad was crystalline and the structure was confirmed by x-ray crystallography (for more details, see Supplementary Note 1 and Supplementary Data 5), indicating (Z)-geometry of the double bond (dr > 98:2). This result is consistent with an anti-attack of the amide bond oxygen on the Ag(I)–NHC-π-activated alkyne (vide infra).

Most crucially, the mild reaction conditions enabled by the N-aryl thiazol-2-ylidene ligands permit this mode of catalysis to be applied to late-stage functionalization to furnish attractive heterocyclic building blocks for medicinal chemistry and agrochemistry research (Fig. 8). Thus, electrophilic cyclization of propargylic amides from acifluorfen (9af, protoporphyrinogen oxidase inhibitor), piperonylic acid (9ag, trans-cinnamate hydroxylase inhibitor), dicamba (9ah, broad spectrum herbicide), febuxostat (9ai, antigout) and probenecid (9aj, antihyperuricemic) delivered cyclization products in good to high yields without modification of the reaction conditions. This successful late-stage diversification highlights the mild conditions of the present protocol with tolerance to an array of sensitive functional groups (halides, cyano, nitro, sulfonamide, aryl ethers, S-heterocycles), demonstrating prospective impact on medicinal chemistry research.

Mechanism

To gain insight into the reaction mechanism of this intriguing transformation, catalytic cycle was studied by DFT computations (Fig. 9). D3 correction has been omitted in optimization. Based on the previous work89,90, our calculation results show that the catalytic cycles for these processes are comprised of three key steps. In the first step, L2Ag will give active catalyst 1′ in the presence of AcOH. The free energy of activation for this step is 20.1 kcal/mol for TS1 (Fig. 9). This step is also the rate-determining one for this reaction, and the calculated free energy of activation is in good agreement with the kinetics of the reaction. After formation of active catalyst 1, the ligand exchange between AcO- and amide Re would give intermediate 2 with a free energy release of −6.7 kcal/mol. In second step, 2′ would generate putative vinyl–silver intermediate 3′ by a 5-exo-dig cyclisation. The free energy of activation for TS2 (Z) and TS22 (E) is 12.7 and 19.6 kcal/mol, respectively. This calculation result for Z/E selectivity is fully consistent with the experiments results. The final step involves a 1,4-H shift that leads to the final product P1. First, two- and three-molecule HOAc-assisted 1,4-H shift (TS31 and TS32) were calculated. The calculated activation free energy of TS31 and TS32 is high (27.4 and 24.5 kcal/mol, See Supplementary Note 3 for details) to occur under the experimental conditions. Another possible pathway involves HOAc and ligand assisted 1,4-H shift. In this pathway, coordination of HOAc and ligand with 3′ would generate intermediate 4′ with a free energy release of −10.0 kcal/mol. The protodemetallation step would form the product P1 and regenerate the silver catalyst and HOAc. This step is exergonic by −30.6 kcal/mol and the free energy barrier is 15.8 kcal/mol.

Fig. 9: DFT-computed free energy profile of Ag–NHC catalyzed cyclization of N-propargylic amides.
figure 9

See Supplementary Note 3 for computational details.

To further evaluate the effect of nitrogen to sulfur replacement in N-aryl thiazol-2-ylidenes, HOMO and LUMO energy levels of carbenes MeIPrS, 6IPrS, 7IPrS and MeIMesS were determined at the B3LYP 6-311 + +g(d,p) level (Fig. 10 and See Supplementary Note 3 for details). It is now recognized that the donating ability of carbenes is closely associated with the HOMO orbital, while the electron acceptance is associated with the LUMO orbital15,16,28,29,30,31,32,33,34,35,81. Computation of frontier orbitals represents the most accurate evaluation of nucleophilicity (higher energy of HOMO) and electrophilicity (lower energy of LUMO) of NHC ligands15,16,28,29,30,31,32,33,34,35,81, while the comparison must be available at the same level of theory.

Fig. 10: HOMO and LUMO energy levels (eV) calculated at B3LYP 6-311 + +g(d,p).
figure 10

See Supplementary Note 3 for details.

The HOMO of MeIPrS (−6.07 eV, σ-bonding orbital) is comparable with IPr (−6.01 eV), which is a routine model for σ-donating NHCs. The HOMO of 6IPrS, 7IPrS and MeIMesS are −6.05 eV, −5.99 eV, −6.02 eV respectively, indicating that these N-aryl thiazol-2-ylidenes are similarly strongly nucleophilic as N-aryl-imidazol-2-ylidenes. Furthermore, the LUMO + 3 of MeIPrS (−0.36 eV, π-accepting orbital) is compared with the standard imidazol-2-ylidene ligands IMes (−0.03 eV), IPr (−0.20 eV) of the corresponding π-accepting orbitals. The LUMO + 3 of 6IPrS, 7IPrS and MeIMesS are −0.33 eV, −0.29 eV, −0.36 eV, respectively (LUMO + 3 due to required symmetry), indicating that N-aryl thiazol-2-ylidenes are as good π-acceptors as the standard imidazolylidene IMes and IPr ligands. In addition, the HOMO-1 (π-donating orbital) in the series of MeIPrS, 6IPrS, 7IPrS and MeIMesS is −6.57 eV, −6.49 eV, −6.44 eV, −6.51 eV, which is in the same range as for the π-donating orbital for the standard imidazolylidene IMes (−6.44 eV) and IPr (−6.55 eV) determined at the same level of theory. Overall, N-aryl thiazol-2-ylidenes can be characterized as π-accepting, σ-donating and sterically-distinct “half umbrella” shaped ligands that are well-poised for electrophilic catalysis91,92,93,94,95.

It should be further noted that in terms of the steric profile, %Vbur of MeIPrS (37.0%, CuCl complex) is much smaller than that of IPr (47.6%, CuCl complex) and CAACEt (43.1%, AuCl complex)79.

Conclusions

In conclusion, although N-heterocyclic carbenes represent a dominant direction in ligand development in the last 20 years, the majority of efforts in catalysis have been almost exclusively limited to N-aryl-imidazolylidenes, such as IPr. In this study, we reported the study on steric and electronic properties of thiazol-2-ylidenes. We presented comprehensive evaluation of steric, electron-donating and π-accepting properties as well as structural characterization of Ag(I) and Cu(I) complexes of N-aryl thiazol-2-ylidenes. The thiazole heterocycle and enhanced π-electrophilicity result in a class of highly active carbene ligands that supersede imidazol-2-ylidenes. We showed that this mode of catalysis can be applied to late-stage drug functionalization to furnish attractive building blocks for medicinal chemistry. The unique electronic properties in combination with steric differentiation in a “half umbrella” shape with the single nitrogen N-wingtip oriented toward the metal open a plethora of possibilities in the development of enhanced arsenal of thiazol-2-ylidene ligands of broad interest in chemical synthesis. Our ongoing studies are focused on expanding the scope of reactions catalyzed by thiazol-2-ylidenes in catalysis using electrophilic group 11 metals as well as comparative studies using other NHC ligands. We believe that the class of N-aryl thiazol-2-ylidenes is well poised to make an impact on catalysis via electrophilic mechanisms.

Methods

General information

For more details, see Supplementary Method 1.

General procedure for the synthesis of Ag(I) complexes

An oven-dried vial equipped with a stir bar was charged with N-Aryl thiazol-2-ylidenes carbene precursors (1.0 equiv), Ag2O (typically, 0.5 equiv) and NaCl (typically, 2.0 equiv). The reaction mixture was placed under a positive pressure of argon and subjected to three evacuation/backfilling cycles under high vacuum. DCM (typically, 0.04 M) was added and the reaction mixture was stirred away from light overnight at room temperature. The reaction mixture was filtered through Celite with DCM as eluent and concentrated under reduced pressure, and dried under high vacuum to afford silver(I) complex. For more details, see Supplementary Method 3.

General procedure for the synthesis of propargylic amides

An oven-dried 100 ml round-bottomed flask equipped with a stir bar was charged with propargylic amine (1.0 equiv), Et3N (typically,1.0 equiv), DMAP (typically,0.02 equiv) and DCM (typically, 0.25 M). The resulting mixture was cooled to 0 °C, and the acid chloride (1.0 equiv) was added. The mixture was stirred for 30 min at 0 °C and 3–12 h at room temperature. H2O (typically, 0.33 M) was added, and the aqueous layer extracted with DCM. The combined organic extracts were washed with saturated NaHCO3, H2O and brine, dried over Na2SO4 and concentrated in vacuo. The crude products were purified by column chromatography on silica gel (EtOAc/hexanes). For more details, see Supplementary Method 7.

General procedure for the cyclisation of propargylic amides

An oven-dried vial equipped with a stir bar was charged with propargylic amides (1.0 equiv), Ag catalyst 4c (typically, 1.0 mol%). The reaction mixture was placed under a positive pressure of argon and subjected to three evacuation/backfilling cycles under high vacuum. Then AcOH (typically, 1.0 equiv) and DCM (typically, 1.0 M) was added and the reaction mixture was stirred at room temperature for 8 h. The volatiles were removed in vacuo and the products were purified by column chromatography on silica gel (EtOAc/hexanes). For more details, see Supplementary Method 9.

Compound characterization

See supplementary note 2 for NMR spectra.