Room-temperature reduction of sulfur hexafluoride with metal phosphides

Article abstract of DOI:10.1039/d1cc01943k

Upon treatment with sulfur hexafluoride, alkali metal diphenyl or dicyclohexyl phosphides are oxidized within seconds to tetraphenyl or tetracyclohexyl diphosphines. When bulky di-tert-butylphosphide is employed, fluorophosphine intermediates are detected. This is the first reported reaction of sulfur hexafluoride with metal phosphides, and a rare example of reactivity of sulfur hexafluoride at ambient temperature. This journal is

Full text of DOI:10.1039/d1cc01943k

ChemComm
View Article Online
View Journal | View Issue
COMMUNICATION
Room-temperature reduction of sulfur
hexafluoride with metal phosphides†
Cite this: Chem. Commun., 2021,
57, 7128
Blake S. N. Huchenski and Alexander W. H. Speed
*
Received 13th April 2021,
Accepted 21st June 2021
DOI: 10.1039/d1cc01943k
rsc.li/chemcomm
Upon treatment with sulfur hexafluoride, alkali metal diphenyl or
Select examples of main-group molecules that react with SF6
dicyclohexyl phosphides are oxidized within seconds to tetraphenyl are shown in Scheme 1. Rueping showed reduction of SF6 in
or tetracyclohexyl diphosphines. When bulky di-tert-butylphosphide solution using Murphy’s electron-donor reagent 1.8 Beier
is employed, fluorophosphine intermediates are detected. This is the showed reduction of SF6 using the lithium salt of the TEMPO
first reported reaction of sulfur hexafluoride with metal phosphides, anion 2.9 Braun and Kemnitz showed fluorination of
and a rare example of reactivity of sulfur hexafluoride at ambient N-heterocyclic carbene 3 by SF6 under UV irradiation.10 Hoge
temperature.
showed reduction of SF6 with an electron rich phenolate anion/
protonated phosphazene pair 4.11 Dielmann employed a
Sulfur hexafluoride (SF6) is an industrially used gas, with many nucleophilic super-basic phosphine complex for stoichiometric
applications depending on its density, high dielectric constant, SF6 activation.12 While this manuscript was under revision, a
and relative inertness. An unfortunate side effect of this relative manuscript from Crimmin appeared, showing the reaction of
inertness is that SF6 released into the atmosphere has a lifetime an aluminum(I) complex with SF6.13
of thousands of years, resulting in it having the highest global
Catalytic reactions for functionalization of SF6 have also
warming potential (B23 500 relative to CO2 over a 100 year been reported. Building on earlier stoichiometric work with
time-frame) of any commonly used industrial gas.1 SF6 is used
most extensively in electrical equipment as an insulator and
arc-quencher.2 This application, involving the use of many
thousands of tonnes of SF6, mandates controls on its use and
release.3 Because of the high global warming potential of SF6,
and its common use, chemistry to decompose surplus SF6 is a
sought after goal. The stability of SF6 arises from a kinetic
barrier to decomposition, rather than high thermodynamic
stability. The central sulfur is well shielded by the fluorides
from nucleophilic attack. Harsh conditions, such as exposure
to Lewis acids at high temperature and pressure, or electric-arc
plasmas have been used to decompose sulfur hexafluoride.4
More recently, several relatively mild solution-phase strategies
resulting in stoichiometric reaction of SF6 with a reductant
have been disclosed. Low valent early metal complexes, and
electron rich nickel complexes are fluorinated by SF6.5 SF6 is
reduced by solvated electrons in ammonia solutions.6 The
reaction of SF6 with butyllithium has been reported in a patent,
however limited analysis was provided.7
Department of Chemistry, Dalhousie University, Halifax, Nova Scotia, B3H 4R2,
Canada. E-mail: aspeed@dal.ca
† Electronic supplementary information (ESI) available. See DOI: 10.1039/
Scheme 1 Selected stoichiometric systems for SF6 degradation.
d1cc01943k
7128 | Chem. Commun., 2021, 57, 7128–7131
This journal is © The Royal Society of Chemistry 2021

View Article Online
Communication
ChemComm
rhodium,14 Braun reported rhodium-complex catalyzed decom- spectroscopy revealed complete consumption of phosphide,
position of SF6 with phosphines as sulfur scavengers and and the formation of tetraphenyldiphosphine 7a. The reaction
silanes as the terminal reductant.15 Several photoredox systems was relatively clean. Analysis of the reaction mixture by
that activate SF6 have been reported. SF6 has recently been used 19F NMR spectroscopy revealed no sharp peaks associated with
by Jamison as a fluorinating reagent for allylic alcohols new products, however excess SF6 is observed in solution.
employing ruthenium and iridium-complex photosensi- Despite this, we were able to account for the fate of the fluoride.
tizers.16 Nagorny showed fluorination of carbohydrates using The volatiles were removed, and the resulting solids were
benzophenone-based photosensitizers.17 Wagenknecht showed extracted with water and dichloromethane. 19F NMR spectro-
phenothiazine-derived photosensitizers allow addition of SF5 scopy of the aqueous layer showed a strong signal attributed to
radicals derived from SF6 to substituted styrenes.18 Despite the KF. The sulfur had multiple fates. Testing the aqueous solution
relative simplicity of the systems shown in Scheme 1, access to for sulfide with an acidic solution of 9 and ferric chloride
most of these SF6-reactive examples require multi-step synth- resulted in the formation the characteristic colour of methylene
esis or relatively harsh conditions, so there is still much room blue 10a, a classic test for sulfide ion (eqn (2), Scheme 2).21
for discovery of simple SF6-reactive systems. Inspired by Diel- Inspection of the 31P NMR spectra revealed minor products
mann’s use of strongly basic phosphines for SF6 functionaliza- corresponding to phosphorus sulfides 10b and 10c.22,23 Com-
tion, we explored if simply employing the conjugate base pound 10d was observed in the NMR spectra acquired in THF,
(i.e. metalation) of a secondary phosphine could afford reactiv- but was insoluble in chloroform or diethyl ether. Compounds
ity with SF6. Phosphides have been used in other small mole- 10b and 10c likely arose from reaction of 7a with sulfur(0)
cule activations, notably alkali di-tert-butylphosphides were during the reaction, while 10d arose from corresponding reac-
shown by Stephan to activate H2.19
tion of phosphide 6a with sulfur(0). We verified independently
We found metal phosphides are highly reactive with SF6. that 7a and 6a formed these compounds upon reaction with
Bubbling SF6 through a commercially available solution of elemental sulfur in THF. We tested for remaining elemental
potassium diphenylphosphide 6a in THF resulted in disappear- sulfur by adding triphenylphosphine to a suspension of the
ance of the deep orange colour within seconds (eqn (1), reaction residue in acetonitrile, and also residue that was freed
Scheme 2).20 Analysis of the reaction mixture by 31P NMR of soluble phosphines by extracting with solvent and centrifu-
ging, but no formation of triphenylphosphine sulfide was
observed in either case.24 The reaction was conducted on a
1 mmol scale with no loss of efficiency, and 7a was isolated
from this reaction in 95% yield.
A temperature rise was observed during this larger reaction.
Metalation of diphenylphosphine 8a with n-BuLi followed by
introduction of SF6 also resulted in formation of 7a showing that
other metals could be used (eqn (3), Scheme 2).25 Dicyclohexyl-
phosphine 8b was also deprotonated by butyllithium and upon
treatment with SF6, oxidative dimerization to 7b was observed
within seconds, indicating the reaction was not limited to aryl
phosphides (eqn (4), Scheme 2).26 Other metalloids were unreac-
tive, phosphide-borane adduct 11a did not show appreciable
reaction,27 and silyl-phosphine TMSPPh2 11b was also unreactive.
These results indicate significant anionic character is necessary
on the phosphorus centre for this reaction to proceed rapidly and
cleanly. One mechanistic scenario involves polar activation of SF6,
where anion 6a attacks a fluorine atom on SF6, precedented by
Dielmann’s observation of nucleophilic reactivity of super-basic
phosphines with SF6.12 While this would initially form fluoropho-
sphine 12a, which was not detected, 12a could react with addi-
tional diphenylphosphide 6a to form 7a (Scheme 3, eqn (1)).
Compound 12a is also known to disproportionate into 7a and
phosphorane 13 (Scheme 3, eqn (2)).28 We did not observe 13 in
the reaction mixture, suggesting if 12a is formed, the reaction
with 6a is faster than disproportionation.
This does not rule out the possibility of formation and subse-
quent reduction of 13 by additional phosphide. An alternate mecha-
nistic possibility involves reduction initiated by single electron
transfer from the phosphide to SF6, generating the phosphinyl
radical (Ph2Pꢀ) and a SF6 radical anion.29
Scheme 2 Reaction of aryl and alkyl phosphides with SF6.
This journal is © The Royal Society of Chemistry 2021
Chem. Commun., 2021, 57, 7128–7131 | 7129

View Article Online
ChemComm
Communication
compound 7b (eqn (4), Scheme 2). Increased steric bulk is known
to stabilize fluorophosphines. Unlike 12a, fluorophosphines 12b
and 12c are stable.33 We sought to generate the appropriate
phosphides to see if 12b and 12c could then be prepared by
treatment with SF6. Caged phosphine 8c was resistant to depro-
tonation by n-BuLi, t-BuLi, or KH in THF over multiple hours, with
the P–H bond remaining intact as ascertained by proton-coupled
31P NMR spectroscopy.34 Di-tert-butylphosphine 8d was deproto-
nated with n-BuLi in THF to generate 6c.35 Upon exposure to SF6,
fluorophosphine 12c was generated, and observed by 31P and
19F NMR spectroscopy, supporting our hypothesis that fluoropho-
sphines are intermediate formed in the reactions of phosphides
with SF6.36 The ability to stop at the fluorophosphine intermediate
was presumably due to the greater steric hinderance of the tert-
butyl groups relative to phenyl or cyclohexyl, however in addition
to 12c, dimer 7c was still observed. Another prominent signal in
the 31P NMR spectrum at +85.2 ppm was observed, this closely
matched that reported for compound 14 in an independent
preparation.37 Compound 14 could be expected to form from
either the reaction of sulfide ion with 12c, or the reaction of sulfur
with 7c followed by rearrangement.
In conclusion, these results represent rare examples of room-
temperature reactivity of SF6. In contrast to some previously reported
systems, the reaction proceeds without the need for light irradiation.
In addition, the reaction only takes seconds, and is to the best of our
knowledge, the first example of a controlled reaction of a commer-
cially available substance with SF6 at ambient temperature. While
this reaction conveniently generates diphosphines from the corres-
ponding phosphides, its utility is limited by the high environmental
burden of SF6, so we did not conduct a large substrate scope. Of
Scheme 3 Further reactivity studies.
The SF6 radical anion would undergo further fragmenta- greatest conceptual importance is the discovery of a new example of
tion.16 Most reported SF6 reduction reactions invoke single an accessible reaction that consumes the potent greenhouse gas SF6.
electron transfer to SF6 as the initial step, so this is a reasonable While the high reactivity of potassium diphenylphosphide limits
scenario, however the reducing agents used in those reactions large scale application of this method for SF6 remediation, one
were weak nucleophiles, in contrast to this system.8,9,11,16,17 potential application could be use of this technique to dispose of
Alternatively, the phosphinyl radical could abstract a fluorine small amounts of SF6 residues from applications such as retinal
atom from a reduced sulfur fluoride intermediate, forming surgery,38 or in field-servicing electrical equipment,2 that would
12a.30 The radical scenarios would generate the same products otherwise be released into the atmosphere. The bleaching of the
in eqn (1), Scheme 3, as proposed for a polar mechanism, since orange diphenylphosphide gives a convenient indication of when
phosphinyl radicals rapidly dimerize in solution to form 7a.31
the reagent is exhausted. Investigation of phosphorus-based systems
Lower oxidation state sulfur fluorides such as SF4 were not for remediation of SF6 capable of catalytic turnover with appropriate
observed by 19F NMR spectroscopy. In general, highly reactive terminal reductants is underway and will be reported in due
free lower fluorides of sulfur have not been observed in SF6 course.39
reduction reactions, with the exception of the pentafluorosul-
Financial support from NSERC (Discovery Grant), and the
fanyl anion (SF5À), which has been observed in Rueping’s, Killam Foundation (B. S. N. H.) are acknowledged. Dr Mike
Hoge’s, and Dielmann’s systems.8,11,12 These systems have Lumsden and Mr Xiao Feng (Dalhousie University) are thanked
diffuse counterions that impart kinetic stability to SF5À, how- for assistance with NMR spectroscopy and mass-spectrometry
À
ever instability of SF5 with simple cations such as potassium respectively. Professors Laura Turculet and Mark Stradiotto
or lithium has been noted.32 We did not detect SF5À, and it is (Dalhousie University) are thanked for providing com-
unlikely this fragile anion would survive with the simple alkali pounds 6a, and 8a–8c. Professor Christopher Algar (Dalhousie
metal cations originating from the phosphides used in University) is thanked for advice on detecting sulphide.
this work.
In the reactions shown in Scheme 2, while 6a did not
produce any compounds that exhibited coupling to 19F in the
Conflicts of interest
31P spectrum, traces of fluoride-containing phosphorus com-
pounds were observed in the reaction that formed cyclohexyl There are no conflicts of interest to declare.
7130 | Chem. Commun., 2021, 57, 7128–7131
This journal is © The Royal Society of Chemistry 2021

View Article Online
Communication
ChemComm
23 Y. Sato, M. Nishimura, S.-I. Kawaguchi, A. Nomoto and A. Ogawa,
Chem. – Eur. J., 2019, 25, 6797.
24 C. T. Tran, P. G. Williard and E. Kim, J. Am. Chem. Soc., 2014,
136, 11874.
Notes and references
1 M. Maiss and C. A. M. Brenninkmeijer, Environ. Sci. Technol., 1998,
32, 3077.
2 L. G. Christophorou, J. K. Olthoff and R. J. van Brunt, IEEE Electr.
Insul. Mag., 1997, 13, 20.
3 Catalog of guidelines and standards for the handling and manage-
ment of Sulfur Hexafluoride, U.S. Environmental Protection Agency
Office of Air and Radiation Global Program Division (625 J)
WASHINGTON DC 20460. https://www.epa.gov/sites/production/
files/2016-02/documents/sf6utility7.pdf.
25 The phosphides are depicted as monomeric, unsolvated structures,
but their solution behaviour may be more complex, see:
I. J. Colquhoun, H. C. E. McFarlane and W. McFarlane, J. Chem.
Soc., Chem. Commun., 1982, 220.
26 R. A. Bartlett, M. M. Olmstead and P. P. Power, Inorg. Chem., 1986,
25, 1243.
27 Small changes in the 31P NMR spectra were observed after SF6
addition, but no dramatic changes like the alkali metal phosphide
reactions.
4 J. R. Case and F. Nyman, Nature, 1962, 193, 473.
¨
5 (a) B. G. Harvey, A. M. Arif, A. Glockner and R. D. Ernst, Organome-
tallics, 2007, 26, 2872; (b) P. Holze, B. Horn, C. Limberg,
C. Matlachowski and S. Mebs, Angew. Chem., Int. Ed., 2014, 53, 2750.
6 (a) G. C. Demitras and A. G. MacDiarmid, Inorg. Chem., 1964,
3, 1198; (b) L. Brewer, C.-A. Chang and B. King, Inorg. Chem.,
1970, 9, 814.
28 F. Seel, K. Rudolph and W. Gombler, Angew. Chem., Int. Ed. Engl.,
1967, 6, 708.
29 Mixing 6a and benzophenone at THF in room-temperature resulted
in the immediate formation of a persistent deep blue-green colour,
characteristic of the ketyl radical anion.
7 The latter patent claims LiSF5 was prepared as an aqueous solution
in D2O, inconsistent with other reports of the preparation of SF5
anion, and its low stability (a) S. Nishino and H. Shima, Method for
producing fluorinated organic compound, Jp. Pat., JP2009292749A,
2009; (b) S. Nishino and H. Shima, Cracking method of sulfur
hexafluoride, Jp. Pat., JP2010013342A, 2010.
8 (a) M. Rueping, P. Nikolaienko, Y. Lebedev and A. Adams, Green
Chem., 2017, 19, 2571; (b) J. A. Murphy, J. Garnier, S. R. Park,
F. Schoenebeck, S.-Z. Shou and A. T. Turner, Org. Lett., 2008,
10, 1227.
30 Phosphides rapidly reacted with common radical traps such as
TEMPO. We lack access to an EPR spectrometer, or the appropriate
equipment to conduct low-temperature NMR studies on this two-
phase system, so we cannot differentiate the polar and radical
scenarios in greater detail. We look forward to results from any
researchers who wish to attempt further mechanistic studies.
31 (a) Y. Sueishi and Y. Nishihara, J. Chem. Res., 2001, 84;
(b) U. Fischbach, M. Trincado and H. Gru¨tzmacher, Dalton Trans.,
2017, 46, 3443.
32 K. Matsumoto, Y. Haruki, S. Sawada, S. Yamada, T. Konno and
R. Hagiwara, Inorg. Chem., 2018, 57, 14882.
ˇ
9 G. Iakobson, M. Posta and P. Beier, J. Fluorine Chem., 2018, 213, 51.
10 P. Tomar, T. Braun and E. Kemnitz, Chem. Commun., 2018, 54, 9753.
11 (a) R. F. Weitkamp, B. Neumann, H.-G. Stammler and B. Hoge,
Chem. – Eur. J., 2021, 27, 6460; (b) R. F. Weitkamp, B. Neumann,
H.-G. Stammler and B. Hoge, Chem. – Eur. J., 2021, 27, 6465.
12 F. Buß, C. Mu¨ck-Lichtenfeld, P. Mehlmann and F. Dielmann,
Angew. Chem., Int. Ed., 2018, 57, 4951.
33 N. Fey, M. Garland, J. P. Hopewell, C. L. McMullin, S. Mastroianni,
A. G. Orpen and P. G. Pringle, Angew. Chem., Int. Ed., 2012, 51,
118.
34 Only deprotonation of borane adducts of the cage have been
reported. See: D. L. Dodds, Joelle Floure, M. Garland,
M. F. Haddow, T. R. Leonard, C. L. McMullin, A. G. Orpen and
P. G. Pringle, Dalton Trans, 2011, 40, 7137.
13 D. J. Sheldon and M. R. Crimmin, Chem. Commun., 2021, Advance
Article, DOI: 10.1039/D1CC02838C.
35 R. A. Jones, A. L. Stuart and T. C. Wright, J. Am. Chem. Soc., 1983,
105, 7459.
´
´
14 L. Zamostna, T. Braun and B. Braun, Angew. Chem., Int. Ed., 2014,
53, 2745.
36 The products of this reaction were not isolated. The literature
31P shift and coupling constant for 12c, recorded in CDCl3, are
´
´
15 L. Zamostna and T. Braun, Angew. Chem., Int. Ed., 2015, 54, 10652.
16 T. A. McTeague and T. F. Jamison, Angew. Chem., Int. Ed., 2016,
55, 15072.
1
+214.9 ppm (d, JPF = 865 Hz). See ref. 25.
37 S. Yogendra, S. S. Chitnis, F. Hennersdorf, M. Bodensteiner,
R. Fischer, N. Burford and J. J. Weigand, Inorg. Chem., 2016,
55, 1854. Compound 14 is reported to have a 31P chemical shift of
+86.6 ppm in CD2Cl2 in this work.
17 S. Kim, Y. Khomutnyk, A. Bannykh and P. Nagorny, Org. Lett., 2021,
23, 190.
18 (a) D. Rombach and H.-A. Wagenknecht, Angew. Chem., Int. Ed.,
2020, 59, 300; (b) D. Rombach and H.-A. Wagenknecht,
ChemCatChem, 2018, 10, 2955.
38 SF6 bubbles are used to hold eye-structures in place after certain
¨
forms of eye surgery. See: U. Mester, P. Kroll, B. Volker and
19 M. Xu, A. R. Jupp, Z.-W. Qu and D. W. Stephan, Angew. Chem, Int.
Ed., 2018, 57, 11050.
20 A video of the transformation is available in the ESI†.
21 J. D. Cline, Limnol. Oceanogr., 1969, 14, 454.
22 M. Gruber, P. G. Jones and R. Schmutzler, Chem. Ber., 1990, 123, 1313.
I. Kreissig, Ophthalmologica, 1983, 186, 151.
39 B. S. N. Huchenski and A. W. H. Speed, A pre-print of this manu-
script was uploaded to ChemRXIV, Room-Temperature Reduction of
Sulfur Hexafluoride with Metal Phosphides, ChemRxiv, 2021, DOI:
10.26434/chemrxiv.14399213.v1.
This journal is © The Royal Society of Chemistry 2021
Chem. Commun., 2021, 57, 7128–7131 | 7131