Thermal Rearrangement of Thiocarbonyl-Stabilised Triphenylphosphonium Ylides Leading to (Z)-1-Diphenylphosphino-2-(phenylsulfenyl)alkenes and Their Coordination Chemistry

While thiocarbonyl-stabilised phosphonium ylides generally react upon flash vacuum pyrolysis by the extrusion of Ph3PS to give alkynes in an analogous way to their carbonyl-stabilised analogues, two examples with a hydrogen atom on the ylidic carbon are found to undergo a quite different process. The net transfer of a phenyl group from P to S gives (Z)-configured 1-diphenylphosphino-2-(phenylsulfenyl)alkenes in a novel isomerisation process via intermediate λ5-1,2-thiaphosphetes. These prove to be versatile hemilabile ligands with a total of seven complexes prepared involving five different transition metals. Four of these are characterised by X-ray diffraction with two involving the bidentate ligand forming a five-membered ring metallacycle and two with the ligand coordinating to the metal only through phosphorus.


Introduction
The thermal extrusion of Ph 3 PO from carbonyl-stabilised triphenylphosphonium ylides 1 is a well-established synthetic route to functionalised alkynes 2 (Scheme 1) [1][2][3].The process proceeds particularly well using flash vacuum pyrolysis (FVP), and we have found that the phosphorus to carbonyl coupling constant 2 J P-CO provides a diagnostic parameter for the likely success of the reaction, with ylides for which 2 J P-CO < 11 Hz usually providing the alkynes in high yield [4].By way of contrast, the thermal behaviour of the corresponding thiocarbonyl-stabilised ylides 3 has only been examined in a few cases and Bestmann and Schaper found that heating the ylides 3 above their melting point resulted in a bimolecular process with the loss of Ph 3 P and Ph 3 PS to give tetrasubstituted thiophenes 4 [5].Some time ago, we described a preliminary study in which FVP of the ylides 5 was found to proceed as expected by analogy with the carbonyl analogues 1 with the loss of Ph 3 PS to give alkynes 2 for R 1 ̸ = H, but when R 1 was hydrogen, a quite different process was observed: rearrangement with the transfer of a phenyl group from P to S giving the potentially useful bidentate proligands 6 [6].In this paper, we describe in more detail the synthesis and structure of these novel phosphinovinyl sulfides as well as their coordination chemistry.Scheme 1. Previously reported thermal reactivity of carbonyl-and thiocarbonyl-stabilised phosphonium ylides.

Results and Discussion
Synthetic access to thiocarbonyl-stabilised ylides 5 is available using various methods including the treatment of non-stabilised ylides with dithioesters or dithiocarbonates [7], or activation of the corresponding carbonyl-stabilised ylides with triflic anhydride [8] or POCl3 [9] followed by treatment with sodium sulfide.For the current study, we used the direct reaction of carbonyl-stabilised ylides 1 with Lawesson's reagent introduced by Capuano and coworkers [10], and in this way, we prepared the five examples 7-11 (Figure 1) from their carbonyl analogues.Compounds 7-9 are already known [8] andgave analytical and spectroscopic data in agreement with the reported values.The new compounds 10 and 11 were fully characterised and showed distinctive NMR signals confirming the presence of P=CH-C=S [10 δP +5.0; δH 5.22 (d, 2 J 34 Hz, P=CH); δC 81.3 (d, 1 J 118 Hz, P=CH), 214.4 (d, 2 J 4 Hz, C=S).11 δP +8.1; δH 5.18 (d, 2 J 32 Hz, P=CH); δC 84.1 (d, 1 J 113 Hz, P=CH), 200.5 (d, 2 J 4 Hz, C=S)].In fact, the phosphorus coupling extended throughout the structures with all carbon signals except the CH3 of t-butyl observed as doublets in the 13 C NMR spectra.This is discussed further below in the context of a comparison of the coupling pattern before and after the thermal rearrangement.The structure of 10 was also confirmed by X-ray diffraction (CSD RefCode: AJOMUI) and this was described in our earlier communication [6].We might also note at this point that compound 10 has a particularly pungent and unpleasant smell.Although it is not very volatile, it is extremely persistent and inadvertent contact with equipment or work surfaces contaminated with 10, even after several months, results in the release of its characteristic smell.We speculate that this may be due to slow hydrolysis of the P=C bond to give Ph3PO and release thiopinacolone, t-BuC(=S)Me.
When the thiocarbonyl ylides 7, 8 and 9 were subjected to FVP, there was complete reaction at a furnace temperature of 650 °C to give Ph3PS (δP +43) at the furnace exit and the expected alkynes 2, 2,2-dimethylpent-3-yne, 1-phenylpropyne and 1-phenyloctyne, in the cold trap.Thus, these compounds behave in a similar way to their carbonyl analogues but react more readily than the latter, which require a temperature of 750 °C for complete reaction [11].The higher reactivity of thiocarbonyl-as compared to carbonyl-stabilised ylides is a feature that we have already noted and quantified in a series of kinetic studies on the pyrolysis of carbamoyl and thiocarbamoyl ylides [12][13][14].
When the two ylides 10 and 11 were subjected to FVP, the reaction was also complete at 650 °C, but the process involved turned out to be completely different.In each case, only a single main product was obtained, which was isomeric with the starting material.In the case of 10, the product 12 was obtained in good yield and in a pure form as a Scheme 1. Previously reported thermal reactivity of carbonyl-and thiocarbonyl-stabilised phosphonium ylides.

Results and Discussion
Synthetic access to thiocarbonyl-stabilised ylides 5 is available using various methods including the treatment of non-stabilised ylides with dithioesters or dithiocarbonates [7], or activation of the corresponding carbonyl-stabilised ylides with triflic anhydride [8] or POCl 3 [9] followed by treatment with sodium sulfide.For the current study, we used the direct reaction of carbonyl-stabilised ylides 1 with Lawesson's reagent introduced by Capuano and coworkers [10], and in this way, we prepared the five examples 7-11 (Figure 1) from their carbonyl analogues.

Results and Discussion
Synthetic access to thiocarbonyl-stabilised ylides 5 is available using various methods including the treatment of non-stabilised ylides with dithioesters or dithiocarbonates [7], or activation of the corresponding carbonyl-stabilised ylides with triflic anhydride [8] or POCl3 [9] followed by treatment with sodium sulfide.For the current study, we used the direct reaction of carbonyl-stabilised ylides 1 with Lawesson's reagent introduced by Capuano and coworkers [10], and in this way, we prepared the five examples 7-11 (Figure 1) from their carbonyl analogues.In fact, the phosphorus coupling extended throughout the structures with all carbon signals except the CH3 of t-butyl observed as doublets in the 13 C NMR spectra.This is discussed further below in the context of a comparison of the coupling pattern before and after the thermal rearrangement.The structure of 10 was also confirmed by X-ray diffraction (CSD RefCode: AJOMUI) and this was described in our earlier communication [6].We might also note at this point that compound 10 has a particularly pungent and unpleasant smell.Although it is not very volatile, it is extremely persistent and inadvertent contact with equipment or work surfaces contaminated with 10, even after several months, results in the release of its characteristic smell.We speculate that this may be due to slow hydrolysis of the P=C bond to give Ph3PO and release thiopinacolone, t-BuC(=S)Me.
When the thiocarbonyl ylides 7, 8 and 9 were subjected to FVP, there was complete reaction at a furnace temperature of 650 °C to give Ph3PS (δP +43) at the furnace exit and the expected alkynes 2, 2,2-dimethylpent-3-yne, 1-phenylpropyne and 1-phenyloctyne, in the cold trap.Thus, these compounds behave in a similar way to their carbonyl analogues but react more readily than the latter, which require a temperature of 750 °C for complete reaction [11].The higher reactivity of thiocarbonyl-as compared to carbonyl-stabilised ylides is a feature that we have already noted and quantified in a series of kinetic studies on the pyrolysis of carbamoyl and thiocarbamoyl ylides [12][13][14].
When the two ylides 10 and 11 were subjected to FVP, the reaction was also complete at 650 °C, but the process involved turned out to be completely different.In each case, only a single main product was obtained, which was isomeric with the starting material.In the case of 10, the product 12 was obtained in good yield and in a pure form as a In fact, the phosphorus coupling extended throughout the structures with all carbon signals except the CH 3 of t-butyl observed as doublets in the 13 C NMR spectra.This is discussed further below in the context of a comparison of the coupling pattern before and after the thermal rearrangement.The structure of 10 was also confirmed by X-ray diffraction (CSD RefCode: AJOMUI) and this was described in our earlier communication [6].We might also note at this point that compound 10 has a particularly pungent and unpleasant smell.Although it is not very volatile, it is extremely persistent and inadvertent contact with equipment or work surfaces contaminated with 10, even after several months, results in the release of its characteristic smell.We speculate that this may be due to slow hydrolysis of the P=C bond to give Ph 3 PO and release thiopinacolone, t-BuC(=S)Me.
When the thiocarbonyl ylides 7, 8 and 9 were subjected to FVP, there was complete reaction at a furnace temperature of 650 • C to give Ph 3 PS (δ P +43) at the furnace exit and the expected alkynes 2, 2,2-dimethylpent-3-yne, 1-phenylpropyne and 1-phenyloctyne, in the cold trap.Thus, these compounds behave in a similar way to their carbonyl analogues but react more readily than the latter, which require a temperature of 750 • C for complete reaction [11].The higher reactivity of thiocarbonyl-as compared to carbonyl-stabilised ylides is a feature that we have already noted and quantified in a series of kinetic studies on the pyrolysis of carbamoyl and thiocarbamoyl ylides [12][13][14].
When the two ylides 10 and 11 were subjected to FVP, the reaction was also complete at 650 • C, but the process involved turned out to be completely different.In each case, only a single main product was obtained, which was isomeric with the starting material.In the case of 10, the product 12 was obtained in good yield and in a pure form as a crystalline solid after preparative TLC.This showed a 31 P NMR signal at −19.8 ppm, in the region expected for an alkyldiphenylphosphine, and the single P-CH= hydrogen gave a 1 H NMR singlet at 6.94 ppm (Scheme 2).
Molecules 2024, 29, x FOR PEER REVIEW 3 of 13 crystalline solid after preparative TLC.This showed a 31 P NMR signal at −19.8 ppm, in the region expected for an alkyldiphenylphosphine, and the single P-CH= hydrogen gave a 1 H NMR singlet at 6.94 ppm (Scheme 2).
The 13 C NMR spectrum, and particularly the pattern of P-C coupling, was particularly informative and showed major changes from the values in 10 (Figure 2).The values of JP-C for the thiocarbonyl ylides 10 and 11 are consistent with those well established for the corresponding carbonyl ylides [11] and carbamoyl/thiocarbamoyl ylides [12][13][14].However, in the rearranged product 12, the values around P-Ph are more similar to those in Ph3P, and the much higher coupling to P-C=C as compared to P-C=C as well as the absence of coupling to P-CH are surprising features.As already mentioned in our preliminary communication [6], the structure of 12 was confirmed by X-ray diffraction (CSD RefCode: AJOMOC).
When we examined the corresponding pyrolysis of the thioacetyl ylide 11, the corresponding reaction occurred at the same temperature, but the product was now a liquid formed in lower yield and containing some impurities including Ph3P, Ph3PO and Ph3PS.More interestingly, while it was predominantly (6.5:1) the (Z)-isomer 13 (δP −22.7), signals attributed to the (E)-isomer 14 (δP −25.4) were also apparent (Scheme 3).Attempts to purify this by repeated Kugelrohr distillation under reduced pressure led instead to isomerisation to give a 1:1 mixture of 13 and 14.At first sight, it might seem surprising that the product is obtained mainly as the less thermodynamically stable isomer from pyrolysis at 650 °C but then isomerises to the more stable isomer upon simple distillation at 90 °C, but this only serves to emphasise the mild nature of the flash vacuum pyrolysis technique.In addition, it has been shown that under FVP conditions, (Z)-alkenes do not normally isomerise to the (E)-isomer to any great extent at temperatures as low as 650 °C with the degree of conversion of (Z)-to (E)-stilbene, for example, being determined as 12%, in good agreement with our results [15].The presence of aromatic impurities even after distillation prevented full assignment of the 13 C NMR spectra for 13 and 14, but the key signals and the values of the phosphorus coupling constants (Figure 2) were in good agreement with those for 12 while also showing significant differences between the two isomers.The detailed form of the signals for P-CH and P-C=C-Me in the two isomers was Scheme 2. FVP of 10 to give 12.
The 13 C NMR spectrum, and particularly the pattern of P-C coupling, was particularly informative and showed major changes from the values in 10 (Figure 2).
Molecules 2024, 29, x FOR PEER REVIEW 3 of 13 crystalline solid after preparative TLC.This showed a 31 P NMR signal at −19.8 ppm, in the region expected for an alkyldiphenylphosphine, and the single P-CH= hydrogen gave a 1 H NMR singlet at 6.94 ppm (Scheme 2).
The 13 C NMR spectrum, and particularly the pattern of P-C coupling, was particularly informative and showed major changes from the values in 10 (Figure 2).The values of JP-C for the thiocarbonyl ylides 10 and 11 are consistent with those well established for the corresponding carbonyl ylides [11] and carbamoyl/thiocarbamoyl ylides [12][13][14].However, in the rearranged product 12, the values around P-Ph are more similar to those in Ph3P, and the much higher coupling to P-C=C as compared to P-C=C as well as the absence of coupling to P-CH are surprising features.As already mentioned in our preliminary communication [6], the structure of 12 was confirmed by X-ray diffraction (CSD RefCode: AJOMOC).
When we examined the corresponding pyrolysis of the thioacetyl ylide 11, the corresponding reaction occurred at the same temperature, but the product was now a liquid formed in lower yield and containing some impurities including Ph3P, Ph3PO and Ph3PS.More interestingly, while it was predominantly (6.5:1) the (Z)-isomer 13 (δP −22.7), signals attributed to the (E)-isomer 14 (δP −25.4) were also apparent (Scheme 3).Attempts to purify this by repeated Kugelrohr distillation under reduced pressure led instead to isomerisation to give a 1:1 mixture of 13 and 14.At first sight, it might seem surprising that the product is obtained mainly as the less thermodynamically stable isomer from pyrolysis at 650 °C but then isomerises to the more stable isomer upon simple distillation at 90 °C, but this only serves to emphasise the mild nature of the flash vacuum pyrolysis technique.In addition, it has been shown that under FVP conditions, (Z)-alkenes do not normally isomerise to the (E)-isomer to any great extent at temperatures as low as 650 °C with the degree of conversion of (Z)-to (E)-stilbene, for example, being determined as 12%, in good agreement with our results [15].The presence of aromatic impurities even after distillation prevented full assignment of the 13 C NMR spectra for 13 and 14, but the key signals and the values of the phosphorus coupling constants (Figure 2) were in good agreement with those for 12 while also showing significant differences between the two isomers.The detailed form of the signals for P-CH and P-C=C-Me in the two isomers was The values of J P-C for the thiocarbonyl ylides 10 and 11 are consistent with those well established for the corresponding carbonyl ylides [11] and carbamoyl/thiocarbamoyl ylides [12][13][14].However, in the rearranged product 12, the values around P-Ph are more similar to those in Ph 3 P, and the much higher coupling to P-C=C as compared to P-C=C as well as the absence of coupling to P-CH are surprising features.As already mentioned in our preliminary communication [6], the structure of 12 was confirmed by X-ray diffraction (CSD RefCode: AJOMOC).
When we examined the corresponding pyrolysis of the thioacetyl ylide 11, the corresponding reaction occurred at the same temperature, but the product was now a liquid formed in lower yield and containing some impurities including Ph 3 P, Ph 3 PO and Ph 3 PS.More interestingly, while it was predominantly (6.5:1) the (Z)-isomer 13 (δ P −22.7), signals attributed to the (E)-isomer 14 (δ P −25.4) were also apparent (Scheme 3).
Molecules 2024, 29, x FOR PEER REVIEW 3 of 13 crystalline solid after preparative TLC.This showed a 31 P NMR signal at −19.8 ppm, in the region expected for an alkyldiphenylphosphine, and the single P-CH= hydrogen gave a 1 H NMR singlet at 6.94 ppm (Scheme 2).
The 13 C NMR spectrum, and particularly the pattern of P-C coupling, was particularly informative and showed major changes from the values in 10 (Figure 2).The values of JP-C for the thiocarbonyl ylides 10 and 11 are consistent with those well established for the corresponding carbonyl ylides [11] and carbamoyl/thiocarbamoyl ylides [12][13][14].However, in the rearranged product 12, the values around P-Ph are more similar to those in Ph3P, and the much higher coupling to P-C=C as compared to P-C=C as well as the absence of coupling to P-CH are surprising features.As already mentioned in our preliminary communication [6], the structure of 12 was confirmed by X-ray diffraction (CSD RefCode: AJOMOC).
When we examined the corresponding pyrolysis of the thioacetyl ylide 11, the corresponding reaction occurred at the same temperature, but the product was now a liquid formed in lower yield and containing some impurities including Ph3P, Ph3PO and Ph3PS.More interestingly, while it was predominantly (6.5:1) the (Z)-isomer 13 (δP −22.7), signals attributed to the (E)-isomer 14 (δP −25.4) were also apparent (Scheme 3).Attempts to purify this by repeated Kugelrohr distillation under reduced pressure led instead to isomerisation to give a 1:1 mixture of 13 and 14.At first sight, it might seem surprising that the product is obtained mainly as the less thermodynamically stable isomer from pyrolysis at 650 °C but then isomerises to the more stable isomer upon simple distillation at 90 °C, but this only serves to emphasise the mild nature of the flash vacuum pyrolysis technique.In addition, it has been shown that under FVP conditions, (Z)-alkenes do not normally isomerise to the (E)-isomer to any great extent at temperatures as low as 650 °C with the degree of conversion of (Z)-to (E)-stilbene, for example, being determined as 12%, in good agreement with our results [15].The presence of aromatic impurities even after distillation prevented full assignment of the 13 C NMR spectra for 13 and 14, but the key signals and the values of the phosphorus coupling constants (Figure 2) were in good agreement with those for 12 while also showing significant differences between the two isomers.The detailed form of the signals for P-CH and P-C=C-Me in the two isomers was Attempts to purify this by repeated Kugelrohr distillation under reduced pressure led instead to isomerisation to give a 1:1 mixture of 13 and 14.At first sight, it might seem surprising that the product is obtained mainly as the less thermodynamically stable isomer from pyrolysis at 650 • C but then isomerises to the more stable isomer upon simple distillation at 90 • C, but this only serves to emphasise the mild nature of the flash vacuum pyrolysis technique.In addition, it has been shown that under FVP conditions, (Z)-alkenes do not normally isomerise to the (E)-isomer to any great extent at temperatures as low as 650 • C with the degree of conversion of (Z)-to (E)-stilbene, for example, being determined as 12%, in good agreement with our results [15].The presence of aromatic impurities even after distillation prevented full assignment of the 13 C NMR spectra for 13 and 14, but the key signals and the values of the phosphorus coupling constants (Figure 2) were in good agreement with those for 12 while also showing significant differences between the two isomers.The detailed form of the signals for P-CH and P-C=C-Me in the two isomers was at first sight surprising.However, this could be explained by coincidental equivalence of some H-H and P-H coupling constants, and the observed patterns (see Supplementary Materials, Figure S14) could be satisfactorily simulated using the values shown in Figure 3.
Molecules 2024, 29, x FOR PEER REVIEW 4 of 13 at first sight surprising.However, this could be explained by coincidental equivalence of some H-H and P-H coupling constants, and the observed patterns (see Supplementary Materials, Figure S14) could be satisfactorily simulated using the values shown in Figure 3.Although the 1,2-arrangement of phosphine and sulfide functions on an alkene double bond, particularly in the (Z)-configuration, gives potentially valuable "hemilabile" proligands, few such compounds seem to be known (Figure 4).Chlorinated compounds such as 15 [16], 16 [17] and 17 [18] have been prepared as mixtures of (E)-and (Z)-isomers by the addition of phosphorus compounds to alkynes.The simpler disubstituted alkenes 18 [19], 19 [20] and 20 [21] have also been prepared but these are the (E)-isomers as shown.More recently, the (Z)-configured vinylphosphonates 21 containing sulfide, selenide and telluride functions [22] as well as the tellurovinylphosphine oxides 22 [23] have also been reported.In terms of the mechanism of this new thermal rearrangement, we envisage attack of the nucleophilic sulfur at phosphorus to give a transient λ 5 -1,2-thiaphosphete, which is of course the same intermediate involved in the extrusion of Ph3PS to give alkynes as observed for 7-9.This behaviour is also consistent with that of the isolable thiaphosphete 23, which fragments with the loss of an alkyne to give the benzoxaphosphole P-sulfide [24].However, perhaps due to the relief of steric congestion, the thiaphosphetes derived from 10 and 11 instead undergo what is effectively a reductive elimination at phosphorus to give the (Z)-phosphinovinyl sulfides 12 and 13 (Scheme 4).Although the 1,2-arrangement of phosphine and sulfide functions on an alkene double bond, particularly in the (Z)-configuration, gives potentially valuable "hemilabile" proligands, few such compounds seem to be known (Figure 4).Chlorinated compounds such as 15 [16], 16 [17] and 17 [18] have been prepared as mixtures of (E)-and (Z)-isomers by the addition of phosphorus compounds to alkynes.The simpler disubstituted alkenes 18 [19], 19 [20] and 20 [21] have also been prepared but these are the (E)-isomers as shown.More recently, the (Z)-configured vinylphosphonates 21 containing sulfide, selenide and telluride functions [22] as well as the tellurovinylphosphine oxides 22 [23] have also been reported.
at first sight surprising.However, this could be explained by coincidental equivalence of some H-H and P-H coupling constants, and the observed patterns (see Supplementary Materials, Figure S14) could be satisfactorily simulated using the values shown in Figure 3.Although the 1,2-arrangement of phosphine and sulfide functions on an alkene double bond, particularly in the (Z)-configuration, gives potentially valuable "hemilabile" proligands, few such compounds seem to be known (Figure 4).Chlorinated compounds such as 15 [16], 16 [17] and 17 [18] have been prepared as mixtures of (E)-and (Z)-isomers by the addition of phosphorus compounds to alkynes.The simpler disubstituted alkenes 18 [19], 19 [20] and 20 [21] have also been prepared but these are the (E)-isomers as shown.More recently, the (Z)-configured vinylphosphonates 21 containing sulfide, selenide and telluride functions [22] as well as the tellurovinylphosphine oxides 22 [23] have also been reported.In terms of the mechanism of this new thermal rearrangement, we envisage attack of the nucleophilic sulfur at phosphorus to give a transient λ 5 -1,2-thiaphosphete, which is of course the same intermediate involved in the extrusion of Ph3PS to give alkynes as observed for 7-9.This behaviour is also consistent with that of the isolable thiaphosphete 23, which fragments with the loss of an alkyne to give the benzoxaphosphole P-sulfide [24].However, perhaps due to the relief of steric congestion, the thiaphosphetes derived from 10 and 11 instead undergo what is effectively a reductive elimination at phosphorus to give the (Z)-phosphinovinyl sulfides 12 and 13 (Scheme 4).In terms of the mechanism of this new thermal rearrangement, we envisage attack of the nucleophilic sulfur at phosphorus to give a transient λ 5 -1,2-thiaphosphete, which is of course the same intermediate involved in the extrusion of Ph 3 PS to give alkynes as observed for 7-9.This behaviour is also consistent with that of the isolable thiaphosphete 23, which fragments with the loss of an alkyne to give the benzoxaphosphole P-sulfide [24].However, perhaps due to the relief of steric congestion, the thiaphosphetes derived from 10 and 11 instead undergo what is effectively a reductive elimination at phosphorus to give the (Z)-phosphinovinyl sulfides 12 and 13 (Scheme 4).
at first sight surprising.However, this could be explained by coincidental equivalence of some H-H and P-H coupling constants, and the observed patterns (see Supplementary Materials, Figure S14) could be satisfactorily simulated using the values shown in Figure 3.Although the 1,2-arrangement of phosphine and sulfide functions on an alkene double bond, particularly in the (Z)-configuration, gives potentially valuable "hemilabile" proligands, few such compounds seem to be known (Figure 4).Chlorinated compounds such as 15 [16], 16 [17] and 17 [18] have been prepared as mixtures of (E)-and (Z)-isomers by the addition of phosphorus compounds to alkynes.The simpler disubstituted alkenes 18 [19], 19 [20] and 20 [21] have also been prepared but these are the (E)-isomers as shown.More recently, the (Z)-configured vinylphosphonates 21 containing sulfide, selenide and telluride functions [22] as well as the tellurovinylphosphine oxides 22 [23] have also been reported.In terms of the mechanism of this new thermal rearrangement, we envisage attack of the nucleophilic sulfur at phosphorus to give a transient λ 5 -1,2-thiaphosphete, which is of course the same intermediate involved in the extrusion of Ph3PS to give alkynes as observed for 7-9.This behaviour is also consistent with that of the isolable thiaphosphete 23, which fragments with the loss of an alkyne to give the benzoxaphosphole P-sulfide [24].However, perhaps due to the relief of steric congestion, the thiaphosphetes derived from 10 and 11 instead undergo what is effectively a reductive elimination at phosphorus to give the (Z)-phosphinovinyl sulfides 12 and 13 (Scheme 4).Somewhat similar processes are the transfer of Ph from Se to O in acylselenonium ylides such as 24 to give 25 observed by Rakitin [25] (Scheme 5), the transformation of 26 into 27 postulated by Zbiral in the interaction of ylides Ph 3 P=CHR with benzyne [26] and the rearrangement of the ylide-containing N-heterocyclic carbene 28 via 29 to give 3-phosphinoindole 30 [27].
With the two new hemilabile proligands in hand, we now set about exploring their coordination chemistry.In our preliminary communication [6], the formation of the square planar platinum complex 31 from 12 was described along with its X-ray structure determination (CSD RefCode: AJONAP).We were also successful in obtaining complexes of 12 with a wide range of other standard transition metal reagents (Scheme 6).Reaction of the starting materials in CH2Cl2 followed by partial evaporation and precipitation with diethyl ether gave the new complexes 32-36 in moderate to good yield as crystalline solids, giving the expected microanalytical data and 1 H and 31 P NMR spectra.From the analytical data, it was clear that the complexes had formed with the expected stoichiometry according to the metal source employed, with 12 acting as a bidentate ligand in the square planar platinum(II) and palladium(II) complexes 31 and 32 and the cationic ruthenium(II) complex 36, but as a monodentate ligand through the more strongly donating phosphorus atom in palladium(II) complex 33, gold(I) complex 34 and iridium(III) complex 35.In addition to the compound 31 already confirmed by X-ray Scheme 5. Some mechanistic precedents.
With the two new hemilabile proligands in hand, we now set about exploring their coordination chemistry.In our preliminary communication [6], the formation of the square planar platinum complex 31 from 12 was described along with its X-ray structure determination (CSD RefCode: AJONAP).We were also successful in obtaining complexes of 12 with a wide range of other standard transition metal reagents (Scheme 6).
With the two new hemilabile proligands in hand, we now set about exploring their coordination chemistry.In our preliminary communication [6], the formation of the square planar platinum complex 31 from 12 was described along with its X-ray structure determination (CSD RefCode: AJONAP).We were also successful in obtaining complexes of 12 with a wide range of other standard transition metal reagents (Scheme 6).Reaction of the starting materials in CH2Cl2 followed by partial evaporation and precipitation with diethyl ether gave the new complexes 32-36 in moderate to good yield as crystalline solids, giving the expected microanalytical data and 1 H and 31 P NMR spectra.From the analytical data, it was clear that the complexes had formed with the expected stoichiometry according to the metal source employed, with 12 acting as a bidentate ligand in the square planar platinum(II) and palladium(II) complexes 31 and 32 and the cationic ruthenium(II) complex 36, but as a monodentate ligand through the more strongly donating phosphorus atom in palladium(II) complex 33, gold(I) complex 34 and iridium(III) complex 35.In addition to the compound 31 already confirmed by X-ray Scheme 6. Formation of transition metal complexes from 12.
Reaction of the starting materials in CH 2 Cl 2 followed by partial evaporation and precipitation with diethyl ether gave the new complexes 32-36 in moderate to good yield as crystalline solids, giving the expected microanalytical data and 1 H and 31 P NMR spectra.From the analytical data, it was clear that the complexes had formed with the expected stoichiometry according to the metal source employed, with 12 acting as a bidentate ligand in the square planar platinum(II) and palladium(II) complexes 31 and 32 and the cationic ruthenium(II) complex 36, but as a monodentate ligand through the more strongly donating phosphorus atom in palladium(II) complex 33, gold(I) complex 34 and iridium(III) complex 35.In addition to the compound 31 already confirmed by X-ray diffraction [6], we were able to obtain X-ray structures for gold complex 34 and iridium complex 35 (see below).
Although the methyl proligand 13 was available in lower quantity and purity, we were able to prepare its palladium dichloride complex 37 analogous to 32 and also determined its X-ray structure (Scheme 7).
diffraction [6], we were able to obtain X-ray structures for gold complex 34 and iridium complex 35 (see below).
Although the methyl proligand 13 was available in lower quantity and purity, we were able to prepare its palladium dichloride complex 37 analogous to 32 and also determined its X-ray structure (Scheme 7).Scheme 7. Formation of palladium complex 37 and structure of a related iron complex.
The structures of the complexes 34, 35 and 37 together with the numbering systems used are shown below (Figure 5) with that of 31 [6] also included for comparison.The structures of the complexes 34, 35 and 37 together with the numbering systems used are shown below (Figure 5) with that of 31 [6] also included for comparison.
Molecules 2024, 29, x FOR PEER REVIEW 6 of 1 diffraction [6], we were able to obtain X-ray structures for gold complex 34 and iridium complex 35 (see below).Although the methyl proligand 13 was available in lower quantity and purity, w were able to prepare its palladium dichloride complex 37 analogous to 32 and also deter mined its X-ray structure (Scheme 7).Scheme 7. Formation of palladium complex 37 and structure of a related iron complex.
The structures of the complexes 34, 35 and 37 together with the numbering system used are shown below (Figure 5) with that of 31 [6] also included for comparison.For comparison, we also show (Figure 6) the previously reported structures of 10 [6] and 12 [6], and the iron(II) complex 38 [28], which, although made in a quite different way, contains the phosphinovinylthiolate corresponding to 12 as an anionic bidentate ligand.
For comparison, we also show (Figure 6) the previously reported structures of 10 [6] and 12 [6], and the iron(II) complex 38 [28], which, although made in a quite different way, contains the phosphinovinylthiolate corresponding to 12 as an anionic bidentate ligand.The key structural parameters for complexes 34, 35 and 37 and, for comparison, those for 10, 12, 31 and 38 are presented in Table 1.In all seven structures, the P-C-C-S fragment is quite accurately planar with a torsion angle of <10° in every case.
Table 1.Geometric parameters for 10 [6], 12 [6], 31 [6], 34, 35, 37 and, for comparison, 38 [28].If we first compare 10 and 12, the major change in geometry associated with the transformation of P=C-C=S into P-C=C-S is clear.However, comparing the structural parameters of 12 with those of its complexes 31, 34 and 35 as well as the related thiolate complex 38 shows a remarkable degree of consistency.As expected, the complexes 34 and 35 where the metal binds only to phosphorus have bond lengths around the =C-SPh that are relatively unaffected, while the bidentate binding to platinum in 31 results in the significant lengthening of C-S and shortening of C-P.While the smaller size of AuCl means the ligand 12 can retain the orientation of the PPh2 group, coordination to the much larger Cp*IrCl2 requires the PPh2 group to rotate, placing the phenyl groups facing towards SPh.The key structural parameters for complexes 34, 35 and 37 and, for comparison, those for 10, 12, 31 and 38 are presented in Table 1.In all seven structures, the P-C-C-S fragment is quite accurately planar with a torsion angle of <10 • in every case.

Bond lengths (Å)
Table 1.Geometric parameters for 10 [6], 12 [6], 31 [6], 34, 35, 37 and, for comparison, 38 [28].If we first compare 10 and 12, the major change in geometry associated with the transformation of P=C-C=S into P-C=C-S is clear.However, comparing the structural parameters of 12 with those of its complexes 31, 34 and 35 as well as the related thiolate complex 38 shows a remarkable degree of consistency.As expected, the complexes 34 and 35 where the metal binds only to phosphorus have bond lengths around the =C-SPh that are relatively unaffected, while the bidentate binding to platinum in 31 results in the significant lengthening of C-S and shortening of C-P.While the smaller size of AuCl means the ligand 12 can retain the orientation of the PPh 2 group, coordination to the much larger Cp*IrCl 2 requires the PPh 2 group to rotate, placing the phenyl groups facing towards SPh.The similarity between the geometry of the neutral ligand 12 in complexes such as 31 and the anionic enethiolate in 38 is also notable, with only the =C-S length being significantly shorter in the latter.The angles within the five-membered ring in complexes 31, 37 and 38 are also remarkably consistent.

General Experimental Details
NMR spectra were recorded on solutions in CDCl 3 unless otherwise stated using Bruker instruments, and chemical shifts are given in ppm to high frequency from Me 4 Si for 1 H and 13 C and H 3 PO 4 for 31 P with coupling constants J in Hz.The 13 C NMR spectra are referenced to the solvent signal at 77.0 (CDCl 3 ).IR spectra were recorded on a Perkin Elmer 1420 instrument.Elemental analysis was conducted using a Carlo Erba CHNS analyser.Mass spectra were obtained using a Micromass instrument and the ionisation method used is noted in each case.Preparative TLC was carried out using 1.0 mm layers of Merck alumina 60 G containing 0.5% Woelm fluorescent green indicator on glass plates.Melting points were recorded on a Gallenkamp 50W melting point apparatus or a Reichert hot-stage microscope.
Flash vacuum pyrolysis (FVP) was carried out in a conventional flow system by subliming the starting material through a horizontal quartz tube (30 × 2.5 cm) externally heated by tube furnace to 650 • C and maintained at a pressure of 2-5 × 10 −2 Torr by a rotary vacuum pump.Products were collected in a liquid N 2 cooled U-shaped trap and purified as noted.
General organic and inorganic reagents and solvents were obtained from standard suppliers and used as received.Dry THF was prepared by storage over sodium wire.Start [34] were prepared by the reported methods.

Preparation of Thioacetylmethylenetriphenylphosphorane 11
A solution of acetylmethylenetriphenylphosphorane (8.0 g, 25 mmol) and Lawesson's reagent (5.1 g,12.6 mmol) in toluene (300 mL) was heated under reflux under nitrogen for 3 h.The mixture was allowed to cool to RT and the solution was poured off leaving an insoluble oily residue and evaporated.Recrystallisation of the resulting solid from ethyl acetate gave the product (4.62 g, 55%) as pale-yellow crystals, mp 172-174   FVP of the ylide 10 (0.50 g, 1.33 mmol) was performed at 650 • C and 3.8 × 10 −2 Torr and was complete within 1 h.Preparative TLC (silica, diethyl ether) of the crude material gave the product 12 (0.41 g, 82%) as pale-yellow plates, mp 121-123

Figure 1 .
Figure 1.Thiocarbonyl ylides used in this study.Compounds 7-9 are already known [8] andgave analytical and spectroscopic data in agreement with the reported values.The new compounds 10 and 11 were fully characterised and showed distinctive NMR signals confirming the presence of P=CH-C=S [10 δP +5.0; δH 5.22 (d, 2 J 34 Hz, P=CH); δC 81.3 (d, 1 J 118 Hz, P=CH), 214.4 (d, 2 J 4 Hz, C=S).11 δP +8.1; δH 5.18 (d, 2 J 32 Hz, P=CH); δC 84.1 (d, 1 J 113 Hz, P=CH), 200.5 (d, 2 J 4 Hz, C=S)].In fact, the phosphorus coupling extended throughout the structures with all carbon signals except the CH3 of t-butyl observed as doublets in the 13 C NMR spectra.This is discussed further below in the context of a comparison of the coupling pattern before and after the thermal rearrangement.The structure of 10 was also confirmed by X-ray diffraction (CSD RefCode: AJOMUI) and this was described in our earlier communication[6].We might also note at this point that compound 10 has a particularly pungent and unpleasant smell.Although it is not very volatile, it is extremely persistent and inadvertent contact with equipment or work surfaces contaminated with 10, even after several months, results in the release of its characteristic smell.We speculate that this may be due to slow hydrolysis of the P=C bond to give Ph3PO and release thiopinacolone, t-BuC(=S)Me.When the thiocarbonyl ylides 7, 8 and 9 were subjected to FVP, there was complete reaction at a furnace temperature of 650 °C to give Ph3PS (δP +43) at the furnace exit and the expected alkynes 2, 2,2-dimethylpent-3-yne, 1-phenylpropyne and 1-phenyloctyne, in the cold trap.Thus, these compounds behave in a similar way to their carbonyl analogues but react more readily than the latter, which require a temperature of 750 °C for complete reaction[11].The higher reactivity of thiocarbonyl-as compared to carbonyl-stabilised ylides is a feature that we have already noted and quantified in a series of kinetic studies on the pyrolysis of carbamoyl and thiocarbamoyl ylides[12][13][14].When the two ylides 10 and 11 were subjected to FVP, the reaction was also complete at 650 °C, but the process involved turned out to be completely different.In each case, only a single main product was obtained, which was isomeric with the starting material.In the case of 10, the product 12 was obtained in good yield and in a pure form as a

Figure 1 .
Figure 1.Thiocarbonyl ylides used in this study.Compounds 7-9 are already known[8] andgave analytical and spectroscopic data in agreement with the reported values.The new compounds 10 and 11 were fully characterised and showed distinctive NMR signals confirming the presence of P=CH-C=S [10 δ P +5.0; δ H 5.22 (d, 2 J 34 Hz, P=CH); δ C 81.3 (d,1 J 118 Hz, P=CH), 214.4 (d, 2 J 4 Hz, C=S).11 δ P +8.1; δ H 5.18 (d, 2 J 32 Hz, P=CH); δ C 84.1 (d,1 J 113 Hz, P=CH), 200.5 (d, 2 J 4 Hz, C=S)].In fact, the phosphorus coupling extended throughout the structures with all carbon signals except the CH 3 of t-butyl observed as doublets in the 13 C NMR spectra.This is discussed further below in the context of a comparison of the coupling pattern before and after the thermal rearrangement.The structure of 10 was also confirmed by X-ray diffraction (CSD RefCode: AJOMUI) and this was described in our earlier communication[6].We might also note at this point that compound 10 has a particularly pungent and unpleasant smell.Although it is not very volatile, it is extremely persistent and inadvertent contact with equipment or work surfaces contaminated with 10, even after several months, results in the release of its characteristic smell.We speculate that this may be due to slow hydrolysis of the P=C bond to give Ph 3 PO and release thiopinacolone, t-BuC(=S)Me.When the thiocarbonyl ylides 7, 8 and 9 were subjected to FVP, there was complete reaction at a furnace temperature of 650 • C to give Ph 3 PS (δ P +43) at the furnace exit and the expected alkynes 2, 2,2-dimethylpent-3-yne, 1-phenylpropyne and 1-phenyloctyne, in the cold trap.Thus, these compounds behave in a similar way to their carbonyl analogues but react more readily than the latter, which require a temperature of 750 • C for complete reaction[11].The higher reactivity of thiocarbonyl-as compared to carbonyl-stabilised ylides is a feature that we have already noted and quantified in a series of kinetic studies on the pyrolysis of carbamoyl and thiocarbamoyl ylides[12][13][14].When the two ylides 10 and 11 were subjected to FVP, the reaction was also complete at 650 • C, but the process involved turned out to be completely different.In each case, only a single main product was obtained, which was isomeric with the starting material.In the case of 10, the product 12 was obtained in good yield and in a pure form as a crystalline solid after preparative TLC.This showed a 31 P NMR signal at −19.8 ppm, in the region

Scheme 4 .
Scheme 4. Mechanism proposed for the thermal rearrangement of 10 and 11.Somewhat similar processes are the transfer of Ph from Se to O in acylselenonium ylides such as 24 to give 25 observed by Rakitin[25] (Scheme 5), the transformation of 26

Scheme 4 .
Scheme 4. Mechanism proposed for the thermal rearrangement of 10 and 11.Somewhat similar processes are the transfer of Ph from Se to O in acylselenonium ylides such as 24 to give 25 observed by Rakitin[25] (Scheme 5), the transformation of 26

Scheme 4 .
Scheme 4. Mechanism proposed for the thermal rearrangement of 10 and 11.Somewhat similar processes are the transfer of Ph from Se to O in acylselenonium ylides such as 24 to give 25 observed by Rakitin [25] (Scheme 5), the transformation of 26 Scheme 4. Mechanism proposed for the thermal rearrangement of 10 and 11.
3.4.7.(Z)-1-Diphenylphosphino-2-phenylthiopropene Palladium Dichloride Complex 37 analysed the data and wrote the paper.All authors have read and agreed to the published version of the manuscript.