Reaction of lithium ( 2 , 4 , 6-tri-tert-butylphenyl ) silylphosphides with haloforms

The reaction of lithium (tert-butyldimethylsilyl)(2,4,6-tri-tert-butylphenyl)phosphide with chloroform afforded (Z)-2-(tert-butyldimethylsilyl)-2-chloro-1-(2,4,6-tri-tert-butylphenyl)-1phosphaethene. The NMR study revealed a secondary phosphine resulting from a formal insertion of a dichlorocarbene to the P-Si bond as a reaction intermediate. The reaction is specific to the reactants and substrates. The use of bromoform gave a bromosilylphosphine and a 2bromo-1-phosphaethene. A less hindered trimethylsilylphosphide afforded bis(2,4,6-tri-tertbutylphenyl)diphosphene.


Introduction
The sterically protected 2-phospha-1-silylethenyllithium A is a convenient reagent for the introduction of the phosphaethenyl and phosphaethenylene moieties (Scheme 1). 1 The lithium reagents are prepared in situ from the corresponding chloro-or bromophosphaethenes B by halogen metal exchange and can react with various electrophiles.The reactions with ketones or aldehydes lead to 1,3-rearrangement of the silyl group followed by elimination to give phosphaallenes.1b,1c,1d The 2-phospha-1-silylethenyllithiums bearing a leaving group on the silicon are employed to the synthesis of phosphasilaallene 1e and phosphinidene oxasiletane.1f The precursors, 2-halo-1-phospha-2-silylethenes B, are generally synthesized by the metallation of 2,2-dihalo-1-phosphaethene followed by the reaction with silyl halides.1a,1b,1c Herein, we report formation of 2-chloro-1-phospha-2-silylethene B by the reaction of a lithium silylphosphide with chloroform.The reaction was unexpectedly found in the course of the synthetic study of phosphaalkenes by the reaction of the lithium silylphosphides with carbonyl compounds.The phosphaethene B was found in the samples mixed with CDCl3 for the NMR monitoring of the reaction.Further investigation revealed that the reaction is specific to the substrate and the reactant, but the NMR study allowed observation of a reaction intermediate, which deepened understanding of the reaction mechanism.Scheme 1.Reactions of 2-phospha-1-silylethenyllithium.

Results and Discussion
The lithiation of the sterically protected primary phosphine 1 2 followed by the addition of tertbutylchlorodimethylsilane gave silylphosphine 2a (Scheme 2). 3 The lithium silylphosphide 3a 3 was generated by the lithiation of 2a with butyllithium and was allowed to react with chloroform at -78 °C.The 31 P NMR spectrum of the sample taken from the cold reaction mixture indicated to consist of 4a (P 297.0 (s))/5a (33.8(s))/6a (-21.9 (d, 1 JPH 231.9 Hz))/2a (-137.5 (d, 1 JPH 207.5 Hz)) in a ratio of 3/2/9/10.The compound of P -21.9, which was assigned as a secondary phosphine 6a based on JPH value and the reaction mechanism (vide infra, Scheme 5), gradually disappeared during the measurement with the growing signal intensity of 4a.Thus, the mixture was warmed to 20 °C for the completion of the reaction and butyllithium was added to the mixture at -78 °C for the regeneration of the silylphosphide from 2a.The 31 P NMR spectrum of the cold sample consisted of the four compounds 4a/5a/6a/2a in a ratio of 10/3/5/5 and 6a was again converted to 4a.The mixture was warmed and purified to give 4a in 47%.The structure of 4a was confirmed by 1 H, 13 C, and 31 P NMR and mass spectroscopy.The geometry of 4a was determined to be (Z)-form by comparison with NMR data of (E)-and (Z)-forms of the trimethylsilyl derivatives 4c.1a,1b 1 H and 13 C NMR signals of the SiCH3 groups of (Z)-4a and (Z)-4c showed the coupling with 31 P nucleus typical of those cis to the lone pair of the low coordinated phosphorus, 4 which are not observed for (E)-4c ((Z)-4a. 1 H NMR  0.26 (d, JPH 1.5 Hz), 13 C NMR  -4.7 (d, JPC 11.0 Hz).(Z)-4c. 1 H NMR  0.26 (d, JPH 1.10 Hz). 13 C NMR −1.27 (d, JPC 9.16 Hz)).Scheme 2. Reaction of lithium (tert-butyldimethylsilyl)(2,4,6-tri-tert-butylphenyl)phosphide with chloroform.
To gain further insight into the reaction, we investigated a similar reaction with different substrates (Scheme 3, Table 1).The reaction of the phosphide 3a with bromoform gave bromophosphine 5b as a main product with small amount of phosphaethene 7b.1a,5 The reaction of the less hindered trimethylsilylphosphide 3b 6 with chloroform or bromoform gave diphosphene 8 7 as a main isolable product with a trace amount of 4. The reaction was substratespecific rather than general.The phosphaethene 4a gives phosphaethenyl anion by halogen metal exchange similarly to the trimethylsilyl derivative 4c and 4d, or by desilylation catalyzed by fluoride ion, and thus 4a is expected to be a potential source of the Mes*PC unit (Scheme 4).Scheme 3. Reaction of the silylphosphides with haloforms.The 31 P NMR spectra of the reaction mixture strongly suggest the secondary phosphine 6a as a reaction intermediate to 4a.Taking the structure of the product 4a and the reaction condition into consideration, the reaction intermediate is 6a and the formation of 4a can be explained as shown in scheme 5.The phosphide 3a deprotonates chloroform to give silylphosphine 2a and trichloromethyllithium.Trichloromethyllithium works as a carbenoid or is converted to dichlorocarbene to give the P-Si insertion product 6a by the direct insertion or by the formation of the phosphonium salt followed by silyl migration.The formal insertion of carbenes to the P-H bond of sterically protected primary phosphine has been reported.1a,5 The dehydrochlorination of 6a under the basic condition gives 4a.The silylphosphide 3a prefers the halogen-metal exchange to the deprotonation in the reaction with bromoform because of the lower acidity of the C-H and higher reactivity of the C-Br bond.The formation of phosphaethene 7b can be rationalized by the formal insertion of the bromocarbene to the P-Si bond of 5b followed by the elimination of the bromotrimethylsilane.The P-Si bond or phosphorus lone pair of 5b is less reactive toward the vacant orbital of the electrophilic carbene than that of silylphosphine 2a because of the inductive effect of the bromo group.Actually, the reaction of 3a with carbon tetrachloride gave chlorophosphine 5a as the sole product without formation of phosphaalkenes.Phosphine 5a is more inert to the carbene as expected from higher electronegativity of chlorine.The absence of the silyl substituted phosphaethene 4b excludes the formation of 7b by the desilylation of 4b.A less hindered trimethylsilylphosphide 3b undergoes halogen-metal exchange to give chlorosilylphosphine 5c or bromosilylphosphine 5d.5c and 5d are more reactive than the more hindered tert-butyldimethylsilyl derivatives 5a and 5b to give diphosphene 8 as reported. 8Scheme 5. A plausible reaction mechanism.Minor or missing products are shown in faint fonts.
The reason for the reactivity difference between the tert-butyldimethylsilyl and the trimethylsilyl derivatives is not clear so far.It is not plausible that the difference of the silyl groups leads to marked difference of the ratio between the chlorinated and the protonated products.In the course of the various reactions of the trimethylsilylphosphide, we often encountered the desilylated products which do not appear in the similar reaction of the tertbutyldimethylsilyl derivative.Thus, it seems the tert-butyl group on the silicon hinders nucleophilic attack on the silicon, so that the tert-butyldimethylsilyl derivative exhibits unique reactivity, while the trimethylsilyl derivatives undergo desilylation and various reactions.As for the reactivity difference within haloforms, the relative acidities can give an explanation for the diversity of the reaction products.

Conclusions
We found a unique reaction of a silylphosphide with a carbene or carbenoid to afford 2-chloro-2silyl-1-phosphaethene.The reaction is specific to tert-butyldimethylsilyl derivative 3a and chloroform because this combination only affords the nucleophilic silylphosphine that is reactive to the carbene.The reaction is unique not only because of the direct formation of the synthetically useful 2-chloro-2-silyl-1-phosphaethene, but also because of the observation of the reaction intermediate.The reaction itself suffers from severe limitation of the substrates; the behavior of the series of the substrates shown in this report is suggestive of the development of a broad range of the reactions having analogous mechanism.

Experimental Section
General. 1 H, 13 C, and 31 P NMR spectra were measured on a JEOL FX90Q spectrometer. 1 H and 13 C NMR chemical shifts are expressed as  from external tetramethylsilane and calibrated to the residual proton of the deuterated solvents ( 7.25 for chloroform-d) or the carbon of the deuterated solvent ( 77.0 for chloroform-d). 31P NMR chemical shifts are expressed as  from external 85% H3PO4.Mass spectra were measured on a JEOL D-300 spectrometer.Fuji Silysia BW-300 was used for the flash column chromatography.All reactions were carried out under argon unless otherwise specified.Anhydrous tetrahydrofuran was distilled from sodium diphenylketyl under argon just prior to use.

Table 1 .
Reaction of the silylphosphides with haloforms Scheme 4. Reaction of 4a.