A polychloride-enabled synthesis of [NEt₃Me][PCl₆] serving as a potential PCl₃-storage and PCl₅-reagent

Abstract

Reaction of the ionic liquid [NEt₃Me][Cl₃] with white phosphorus (P₄) gives, quantitatively, hexachlorophosphate [NEt₃Me][PCl₆]. This compound shows similar reactivity as PCl₅, as confirmed for the reaction with phenol, carboxylic acids and ammonium chloride. At elevated temperature, [NEt₃Me][PCl₆] releases PCl₃ and can therefore be used as a potential PCl₃-storage material.

---

Among all P₁ building blocks, phosphorus(III) chloride (PCl₃) is the most important, being produced on an industrial scale of approximately 350.000 t per a.¹ This base chemical is widely employed for the synthesis of pharmaceuticals, agrochemicals, flame retardants and valuable organophosphorus compounds, such as PPh₃.^1,2 However, the use of PCl₃ has been under debate,^1,3–6 partly because it is a highly reactive and hazardous chemical that has caused severe accidents in the past.¹ Moreover, the industrial synthesis of PCl₃ is typically performed by the exothermic chlorination of molten white phosphorus (P₄) with gaseous chorine at elevated temperatures.² This procedure is “widely considered to be problematic” yet unavoidable as there is currently no useful alternative.³

Recently, Riedel and co-workers proposed trichlorides of the type [NEt₃Me][Cl₃] (1) as a reversible chlorine storage and chlorination agent that could pave the way to a new and improved chlorine technology.⁷ Using [NEt₃Me][Cl₃], new synthetic processes for base chemicals such as phosgene (COCl₂) could be realized under mild conditions.⁸ In general, [NEt₃Me][Cl₃] shows an enhanced reactivity compared to elemental chlorine, while enabling the stabilization of anionic species.⁷ This unique reactivity was demonstrated for the synthesis of the ammonium salt [NEt₃Me]₂[SCl₆], which contains the first hexachlorosulphate(IV) anion reported in the literature.⁹ Most importantly, [NEt₃Me][Cl₃] is a room-temperature ionic liquid with a low chlorine vapor pressure, serving as a safer and easier-to-handle alternative to elemental chlorine.^10,11 Therefore, we investigated the use of [NEt₃Me][Cl₃] as a safe and simple agent for the preparation and handling of PCl₃.

In contrast to the conversion of sulphur (S₈) with [NEt₃Me][Cl₃] to [NEt₃Me]₂[SCl₆], the reaction of red phosphorus (P_red) with the trichloride under ambient conditions turned out to be extremely exothermic and afforded no isolable products.⁹ Consequently, we attempted the reaction of P₄ with [NEt₃Me][Cl₃] under milder and inert reaction conditions. Tetrachloromethane (CCl₄) as a non-polar and rather inert reaction medium was chosen to prevent chlorination of the solvent. Thus, [NEt₃Me][Cl₃] was added dropwise to the suspension of P₄ in CCl₄ at T = −20 °C. While heating the reaction mixture slowly up to room temperature, the precipitation of a colourless powder was observed (Scheme 1).

Scheme 1 Reaction of P₄ with [NEt₃Me][Cl₃] (1) to form 2.

³¹P NMR spectra of the precipitate revealed a single resonance at δ(ppm) = −295.2. This signal is characteristic for the [PCl₆]⁻ anion in solution.^12–16 The same reactivity is observed when performing the reaction in dichloromethane (CH₂Cl₂) instead of CCl₄. After removing the solvent, the salt [NEt₃Me][PCl₆] (2) can be isolated quantitatively, based on the used amount of white phosphorus. Crystallization from both CH₂Cl₂ and acetonitrile (CH₃CN) gave colourless crystals that were analysed by means of single crystal X-ray diffraction. The molecular structure of the [NEt₃Me][PCl₆] salt 2 is depicted in Fig. 1, selected bond lengths and angles are shown in the caption of the figure.

Fig. 1 Molecular structure of 2 in the crystal. Displacement ellipsoids are shown at the 50% probability level. All rotational disorders are depicted, with the most occupied position of the [PCl₆]⁻ anion and one of the positions of the [NEt₃Me]⁺ cation highlighted. Selected bond lengths (Å) and angels (°): P–Cl 2.11(2) Å, Cl–P–Cl 90°.

The crystallographic characterization of the product confirms the formation of [NEt₃Me][PCl₆], which crystallizes in the cubic space group Pm3̄m. The highly symmetric structure includes an octahedral [PCl₆]⁻ anion with Cl–P–Cl bond angles of 90° and P–Cl bond lengths of 2.11(2) Å, which is in the range of classical P–Cl-single bonds.‡ The cation [NEt₃Me]⁺ is highly disordered, such that the methyl and ethyl groups are undistinguishable.§ The Raman spectrum of [NEt₃Me][PCl₆] also points to the octahedral symmetry in accordance with the reported spectra of the anion (Fig. S5, ESI†).¹² The ESI-TOF mass spectrum of 2 shows a pattern that is similar to the one of the [PF₆]⁻ anion under ambient conditions, along with molecular ion peaks for oxidation products, such as [PO₂Cl₂]⁻ (Fig. S7, ESI†).¹⁷

The ammonium hexachlorophosphate reported here is one of very few examples of this compound class. In fact, with less than 20 crystallogaphically characterized structures containing the [PCl₆]⁻ anion, it is highly underrepresented in literature, particularly when compared to the [PF₆]⁻ anion, which is one of the most common weakly-coordinating anions, with over 25 thousand known structures. The synthesis described above is the first direct synthetic pathway to a [PCl₆]⁻ salt starting from white phosphorus. Other known synthetic routes towards [PCl₆]⁻ salts start either from phosphorus pentachloride (PCl₅), or by using phosphorus chlorides or polyphosphorus compounds.^13–16 Brockner and coworkers reported on the reactions of imidazolium chloride and pyridinium chloride salts with PCl₅ yielding the respective [PCl₆]⁻ salts.¹² In contrast, we found that the reaction of PCl₅ with [NEt₃Me][Cl] does not give [NEt₃Me][PCl₆] (Scheme 2a), while the reaction of [NEt₃Me][Cl₃] with PCl₃ does result in the formation of [PCl₆]⁻ (Scheme 2b).

Scheme 2 Alternative synthesis of compound 2.

It should be noted that PCl₅ exists as the ionic species [PCl₄[PCl₆]] in the solid.^18,19 In solution, PCl₅ exists primarily as PCl₅, while in polar solvents a partial dissociation into [PCl₄]⁺ and [PCl₆]⁻ has been observed.²⁰ In contrast, for [NEt₃Me][PCl₆], only [PCl₆]⁻ species could be observed by ³¹P NMR spectroscopy.

In 2009, Manners and coworkers showed that the [PCl₆]⁻ anion plays a key role in the PCl₅-initiated living cationic polymerization of a phosphoranimine.²¹ Inspired by this work, we further investigated the reactivity of [NEt₃Me][PCl₆]. Reacting [NEt₃Me][PCl₆] with phenol in CD₂Cl₂ produces PCl(OPh)₄, which shows a characteristic resonance at δ(ppm) = −23.7 in the ³¹P{¹H} NMR spectrum (Scheme 3a and Fig. S13, ESI†). PCl(OPh)₄ is the known product of the reaction of phenol and PCl₅, while the reaction of phenol with PCl₃ yields the phosphite P(OPh)₃.²²

Scheme 3 [NEt₃Me][PCl₆] as a synthon for important reactions of PCl₅. Ad = adamantyl.

Thus, this observation reveals a reactivity of [NEt₃Me][PCl₆] similar to the one of PCl₅. Interestingly, HCl, which is liberated during the reaction of phenol with [NEt₃Me][PCl₆], adds to the resulting chloride salt giving the bichloride based ionic liquid [NEt₃Me][Cl(HCl)_n] (3, n = 1–4) as the only byproduct. Formation of the bichloride ionic liquid has recently been investigated by us.²³ The bichloride can be detected by means of ¹H NMR spectroscopy as a singlet resonance at δ(ppm) = 11.44. This shift slightly varies depending on the loading of the salt with HCl.²¹ Remarkably, the bichloride salt can be separated from the product mixture by extracting PCl(OPh)₄ and can further be used in several classical reactions in which HCl is used. With this process in hand, a combined trichloride/bichloride process can be envisioned.²³

Further studies confirmed the similar reactivity of [NEt₃Me][PCl₆] to PCl₅ and the possibilities to use the combined process. Carboxylic acids were found to react with [NEt₃Me][PCl₆] in a similar fashion as PCl₅, giving the respective acyl chlorides with POCl₃ and, again, the bichloride [NEt₃Me][Cl(HCl)_n] (¹H NMR (CD₂Cl₂) δ(ppm) = 12.63) as the only byproducts, as demonstrated with 1-adamantanecarboxylic acid (Scheme 3b).²

We subsequently turned our attention to a vast industrial process: the conversion of PCl₅ with NH₄Cl to hexachlorophosphazene. This compound is used, for example, for the synthesis of poly(dichlorophosphazene).^24,25 It turned out that hexachlorophosphazene can also be easily obtained by reacting NH₄Cl with [NEt₃Me][PCl₆] (Scheme 3c). The products can be separated by extracting hexachlorophosphazene with non-polar solvents. After extraction, the ionic liquid [NEt₃Me][Cl(HCl)_n] (¹H NMR (CD₂Cl₂) δ(ppm) = 12.83) remains as a residue, which can be reconverted, e.g., by electrolysis, to [NEt₃Me][Cl₃] and H₂, or used for further applications.²³

During this work, we noticed that [NEt₃Me][PCl₆] releases PCl₃ upon heating to temperatures above T = 100 °C. Thus, while [NEt₃Me][PCl₆] is stable at room temperature for a rather long period of time, it can be used as a PCl₃ storage material as well. The release of PCl₃ from 2 can be performed in CH₂Cl₂ with a conversion of approximately 50% based on the amount of [NEt₃Me][PCl₆] used. The reaction mixture noticeably turns orange during the release of PCl₃. Raman spectroscopic analysis reveals that red phosphorus is formed as a byproduct during this process (Fig. S21, ESI†).¶ The release of PCl₃ also proceeds in different chlorinated solvents, such as tetrachloromethane, tetrachloroethane, or dichloromethane. However, due to the formation of azeotropes, the separation of PCl₃ from the solvents turned out to be rather challenging. As PCl₃ is one of the most used phosphorus precursors in chemical industry, the direct synthesis of PCl₃ without elemental chlorine could become industrially valuable (vide supra). To overcome the problem of the separation of PCl₃ it is possible to react stochiometric amounts of [NEt₃Me][Cl₃] and white phosphorus in a closed system (Scheme 4).

Scheme 4 Direct synthesis of PCl₃ from white phosphorus and [NEt₃Me][Cl₃].

When the mixture is heated to T = 120 °C, PCl₃ is produced and can be condensed under vacuum into a flask cooled with liquid nitrogen (Fig. S23, ESI†). This reaction can be scaled up to a multiple gram scale with yields of up to 74%. The reaction of [NEt₃Me][Cl₃] with white phosphorus can therefore be recommended as an alternative route to the preparation of PCl₃, not only for laboratory use, but also from an industrial point of view, as the storage and handling of gaseous chlorine involved in the conventional synthesis of PCl₃ can thus be avoided.

We have synthesized and characterized the novel ammonium hexachlorophosphate [NEt₃Me][PCl₆]. This salt is accessible directly from white phosphorus and the trichloride [NEt₃Me][Cl₃] and can be used efficiently as a synthon for PCl₅. This was shown for the synthesis of PCl(OPh)₄, acid chlorides and hexachlorophosphazene with the advantage of the direct accessibility from white phosphorus and without the use of gaseous chlorine. The byproduct of the reactions is the bichloride based ionic liquid [NEt₃Me][Cl(HCl)_n], which could also be further used for chemical conversions in which HCl is needed or be electrochemically recycled to [NEt₃Me][Cl₃]. Moreover, we demonstrated that [NEt₃Me][PCl₆] can be used as a PCl₃ storage material, as the reaction of white phosphorus with [NEt₃Me][Cl₃] in a closed system gives PCl₃, without the use of any gaseous starting materials.

We acknowledge financial support by the Werner Siemens-Stiftung and the Core Facility BioSupraMol supported by the DFG.

Data availability

The data supporting this article have been included as part of the ESI.† Furthermore, crystallographic data for [NEt₃Me][PCl₆] has been deposited under the CCDC number 2379260.

Conflicts of interest

There are no conflicts to declare.

Notes and references

1. J.-L. Montchamp, Phosphorus, Sulfur Silicon Relat. Elem., 2013, 188, 66.
2. G. Bettermann, W. Krause, G. Riess and T. Hofmann, Phosporus Compounds, Inorganic in Ullmann‘s Encyclopedia of Industrial Chemistry, Wiley-VCH, Weinheim, 2012.
3. M. Donath, K. Schwedtmann, T. Schneider, F. Hennersdorf, A. Bauzá, A. Frontera and J. J. Weigand, Nat. Chem., 2022, 14, 384.
4. A. R. Jupp, S. Beijer, G. C. Narain, W. Schipper and J. C. Slootweg, Chem. Soc. Rev., 2021, 50, 87.
5. B. M. Cossairt, N. A. Piro and C. C. Cummins, Chem. Rev., 2010, 110, 4164.
6. D. J. Scott, Angew. Chem., Int. Ed., 2022, 61, e202205019.
7. M. Kleoff, P. Voßnacker and S. Riedel, Angew. Chem., Int. Ed., 2023, 62, e202216586.
8. P. Voßnacker, A. Wüst, T. Keilhack, C. Müller, S. Steinhauer, H. Beckers, S. Yogendra, Y. Schiesser, R. Weber, M. Reimann, R. Müller, M. Kaupp and S. Riedel, Sci. Adv., 2021, 7, eabj5186.
9. P. Voßnacker, A. Wüst, C. Müller, M. Kleoff and S. Riedel, Angew. Chem., Int. Ed., 2022, 61, e202209684.
10. S. Rathmann, F. Braune, M. Kleoff, P. Voßnacker, B. Pölloth, C. Müller and S. Riedel, J. Chem. Educ., 2024, 101, 1394.
11. P. Voßnacker, N. Schwarze, T. Keilhack, M. Kleoff, S. Steinhauer, Y. Schiesser, M. Paven, S. Yogendra, R. Weber and S. Riedel, ACS Sustainable Chem. Eng., 2022, 10, 9525.
12. M. Gjikaj, J.-C. Leye, T. Xie and W. Brockner, CrystEngComm, 2010, 12, 1474.
13. G. S. Harris and D. S. Payne, J. Chem. Soc., 1956, 3038.
14. K. B. Dillon and A. W. G. Platt, Phosphorus, Sulfur Silicon Relat. Elem., 1984, 19, 299.
15. A. S. Muir, Polyhedron, 1991, 10, 2217.
16. B. Neumüller and K. Dehnicke, Z. Anorg. Allg. Chem., 2005, 631, 1471.
17. Y. Lu, F. L. King and D. C. Duckworth, J. Am. Soc. Mass Spectrom., 2006, 17, 939.
18. H. M. Powell, D. Clark and A. F. Wells, Nature, 1940, 145, 149.
19. D. Clark, H. M. Powell and A. F. Wells, J. Chem. Soc., 1942, 642.
20. H. C. Suter, H. C. Knachel, V. P. Petro, J. H. Howatson and S. G. Shore, J. Am. Chem. Soc., 1973, 1474.
21. V. Blackstone, A. J. Lough, M. Murray and I. Manners, J. Am. Chem. Soc., 2009, 131, 3658–3667.
22. L. W. Dennis, V. J. Bartuska and G. E. Maciel, J. Am. Chem. Soc., 1982, 104, 230.
23. G. H. Dreyhsig, P. Voßnacker, M. Kleoff, H. Baunis, N. Limberg, M. Lu, R. Schomäcker and S. Riedel, Sci. Adv., 2024, 10, eadn5353.
24. J. Emsley and P. B. Udy, J. Chem. Soc. A, 1970, 3025.
25. M. Gorlov, N. Bredov, A. Esin, I. Sirotin, M. Soldatov, V. Oberemok and V. V. Kireev, Int. J. Mol. Sci., 2021, 22, 5958.

---

Footnotes

† Electronic supplementary information (ESI) available. CCDC 2379260. For ESI and crystallographic data in CIF or other electronic format see DOI: https://doi.org/10.1039/d4cc04319g

‡ The structure shows a [PCl₆]⁻ octahedron where the main position is 62.5% occupied. In addition, there are three further (rotational) disorders, each of them is occupied to 12.5%. In total there are six “main positions” of the chlorides with 0.75 occupation factor and 12 more with occupation factor 0.125. All four orientations of the [PCl₆]⁻ anion have an octahedral symmetry with only 90° angles and equal P–Cl bond lengths within the error tolerance. For details of the different orientations of the [PCl₆]⁻ anion see the ESI.† The comparably large size of the anion and the cation, combined with a highly delocalized charge of the anion, results in a pseudo-binary CsCl type packing that is caused by coulombic interactions with no preferential orientation of the anion. This is assumed to yield the observed rotational disorders.

§ The same applies to the [NEt₃Me] cation. The methyl and ethyl fragments are indistinguishable. Lower space group types show the same statistical occupation and even a triclinic solution with or without forced superstructure (no sign of superstructure in reciprocal space) does not provide any alignment. The central carbon atoms (bounded to N1) are disordered on 12 positions (in total 4 atoms, thus occupation factor 1/3), while the outer carbon atoms are disordered on six positions and thus, since statistically only 3 atoms, the occupation factor of these atoms is 0.5. A respective modelling of hydrogen atoms does not comprise any physically meaningful information and was thus omitted. The nitrogen and the phosphorus atom are the only fully occupied atoms.

¶ It has once been reported that red phosphorus can be produced from white phosphorus in boiling PBr₃ leading to the so called “light-red phosphorus”: R. Schenk, Z. Elektrochem., 1905, 8, 1.

---