Increasing the Brønsted acidity of Ph₂PO₂H by the Lewis acid B(C₆F₅)₃. Formation of an eight-membered boraphosphinate ring [Ph₂POB(C₆F₅)₂O]₂

Abstract

Autoprotolysis of the metastable acid (C₆F₅)₃BOPPh₂OH, prepared in situ by the reaction of the rather weak Brønsted acid Ph₂PO₂H with the strong Lewis acid B(C₆F₅)₃, gave rise to the formation of the eight-membered ring [Ph₂POB(C₆F₅)₂O]₂ and C₆F₅H. The conjugate base was isolated as stable sodium crown ether salt [Na(15-crown-5)][Ph₂PO₂B(C₆F₅)₃].

---

Lewis acids can significantly increase the acidity of Brønsted acids.¹ This principle is operative in the prototypical Lewis pair complex (C₆F₅)₃BOH₂, the adduct of the electron pair acceptor B(C₆F₅)₃ and the electron pair donor H₂O.² In MeCN, the acidity of (C₆F₅)₃BOH₂ (pK_a = 8.4) is very similar to that of HCl (pK_a = 8.5).³ Thus, (C₆F₅)₃BOH₂ is a strong acid that readily protonates basic organic⁴ and organometallic compounds.^2,5 Diphenylphosphinic acid, Ph₂PO₂H, is a rather weak acid. As it is well-known that B(C₆F₅)₃ forms Lewis pair complexes with phosphine oxides,⁶ we were curious to study if B(C₆F₅)₃ will also increase the Brønsted acidity of Ph₂PO₂H.

Upon dissolving Ph₂PO₂H and B(C₆F₅)₃ in CDCl₃, multinuclear NMR spectroscopy indeed indicates the formation of a single product that was assigned to (C₆F₅)₃BOPPh₂OH (1) (Scheme 1). The ³¹P NMR spectrum (CDCl₃) of 1 shows signal at δ = 42.1 ppm that differs substantially from that of Ph₂PO₂H (33.9 ppm). The ¹¹B NMR spectrum (CDCl₃)₃ of 1 exhibits a broad signal at δ = −1.3 ppm, which is significantly different from that of B(C₆F₅)₃ (59.0 ppm). Solutions in CDCl₃ show a limited stability and all attempts to isolate 1 by removal of the solvents failed. However, these solutions are stable at r.t. for 2 h; within this time NMR spectroscopy gave no evidence for the formation of other species. While the acid 1 could not be isolated, the reaction of Ph₂PO₂Na, B(C₆F₅)₃ and 15-crown-5 provided the indefinitely stable, conjugate base [Na(15-crown-5)][Ph₂PO₂B(C₆F₅)₃] (2), which was obtained as colourless crystals in 73% yield (Scheme 1). The ³¹P and ¹¹B NMR spectra (THF-d₈) gave signals at δ = 22.3 and −2.7 ppm. The molecular structure of 2 reveals that the [Na(15-crown-5)]⁺ ion and the [Ph₂PO₂B(C₆F₅)₃]⁻ ion are associated by a Na⋯O contact (Fig. 1). When a solution of 1 in CDCl₃ was kept standing for a few hours or heated under reflux for a few minutes NMR spectroscopy indicates the formation of new species, which were identified as the eight-membered boraphosphinate ring [Ph₂POB(C₆F₅)₂O]₂ (3) and C₆F₅H. On a preparative scale, 3 was isolated in 76% yield when a solution of 1 prepared in situ from Ph₂PO₂H and B(C₆F₅)₃, in toluene was heated overnight under reflux (Scheme 1). This reactivity resembles the autoprotolysis of (C₆F₅)₃BOH₂ at elevated temperatures giving rise to the formation of [(C₆F₅)₂BOH]₃ and C₆F₅H.⁷ The eight-membered boraphosphinate ring 3 seems to be the first member of this compound class, however, we note the closely related series of cubic boraphosphonate cages in the literature comprising similar eight-membered ring subunits within the cage structure.⁸ The ³¹P and ¹¹B NMR spectra (CDCl₃) of 3 revealed signals at δ = 37.8 and 6.3 ppm, but no coupling information. The molecular structure of 3 comprises a strongly puckered B₂P₂O₄ ring (puckering factor = 0.890), whereas isolobal eight-membered siloxane rings are usually almost planar (Fig. 2).⁹ The bond parameters of 3 are very similar to those of the cubic boraphosphonate cages.⁸ In a failed attempt to isolate 1 by crystallisation, a small crop of single crystals 4 was isolated, which turned out to be a hydrogen-bonded complex between two molecules of (C₆F₅)₃BOH₂ and the disiloxadiphosphinate [Ph₂P(O)OSiMe₂]₂O. The formation of 4 can be rationalized by the accidental cleavage of silicon grease used to seal the joints and stopcocks (Scheme 2).¹⁰ The facile cleavage of siloxanes is remarkable and points to the high Brønsted acidity of 1. Variation of the stoichiometric ratio of the reactants gave no other product than 4. The O⋯O donor acceptor distances (2.542(5), 2.684(4), 2.681(4), 2.559(4) Å) are indicative of medium strength hydrogen bonding.¹¹ The ³¹P, ²⁹Si and ¹¹B NMR spectra (THF-d₈) of 4 show signals at δ = 32.4, −23.9 and 3.4 ppm. The molecular structure of 4 comprises a novel hydrogen bond motif featuring two BOH₂ hydrogen bond donors and two PO hydrogen acceptors (Fig. 3). The hydrogen bond motif can be described as binary graph set R₄⁴(8)¹² and is strongly reminiscent to that of (Ph₃SiOH)₄,¹³ in which four silanol groups serve as donors and acceptors.

Scheme 1 Formation and reactivity of 1 and its stable sodium salt 2.

Fig. 1 Molecular structure of 2 showing 30% probability ellipsoids and the crystallographic numbering scheme. Selected bond parameters [Å, °]: B1–O1 1.508(2), P1–O1 1.544(1), P1–O2 1.482(1), Na1–O2 2.211(2), B1–O1–P1 134.5(1).

Fig. 2 Molecular structure of 3 showing 30% probability ellipsoids and the crystallographic numbering scheme. Selected bond parameters [Å, °]: B1–O1 1.501(3), B1–O4 1.507(3), B2–O2 1.506(3), B2–O3 1.513(3), P1–O1 1.538(2), P1–O2 1.537(2), P2–O3 1.534(2), P2–O4 1.539(2), B1–O1–P1 129.2(1), B1–O4–P2 129.1(1), B2–O2–P1 128.8(1), B2–O3–P2 134.2(1).

Scheme 2 Reactivity of 1 towards polymeric group 14 oxides (Me₂SiO)_n and (Me₂SnO)_n.

Fig. 3 Molecular structure of 4 showing 30% probability ellipsoids and the crystallographic numbering scheme. Selected bond parameters [Å, °]: B1–O1 1.562(5), B2–O2 1.556(5), P1–O3 1.492(3), P1–O4 1.558(4), P2–O6 1.567(3), P2–O7 1.502(3), Si1–O4 1.650(4), Si1–O5 1.613(4), Si2–O5 1.611(4), Si2–O6 1.668(3), P1–O4–Si1 149.7(2), P2–O6–Si2 143.8(2), Si1–O5–Si2 159.1(3), O1⋯O3 2.542(5), O1⋯O7 2.684(4), O2⋯O3 2.681(4), O2⋯O7 2.559(4).

To provide a quantitative description of the Brønsted acidity increase upon going from Ph₂PO₂H to 1 and to reveal the corresponding electronic structure changes we carried out DFT calculations of these acids and the conjugate bases with use of the Gaussian09 package.¹⁴ The optimized molecular geometries agree well with the experimental data for 2 (Fig. 1), [Ph₂PO₂]⁻ and Ph₂PO₂H¹⁵ (Table S2, see ESI†). The difference in the dissociation enthalpies of Ph₂PO₂H and 1 (eqn (1) and (2)) ΔΔH = ΔH₁ − ΔH₂ is estimated at the M052X/6-31+G** level of theory as 34.0 kcal mol⁻¹ (gas phase) and 14.1 kcal mol⁻¹ (MeCN solution). These values are indicative of much higher Brønsted acidity of 1 as compared to that of the Ph₂PO₂H. Our calculations of atomic charges show that the O–H bond becomes more polar upon going from Ph₂PO₂H to 1 (Table S3, see ESI†). Calculated deformation electron densities (DED) reveal a weakening of the O–H covalent bonding upon coordination of B(C₆F₅)₃ to Ph₂PO₂H (Fig. S35, see ESI†).

These changes in the electronic structures explain the increased Brønsted acidity of 1. The ΔH₁ − ΔH₂ enthalpy change is equal to the ΔH₃ − ΔH₄ difference in the B–O bond dissociation energies in the [Ph₂PO₂B(C₆F₅)₃]⁻ anion and 1 (eqn (3) and (4)).

Ph₂PO₂H ⇄ [Ph₂PO₂]⁻ + H⁺ ΔH₁ (1)

(C₆F₅)₃BOPPh₂OH ⇄ [Ph₂PO₂B(C₆F₅)₃]⁻ + H⁺ ΔH₂ (2)

[Ph₂PO₂B(C₆F₅)₃]⁻ ⇄ [Ph₂PO₂]⁻ + B(C₆F₅)₃ ΔH₃ (3)

(C₆F₅)₃BOPPh₂OH ⇄ Ph₂PO₂H + B(C₆F₅)₃ ΔH₄ (4)

The B–O bond in [Ph₂PO₂B(C₆F₅)₃]⁻ is expected, therefore, to be stronger than that in 1. Indeed, the DED maps (Fig. S36, see ESI†) demonstrate a higher B–O deformation density in the anion. This stabilization of [Ph₂PO₂B(C₆F₅)₃]⁻ also contributes to the higher Brønsted acidity of 1. To compare the acidities of Ph₂PO₂H and 1 with those of other acids we calculated¹⁶ the pK_a values in the gas phase and MeCN solution for a series of 15 compounds with tabulated experimental data in the ranges of pK_a(gas) = 209–251 and pK_a(MeCN) = 0–30 (Tables S4 and S5, see ESI†). On the basis of the linear regressions between experimental and calculated pK_a values (Fig. S37 and S38, see ESI†) the expected pK_a values for Ph₂PO₂H and 1 were found to be, respectively, 239.2 and 214.4 in the gas phase and 20.5 and 9.4 in MeCN solution. The gas-phase acidity of 1 appears to be stronger than that of CF₃SO₃H (pK_a(gas) 219.6)¹⁷ while in MeCN solution 1 is comparable with HCl and tosylic acid (pK_a(MeCN) 8.5³ and 8.6,¹⁸ respectively).

In light of the remarkable siloxane bond cleavage, we have started to elaborate the reactivity of 1 towards other element oxides. Indeed, the reaction of polymeric (Me₂SnO)_n with 1 rapidly occurred at r.t. and produced the eight-membered Sn₂P₂O₄ heterocycle [Me₂Sn(OPPh₂O)₂SnMe₂][HOB(C₆F₅)₃]₂ (5) in 83% yield (Scheme 2). The ¹¹⁹Sn and ³¹P MAS NMR spectra show broad signals at δ = −180.5 ppm and 31.2 ppm. Freshly prepared solutions of phase-pure 5 (checked by powder diffraction) in CDCl₃ shows four ¹¹⁹Sn NMR signals and three ³¹P NMR signals, which point to a reversible dynamic process that is not yet understood in full detail (see ESI†). Similar solution behaviour was observed for the related heterocycles [R₂Sn(OPPh₂O)₂SnR₂][O₃SCF₃]₂ (R = Ph, t-Bu), which were obtained by the reaction of (Ph₂SnO)_n or (t-Bu₂SnO)₃ with Ph₂PO₂H and triflic acid.¹⁹ On a longer time scale (several weeks) 5 shows signs of irreversible decomposition in solution and in the solid-state. In both states, the same unassigned decomposition product with a ¹¹⁹Sn chemical shift of δ = 71.1 ppm slowly forms. The molecular structure of 5 contains a strongly puckered Sn₂P₂O₄ ring (puckering factor = 0.888)⁹ that resembles that of the slightly less puckered [t-Bu₂Sn(OPPh₂O)₂Snt-Bu₂][O₃SCF₃]₂ (puckering factor = 0.921) (Fig. 4).¹⁹ The spatial arrangement of the Sn atoms is distorted trigonal bipyramidal (geometrical goodness = 89.7°)²⁰ and defined by a C₂O₃ donor set. The Sn–O bond lengths within the ring (2.040(2) and 2.156(2) Å) are shorter than that of the exocyclic HOB(C₆F₅)₃ moiety (2.231(2) Å). The same trend was observed for [t-Bu₂Sn(OPPh₂O)₂Snt-Bu₂][O₃SCF₃]₂,¹⁹ in which the endocyclic Sn–O bonds (2.045(3) and 2.173(4) Å) are shorter than the Sn–O bond length related with the triflate moiety (2.303(1) Å). It might be speculated that the longer Sn–O bonds are subject to electrolytic dissociation, which could explain the dynamic behaviour in solution. We finally studied the reactivity of 1 towards Ph₄Sn, which proceeded with facile phenyl group cleavage providing Ph₃SnOPPh₂OB(C₆F₅)₃ in 86% yield (Scheme 3). This reaction closely resembles the quantitative reaction of Ph₄Sn with triflic acid giving rise to the formation Ph₃SnO₃SCF₃.²¹ The ¹¹⁹Sn NMR spectrum (CDCl₃) of 6 shows a doublet centred at δ = −59.6 ppm with a ²J(¹¹⁹Sn–O–³¹P) coupling of 146 Hz, which suggests that the Sn atoms are tetracoordinate in solution (Fig. 5). In the solid-state, 6 comprises a 1D coordination polymer with distorted trigonal bipyramidal Sn atoms (geometrical goodness = 51.6°)²⁰ defined by a C₃OF donor set.

Fig. 4 Molecular structure of 5 showing 30% probability ellipsoids and the crystallographic numbering scheme. Selected bond parameters [Å, °]: B1–O1 1.512(3), P1–O2 1.538(2), P1–O3a 1.515(2), Sn1–O1 2.231(2), Sn1–O2 2.040(2), Sn1–O3a 2.156(2), B1–O1–Sn1 133.5(2), P1–O2–Sn1 139.5(1), P1–O3a–Sn1a 138.1(1).

Scheme 3 Phenyl group cleavage in Ph₄Sn using 1.

Fig. 5 Molecular structure of 6 showing 30% probability ellipsoids and the crystallographic numbering scheme. Selected bond parameters [Å, °]: B1–O1 1.527(2), P1–O1 1.521(1), P1–O2 1.527(1), Sn1–O2 2.058(1), Sn1–F23a 3.392(3), B1–O1–P1 139.9(1), P1–O2–Sn1 142.29(7).

The Brønsted acidity of Ph₂PO₂H was significantly increased upon addition of the Lewis acid B(C₆F₅)₃ giving rise to (C₆F₅)₃BOPPh₂OH (1) in solution. Unlike its conjugate base [Na(15-crown-5)][Ph₂PO₂B(C₆F₅)₃] (2), the acid 1 is thermally unstable and undergoes autoprotolysis and formation of the boraphosphinate ring [Ph₂POB(C₆F₅)₂O]₂ (3) and C₆F₅H. Despite its limited life span, 1 can be used for synthetic purposes, as was demonstrated for two examples from organotin chemistry. The stable water adduct (C₆F₅)₃BOH₂ is known to bind up to two additional water molecules via hydrogen bonding, e.g. (C₆F₅)₃BOH₂·2H₂O,²² which adversely affects the stoichiometric control of protonation reactions. Moreover, the various related anions, e.g. [HOB(C₆F₅)₃]⁻, [HO{B(C₆F₅)₃}₂]⁻ and [O{B(C₆F₅)₃}₂]²⁻,^2,4 suggest that hydroxide and oxide ions may be also transferred upon protonation. These adverse properties have not been observed for 1. We are currently investigating if the acidity of other Brønsted acids, such sulfinic and sulfonic acids, may be also increased by applying the same concept.

The Deutsche Forschungsgemeinschaft (DFG) is gratefully acknowledged for financial support. The theoretical part of this work was supported by the Russian Science Foundation (Project 14-13-00832). We thank Dr Daniel Himmel (Universität Freiburg) for an invaluable discussion about pK_a value calculations.

Notes and references

1. Superacid Chemistry, ed. G. A. Olah, G. K. S. Prakash, A. Molnár and J. Sommer, John Wiley & Sons, Inc., Hoboken, New Jersey, 2nd edn, 2009.
2. A. A. Danopoulos, J. R. Galsworthy, M. L. H. Green, S. Cafferkey, L. H. Doerrer and M. B. Hursthouse, Chem. Commun., 1998, 2529.
3. C. Bergquist, B. M. Bridgewater, C. J. Harlan, J. R. Norton, R. A. Friesner and G. Parkin, J. Am. Chem. Soc., 2000, 122, 10581.
4. (a) R. Duchateau, R. A. van Santen and G. P. A. Yap, Organometallics, 2000, 19, 809; (b) R. T. Stibrany and P. Brant, Acta Crystallogr., Sect. C: Cryst. Struct. Commun., 2001, 57, 644; (c) M. J. Drewitt, M. Niedermann and M. C. Baird, Inorg. Chim. Acta, 2002, 340, 207; (d) A. Di Saverio, F. Focante, I. Camurati, L. Resconi, T. Beringhelli, G. D'Alfonso, D. Donghi, D. Maggioni, P. Mercandelli and A. Sironi, Inorg. Chem., 2005, 44, 5030; (e) F. Focante, I. Camurati, L. Resconi, S. Guidotti, T. Beringhelli, G. D'Alfonso, D. Donghi, D. Maggioni, P. Mercandelli and A. Sironi, Inorg. Chem., 2006, 45, 1683; (f) F. Focante, P. Mercandelli, A. Sironi and L. Resconi, Coord. Chem. Rev., 2006, 250, 170.
5. (a) G. S. Hill, L. Manojlovic-Muir, k. W. Muir and R. J. Puddephatt, Organometallics, 1997, 16, 525; (b) L. Doerrer and M. L. H. Green, J. Chem. Soc., Dalton Trans., 1999, 4325; (c) D. Neculai, H. W. Roesky, A. M. Neculai, J. Magull, B. Walfort and D. Stalke, Angew. Chem., Int. Ed., 2002, 41, 4294; (d) G. Schatte, T. Chivers, H. M. Tuononen, R. Suontamo, R. Laitinen and J. Valkonen, Inorg. Chem., 2005, 44, 443; (e) J. Sánchez-Nieves, L. M. Frutos, P. Royo, O. Castaño and E. Herdtweck, Organometallics, 2005, 24, 2004; (f) J. Sánchez-Nieves, L. M. Frutos, P. Royo, O. Castaño, E. Herdtweck and M. E. G. Mosquera, Inorg. Chem., 2010, 49, 10642.
6. (a) M. A. Beckett, D. S. Brassington, S. J. Coles and M. B. Hursthouse, Inorg. Chem. Commun., 2000, 3, 530; (b) M. A. Beckett, D. S. Brassington, M. E. Light and M. B. Hursthouse, J. Chem. Soc., Dalton Trans., 2001, 1768; (c) A. J. P. Cardenas, B. J. Culotta, T. H. Warren, S. Grimme, A. Stute, R. Fröhlich, G. Kehr and G. Erker, Angew. Chem., Int. Ed., 2011, 50, 7567.
7. (a) J. Tian, S. Wang, Y. Feng, J. Li and S. Collins, J. Mol. Catal. A: Chem., 1999, 144, 137; (b) T. Beringhelli, G. D'Alfonso, D. Donghi, D. Maggioni, P. Mercandelli and A. Sironi, Organometallics, 2003, 22, 1588; (c) A. Y. Timoshkin and G. Frenking, Organometallics, 2008, 27, 371.
8. (a) K. Diemert, U. Englert, W. Kuchen and F. Sandt, Angew. Chem., Int. Ed., 1997, 36, 241; (b) M. G. Walawalkar, R. Murugavel, H. W. Roesky and H.-G. Schmidt, Organometallics, 1997, 16, 516; (c) M. G. Walawalkar, R. Murugavel, H. W. Roesky and H.-G. Schmidt, Inorg. Chem., 1997, 36, 4202; (d) J. Mortier, I. D. Gridnev and P. Guénot, Organometallics, 2000, 19, 4266; (e) J. Tönnemann, R. Scopelliti, K. O. Zhurov, L. Menin, S. Dehnen and K. Severin, Chem. – Eur. J., 2012, 18, 9939–9945; (f) Q. Liu, N. D. Contrella, A. S. Filatov and R. F. Jordan, Organometallics, 2015, 34, 254.
9. J. Beckmann, D. Dakternieks, A. E. K. Lim, K. F. Lim and K. Jurkschat, THEOCHEM, 2006, 761, 177.
10. I. Haiduc, Organometallics, 2004, 23, 3.
11. T. Steiner, Angew. Chem., Int. Ed., 2002, 41, 49.
12. J. Bernstein, R. E. Davis, L. Shimoni and N.-L. Chang, Angew. Chem., Int. Ed., 1995, 34, 1555.
13. K. F. Bowes, C. Glidewell and J. N. Low, Acta Crystallogr., Sect. C: Cryst. Struct. Commun., 2002, 58, o409.
14. M. J. Frisch, et al., Gaussian09, Revision B.01, Gaussian, Inc., Wallingford, CT, 2010. See ESI† for the computational details.
15. (a) J. Guo, W.-K. Wong and W.-Y. Wong, Polyhedron, 2005, 24, 927; (b) K. A. Lyssenko, G. V. Grintselev-Knyazev and M. Yu. Antipin, Mendeleev Commun., 2002, 12, 128.
16. pK_a(gas) were calculated on the basis of the M052X/6-31+G** sums of the electronic and thermal Gibbs free energies; pK_a(MeCN) were estimated at the same level of theory using the SMD solvation model: A. V. Marenich, C. J. Cramer and D. G. Truhlar, J. Phys. Chem. B, 2009, 113, 6378 . See ESI† for the details of calculations.
17. pK_aexp(gas) were obtained from the gas-phase acidity (GA) values: J. E. Bartmess, in NIST Chemistry WebBook, NIST Standard Reference Database Number 69, ed. P. J. Linstrom and W. G. Mallard, NIST, Gaithersburg, MD, 20899, http://webbook.nist.gov.
18. A. Kütt, I. Leito, I. Kaljurand, L. Soovali, V. M. Vlasov, L. M. Yagupolskii and I. A. Koppel, J. Org. Chem., 2006, 71, 2829.
19. J. Beckmann, D. Dakternieks, A. Duthie and C. Mitchell, Organometallics, 2003, 22, 2161.
20. U. Kolb, M. Beuter and M. Dräger, Inorg. Chem., 1994, 33, 4522.
21. (a) M. Schmeisser, P. Sartori and B. Lippsmeier, Chem. Ber., 1970, 103, 868; (b) J. Beckmann, E. Lork and O. Mallow, Main Group Met. Chem., 2012, 35, 183.
22. X. Wang and P. P. Power, Angew. Chem., Int. Ed., 2011, 50, 10965.

---

Footnote

† Electronic supplementary information (ESI) available: Experimental section, NMR spectra, X-ray crystallography, X-ray powder diffraction, computational details, additional references. CCDC 1480495 and 1411098–1411113. For ESI and crystallographic data in CIF or other electronic format see DOI: 10.1039/c6cc06102h

---