“While the holy grail remains […] finding a method to catalytically activate and functionalize P₄ to yield new organophosphorus compounds, fundamental research lies in developing procedures to control the reactivity of elemental phosphorus.”¹

That sentence – written just a few years ago by Ragogna and coworkers – succinctly describes both our motivation, and the biggest problem we faced while pursuing this project. We were seeking to find a direct and catalytic method to functionalize white phosphorus (P₄). Why? P₄ is the common phosphorus source for almost all organophosphorus compounds, including all manner of industrially-important phosphines and phosphonium salts, among many other species. State-of-the-art processes to functionalize P₄ rely on an indirect two-step procedure involving oxidation with toxic chlorine gas to form corrosive phosphorus chlorides. Nucleophilic substitution of the chloride ions then leads to formation of the desired organophosphorus compounds. Far better would be a catalytic and chlorine free method to achieve the same overall transformation in a single step. 

Our initial concept was based on the ability of P₄ to serve as an excellent radical trap.^2–4 If we could catalytically generate carbon-centered radicals in the presence of P₄, this would lead to the formation of new P—C bonds. We chose visible-light-photocatalysis as our “radical supplier”, which was for two reasons. First, P₄ is transparent in the visible region, meaning visible light irradiation does not decompose P₄. Second, photocatalysis is an unusually mild method to generate radicals, which should minimize any side-reactions with P₄ (the chemistry of P₄ can be notoriously unselective). We believed that photocatalytic reductive dehalogenation would fit well with our idea (Fig 1a): in the presence of a photocatalyst and a sacrificial electron donor, aryl halides would be reduced to form carbon centred radicals, which could be trapped by P₄ to yield arylphosphines or phosphonium salts. 

With this idea in mind we began the project, and tried to employ several well-established and well-known photocatalytic systems. Unfortunately, while P₄ was consumed in almost all cases, controlling its reactivity proved challenging. In fact, most ³¹P{¹H} NMR spectra and GC-MS analyses did not show any phosphorus containing compounds at all! Instead, a precipitate was almost always formed, and we realised that P₄ was probably being reduced directly by the photocatalyst to form the anion P₄^∙, which then induces polymerization of further P₄ (see Fig.1b). Nevertheless, we persisted in our investigation, and were ultimately rewarded with the discovery that using [Ir(dtbby)(ppy)₂]PF₆ ([1]PF₆) as catalyst and Et₃N as electron donor solved our problem. The in-situ formed reduced photocatalyst [1] preferentially dehalogenates iodobenzene derivatives to form the desired carbon-centered radicals, and only traces of P₄ seem to be reduced. Thus, irradiation of a solution containing [1]⁺, triethylamine, iodobenzene and P₄ in a benzene/acetonitrile mixture with blue LEDs yielded the corresponding tetraarylphosphonium salt, [Ph₄P]I. Moreover, substrate screening showed that differently-substituted triarylphosphines and tetraarylphosphonium salts could be synthesized by this method. With these results we were able to validate our reaction design principle, and to demonstrate the first example of an effective catalytic system for the direct organofunctionalisation of P₄. 

For more information, please read our paper: https://www.nature.com/articles/s41929-019-0378-4

1.         Graham, C. M. E., Macdonald, C. L. B., Power, P. P., Brown, Z. D. & Ragogna, P. J. Transition Metal Functionalization of P₄ Using a              Diarylgermylene Anchor. Inorg. Chem. 56, 9111–9119 (2017).

2.         Ghosh, S. K., Cummins, C. C. & Gladysz, J. A. A direct route from white phosphorus and fluorous alkyl and aryl iodides to the                    corresponding trialkyl- and triarylphosphines. Org. Chem. Front. 5, 3421–3429 (2018).

3.         Cossairt, B. M. & Cummins, C. C. Radical synthesis of trialkyl, triaryl, trisilyl and tristannyl phosphines from P₄. New J. Chem. 34,              1533–1536 (2010).

4.         Barton, D. H. R. & Vonder Embse, R. A. The invention of radical reactions. Part 39. The reaction of white phosphorus with carbon-              centered radicals. An improved procedure for the synthesis of phosphonic acids and further mechanistic insights. Tetrahedron 54,              12475–12496 (1998).