1,2,3-Triazolium-Derived Mesoionic Carbene Ligands Bearing Chiral Sulfur-Based Moieties: Synthesis, Catalytic Properties, and Their Role in Chirality Transfer.

1,2,3-Triazole-derived mesoionic carbenes (MICs) having a chiral sulfur functional group at the C5 position are easily available through a CuAAC between chiral alkynyl sulfoxides and different azides. The MICs form complexes with several metals (Au, Ag, Ir, Rh, and Ru) that are enantiomerically pure. Moreover, enantiomerically pure MIC sulfinilimines are obtained from the corresponding sulfoxide retaining the chirality. Through this article, the participation of sulfoxide moieties in different catalytic and chirality transfer processes, as well as in discovering mechanistically new processes will be shown. The role of the sulfur chiral moiety in catalytic cycloisomerization and cycloisomerization-dimerization processes using Au-MIC catalysts is dual. The sulfur functional group either stabilizes intermediates in the catalytic cycle, allowing for the reaction to occur or significantly increases the selectivity of the cyclization processes. 1,2,3-Triazole MICs having chiral sulfoxides at C5 are extremely efficient in preparing chiral at the metal complexes by C-H insertion processes. The chiral at the metal half-sandwich complexes, having the enantiopure sulfur chiral group unaltered, experiences different reactions with complete retention of the configuration. Finally, mechanistically new processes, like the desulfinilation of 1,2,3-triazolium salts in Ag-MIC complexes have been uncovered. These still-nascent classes of compounds will offer opportunities for the discovery of novel catalytic applications and to study new mechanistically sound processes.

Introduction

The preparation of stable carbenes I in 1988 by Bertrand[1] and the subsequent report by Arduengo describing the synthesis of the first stable N-heterocyclic carbenes (NHCs) II (Figure )[2] in 1991 opened up an entirely new research field. Stable carbenes[3] have found extensive application in fields as diverse as organocatalysis,[4] medicinal,[5] or materials chemistry.[6] However, it is in their use as ligands for transition metals that these species are important.[7] In fact, their availability and the possibilities of modification of the structure including steric electronic properties of the carbene ligands allow the modulation of the electronic properties of the metal center. This fact is essential in the efficient use of transition metal complexes having carbene ligands in catalysis.

Figure 1

Three examples of stable carbenes.

In this regard, Albretch[8] introduced 1,2,3-triazolilylidene carbenes III as a new type of mesoionic carbene (MIC) in 2008. Since precursor heterocycles for these MICs are prepared through the efficient and regioselective Cu(I)-catalyzed Huisgen cycloaddition (CuAAC) between a terminal alkyne and an azide,[9] these ligands can be prepared almost “a la carte” to fulfill specific electronic and steric properties (Scheme ). The CuAAC allows for the placement in the 1,2,3-triazole ring of a large number of functional groups that can be introduced in the core of the heterocycle through the terminal alkyne, the azide, or both. The MIC ligand is coordinated to different transition metals using several procedures. The most frequent method uses the coordination to silver to form the corresponding Ag–MIC followed by transmetalation. Therefore, the last 10 years have witnessed an explosive outgrowth in the preparation of new structural types of MICs and their use as ligands in transition-metal-mediated catalysis.[10]

Scheme 1

Synthesis of Metal-Stabilized 1,2,3-Triazolilyliden Carbenes

In spite of the fact that sulfoxides are the alternative choice when considering the scarce chiral functional groups based on elements different from the chiral-carbon groups,[11] its incorporation in a 1,2,3-triazolylidene MIC following a CuAAC reaction was unknown at the beginning of our work in this field. The precursors of these heterocycles, namely the enantiopure ethynyl sulfoxides are easy to make through the Andersen method, which involves the reaction between acetylenic Grignard reagents and enantiopure menthyl sulfinate (Scheme ).[12] Nevertheless, placing an enantiomerically pure sulfoxide moiety in a 1,2,3-triazolylidene carbene requires several problems to be addressed. The main concern was the configurational stability of the sulfoxide center in the conditions required to synthetize the heterocycle. Other problems may arise from the compatibility of the sulfoxide moiety with the strong alkylating agents required to obtain the 1,2,3-triazolylidene precursor,[13] and finally the coordinating nature of the sulfoxide moiety may compete with the emerging carbene during the coordination of such ligands to the metal.

Scheme 2

Andersen’s Synthesis of Enantiopure Ethynyl Sulfoxides

This perspective will present the solution of these problems, the extension of this approach to triazole containing sulfoximines, the scarce examples of triazolilylidene carbenes containing sulfur-based chiral (enantiopure or racemic) functional groups, and their emerging applications in catalysis.

Synthesis

Hybrid ligands composed of classic NHCs and sulphur-based functional groups have been repeatedly investigated.[14] Placing hemilabile functional groups based on sulphur into a 1,2,3-triazolilylidene heterocyclic system is straight due to the versatility of the CuAAC process, where the sulphur functional group may be introduced either through the azide, the alkyne, or both. Following this approach, several 1,2,3-triazolilylidenes having aliphatic thioether moieties at the C4 of the triazole core were prepared.[15] Metalation (Pd, Rh, Au) of these compounds showed no coordination of the pendant thioether to the metal center. The synthesis of thioether functionalized ligands 1-[2-(methylthio)phenyl]-4-phenyl-1H-1,2,3-triazole and 3-methyl-1-[2-(methylthio)phenyl]-4-phenyl-1H-1,2,3-triazol-5-ylidene) has been also reported.[16]

Sarkar reported in 2014 the first heteroleptic ligand formed by an aromatic thioether and one 1,2,3-triazolilylidene MIC.[17] Thus, Ir complex 1, and Ru and Os-complexes 2 were prepared from the corresponding 1,2,3-triazolium salts using the sequence Ag-complexation–transmetalation on the C5-position of the heterocycle, with the concomitant coordination of the arylthioether group. This route provided the desired complexes in high yields (Scheme ).

Scheme 3

Synthesis of Complexes Having 1,2,3-Triazolilydenes with Aromatic Thioether MICs as Ligands

Contrary to sulphide-containing NHCs,[14] the NHCs having sulfoxide functional groups have been much less studied. The first example of a metal–NHC complex having a sulfoxide was the CSC-pincer compound 3 reported by Huyhn.[18] This NHC-pincer ligand has the sulfoxide group as a bridge of two dibenzimidazolium salts. The synthetic approach to ligand 4 involved the oxidation of the sulphide group of compound 5 using H2O2. Subsequent reaction with Pd(OAc)2 in dimethyl sulfoxide at 80 °C afforded the pincer complex 3. It should be noted that the sulphoxide group of 3 did not coordinate the Pd-center (X-ray) (Scheme ).

Scheme 4

Synthesis of Pseudo Pincer 3

Cardenas reported[19] the synthesis of Pd-imidazolilylidenes having a sulfoxide moiety in a chain attached to N3. Thus, alkylation of imidazoles 6 with racemic bromosulphoxide 7 under microwave irradiation conditions (MW) formed imidazolium salts 8 that were coordinated to Pd using Ag2O in MeCN to form complexes 9 (Scheme ). The structure of complexes [9aNCMe]+ and (9d)2 were unambiguously determined by X-ray diffraction. The Pd complex [9aNCMe]+ has a nondistorted square-planar geometry with a distance between Pd and S atoms of 3.34(2) Å, which is lower than the sum of van der Waals radii, suggesting a weak interaction between both centers, while the sulfoxide moiety of the dimeric complex (9d)2 shows no interaction between the sulfoxide group and the metal center.

Scheme 5

Synthesis of Palladium Sulfinyl Imidazolilydenes

It should be noted that in the examples above, the sulfoxide moiety is racemic. At the beginning of our work, procedures to prepare enantiomerically pure 1,2,3-triazolilylidene ligands having chiral sulfoxide moieties were unknown. We devised a methodology suitable to access different enantiomerically pure 1,2,3-triazolilyliden-ligands having in mind the possibility of introducing diversity[20] in the prepared molecules. Within this scheme, considering that the reagents have to be easily available and that the reaction to be used to build the heterocyclic ring will be the CuAAC between one alkynylsulfoxide and one azide, the following structural variants in the reagents can be taken into account (Scheme ):

Scheme 6

Structural Diversity in the Synthesis of Enantiopure Sulfinyl 1,2,3-Triazolilydenes

The nature of the ethynyl sulfoxide (a).

The azide (b).

The nuclearity (c).

Post-functionalization of the sulfoxide (ca transformation in sulfoximines) (d).

Change on the catalytic system to effect the AAC (translated into the regiochemistry of the AAC) (e).

Routes (a) and (b) shown in Scheme were tested by synthetizing enantiopure 1,2,3-triazole-containing sulfoxide moieties by the reaction of alkynylsulfoxides 10 and azides 11. Examples shown in Scheme demonstrate that the synthesis of triazoles 12 is compatible with alkyl and aryl azides as well as with differently substituted sulfoxides. Especially relevant in these experiments was the preparation of triazoles containing an additional chiral center. One single enantiomer was obtained in both cases confirming that the stereochemistry of both centers was maintained during the cycloaddition process.[21] Methylation of triazoles 12 occurred uneventfully in the presence of Me3OBF4, and the gold complexes 13 were prepared through the corresponding silver carbene complexes as intermediates (Ag2O, NMe4Cl) and subsequent transmetalation with [AuCl(Me2S)]. Yields were high in most cases. Complexes 13 were characterized by X-ray diffraction. Interestingly, complex 13 (R = Me, R1 = (−)-α-phenylethyl) shows, in the solid state, a gold(I)–gold(I) distance of 3.223 Å that reflects unsupported strong aurophilic interactions. The two C–Au–Cl units involved in the interaction show a crossed arrangement with a C–Au–Au–C torsion angle of 66.5°.

Scheme 7

Synthesis of Gold Sulfinyl 1,2,3-Triazolilydenes

The subsequent transformation of triazoles 12 into sulfoximines 15 (route d in Scheme ) was affected by the chemoselective oxidation with m-CPBA to form the corresponding sulfonyl triazole 14 and subsequent imination[22] with PhI=NTs/Cu(OTf)2 to form the racemic sulfoximine 15 in an excellent yield (a in Scheme ).[23] Enantiopure sulfoximines 15 were prepared avoiding the oxidation step by the reaction of triazole sulfoxides 12 and PhI=NTs/Cu(OTf)2. Subsequent methylation yielded the enantiopure salts 16 (b in Scheme ).[24]

Scheme 8

Synthesis of 1,2,3-Sulfonyl- and 1,2,3-Sulfoximidoyl-Substituted Triazoles

An alternate route to 1,2,3-sulfoximidoyl-substituted triazoles was reported by Bolm using a Huisgen 1,3-dipolar cycloaddition between organoazides and sulfoximidoyl alkynes 17 (Scheme ).[25] In these cases, the sulfoximidoyl alkynes were prepared by copper-catalyzed imination of the corresponding sulfoxides with PhI=NTs. The cycloaddition reaction was regioselective favoring the depicted product, although variable amounts of the regioisomeric triazoles were formed. Compounds 18 were not used in metal complexation. A related procedure to prepare fully substituted 1,2,3-triazolyl-5-sulfoximines by Cu(I)-catalyzed three component coupling of sulfoximines, alkynes, and azides has been reported recently.[26]

Scheme 9

Bolm’s Synthesis of 1,2,3-Sulfoximidoyl-Substituted Triazoles

Catalysis

The metal complexes derived from 1,2,3-MICs containing sulfoxide or sufoximidoyl moieties have been used in different catalytic processes. Thus complexes 3 (Scheme ), 19, and 20 were good catalysts for the Heck reaction (Scheme ).[18] Essentially, the reactions of bromoaryl derivatives 21 and t-Bu-acrylate gave the expected products with the above catalysts. Double adducts 22 were obtained when 2,6-dibromopiridine was used. What is interesting is that both the coordinated cationic sulfide 20 and the uncoordinated sulfide 19 behave identically to the uncoordinated sulfoxide catalyst 3, which indicates that the sulfur-derived moiety is not participating in the catalytic process (Scheme ).

Scheme 10

Sulfur-Containing Palladium Benzimidazolilylidenes as Catalysts in the Heck Reaction

Complexes 9a–d were used as catalysts in the acetoxylation reaction of toluene-promoted (diacetoxyiodo)benzene as the oxidant in AcOH/H2O mixtures. The yields in acetoxylated toluene derivatives ranged from acceptable to good (Scheme ).[19]

Scheme 11

Sulfinyl Palladium-Imidazolilylidenes as Catalysts

Enantiomerically pure Au–MIC 13 were tested in the cycloisomerization of 1,6-enynes 23. While these complexes were very efficient in the formation of cyclic products 24 and 25, it soon became evident that sulfoxide moieties were required for the cycloisomerization to occur. Thus, Au–MICs 27 lacking the sulfur functional group did not promote the cycloisomerization of enyne 23a, while the analogous product 13a having the sulfoxide moiety instead of the benzyl group formed exclusively the cyclized product 24 in essentially quantitative yields. Moreover, while the sulfoxide moiety is essential for the activity of the catalyst, the bulkiness of the substituent at nitrogen determines the 5-exo versus 6-endo ratio of cyclized products (Scheme ). The bulkier the substituent at nitrogen, the higher is the selectivity favoring the 5-exo cyclization product. For example, catalyst 13b formed a 1:1.4 mixture of compounds 24 and 25 while the analogous reaction with catalyst 13a formed exclusively compound 24 in essentially quantitative yields.

Scheme 12

Enantiopure Gold–MICs as Catalysts in 1,6-Enyne Cycloisomerizations

The proposed catalytic cycle[21] to explain these observations involve an initial formation of gold carbene complexes A and B by reaction of the terminal alkyne and the catalyst (internal alkynes do not react under these conditions). This is the standard mechanism for Au-promoted carbocyclizations,[27] and at this stage the sulfoxide moiety should be a spectator. The 5-exo versus 6-endo cyclization selectivity should be related to the bulkiness of the carbon linked to N1. Since intermediate C is much more crowded than D, the bulkier α-phenylethyl moiety of complex 13a should favor the formation of A leading exclusively to the 5-exo product 24 (Scheme ). On the other hand, the benzyl moiety of complex 13b is less sterically demanding, leading to mixtures of two regioisomers. This mechanistic hypothesis is congruent with the results obtained and gives the apparent ancillary moiety (the sulfoxide group) a key role unprecedented in the literature. In fact, the lack of catalytic activity observed for 27 (Scheme ) indicated a clear involvement of the sulfoxide oxygen to stabilize Au–carbene intermediates, very probably by the interaction of either the sulfur lone pair or oxygen with carbene carbon or the gold center.[28] This interaction stabilizes intermediates A and B and allows the reaction to proceed (Scheme ).

Scheme 13

Mechanistic Hypothesis for the Au-Catalyzed Carbocyclization of Enynes Promoted by Catalysts 13

These results show that the sulfoxide group of the enantiopure MICs 13 has a role in the catalytic cycles in contrast with the results reported by other authors. In fact, no differences in the catalytic activity were reported for isostructural catalysts having a sulfide, a coordinated sulfide, or non-coordinated sulfoxide functional groups.[18]

The catalytic activity of Au–MICs 13 was also tested in the cycloisomerization of enynes tethered to an aromatic ring.[29] The reaction of enynes 28 with catalyst 13a in the presence of NaBArF yielded mixtures of compounds 29 and 30 differing in the position of one double bond. These compounds are formed through a cyclization dimerization cascade in excellent yields. The ratio 29:30 is temperature-dependent. Compounds 29 are the thermodynamic products and are formed at rt, while compounds 30 are the kinetic products and are formed at low temperatures. Compound 30 converts to the thermodynamic 29 on heating at rt in the presence of catalyst 13a. Again, the use of a simple Au-catalyst like Ph3PAuCl/AgSbF6 formed a different product, namely naphthalene 31 (Scheme ). Clearly, the role of the sulfoxide group in determining the nature of the reaction products is decisive. In all cases the obtained products were racemic. Therefore, albeit the sulfoxide group determines the effectiveness of the reaction, it is not able to exercise any enantio-discrimination. It should be noted that the use of stoichiometric amounts of NaBArF in the cycloisomerization–dimerization cascade of enyne 28a formed quantitatively product 29. The mechanism of these reactions should be markedly different from the Au-promoted reactions.

Scheme 14

Catalyzed Cycloisomerization of Aromatic Enynes

The mechanism to explain these findings is depicted in Scheme . The initial reaction of the diene system of 32 with Au(I)-catalysis should form the cyclized intermediate 33 through transition state 34. Deauration of 33 would yield 30 that may form the allylic isomerization product 29 depending on the reaction conditions. It should be noted (see below) that the participation of the sulfoxide moiety in the stabilization of intermediates 34 is decisive to determine the nature and the yield of the final products (Scheme ).

Scheme 15

Mechanistic Hypothesis for the Catalytic Cyclization of Diene 32

The effect of the sulfoxide group on the MICs in determining the efficiency and selectivity of the cyclization processes is also seen in the reaction of enynes 28a,b, and benzaldehyde. In fact, catalysts 13a in the presence of NaBArF formed exclusively compounds 35 as single isomers in excellent yields, while Ph3PAuCl yielded mixtures of diasteromers 35 and 36 in variable low yields (Scheme ).

Scheme 16

Influence of the Sulfoxide Group of the MICs in the Selectivity of the Cycloisomerization of Aromatic Tethered Dienes

These differences in selectivity and reactivity were explained by the formation of cationic species 37 through intermediate 38. This intermediate is trapped by the aldehyde to form Au-species 39 that yielded the final products 35a and 36a upon cationic cyclization and deauration. It is clear that the role of the sulfinyl moiety in the stabilization of intermediate 37 is to produce the syn-isomer 35a. In the absence of this stabilizing factor, both the yields and stereochemistry of the reaction fall dramatically (Scheme ). This hypothesis requires that the cyclization of 28a and benzaldehyde through chair-like transition state 39 has to be reversible.[30] Thus, stabilization of the reactive intermediates by the sulfoxide group accounts for the catalytic activity of MIC–Au catalysts 13 against dienes 28 and 32.

Scheme 17

Mechanistic Hypothesis for the Cyclization of Enyne 28a and Benzaldehyde

Chirality Transfer in Organometallic Complexes

The possibility of affecting the transfer of the chirality residing in the sulfoxide chiral moiety of the MIC ligand to a metal center is especially attractive. Methods to prepare chiral at metal half-sandwich complexes have been extensively developed,[31] including the sequence coordination of the metal to a chiral triazolinylidene and imidazolinylidene carbenes and subsequent diastereoselective C–H insertion,[32] as well as the use of C2-chiral NHCs[33] and sugar-derived NHCs[34] in analogous processes. By contrast, the use of chiral sulfoxide ligands or reagents to achieve the synthesis of chiral at the metal complexes has been scarcely studied.

Optically active octahedral Ru-complexes were obtained by reaction of cis- or trans-Ru(bpy)2Cl, with either (R)-(+)- or (S)-(−)-methyl p-tolyl sulfoxide.[35] An analogous procedure was used to prepare Ru–bis(diimine) sulfoxide complexes,[36] and Ru–trispyridine complexes.[37] The preparation of achiral half-sandwich complexes has been reported for Rh(III)-complexes having an achiral sulfoxide bound to the metal within a naphthylindenyl moiety, which confers planar chirality to the complex.[38] Subsequently, these complexes were resolved and used to prepare the corresponding chiral at the metal complexes by replacing the chloride ligands by a phenylpyridine ligand (Scheme ).[39]

Scheme 18

Sulfinyl Chiral at the Metal Half-Sandwich Rh(III) complex

MIC ligands containing chiral sulfoxide moieties were used to develop a general method to prepare chiral at the metal half-sandwich complexes (M = Ir, Rh).[24] Triazolium salts 41 were reacted with [MCl2Cp*]2 (M = Ir, Rh) to form the corresponding dichloro derivatives 42 using Ag2O in the presence of NH4Cl and 4 Å molecular sieves. Treatment of complexes 42 with NaOAc formed the corresponding cyclometalated chiral at metal complexes 43 in good yields and as single enantiomers (Scheme ). It should be noted that the sense of the asymmetric induction is independent on the bulkiness of the sulfoxide substituent and on the nature of the metal. Moreover, the use of sulfoximine salts 16 as the elements of asymmetry in the starting MIC precursors also produces excellent levels of asymmetric induction in the formation of complexes analogous to 43 while maintaining good chemical yields.

Scheme 19

Synthesis of Enantiopure Chiral at the Metal Half-Sandwich Complexes

Enantiopure complexes 43 (M = Ir) can be transformed into cationic complexes 44 by reaction with NaPF6 maintaining the configurational integrity of the metal center (Scheme ).

Scheme 20

Cationic Enantiopure Chiral at the Metal Half-Sandwich Complexes

Additionally, the insertion of symmetrical alkynes into the complexes 43 formed the corresponding alkyne-insertion products 45 in excellent yields while maintaining the integrity of the configuration of the metal center. As expected, the insertion of unsymmetrical alkynes (methyl propiolate) produced the mixture of regioisomeric inserted alkynes 46a and 46b, with both regioisomers being enantiomerically pure compounds. The configuration at the metal also remains unaltered in this case (Scheme ).

Scheme 21

Insertion of Alkynes Into Complexes 43

The formation of six-membered metallacycles by C–H activation occurred with a considerable loss of stereoselectivity. Thus, complexes 47 proportionate the corresponding six-membered metallacycles 48 in an excellent chemical yield but with poor estereoselectivity (3:2) (Scheme ). This is a consequence of the intermediacy of a seven-membered transition state,[40] which places the sulfoxide away from the reactive center. The putative influence of the sulfoxide in the stabilization of reaction intermediates looks clear.

Scheme 22

Synthesis of Six-Membered Metallacycles

The stereochemistry of these reactions was determined by a combined circular dichroism (CD)-X-ray study. Circular dichroism has been scarcely used to establish the absolute configuration at the metal center of half-sandwich metal carbene complexes.[34,41] For example, the CD spectra of complexes 43 show one main negative absorption centered around 250 nm. X-ray analysis of complex 43 (R1 = H, R2 = Br) established an S absolute configuration at the Ir(III) chiral center. Therefore, the negative Cotton effect around 250 nm was correlated to the S absolute configuration around the metal center. These results were translated to cationic complexes 44 which showed a negative Cotton effect also around 250 nm. This pointed to a complete retention of the configuration at the metal center upon the formation of the cationic complex by displacement of a chlorine ligand by MeCN. This is due to the participation of the sulfoxide moiety in the SN1-like mechanism leading to the cationic complexes (Figure ).[42]

Figure 2

CD-spectra of complexes 43–45.

The methodology CD-X-ray diffraction to determine the configurational fate of the metal center during alkyne insertion was also effective. Thus, metallacycle 45 has two weak negative absorptions centered at 232.8 and 204.4 nm, and X-ray diffraction analysis confirmed that the metal center retains its configuration. Again, the retention of the configuration at the metal center can be traced to the participation of the sulfoxide moiety during the dissociative process prior to the coordination of the alkyne (Figure ).[43]

Clearly, the results above show the crucial participation of the sulfoxide group in the chirality transfer processes, both during the formation of the enantiopure metal center and during the transformation of these enantiopure centers. The combination of CD-X-ray is a powerful methodology to study these processes in the metal center.

Desulfinylation Reactions: Novel Reaction Pathways

1,2,3-Triazole MICs having sulfoxide moieties are also interesting substrates to study new processes. The lability of some groups attached to the 1,2,3-triazole MICs has been observed by Albrecht.[44] Thus, during the formation of Ru-complexes from 4-methoxycarbonyl-1,2,3-triazolium salts, partial decarboxylation of the ester group in the precursor triazolium salt was detected, which allowed for the preparation of C4-unsubstituted Ru-complexes. Mechanistic studies were not pursued. Subsequent work by these authors[45] resulted in a method to prepare C4-unsubstituted Ru(II) and Au(I) complexes by decarboxylation of the corresponding 4-carboxy-MICs during the coordination process (Scheme ).

Scheme 23

Decarboxylation of a 1,2,3-Triazolium Salt during a Metalation Process

The preparation of silver complexes 49 having two C4-sulfoxide substituted 1,2,3-triazole MICs ligands occurred uneventfully in excellent yields using the standard Ag2O/Me4NCl conditions for the metalation of the C5 position of the heterocyclic ring. These complexes were isolated and characterized. However, when complexes 49 were subjected to reaction with MeOH, removal of the sulfinyl group was observed with the concomitant formation of three new Ag(I) complexes 50, having the C4 position unsubstituted.[46] Regioisomeric Au(I) complexes 51/52 were obtained from the mixture of the Ag(I) complexes (Scheme ). X-ray diffraction of complexes 51d and 52d secured the regioisomeric nature of complexes 50. Additionally, the reaction occurs with other primary and secondary but not tertiary alcohols.

Scheme 24

Silver MIC Carbenes and Their Desulfinilation Reactions

To determine whether the desulfinilation occurred in the silver complexes 49 or in the free triazolium salts formed by dissociation of these silver complexes, the free salt 53 was submitted to treatment with MeOH forming the unsubstituted triazolium salt 54 together with methyl sulfinate (Scheme ). The existence of equilibrium between the Ag–MIC 49, the free carbene species, and the corresponding Ag–monocarbenes was demonstrated by the crystallization of silver monocarbene 56 from a solution of 55. Moreover, the structure of this silver monocarbene was resolved by X-ray diffraction.

Scheme 25

Desulfinylation of Salt 53 and Formation of Ag(I)-Complex 56

The isolation of Ag–monocarbene 56 coupled to the desulfinylation of 53 is coherent with the desulfinylation reaction occurring through the free carbene A, analogous to the one derived from 53 (Figure ). Extensive deuteration experiments and density functional theory (DFT) calculations allow the proposal of a mechanism for this unprecedented desulfinylation reaction. MeOH was modeled as one MeOH solvated by another two explicit MeOH molecules.[47] The structure of TS-B is depicted in Figure . In this TS, the MeO– having two MeOH molecules experiences the nucleophilic attack to the sulfur in a SN2 type reaction. The participation of two molecules of MeOH proportionates the lower energy TS computed for this rate determining step.

Figure 3

Computed reaction pathway for the desulfinilation reaction in C4-sulfinyl-1,2,3-triazolium salts. All values correspond to Gibbs energies in DCM (dichloromethane) (kcal mol–1). DFT calculations were carried out using the M06 density functional,[48] with an ultrafine grid as implemented in Gaussian 09.[49] Gaussian 09 performs well for main-group chemistry and noncovalent interactions.[50] All intermediates and transition states were fully optimized in DCM solution (ε = 8.93) using the continuum method SMD.[51] The 6-31G** basis set (BS1) was used. Final single-point calculations were performed with the 6-311++G** basis set (BS2).[52] Transition states were identified by having one imaginary frequency in the Hessian matrix. It was confirmed that transition states connect with the corresponding intermediates by means of application of the eigenvector of the imaginary frequency and subsequent optimization of the resulting structures.

The proposed reaction mechanism is shown in Figure and entails three main steps: (i) protonation of the free carbene, (ii) desulfinylation via nucleophilic attack, and (iii) carbene isomerization. First, the free carbene A (zero of energies) is protonated by methanol via TS-A (4.5 kcal mol–1) to form the triazolium cation B and [MeO–(MeOH)2]. The solvated methoxide performs a nucleophilic attack on the sulfur center via TS-B (14.3 kcal mol–1), thus breaking the S–C4 bond. Release of the sulfinate produces carbene C, which lacks substituents at the ring carbon atoms. Carbene C can then isomerize to E via the protonated species D1 and D2 involving TS-C (4.1 kcal mol–1) and TS-D (1.9 kcal mol–1). D1 and D2 are computationally less favored than C and E due to entropic effects and poor stabilization of charged species in the low polar DCM solvent. Overall, the largest activation energy is 14.3 kcal mol–1 (from A to TS-B), which corresponds to the nucleophilic attack step.

This mechanism is consistent with the experimental data. The low activation energy agrees with the mild experimental conditions, and the rate-determining nucleophilic attack explains the slower reaction rates observed with bulkier alcohols. In the presence of strong electrophiles such as [Ag+] and [Au+], the carbenes C and E would readily react with Au(I) or Ag(I) giving unsubstituted MIC-complexes. The exoergic desulfinylation from B to D2 (ΔΔG = −2.8 kcal mol–1) is also in line with the no participation of Ag in the process. Moreover, the deuteration experiments carried out for these processes show deuterium incorporation in the C3 and C4 positions of the triazole ring, which is consistent with the facile isomerization through proton transfer steps in the species D1 and D2.

Outlook

The results presented above show the potential of MICs derived from 1,2,3-triazoles having a chiral sulfur functional group at the C5 position. While for some applications like anion recognition the effect of the sulfur groups is to decrease the electronic density of the triazole ring,[23] the participation of these groups in determining the nature and stereochemistry of the final product is decisive. In catalytic cycloisomerization processes and cycloisomerization–dimerization processes, the role of the sulfoxide moiety is to stabilize intermediates in the catalytic cycle, allowing for the reaction to occur or significantly increasing the selectivity of the cyclization processes. Meanwhile, during chirality transfer processes to the newly formed chiral metal center, the sulfur functional group determines the chirality of this center which is formed in a completely diastereoselective fashion. Moreover, the chiral half-sandwich complexes containing the sulfur chiral group unaltered, experience different reactions with complete retention of the configuration. Finally, mechanistically new processes, like the desulfinilization of 1,2,3-triazolium salts have been uncovered. Clearly, these classes of compounds also offer new opportunities for the discovery and study of new mechanistically relevant reactions. Overall, the sulfur chiral moieties play a pivotal role in these processes. In addition, hypothetically the reversible conversion of the S=O functional group into its reduced and oxidized forms may be possible, which opens up substantial opportunities both in catalysis and in the building of new structures.[53]

The reactions in Scheme show that much remains to be done in this field, including the multinuclear systems, and the synthesis and application of the regioisomeric 1,2,3-triazole MICs having the sulfur chiral group in C4. The catalytic opportunities of these new chiral metal complexes that include the chiral at the metal complexes have been just devised. Again, these new catalysts offer opportunities for asymmetric synthesis to be disclosed during the next few years. Finally, the disentangling of the role of functional groups having more than one mode of coordination to the metal opens up new opportunities for mechanistic studies (both experimental and computational).