Mild electrophilic trifluoromethylation of secondary and primary aryl- and alkylphosphines using hypervalent iodine(III)–CF₃ reagents

Abstract

A direct, mild and efficient trifluoromethylation of primary and secondary phosphines is achieved with easily accessible, cheap hypervalent iodine compounds acting as electrophilic CF₃-transfer reagents.

---

Organophosphorus compounds containing a trivalent P-centre are of great importance in many areas of chemistry—in particular as ligands in transition-metal complexes. The introduction of small, strongly electron withdrawing substituents at phosphorus is an important aspect since it allows the bonding behavior of the P-donor atom towards the metal atom in a complex to be altered. Thus, the introduction of a CF₃ fragment on phosphorus(III) is interesting for the reason that the smallest member of the perfluoroalkyl series leads to phosphine ligands with altered σ-donating and π-accepting properties. Such phosphines can thus serve as surrogates of CO, NO or PF₃ from an electronic point of view at the expense of an increased steric demand.¹ The direct, late stage introduction of a CF₃ functional group at phosphorus is a task with little precedence in the literature. The first reports on the synthesis of R₂P(CF₃) date back to the middle of the last century and utilize trifluoroiodomethane in a radical reaction with suitable phosphorus precursors such as Me₃P ² or Ph₂P–PPh₂.³ The concomitantly formed equimolar amount of Me₄P⁺ I⁻ or Ph₂PI and the handling of gaseous ICF₃ render this method problematic, inefficient and somewhat limited with respect to generality.

Alternatively, nucleophilic trifluoromethylation reactions using Me₃SiCF₃ (known as Ruppert–Prakash reagent) have been reported for P-fluorophosphines and -phosphates,⁴P-cyano phosphines,⁵ and fluorinated phosphazenes.⁶ These methods require rather uncommon starting materials which are not commercially available and have to be accessed in multistep procedures. The synthesis of P(CF₃)₃ from cheap P(OPh)₃ and Me₃SiCF₃ rather represents an exception.⁷

Umemoto has shown that S-trifluoromethyl dibenzothiophenium salts as electrophilic trifluoromethylating reagents give P-trifluoromethyltriphenylphosphonium salts starting from triphenylphosphine.⁸ However, the same approach has never been applied to the synthesis of tertiary phosphines from the corresponding secondary derivatives.

We recently succeeded in accessing a new class of electrophilic trifluoromethylating reagents based on hypervalent λ³-organoiodine exhibiting good to excellent reactivity toward several classes of nucleophiles, such as β-keto esters, α-nitro esters and mercaptanes.⁹ Two members of this reagent class, 1 and 2 (Scheme 1), both derived from 2-iodobenzoic acid, are typically applied in electrophilic trifluoromethylations in our laboratory.

Scheme 1 Left: Electrophilic trifluoromethylation reagents 1 and 2. Right: Trifluoromethylation of P(III)-centres using λ³-iodanes 1 or 2.

We have found that these reagents are suited for the formal exchange of a H⁺ with CF₃⁺ at the phosphorus atom of a phosphine, in close analogy to the synthesis of trifluoromethylthioethers from thiols. Thus, mixing equimolar amounts of either one of the reagents 1–2 and diphenylphosphine at ambient or at low temperature (−78 °C) in CH₂Cl₂ gave the desired diphenyl(trifluoromethyl)phosphine (7) in 78% or 74% isolated yield after purification by column chromatography (Table 1, Entries 3 and 4). Similarly, the more basic, nucleophilic and sterically demanding dicyclohexylphosphine reacted readily under the same reaction conditions and was isolated as the corresponding phosphine sulfide (6) in 52% isolated yield after stirring with S₈ to avoid rapid oxidation during work-up and isolation (Entry 1). Interestingly, in addition to diphenylphosphine, the corresponding P-trimethylsilylated derivative¹⁰ underwent trifluoromethylation under the same reaction conditions in comparable yield (7, 69%, Entries 9–11). As a side product, trimethylsilyl 2-iodobenzoate was detected by ¹H and ¹³C{¹H} NMR spectroscopy. Increasing the steric bulk on the ortho-position of the aryl substituents of the phosphine resulted in diminished product yields with both reagents (Table 1, Entries 12 and 13). Notably, the corresponding lithium and potassium phosphides (MPPh₂, M = Li, K) did only produce trace amounts of the trifluoromethylated product as observed by ¹⁹F NMR spectroscopy. Interestingly, mixing primary phosphines such as phenyl- or cyclohexylphosphine with equimolar amounts of 1 in CD₂Cl₂ at ambient temperature resulted in the formation of the corresponding monotrifluoromethylated phosphines exclusively (Table 1, Entries 2 and 8). This constitutes a direct and very convenient synthesis of such secondary racemic P-trifluoromethylated phosphines. Experimental details are provided as Electronic Supplementary Information.†

Table 1 Electrophilic trifluoromethylation of phosphines using 1 or 2

	Substrate	Conditions	Product	Yield^a
a Isolated yields, unless otherwise stated. b S8 as oxidant. c A solution of the phosphine was added to a solution of 2 over the course of 45 min. d Conversion calculated based on ^19F NMR spectroscopy with PhCF3 as internal reference. e Sum of CyPH(CF3) and CyPH2(CF3)^+.
1	Cy2PH	2, CH2Cl2, −78 °C → r.t.	Cy2(CF3)P=S^b	52%
			(6)
2	CyPH2	1, CD2Cl2, r.t.	CyPH(CF3)	54%^d^,^e
			(15)
3	Ph2PH	1, CH2Cl2, r.t.	Ph2P(CF3)	78%
			(7)
4	Ph2PH	2, CH2Cl2, −78 °C → r.t.	Ph2P(CF3)	74%
			(7)
5	Ph2PH	2, MeOH, −78 °C → r.t.	Ph2P(CF3)	65%
			(7)
6	Ph2PH	2, MeCN, −78 °C → r.t.	Ph2P(CF3)	70%
			(7)
7	Ph2PH	2, Toluene, −78 °C → r.t.	Ph2P(CF3)	55%
			(7)
8	PhPH2	1, CD2Cl2, r.t.	PhPH(CF3)	84%^d
			(14)
9	Ph2P(SiMe3)	1, CD2Cl2, r.t.	Ph2P(CF3)	92%^d
			(7)
10	Ph2P(SiMe3)	2, CH2Cl2, −78 °C → r.t.	Ph2P(CF3)	69%
			(7)
11	Ph2P(SiMe3)	2, CD2Cl2, r.t.	Ph2P(CF3)	66%
			(7)
12	(o-Tol)2PH	1, CH2Cl2, r.t.	(o-Tol)2P(CF3)	48%
			(8)
13	(o-Tol)2PH	2, CH2Cl2, −78 °C → r.t.	(o-Tol)2P(CF3)	50%
			(8)
14	(p-Tol)2PH	1, CH2Cl2, r.t.	(p-Tol)2P(CF3)	78%
			(9)
15	(p-Tol)2PH	2, CH2Cl2, −78 °C → r.t.	(p-Tol)2P(CF3)	70%
			(9)
16	(p-Tol)2PH	2, CH2Cl2, 0 °C	(p-Tol)2P(CF3)	66%
		Syringe pump^c	(9)
17	(β-Np)2PH	1, CH2Cl2, 0 °C → r.t.	(β-Np)2P(CF3)	58%
			(10)
18	(β-Np)2PH	2, CH2Cl2, −78 °C → r.t.	(β-Np)2P(CF3)	53%
			(10)
19		1, CH2Cl2, r.t.		41%
20		2, CH2Cl2, −78 °C → r.t.		44%
21	(p-OMePh)2PH	1, CH2Cl2, r.t.	(p-OMePh)2P(CF3)	36%
			(12)
22	(p-OMePh)2PH	2, CH2Cl2, −78 °C → r.t.	(p-OMePh)2P(CF3)	58%
			(12)
23	rac-(o-anisyl)PhPH	2, CH2Cl2, −78 °C → r.t.	rac-(o-anisyl)PhP(CF3)	63%
			(13)

---

At present, the chromatographic separation of the trifluoromethylphosphines from the byproduct deriving from the reagents is not yet optimal and explains the relatively low yields of isolated product in some cases.

From a mechanistic point of view, it seems reasonable to exclude the involvement of phosphides as intermediates. The two reagents 1 and 2 generate bases of very different strengths (a carboxylate and an alcoholate, respectively) and yet afford similar yields. This interpretation is further supported by the fact that, as mentioned above, neither Li or K diphenylphosphide leads to any significant product formation. Furthermore, it can be speculated that a radical pathway might be operating, based on the observation that the reaction of Cy₂PH with 2 leads to the formation of CyP(CF₃)₂ and CyP(CF₃)H in trace amounts, together with the main product Cy₂P(CF₃) as detected by ³¹P{¹H} NMR spectroscopy. This can be rationalized by assuming a homolytic cleavage of a C–P bond (P–Cy) after the attack of Cy₂PH by a CF₃-radical instead of P–H bond cleavage.

In conclusion, we have demonstrated that secondary and tertiary P(III) compounds containing only one CF₃-substituent can be conveniently obtained using hypervalent electrophilic λ³ I–CF₃ compounds starting from readily accessible or even commercially available primary and secondary phosphines. Thus, the method allows for selective alteration of the electronic properties of the phosphorus donor atom. The trifluoromethylated phosphines were found to serve as suitable ligands for transition-metals in several oxidation states forming stable, crystalline complexes with appropriate Pd(II),¹¹Ru(II), Rh(I) and Ir(III) precursors, respectively. We are currently pursuing the application of these newly accessible P–CF₃ ligands in catalytic reactions.¹²

This work was supported by ETH Zürich and the Swiss National Science Foundation (PhD grants to P.E. and I.K.).

Notes and references

1. S. P. Flanagan and P. J. Guiry, J. Organomet. Chem., 2006, 691, 2125.
2. R. N. Haszeldine and B. O. West, J. Chem. Soc., 1956, 3631.
3. (a) W. R. Cullen, Can. J. Chem., 1960, 38, 439; (b) M. A. A. Beg and H. C. Clark, Can. J. Chem., 1961, 39, 564; (c) M. A. A. Beg and H. C. Clark, Can. J. Chem., 1962, 40, 283.
4. I. Tworowska, W. Dabkowski and J. Michalski, Angew. Chem., Int. Ed., 2001, 40, 2898.
5. P. Panne, D. Naumann and B. Hoge, J. Fluorine Chem., 2001, 112, 283.
6. R. P. Singh, A. Vij, R. L. Kirchmeier and J. M. Shreeve, Inorg. Chem., 2000, 39, 375.
7. M. B. Murphy-Jolly, L. C. Lewis and A. J. M. Caffyn, Chem. Commun., 2005, 4479.
8. T. Umemoto and S. Ishihara, J. Am. Chem. Soc., 1993, 115, 2156.
9. (a) P. Eisenberger, S. Gischig and A. Togni, Chem.–Eur. J., 2006, 12, 2579; (b) I. Kieltsch, P. Eisenberger and A. Togni, Angew. Chem., Int. Ed., 2007, 46, 754.
10. S. E. Vaillard, C. Mück-Lichtenfeld, S. Grimme and A. Studer, Angew. Chem., Int. Ed., 2007, 46, 6533.
11. (a) M. A. A. Beg and H. C. Clark, Can. J. Chem., 1962, 40, 283; (b) A. J. Rest, J. Chem. Soc., 1968, 2212; (c) M. A. A. Beg and M. A. Qaiser, Thermochim. Acta, 1992, 210, 123.
12. See, e.g.: S. White, B. L. Bennett and D. M. Roddick, Organometallics, 1999, 18, 2536.

---

Footnote

† Electronic supplementary information (ESI) available: Detailed experimental procedures including characterization and spectroscopic data for compounds 6–15. See DOI: 10.1039/b801424h

---