An efficient, scalable synthesis of ferrocenylphosphine and dichloroferrocenylphosphine

Abstract

A new synthetic route to FcPH₂ and FcPCl₂ (Fc = ferrocenyl) is presented. This method avoids the challenging monolithiation of ferrocene, as well as any tedious purification steps. All reactions are high yielding and easily conducted on a relatively large scale, using economical and commercially available synthetic precursors.

---

Since its discovery in the 1950s,^1,2 which kickstarted research into the field of organometallic chemistry,³ ferrocene has cemented itself as one of the most important organometallic compounds due to its high stability and the relative ease with which it can be functionalised. The Cambridge structural database currently contains over 2500 entries for phosphorus functionalised ferrocene derivatives. These include strained ferrocenophanes serving as precursors for ring opening polymerisation,^4,5 and chiral phosphine ligands with a ferrocene backbone.⁶ Ferrocene is an interesting group to introduce due to its redox activity and strong electron donating properties, which allow it to stabilise electron deficient species. In fact, the first “all-carbon” phosphenium cation to be isolated was stabilised by two Fc substituents.⁷

Despite all of this, the archetypal ferrocenyl phosphorus synthons, FcPH₂ and FcPCl₂, remain significantly underutilised. This is due largely to a lack of convenient synthetic routes to these molecules. So far, the main hurdle appeared to be the monolithiation of ferrocene, which is notoriously challenging and requires the use of highly pyrophoric tBuLi.^8–10

Arguably, the best synthesis of FcPCl₂ was published by Pietschnig and Niecke in 1997.¹¹ In their protocol (Scheme 1, top) ferrocene is monolithiated with tBuLi, followed by the addition of (Et₂N)₂PCl. The resulting FcP(NEt₂)₂ is converted to FcPCl₂ by reaction with excess PCl₃. The work up for the reaction is rather laborious, involving removal of excess PCl₃ and the Et₂NPCl₂ byproduct by distillation, then vacuum distillation of the FcPCl₂ to remove the various ferrocene containing sideproducts. At 52%, the yield is acceptable but not ideal, and the use of tBuLi makes the reaction dangerous to conduct on a large scale. This problem is further exacerbated by the need for careful control of the lithiation reaction conditions.

Scheme 1 Previous syntheses of FcPCl₂.^11–13

Another literature method utilises the Friedel–Crafts reaction of ferrocene with Me₂NPCl₂ (Scheme 1, bottom).^12,13 The yields are low (11%) and repeated attempts to improve on this reaction demonstrated the yields to be highly variable and often much lower.^14,15 Nonetheless, this method was still being used relatively recently.^15,16

The most recent synthesis of ferrocenylphosphine, FcPH₂, was reported by Wright and co-workers in 2009. In this procedure diethyl ferrocenylphosphonate is reduced by LiAlH₄/Me₃SiCl.¹⁷ The diethyl ferrocenylphosphonate was itself prepared by monolithiation of ferrocene, followed by reaction with chlorodiethylphosphate (Scheme 2).¹⁸ FcPH₂ was obtained in relatively good overall yields (50–60%), but the procedure requires the use of tBuLi and column chromatography, which hinders scale-up of the reaction.

Scheme 2 Recent synthesis of FcPH₂.^17,18

In all of the lithiation based syntheses, the biggest issue has been the selective addition of a single phosphorus moiety to the ferrocene ring, avoiding the disubstituted species formed through overlithiation to 1,1′-dilithioferrocene. An elegant bypass of the lithiation strategy was identified in the regiospecific reaction of ferrocene with P₄S₁₀, which yields the perthiophosphonic anhydride Fc₂P₂S₄ (Scheme 3). The reaction is accomplished by refluxing the two components in xylenes, with the air stable product being collected by filtration in >80% yield.¹⁹ The reaction is relatively facile, and has been scaled to ca. 250 g ferrocene with little difficulty.

Scheme 3 Two-step synthesis of FcPH₂ from ferrocene via the perthiophosphonic anhydride Fc₂P₂S₄. Also shown is the subsequent quantitative transformation into FcPCl₂.

Perthiophosphonic anhydrides (general formula R₂P₂S₄) are well known in the literature, with the most famous example being Lawesson's reagent²⁰ (R = 4-MeOC₆H₄). Although the reactivity of selected perthiophosphonic anhydrides has been studied in some detail,²¹ their use as thionation reagents has been most prominent.^22,23 Nevertheless, no reports on reductions of such species with hydridic reagents have appeared in the literature to date. We have found that reduction of Fc₂P₂S₄ using LiAlH₄ gave the primary phosphine FcPH₂ as a low melting solid in 79% yield (66% overall yield, Scheme 3). This reaction was conducted on a 100 g scale without issue, something which would be very difficult and potentially dangerous to do using tBuLi as in the Wright synthesis.^17,18 Crucially, the reaction proceeds cleanly when 4 eq. LiAlH₄ are used per Fc₂P₂S₄. The FcPH₂ obtained was found to be exceptionally pure by microanalysis, as well as ¹H, ³¹P and ¹³C NMR spectroscopy, with no need for further purification.

With an efficient synthesis of FcPH₂ in hand, it was subsequently found that FcPH₂ can be chlorinated by triphosgene (considered a safer substitute for phosgene gas) or phosgene (commercially supplied as a solution in toluene) to give FcPCl₂ in quantitative yield (Scheme 3). This reaction is simple to conduct, although an efficient fumehood is necessary. The reaction is completed in a few hours with either chlorinating reagent. Removal of volatiles in vacuo affords FcPCl₂ as a red-brown oil, which solidifies on standing at room temperature. The purity of the compound, as determined by microanalysis and multinuclear NMR spectroscopy, was exceptionally high with no need for further purification. This represents a marked improvement over the Pietschnig synthesis which, in addition to requiring purification by vacuum distillation, gave FcPCl₂ in a lower yield (52% vs. 65% overall yield). Although the toxicity of phosgene/triphosgene is a concern in this procedure, we feel these compounds are easier to handle on a large scale than the pyrophoric tBuLi, and have repeatedly carried out this transformation on a relatively large scale (ca. 60 g FcPH₂).

Somewhat surprisingly, the crystal structures of FcPH₂ and FcPCl₂ have not been reported previously.²⁴ Using the pure material obtained via our procedure, crystals suitable for X-ray diffraction were obtained with relative ease. The structures are shown in Fig. 1,^25,26 both compounds are as expected, with P–C bond lengths of 1.818(9) Å for FcPH₂ and 1.778(4) [1.784(4)] Å for FcPCl₂.²⁷

Fig. 1 Solid state structures of FcPH₂ (top) and FcPCl₂ (bottom). Thermal ellipsoids are drawn with 50% probability. Carbon bound H atoms and second molecule in asymmetric unit (FcPCl₂) are omitted for clarity. Selected bond lengths (Å) and angles (°): FcPH₂: C1–P1 1.818(9); FcPCl₂: C1–P1 1.778(4) [1.784(4)], Cl1–P1–Cl2 97.95(8) [98.34(7)].²⁷

Conclusions

The electrophilic substitution reaction of ferrocene with P₄S₁₀ has been used to introduce a single phosphorus moiety onto ferrocene with complete regiospecificity and in high yield. Subsequently, the two archetypal phosphorus synthons FcPH₂ and FcPCl₂ were obtained in high yields and purity through the reduction of Fc₂P₂S₄ with LiAlH₄, followed by chlorination with triphosgene/phosgene. By avoiding the difficult monolithiation of ferrocene and associated use of pyrophoric tBuLi, it was possible to scale this new synthetic route to multi-gram quantities (ca. 60 g).

Experimental

Fc₂P₂S₄

A slight modification of the literature procedure was employed.¹⁹ Ferrocene (37.7 g, 202 mmol) and P₄S₁₀ (20.0 g, 45.0 mmol) were heated under reflux in xylenes (300 mL) for 3 hours before being allowed to cool. Fc₂P₂S₄ was collected by filtration as an air stable, dark orange solid, which was washed with toluene (2 × 100 mL) and diethyl ether (2 × 50 mL) and dried in vacuo. Yield 41.8 g (75.0 mmol, 83%).

FcPH₂

To a vigorously stirred suspension of Fc₂P₂S₄ (6.00 g, 10.7 mmol) in diethyl ether (70 mL), cooled to 0 °C, a suspension of LiAlH₄ (1.62 g, 42.7 mmol) in diethyl ether (30 mL) was added dropwise via cannula and the mixture allowed to warm to RT. Insoluble by-products were removed by filtration and washed with diethyl ether (2 × 20 mL). The orange filtrate and washings were cooled to 0 °C, degassed water (1 mL) was added cautiously and the resulting suspension was warmed to RT. The solid formed was removed by filtration and washed with DCM (2 × 20 mL). The filtrate and washings were collected and volatiles removed in vacuo to yield FcPH₂ as a dark orange, low melting solid (3.67 g, 16.8 mmol, 79%). Crystals suitable for single crystal X-ray diffraction were obtained by melting and recrystallisation of the solid between 25–40 °C.²⁵ Mp 36–37 °C. Found: C 55.16; H 5.25. Calc. for C₁₀H₁₁FeP: C 55.09; H 5.09. IR (thin layer) ν_max/cm⁻¹ 3094s (νCH), 2262s (νPH), 1206s, 1024s, 817vs, 495s. ¹H NMR δ_H (270 MHz; CDCl₃), 4.31–4.29 (4H, m, 4 × CpH), 4.20 (5H, s, 5 × CpH), 3.86 (2H, d, ¹J_HP = 203 Hz, PH₂). ¹³C{¹H} NMR (75 MHz; CDCl₃), 75.7 (d, ²J_CP = 14.0 Hz, CH, subst. Cp ring), 70.7 (d, ³J_CP = 3.6 Hz, CH, subst. Cp ring), 69.3 (s, CH, unsubst. Cp ring), 64.1 (d, ¹J_CP = 4.7 Hz, qC). ³¹P NMR δ_P (121 MHz; CDCl₃), −143.6 (t, ¹J_PH = 203 Hz). ³¹P{¹H} NMR δ_P (109 MHz; CDCl₃), −143.8 (s).

FcPCl₂

To a stirred solution of FcPH₂ (3.00 g, 13.8 mmol) in dichloromethane (70 mL), cooled to −10 °C, a solution of triphosgene (2.7 g, 9.2 mmol) in dichloromethane (50 mL) was added slowly via cannula. Alternatively, a solution of phosgene (15.2 mL of a 1.9 M solution in toluene, 28.9 mmol) was added dropwise over 30 min. The mixture was allowed to warm to RT and stirred overnight. Volatiles were removed in vacuo to give FcPCl₂ as a red-brown oil which solidified on standing at room temperature (3.93 g, 13.7 mmol, 99%). Crystals suitable for single crystal X-ray diffraction were grown from toluene at −35 °C.²⁶ Found: C 41.98; H 3.24. Calc. for C₁₀H₉Cl₂FeP: C 41.86; H 3.16. IR (thin layer) ν_max/cm⁻¹ 3099s (νCH), 1411s, 1163s, 1028s, 825s, 446vs. ¹H NMR δ_H (400 MHz; CDCl₃), 4.61–4.59 (4H, m, 4 × CpH), 4.30 (5H, s, 5 × CpH). ¹³C{¹H} NMR (101 MHz; CDCl₃), 79.5 (d, ¹J_CP = 52.3 Hz, qC), 73.3 (d, J_CP = 4.5 Hz, CH, subst. Cp ring), 71.1 (d, J_CP = 22.5 Hz, CH, subst. Cp ring), 70.0 (s, CH, unsubst. Cp ring). ³¹P NMR δ_P (162 MHz; CDCl₃) 164.6 (s). ³¹P{¹H} NMR δ_P (162 MHz; CDCl₃) 164.7 (s).

Acknowledgements

This work was financially supported by the Engineering and Physical Sciences Research Council (EPSRC) and by COST action (grant CM1302, SIPs).

Notes and references

1. T. J. Kealy and P. L. Pauson, Nature, 1951, 168, 1039–1040.
2. G. Wilkinson, M. Rosenblum, M. C. Whiting and R. B. Woodward, J. Am. Chem. Soc., 1952, 74, 2125–2126.
3. H. Werner, Angew. Chem., Int. Ed., 2012, 51, 6052–6058.
4. C. H. Honeyman, D. A. Foucher, F. Y. Dahmen, R. Rulkens, A. J. Lough and I. Manners, Organometallics, 1995, 14, 5503–5512.
5. A. D. Russell, R. A. Musgrave, L. K. Stoll, P. Choi, H. Qiu and I. Manners, J. Organomet. Chem., 2015, 784, 24–30.
6. Chiral Ferrocenes in Asymmetric Catalysis: Synthesis and Applications, ed. L.-X. Dai and X.-L. Hou, Wiley-VCH, 1st edn, 2010.
7. S. G. Baxter, R. L. Collins, A. H. Cowley and S. F. Sena, J. Am. Chem. Soc., 1981, 103, 714–715.
8. D. Guillaneux and H. B. Kagan, J. Org. Chem., 1995, 60, 2502–2505.
9. R. Sanders and U. T. Mueller-Westerhoff, J. Organomet. Chem., 1996, 512, 219–224.
10. F. Rebiere, O. Samuel and H. B. Kagan, Tetrahedron Lett., 1990, 31, 3121–3124.
11. R. Pietschnig and E. Niecke, Bull. Soc. Chim. Fr., 1997, 134, 605–608.
12. G. P. Sollott and W. R. Peterson, J. Organomet. Chem., 1965, 4, 491–493.
13. G. P. Sollott and W. R. Peterson, J. Organomet. Chem., 1969, 19, 143–159.
14. G. R. Knox, P. L. Pauson and D. Willison, Organometallics, 1992, 11, 2930–2933.
15. S. Parveen, P. Kilian, A. M. Z. Slawin and J. D. Woollins, Dalton Trans., 2006, 2586–2590.
16. N. Nagahora, T. Sasamori, N. Takeda and N. Tokitoh, Organometallics, 2005, 24, 3074–3080.
17. R. J. Less, V. Naseri and D. S. Wright, Organometallics, 2009, 28, 1995–1997 . See also W. Henderson and S. R. Alley, J. Organomet. Chem., 2002, 656, 120–128.
18. O. Oms, F. Maurel, F. Carré, J. Le Bideau, A. Vioux and D. Leclercq, J. Organomet. Chem., 2004, 689, 2654–2661.
19. M. R. St J. Foreman, A. M. Z. Slawin and J. D. Woollins, J. Chem. Soc., Dalton Trans., 1996, 3653–3657.
20. B. S. Pedersen, S. Scheibye, N. H. Nilsson and S.-O. Lawesson, Bull. Soc. Chim. Belg., 1978, 87, 223–228.
21. P. Kilian, J. Marek, R. Marek, J. Touzin, O. Humpa and J. D. Woollins, J. Chem. Soc., Dalton Trans., 1998, 1175–1180.
22. T. Ozturk, E. Ertas and O. Mert, Chem. Rev., 2007, 107, 5210–5278.
23. M. Jesberger, T. P. Davis and L. Barner, Synthesis, 2003, 1929–1958.
24. The structure of the borane adduct (FcPH₂·BH₃) has been reported recently, S. Pandey, P. Lönnecke and E. Hey-Hawkins, Eur. J. Inorg. Chem., 2014, 2456–2465.
25. Crystal data for FcPH₂: C₁₀H₁₁FeP, M_r = 218.01, monoclinic, P21/c, a = 12.658(10), b = 7.528(5), c = 10.432(8) Å, β = 112.071(18)°, V = 921.2(12) ų, T = −180(1) °C, Z = 4, reflections collected 5398, independent 1662 (R_int = 0.1024), R(obs) = 0.0944, wR (all data) = 0.2629. Deposition number CCDC 812314†.
26. Crystal data for FcPCl₂: C₁₀H₉Cl₂FeP, M_r = 286.91, monoclinic, P21/n, a = 12.9787(13), b = 13.2924(12), c = 13.6674(13) Å, β = 110.613(8)°, V = 2206.9(4) ų, T = −100(1) °C, Z = 8, reflections collected 14 488, independent 4007 (R_int = 0.0499), R(obs) = 0.0472, wR (all data) = 0.0779. Deposition number CCDC 1437271†.
27. Value in square brackets for second molecule of FcPCl₂ in the asymmetric unit.

---

Footnote

† Electronic supplementary information (ESI) available. CCDC 812314 and 1437271. For ESI and crystallographic data in CIF or other electronic format see DOI: 10.1039/c5ra27594f

---