Photocatalytic stannylation of white phosphorus

Organophosphorus compounds (OPCs) are highly important chemicals, finding numerous applications in both academia and industry. Herein we describe a simple photocatalytic method for the stannylation of white phosphorus (P4) using a cheap, commercially-available distannane, (Bu3Sn)2, and anthraquinone as a simple photocatalyst. Subsequent ‘one pot’ transformation of the resulting stannylated monophosphine intermediate (Bu3Sn)3P provides direct, convenient and versatile access to valuable OPCs such as acylated phosphines and tetraalkylphosphonium salts.

Qualitative NMR spectra were recorded at room temperature on Bruker Avance III HD 400 (400 MHz) spectrometers and were processed using Topspin 3.2. Chemical shifts δ, are reported in parts per million (ppm); 1 H and 13 C shifts are reported relative to SiMe4 and were calibrated internally to residual solvent peaks, while 31 P shifts and 119 Sn shifts were referenced externally to 85 % H3PO4 (aq.) and Me4Sn, respectively. NMR samples were prepared in the glovebox using NMR tubes fitted with screw caps. Optimization reactions (see sections S5, S6 and S7) and photocatalytic stannylation of P4 to (Bu3Sn)3P and subsequent functionalization to P1 products were analyzed by 31 P{ 1 H} spectra using triphenylphosphine, Ph3PO, as a standard.

S2. Supplementary mechanistic discussion
The proposed mechanism for the photocatalytic stannylation of P4 by (Bu3Sn)2 catalyzed by simple ketones is outlined in Scheme 2 of the primary manuscript. However, it should be noted that in the absence of more comprehensive mechanistic studies it is currently not possible to exclude the involvement of several alternative elementary reaction steps. For example, while the formation of an intermediate stannylated ketyl radical [R2COSnBu3] • is supported by the prior literature, [3] an alternative energy transfer from [R2CO]* to (Bu3Sn)2 resulting in cleavage to Bu3Sn • by a purely outer sphere mechanism has not yet been definitively excluded.
One key question left unanswered at present relates to the precise mechanism by which radical breakdown of the P4 tetrahedron is achieved. Based on our previous studies it can be confidently proposed that each P-P bond cleavage step is likely to begin with attack of a photocatalyticallygenerated Bu3Sn • radical (Scheme S1). [1] The simplest subsequent step would be recombination of the resulting P-centered radical with a second Bu3Sn • radical, thus completing the stannylation of the original P-P moiety (Scheme S1a). Note, however, that this requires bimolecular recombination of two transient radicals whose concentrations are expected to be low.
Alternatively, stannylation of this P-centered radical could be achieved through SH2-type attack on the proposed intermediate [R2COSnBu3] • and/or its pinacol-type dimer, examples of which are known from the literature to exist in equilibrium (Scheme S1b). [3] In this scenario, this equilibrium would act as a stabilized and hence more persistent reservoir of chemically-accessible "Bu3Sn • ".
A third scenario is that SH2-type attack could occur directly on the (Bu3Sn)2 substrate, which should be present in high concentrations throughout the reaction. This scenario would result in a radical chain mechanism analogous to that proposed previously for P4 hydrostannylation (Scheme S1c). [1] In this case the R2CO photocatalyst would formally act as a photosensitizing initiator for the chain reaction, rather than as a strict catalyst per se. However, we currently consider this to be a less likely option, due to several qualitative experimental observations. Firstly, attempts to achieve an analogous chain reaction mechanism using "standard" chemical initiators such as AIBN under thermal conditions (instead of R2CO under photochemical conditions) have thus far been unsuccessful, not leading to any identifiable change by 31 P{ 1 H} NMR spectroscopy (for example, see Figure S1).
Secondly, attempts to initiate the same radical chain under thermal conditions by using catalytic quantities of both AIBN and Bu3SnH, which should definitely result in formation of Bu3Sn • , have also been unsuccessful. These led only to trace hydrostannylation, consistent with the amount of Bu3SnH present and similar to the outcome when no distannane is present, which argues against any efficient radical interception by (Bu3Sn)2 (for example, see Figure S2).
Scheme S1. Illustration of the equipment setup used for photocatalytic reactions at 0.04 mmol scale. Figure S1. Representative 31 P{ 1 H} NMR spectrum for the attempted stannylation of P4 using (Bu3Sn)2 and 12.5 mol% AIBN in toluene, after heating to 80 °C for 16 h. Figure S2. Representative 31 P{ 1 H} NMR spectrum for the attempted stannylation of P4 using (Bu3Sn)2 and 12.5 mol% of both AIBN and Bu3SnH in toluene, after heating to 80 °C for 18 h. The relatively high proportion of (Bu3Sn)2PH and (Bu3Sn)3P relative to Bu3SnPH2 (c.f. ref. [1]) can likely be attributed to Bu3Sn/H ligand scrambling at the elevated reaction temperature and/or the high loading of AIBN relative to Bu3SnH which reduces the number of available H atoms. This is supported by the observation that when the reaction is repeated in the absence of (Bu3Sn)2 the resulting (Bu3Sn)xPH3-x mixture is also heavily weighted towards (Bu3Sn)2PH and (Bu3Sn)3P.

S3. General procedure for photocatalytic functionalization of P4 (0.04 mmol scale) into stannylated phosphine (Bu3Sn)3P using benzophenone
At the start of this project, benzophenone (BP) was chosen as a photocatalyst for the initial reaction optimization due to both its simplicity and the fact that its (photo)reactivity towards hexaalkyldistannanes has been studied previously. [3] To a 10 mL stoppered tube equipped with a stirring bar were added (Bu3Sn)2 (101.1 μL, 5 equiv. based on phosphorus atoms, 20 equiv. based on P4), BP (0.01 mmol, as a stock solution in 149.2 μL benzene, 1 equiv. based on P4) and P4 (0.01 mmol, as a stock solution in 71.3 µL benzene) in benzene as solvent (in total 0.5 mL). The tube was sealed, placed in a water-cooled block (to ensure a near-ambient temperature was maintained, Figure S3), [ Figure S4). Formation of (Bu3Sn)3P was indicated by the characteristic 117/119 Sn-satellited resonance at -346.5 ppm. [1] Figure S3. Illustration of the equipment setup used for photocatalytic reactions at 0.04 mmol scale. Figure S4. Representative 31 P{ 1 H} NMR spectrum for the photocatalytic functionalization of P4 using benzophenone (BP) as a photocatalyst (Table 1, Entry 1). * marks the internal standard Ph3PO (0.02 mmol). ~ marks an unknown Sn-containing side product. # marks unknown side products.       Table S7. Photocatalytic functionalization of P4 to (Bu3Sn)3P: screening of benzophenone derivatives. [

S6. General procedure for photocatalytic functionalization of P4 (0.04 mmol scale) into stannylated phosphine (Bu3Sn)3P using anthraquinone
To a 10 mL stoppered tube equipped with a stirring bar were added (Bu3Sn)2 (101.1 μL, 5 equiv. based on phosphorus atoms, 20 equiv. based on P4), anthraquinone (AQ) (1.0 mg, 0.5 equiv. based on P4) and P4 (0.01 mmol, as a stock solution in 71.3 µL benzene) in acetone as solvent (429 µL, in total 0.5 mL PhH/acetone mixture). The tube was sealed, placed in a water-cooled block (to ensure a near-ambient temperature was maintained, Figure S16), and irradiated with UV light (365 nm, 4.3 V, 700 mA, Osram OSLON SSL 80) for 22 h (unless stated otherwise). Ph3PO (0.02 mmol, stock solution in benzene) was subsequently added to act as an internal standard. The resulting mixture was subjected to 31 P{ 1 H} NMR analysis and showed 83% conversion to the product (Bu3Sn)3P [1] ( Figure S17). Spectroscopic data of (Bu3Sn)3P: 31 P{ 1 H} and 119 Sn{ 1 H} NMR data of the photocatalytically generated (Bu3Sn)3P were extracted from spectra of the crude reaction mixture (see Figure S17 and S18), and are consistent with previous reports. [0] Note that isolation of (Bu3Sn)3P was not pursued due to separation from unreacted (Bu3Sn)2 being complicated by very similar solubilities as well as the high boiling point of both compounds.   Table S8, Entry 1). * marks the internal standard Ph3PO (0.02 mmol). ~ marks an unknown Sncontaining side product. # marks unknown side products.

S7. Optimization of photocatalytic reaction conditions using anthraquinone
In the final stage of optimization, it was found that reducing the (Bu3Sn)2 loading from 20 equiv. to 12 equiv. had only a very minor impact on conversion to (Bu3Sn)3P (from 83% to 79%). Thus, while the reaction with 20 equiv. formally gave the best conversion, the reaction with 12 equiv. was chosen as being optimal for further elaboration into 'one pot' reactions, as it should reduce the formation of stoichiometric, Sn-containing waste.  Table S13) were added to the photocatalytic reaction mixture, which each showed a color change from an orange to a yellow solution while stirring overnight.
The resulting suspension was stirred at room temperature for 16 h. Ph3PO (0.02 mmol, stock solution in benzene) was subsequently added to act as an internal standard. The resulting mixture was subjected to 31 P{ 1 H} NMR analysis and showed 48% conversion of THP ( Figure S28).
Afterwards the 'one-pot' reaction mixture was stirred at 80 °C under air for 16 h to convert the initially formed product THP into its oxidized form THPO ( 31 P{ 1 H} NMR analysis: 38% conversion to THPO, see Figure S29). Figure S28. Quantitative 31 P{ 1 H} NMR (zgig) spectrum of THP generated via photocatalytic stannylation of P4 in benzene/acetone followed by addition of paraformaldehyde in EtOH. * marks the internal standard Ph3PO (0.02 mmol). Figure S29. Quantitative 31 P{ 1 H} NMR spectrum of THPO generated via photocatalytic stannylation of P4 in benzene/acetone followed by addition of paraformaldehyde in EtOH and subsequent oxidation in air. * marks the internal standard Ph3PO (0.02 mmol).

Spectroscopic data of (Ph(O)C)3P:
The NMR data are consistent with previous reports. [
The volatiles were removed under vacuum. n-Hexane was added to the remaining orange oil and the product was precipitated and washed with n-hexane (2 x 10 mL). Recrystallization twice from acetone/n-hexane afforded the desired product [Bn4P]Br (212.2 mg, 56%) as an off-white powder. (crude)

Spectroscopic data of [Bn4P]Br:
The NMR data are consistent with previous reports of both the chloride salt [Bn4P]Cl [6] and [Bn4P]Br. [ Figure S41. 13

S9.4 'One-pot' synthesis and purification of THPC and recycling of Bu3SnCl and (Bu3Sn)2
To a 100 mL stoppered tube equipped with a stirring bar were added (Bu3Sn)2 (1.212 mL, 2.4 mmol, 3 equiv. based on phosphorus atoms, 12 equiv. based on P4), anthraquinone (20.8 mg, 0.1 mmol, 0.5 equiv. based on P4) and P4 (0.2 mmol, as a stock solution in 1.426 mL benzene) in acetone as solvent (2.574 mL, in total 4 mL PhH/acetone mixture). The tube was sealed, placed in a water-cooled block (to ensure a near-ambient temperature was maintained, Figure S31), and irradiated with UV light

Spectroscopic data of THPC:
The NMR data are consistent with previous reports.

Spectroscopic data of the recycled mixture of Bu3SnCl and (Bu3Sn)2:
The NMR data of Bu3SnCl [8] and (Bu3Sn)2 [9] are consistent with previous reports. The 1