Triple para-substitution reactions of B(C₆F₅)₃ and [(C₆F₅)₃PF]⁺ with P(SiMe₃)₃

Abstract

para-Substitution reactions on C₆F₅ rings of Lewis acids have been exploited to achieve triply substituted derivatives. The reaction of B(C₆F₅)₃ with P(SiMe₃)₃ ultimately affords the Lewis acid B(C₆F₄P(SiMe₃)₂)₃1. This species binds Lewis bases affording the adducts LB(C₆F₄P(SiMe₃)₂)₃ (L = MeCN 2, OPEt₃3, PMe₃4, PBu₃5) and reacts with LiMe to give the salt [Li][MeB(C₆F₄P(SiMe₃)₂)₃]·3THF 6. It also reacts with H₂O to give (L)B(C₆F₄PH₂)₃ (L = H₂O 7, MeCN 8). In an analogous fashion, [(C₆F₅)₃PF][B(C₆F₅)₄] was converted to [FP(C₆F₄P(SiMe₃)₂)₃] [B(C₆F₅)₄] 9 and subsequently to [(MeO)P(C₆F₄PH₂)₃][B(C₆F₅)₄] 10.

---

The discovery of frustrated Lewis pairs (FLPs) emerged from the initial exploration of the chemistry derived from the products of the reactions of B(C₆F₅)₃ with bulky phosphines.¹ In the first such systems, these reactions lead to para-substitution of the phosphine on a C₆F₅ ring with concurrent fluoride transfer to boron affording the zwitterions R₃PC₆F₄B(C₆F₅)₂F (Scheme 1). Use of secondary phosphines and replacement of the fluoride on boron by hydride, afforded R₂PHC₆F₄B(C₆F₅)₂H, which proved to be the first metal-free system capable of reversible activation of H₂.¹ This finding led to further studies revealing that bulky phosphines did not react with B(C₆F₅)₃ at all, but the combination effected the homolytic cleavage of H₂.² These early findings set the stage for the plethora of FLP chemistry that has emerged in the last two decades.^3–9

Scheme 1 Selected examples of routes to para-donor substitution on aryl-rings of Lewis acids.

Focusing attention on the para-substitution reactions that provided the original FLP systems, it should be noted that Erker and coworkers¹⁰ had previously reported analogous para-substitution products as early as 1998. In those cases, while classical Lewis acid–base adducts were observed at room temperature, thermal reactions of ylides with B(C₆F₅)₃ afforded the para-substitution products R₃PCHRC₆F₄B(C₆F₅)₂F (Scheme 1). Subsequent to our preparation of R₂PHC₆F₄B(C₆F₅)₂H and quite apart from FLP chemistry, we exploited such species as anionic ligands for transition metal chemistry.^11,12 We also studied the analogous reaction of phosphines with trityl cations affording similar para-substitution products of the form [Ph₂CHC₆H₄PR₃]X.¹³ In 2010, Rieger and coworkers¹⁴ reported an alternative synthetic route, deriving tBu₂PC₆F₄B(C₆F₅)₂ from the reaction of tBu₂PSiMe₃ with B(C₆F₅)₃ (Scheme 1). In 2013, we showed that analogous para-substitution of a C₆F₅ ring in fluorophosphonium cations provided an avenues to the species of the form [Ph₃PC₆F₄PF₂Ph₂][X] (Scheme 1).¹⁵ More recently, we have described the reactions of B(C₆F₅)₃ or the trityl cation with N,N′-dimesityldiamidocarbene (DAC) (Scheme 1).¹⁶ In the former case, Me₂C(CONMes)₂CC₆F₄BF(C₆F₅)₂ was obtained. This species could be reduced to the radical borane derivative [Me₂C(CONMes)₂CC₆F₄B(C₆F₅)₂]˙ while for the trityl analogue, the radical cation salt [Me₂C(CONMes)₂CC₆H₄CPh₂˙][B(C₆F₅)₄] was obtained directly.

The notion of multiple para-substitution reactions has been briefly probed. Using judiciously selected ferrocenyl-bisphosphine we showed that reaction with B(C₆F₅)₃ affording the di-zwitterion Fe(C₅H₄P(iPr₂)₂C₆F₄B(C₆F₅)₂F)₂ (Scheme 1).¹⁷ Multiple substitutions on aryl boranes have drawn lesser attention. In 2008 and 2009, Gabbaï and coworkers^18,19 reported the preparation of the bis- and tris-ammonium substituted boranes [MesB(C₆H₂Me₂NMe₃)₂]²⁺ and [B(C₆H₂Me₂NMe₃)₃]³⁺ derived from the amine-substituted precursors and their use in cyanide binding (Scheme 1). Subsequently, Do and Lee and coworkers,²⁰ reported the synthesis of the tris-phosphine species affording the phosphonium borane [B(C₆H₂Me₂PPh₂Me)₃]³⁺ (Scheme 1). In both studies, these species were prepared via the reaction of a donor substituted aryl lithium reagent with a boron halide. In this communication, we develop a synthetic route targeting species derived from para-substitution on fluoro-arene rings. Initially this is applied to B(C₆F₅)₃ and subsequently to the fluorophosphonium cation, [FP(C₆F₄PR₂)₃]⁺. The Lewis acidity and chemistry of these derivatives is also explored.

Our initial efforts involved a modification of the Reiger protocol.¹⁴ Treatment of B(C₆F₅)₃ with excess tBu₂PSiMe₃ was performed under various conditions, however this gave only the known species (C₆F₅)₂B(C₆F₄PtBu₂). This finding suggests that while this silyl phosphine is sufficiently bulky to effect para-substitution, the Lewis acidity of the product is sufficiently diminished precluding further reactivity.

Undaunted we next probed the reaction of B(C₆F₅)₃ with one equivalent of P(SiMe₃)₃. This gave a ¹⁹F NMR spectrum consistent with a mixture of two products arising from para-substitution and release of Me₃SiF. Addition of a further equivalent gave rise to yet another species with the consumption of one of the initial species. Finally, addition of three equivalents afforded signals consistent with one major product with a minor amount of a second species. These data were interpreted as the consecutive substitution at the para positions of the C₆F₅ rings of B(C₆F₅)₃ (Scheme 2). Indeed, the reaction of B(C₆F₅)₃ with a slight excess of 3 equivalents of P(SiMe₃)₃ afforded the new yellow product 1 after 4 days. This species was isolated in 82% yield. The ³¹P{¹H} NMR spectrum displayed a resonance at δ −162.7 that is a singlet. The ¹⁹F NMR spectrum shows two multiplets at δ −125.0 and δ −130.0 ppm consistent with the presence of C₆F₄ fragments (Fig. 1). The ¹¹B NMR spectrum showed only a broad resonance at δ 60.0 ppm consistent with a three coordinate B atom. These data together with elemental analysis affirm the formulation of 1 as B(C₆F₄P(SiMe₃)₂)₃. Compound 1 proved to be soluble in all common organic solvents.

Fig. 1 Portion of ¹⁹F NMR spectra of reaction mixtures of B(C₆F₅)₃ and P(SiMe₃)₃. Signals arising from mono, di and tri substituted products shown in green, blue, and red shading respectively.

Scheme 2 Synthesis of 1–8.

The para-substitution reactions affording 1 hinge on the steric frustration that precludes B–P bond formation prompting instead attack at the para-carbon of a C₆F₅ ring. Loss of Me₃SiF affords concurrent ring substitution. Interestingly the initially formed product (C₆F₅)₂B(C₆F₄P(SiMe₃)₂) retains sufficient Lewis acidity to allow this process to be repeated on the remaining rings, ultimately affording 1. This stands in contrast to the known reaction where substitution with one PtBu₂ group. In the present case, the installation of a P(SiMe₃)₂ fragment enhances the electrophilicity of the remaining rings, presumably via the beta-silicon effect, allowing subsequent substitutions to proceed.

Performing the Gutman–Beckett Lewis acidity test on 1 reveals an acceptor number of 72 based on the change in the ³¹P NMR chemical shift. As the corresponding acceptor number for B(C₆F₅)₃ is 88, the replacement of the para-fluorine atoms with P(SiMe₃)₂ fragment reduces the Lewis acidity.

Nonetheless, compound 1 reacts with small, sterically unencumbered donors such as acetonitrile and PMe₃ to form the classical Lewis acid–base adducts LB(C₆F₄P(SiMe₃)₂)₃ (L = MeCN 2, OPEt₃3, PMe₃4, PBu₃5) (Scheme 2). The resulting colourless solutions showed ¹¹B NMR resonances at δ −10.8, −2.4, −14.4, and −13.2 ppm respectively, suggest the presence of four coordinate boron. The ¹⁹F chemical shifts for these products differ only slightly. The ³¹P resonance for the P(SiMe₃)₂ groups of 2 was seen at −171.4 ppm. In the case of 4 and 5 the ³¹P resonances were observed at −6.4 and −171.5 ppm and −0.4, −171.9 ppm, respectively with the former signals being attributed to the phosphines coordinated to boron. In the case of 3 the ³¹P resonances were observed at 73.5 and −173.4 ppm, with the former signal arising from the coordinated OPEt₃.

After protracted and exhaustive trails under various conditions, crystals of 4 and 5 were finally obtained via recrystallization from pentane using an excess of PMe₃ and PBu₃, respectively. Single crystal X-ray diffraction studies confirmed the formulation of 4 (Fig. 2) and 5 (Fig. 3). In the case of 4 the P–B bond was found to be 2.051(5) Å, while the C–B–P angles were 114.7(2)°, 112.7(2)°, and 116.8(2)°, affirming the pseudo-pyramidal geometry about boron. Three fluoroarene rings each substituted in the para-position with a P(SiMe₃)₂ fragment complete the coordination sphere of boron. The P–C for the P(SiMe₃)₂ substituents were found to be 1.841(4) Å, 1.844(5) Å, and 1.850(5) Å, while the P–Si bond lengths range from 2.255(2) Å to 2.266(2) Å. The Si–P–Si angles were found to be 105.54(7)°, 105.18(7)°, and 104.89(7)° while the Si–P–C angle range from to ranges from 98.6(1)° to 110.61(15)° consistent with a pseudo pyramidal geometry at the pendant phosphorus atoms. The corresponding data for 5 was very similar, although it is interesting to note that a threefold axis of symmetry was crystallographically imposed on the molecule in the solid state.

Fig. 2 POV-ray depictions of the crystallographic structure of 4. Hydrogen atoms are omitted for clarity. C: black, B: yellow green, F; pink, P: orange, Si: salmon.

Fig. 3 POV-ray depictions of the crystallographic structure of 5. Hydrogen atoms are omitted for clarity. C: black, B: yellow green, F; pink, P: orange, Si: salmon.

In a similar fashion, reaction of 1 with methyllithium (1.6 M solution in diethyl ether) afforded the lithium methyl-borate salt [Li][MeB(C₆F₄P(SiMe₃)₂)₃]·3THF 6 (Scheme 2). The ¹¹B NMR spectrum was again consistent with the presence of a 4-coordinate boron species with a resonance at δ −14.0 ppm while ³¹P and ¹⁹F{¹H} NMR resonances were seen at –177.2, and −130.7 and −131.5 ppm, respectively.

Compound 1 also showed a facile reaction with water or alcohol. In the case of water, the ³¹P NMR spectrum of the reaction mixture showed a triplet at −174.3 ppm with a ¹J_P–H of 209 Hz in C₆H₆. The ¹⁹F NMR spectrum shift slightly from that of 1 consistent with the retention of the C₆F₄ linking units. The corresponding ¹¹B resonance was observed at −1.2 ppm consistent with coordination of water to the boron atom. These data are consistent with the new species 7 as (H₂O)B(C₆F₄PH₂)₃ (Scheme 2). Interestingly, the corresponding reaction with D₂O affording the analogous species (D₂O)B(C₆F₄PD₂)₃7_d. This species exhibited a pentet in the ³¹P{¹H} NMR spectrum at −179.7 ppm with P–D coupling of 32 Hz consistent with the presence of the PD₂ fragments. Notably, trace amounts of the species (D₂O)B(C₆F₄PHD)₃ were also evident from the 1 : 1 : 1 triplet signal at −178.9 ppm in the ³¹P{¹H} NMR spectrum.

Dissolution of 7 in MeCN resulted in replacement of the water affording a new species 8. The observation of a ³¹P NMR triplet at −175.1 ppm with a ¹J_PH of 209 Hz and a ¹¹B NMR signal at −10.8 ppm were consistent with the formulation of 8 as the adduct (MeCN)B(C₆F₄PH₂)₃ (Scheme 2). It is interesting to note that despite the presence of three primary phosphine sites in 7 and 8, no evidence of B–P bond formation is seen. This is presumed to be a result of the diminished Lewis basicity at phosphorus because of the BC₆F₄ substituent, and the steric congestion at boron.

To extend this reactivity to other systems, we recognized the fluorophosphonium Lewis acid cation [(C₆F₅)₃PF][B(C₆F₅)₄] has proved to be more Lewis acidic than B(C₆F₅)₃.²¹ This prompted us to examine the corresponding reaction with an excess of three equivalents of P(SiMe₃)₃ in CH₂Cl₂ for 1 h at −35 °C. Pleasingly, this afforded the subsequent isolation of a new species 9 in 88% yield. This product gave rise to ¹⁹F NMR signals at ¹⁹F resonances at −116.5, −129.1, −133.1, −163.9 and −167.7 ppm consistent with the presence of C₆F₄ in the cation and C₆F₅ in the anion. In addition, a P–F doublet was observed at −115.9 ppm with a FP coupling constant of 1050 Hz. The corresponding ³¹P{¹H} NMR showed a resonance at 62.6 ppm with a PF coupling of 1040 Hz as well as signal at −139.4 ppm. These data were consistent with the formulation of 9 as [PF(C₆F₄P(SiMe₃)₂)₃][B(C₆F₅)₄] (Scheme 3). Similar to the parent cation, efforts to use the Gutmann–Beckett method to assess the Lewis acidity of 9 were not possible due to an F/O exchange reaction.²¹

Scheme 3 Synthesis of 9–10.

Like the reaction generating 7, compound 9 was treated with excess methanol giving rise to the new product 10. This resulted in downfield shift of the ³¹P signal to 36.7 ppm and the appears of a triplet resonance at −164.4 ppm with P–H coupling of 215 Hz. The ¹⁹F resonances attributable to the C₆F₄ fragments were observed at −124.9 and −134.5 ppm. Collectively these data were consistent with the formulation of 10 as [(MeO)P(C₆F₄PH₂)₃][B(C₆F₅)₄] (Scheme 3).

Conclusions

Exploiting the steric bulk and the basicity of P(SiMe₃)₃, it has been shown to effect triple para-substitution on the electron-deficient arene rings of B(C₆F₅)₃ and [FP(C₆F₅)₃]⁺, affording B(C₆F₄P(SiMe₃)₂)₃1 and [FP(C₆F₄P(SiMe₃)₂)₃][B(C₆F₅)₄] 9, respectively. Compound 1 remains Lewis acidic at boron forming adducts with sterically unencumbered donors or anions. Both compounds 1 and 9 react further with H₂O or methanol to give the primary phosphine analogs (L)B(C₆F₄PH₂)₃ (L = H₂O 7, MeCN 8) and [(MeO)P(C₆F₄PH₂)₃] [B(C₆F₅)₄] 10, respectively. This synthetic methodology offers a new protocol to the incorporation of phosphine derived aryl substituents on Lewis acid centres. The potential of approach as an avenue tuning Lewis acidity as well as in formation of extended donor–acceptor arrays is the subject of ongoing study.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

NSERC of Canada is thanked for financial support.

References

1. G. C. Welch, R. R. S. Juan, J. D. Masuda and D. W. Stephan, Science, 2006, 314, 1124–1126.
2. G. C. Welch and D. W. Stephan, J. Am. Chem. Soc., 2007, 129, 1880–1881.
3. D. W. Stephan and G. Erker, Angew. Chem., Int. Ed., 2010, 49, 46–76.
4. D. W. Stephan, Acc. Chem. Res., 2015, 48, 306–316.
5. D. W. Stephan, J. Am. Chem. Soc., 2015, 137, 10018–10032.
6. D. W. Stephan and G. Erker, Angew. Chem., Int. Ed., 2015, 54, 6400–6441.
7. D. W. Stephan, Science, 2016, 354, aaf7229.
8. A. R. Jupp and D. W. Stephan, Trends Chem., 2019, 1, 35–48.
9. D. W. Stephan, Chem, 2020, 6, 1520–1526.
10. S. Doering, G. Erker, R. Fröhlich, O. Meyer and K. Bergander, Organometallics, 1998, 17, 2183–2187.
11. S. L. Granville, G. C. Welch and D. W. Stephan, Inorg. Chem., 2012, 51, 4711–4721.
12. K. F. Firth, J. Möbus and D. W. Stephan, Chem. Commun., 2016, 52, 13967–13970.
13. L. Cabrera, G. C. Welch, J. D. Masuda, P. R. Wei and D. W. Stephan, Inorg. Chim. Acta, 2006, 359, 3066–3071.
14. F. Schulz, V. Sumerin, M. Leskelä, T. Repo and B. Rieger, Dalton Trans., 2010, 39, 1920–1922.
15. L. J. Hounjet, C. B. Caputo and D. W. Stephan, Dalton Trans., 2013, 42, 2629–2635.
16. R. J. Andrews and D. W. Stephan, Chem. – Eur. J., 2020, 26, 7194–7198.
17. A. Ramos, A. J. Lough and D. W. Stephan, Chem. Commun., 2009, 1118–1120.
18. C.-W. Chiu, Y. Kim and F. P. Gabbaï, J. Am. Chem. Soc., 2009, 131, 60–61.
19. C. W. Chiu and F. P. Gabbaï, Organometallics, 2008, 27, 1657–1659.
20. K. C. Song, K. M. Lee, N. V. Nghia, W. Y. Sung, Y. Do and M. H. Lee, Organometallics, 2013, 32, 817–823.
21. C. B. Caputo, L. J. Hounjet, R. Dobrovetsky and D. W. Stephan, Science, 2013, 341, 1374–1377.

---

Footnote

† Electronic supplementary information (ESI) available: Synthetic and spectral details. CCDC 2328349–2328350. For ESI and crystallographic data in CIF or other electronic format see DOI: https://doi.org/10.1039/d4dt00251b

---