Vanadium, niobium and tantalum complexes with terminal sulfur radical ligands

Article abstract of DOI:10.1039/d1dt01956b

Sulfur radicals terminally bound to the metal center can be considered as the one-electron reduction products of complexes with terminal sulfido ligands which serve as the reactive sites in enzymes and precursors. However, there is limited information regarding this kind of metal stabilized sulfur radical, which contrasts the more commonly known metal stabilized thiyl radical. In this work, we report the preparation of vanadium, niobium and tantalum radical complexes in the form of M(O)(S)F2 from the reactions of laser-ablated metal atoms and SOF2 in cryogenic matrixes. Combined with the results from infrared spectroscopy and density functional theory calculations, the sulfur ligand in M(O)(S)F2 is characterized to be a terminally bound radical with the unpaired electron located on the sulfur 3p orbital. Besides this radical complex, calculations also predict the existence of MF2(η2-SO) with a side-on SO ligand, but this less stable isomer is not observed as a result of high exothermicity along with its formation from metal atoms and SOF2 that is large enough to overcome the energy barrier towards the occurrence of M(O)(S)F2.

Full text of DOI:10.1039/d1dt01956b

Dalton
Transactions
View Article Online
View Journal | View Issue
PAPER
Vanadium, niobium and tantalum complexes with
terminal sulfur radical ligands†
Rui Wei, a,b Jingwen Hu, a,b Xiuting Chen a and Yu Gong
Cite this: Dalton Trans., 2021, 50,
11300
a
*
Sulfur radicals terminally bound to the metal center can be considered as the one-electron reduction
products of complexes with terminal sulfido ligands which serve as the reactive sites in enzymes and pre-
cursors. However, there is limited information regarding this kind of metal stabilized sulfur radical, which
contrasts the more commonly known metal stabilized thiyl radical. In this work, we report the preparation
of vanadium, niobium and tantalum radical complexes in the form of M(O)(S)F2 from the reactions of
laser-ablated metal atoms and SOF2 in cryogenic matrixes. Combined with the results from infrared spec-
troscopy and density functional theory calculations, the sulfur ligand in M(O)(S)F2 is characterized to be a
terminally bound radical with the unpaired electron located on the sulfur 3p orbital. Besides this radical
complex, calculations also predict the existence of MF2(η2-SO) with a side-on SO ligand, but this less
stable isomer is not observed as a result of high exothermicity along with its formation from metal atoms
and SOF2 that is large enough to overcome the energy barrier towards the occurrence of M(O)(S)F2.
Received 14th June 2021,
Accepted 6th July 2021
DOI: 10.1039/d1dt01956b
rsc.li/dalton
It has been recently demonstrated that thionyl fluoride
(SOF2) can be used for the introduction of sulfido and oxo
Introduction
Complexes with metal stabilized thiyl moieties have been of inter- groups by atom transfer reactions, and a fluoride stabilized
est in recent years due to their involvement in a number of reac- sulfur-substituted uranyl complex [UVI(O)(S)F2] featuring triple
tions including hydrogen evolution, carbon–sulfur bond for- UuO and UuS bonds was obtained for the first time via the
mation as well as carbon–oxygen and oxygen–hydrogen bond acti- oxidation of U(0) by SOF2 in cryogenic matrixes.17 In view of
vation in methanol and water.1–3 Such highly reactive radicals are the success in the formation of U(O)(S)F2, it is possible to
usually generated via one-electron oxidation of complexes with obtain complexes with sulfur radical ligands if the metal
non-innocent thiolate ligands. Their presence in the reaction center does not have enough valence electrons to afford metal–
system is generally confirmed by electron spin resonance spec- sulfur multiple bonds. Herein, we report the matrix prepa-
troscopy, UV-vis absorption spectroscopy and cyclic voltammetry, ration of vanadium, niobium and tantalum complexes M(O)(S)F2
and the nature of these radicals is further elucidated by density in which the terminal sulfur ligand is characterized to be a
functional theory (DFT) and high level ab initio calculations.4–7 radical. The structures of these radical species were identified
Compared with the more commonly known metal stabilized thiyl by infrared spectroscopy, and DFT calculations were employed
radicals in which the sulfur atom connects a metal center and a to confirm the assignments and shed light on the structure,
functional group, there is limited information regarding metal bonding and formation mechanism associated with the M–S
stabilized sulfur radicals that are terminally bound to the metal radical bond.
center, which also contrasts the rich studies on organic sulfur
radicals.8–10 Metal stabilized sulfur radicals can be considered as
the one-electron reduction products of complexes with terminal
sulfido ligands which serve as the reactive sites in enzymes and
Experimental section
precursors and their structures have been well characterized for a
series of transition and actinide metals.11–16
The vanadium, niobium and tantalum complexes were pre-
pared using an experimental apparatus that has been
described in detail elsewhere.18,19 The 1064 nm Nd:YAG laser
fundamental (10 Hz repetition rate with 6 ns pulse width) was
used for metal evaporation with an output of 15 mJ per pulse.
The laser beam was focused onto a rotating vanadium,
niobium or tantalum (Alfa Aesar) target, and the ablated metal
atoms were codeposited with 0.5% SOF2 in argon or neon onto
a 4 K CsI window for 60 (Ar) or 30 min (Ne). SOF2 was syn-
aDepartment of Radiochemistry, Shanghai Institute of Applied Physics, Chinese
Academy of Sciences, Shanghai 201800, China. E-mail: gongyu@sinap.ac.cn
bSchool of Nuclear Science and Technology, University of Chinese Academy of
Sciences, Beijing 100049, China
†Electronic supplementary information (ESI) available: Additional infrared
spectra and calculation results. See DOI: 10.1039/d1dt01956b
11300 | Dalton Trans., 2021, 50, 11300–11306
This journal is © The Royal Society of Chemistry 2021

View Article Online
Dalton Transactions
Paper
thesized by the reactions of 32 mg of sulfur powder (99.9%,
Sinopharm Chemical Reagent, China) and 127 mg of AgF
(98%, Alfa Aesar) in the presence of 5 μL of H2O in a stainless
steel container. After the mixture was heated at 150 °C for
2 hours, gaseous SOF2 was obtained with a yield of around
40%. 34SOF2 and S18OF2 were prepared following a similar pro-
cedure using 34S powder (99.3%, ISOFLEX), AgF and H2O or
natural sulfur powder, AgF and H218O (97% 18O, Sigma-
Aldrich). SOF2, 34SOF2 and S18OF2 were subjected to freeze–
pump–thaw cycles using liquid nitrogen before use. The
gaseous SOF2/Ar and SOF2/Ne samples were prepared by
mixing SOF2 and high purity argon (99.999%, Xiangkun
Special Gas, China)/neon (99.999%, Dalian Special Gas, China)
using a homemade stainless steel manifold. To prepare the
0.5% (V/V) SOF2/Ar or SOF2/Ne mixture, 500 Pa SOF2 was first
released from the SOF2 container to the manifold which was
then filled with 105 Pa argon/neon gas. After sample codeposi-
tion, the CsI window was rotated 90° for recording the infrared
spectra at 0.5 cm−1 resolution using a Vertex 70 V spectrometer
equipped with a DLaTGS detector. The matrix samples were
annealed at different temperatures using a LakeShore325
temperature controller and cooled back to 4 K for spectral
acquisition. The selected samples were subjected to UV-vis
irradiation using a 250 W mercury arc lamp with the outer
globe removed.
Fig. 1 Infrared spectra in the product regions from the reactions of
laser-ablated vanadium atoms and 0.5% SOF2 in argon matrixes at 4 K
(a) after codeposition for 60 min; (b) after annealing at 25 K; (c) after λ >
220 nm UV-vis irradiation for 20 min; and (d) after annealing at 30 K.
16O/18O isotopic frequency ratio of 1.0418 (Table S1†). Both the
band position and isotopic ratio are comparable to those for
the V–O stretching vibration of VO (983.6 cm−1 16O/18O:
,
1.0452) and OVF2(η2-SO) (1043.3 cm−1
,
16O/18O: 1.0443)
Theoretical calculations were performed to help assign the
product structures. Density functional theory (DFT) calcu-
lations were carried out using the Gaussian 09 program.20 The
hybrid B3LYP functional21,22 was employed in combination
with the 6-311+G(3df) basis set for O, F, S and V and the 28
and 60 electron core SDD pseudopotentials for Nb and Ta.23–27
All the geometrical parameters were fully relaxed and the har-
monic vibrational frequencies were analytically obtained. Zero
point energy was included in the relative energies. The tran-
sition state for the isomerization process from MF2(η2-SO) to
M(O)(S)F2 was optimized using the synchronous transit-guided
quasi-Newton method implemented in the Gaussian 09
program. Natural bond orbital (NBO) analysis was performed
at the B3LYP/6-311+G(d)/SDD level using the NBO6 program.28
molecules,29,30 and this band is therefore assigned to the V–O
stretch of the new product molecule. Note that a weak band at
999.4 cm−1 was also observed as a result of ∼25% S16OF2 in
the S18OF2 sample (Fig. 2). The absence of any band between
999.4 and 959.3 cm−1 suggests the involvement of only one VO
subunit in this mode. For the 770.1 and 704.0 cm−1 bands,
they are located in the region where vanadium fluorides
absorb.31 Less than 2 cm−1 shifts were observed upon 18O sub-
stitution which is consistent with the shifts observed for the
Results and discussion
Fig. 1 shows the infrared spectra of the reactions of vanadium
and 0.5% SOF2 in excess argon. Besides the intense SOF2 pre-
cursor absorptions, three new vanadium dependent absorp-
tions were observed at 999.4, 770.1 and 704.0 cm−1 after
sample deposition, and they were slightly sharpened after
sample annealing at 25 K. The intensities of all three absorp-
tions increased dramatically when the sample was exposed to
UV-vis irradiation, and they grew slightly upon further sample
annealing at 30 K. The identical behaviours of these bands
throughout the experiments indicate that they should arise
from different vibrational modes of the same vanadium and
SOF2 reaction product. The 999.4 cm−1 absorption shifted to
Fig. 2 Infrared spectra from the reaction of laser-ablated vanadium
atoms and isotopically labelled SOF2 in solid argon at 4 K. (a) V + 0.5%
SOF2 and (b) V + 0.5% S18OF2. The spectra were recorded after λ >
220 nm UV-vis irradiation followed by annealing at 30 K. P and 18P
959.3 cm−1 in the experiment with S18OF2, resulting in an denote the V(O)(S)F2 and V(18O)(S)F2 absorptions.
This journal is © The Royal Society of Chemistry 2021
Dalton Trans., 2021, 50, 11300–11306 | 11301

View Article Online
Paper
Dalton Transactions
antisymmetric and symmetric F–V–F stretches of the VF2
subunit in OVF2(η2-SO).30 We therefore assign the 770.1 and
704.0 cm−1 bands to antisymmetric and symmetric F–V–F
stretching vibrational modes of the product. The observation
of the 999.4, 770.1 and 704.0 cm−1 bands leads us to propose a
structure containing an OVF2 moiety. Since these three bands
show some difference in position from those of OVF2,31 there
should be a sulfur atom terminally bound to the vanadium
center that perturbs the vibrations of OVF2. This is in agree-
ment with the fact that the new product is formed by the reac-
tion of vanadium and SOF2. Hence, we assign the three bands
at 999.4, 770.1 and 704.0 cm−1 to the V(O)(S)F2 complex. A
complementary experiment on the reaction of vanadium and
SOF2 was carried out in neon, and three absorptions appearing
at 1004.2, 778.7 and 712.8 cm−1 are assignable to V(O)(S)F2
(Fig. S1, ESI†). The band positions parallel those observed in
argon, and the neon-to-argon shifts are within the typical
values for the V–O and F–V–F stretches.30
Fig. 4 Infrared spectra in the product regions from the reactions of
laser-ablated tantalum atoms and 0.5% SOF2 in argon matrixes at 4 K (a)
after codeposition for 60 min; (b) after annealing at 25 K; (c) after λ >
220 nm UV-vis irradiation for 20 min; and (d) after annealing at 30 K.
The infrared spectra from the reactions of laser-ablated
niobium and tantalum with SOF2 are shown in Fig. 3 and 4.
Two sets of product absorptions were observed at 961.4, 689.1,
and 664.6 and 955.2, 674.4, and 672.8 cm−1 respectively, and
their behaviours upon sample annealing and UV-vis
irradiation are the same as those for V(O)(S)F2. Following the
vanadium case, these new bands are readily assigned to the
Nb(O)(S)F2 and Ta(O)(S)F2 complexes, which is supported by
the results with S18OF2 (Fig. 5 and 6). An addition niobium
experiment with 34SOF2 was carried out to probe the influence
of sulfur on these infrared absorptions (Fig. 5). As listed in
Table S1,† no 34S shift was observed for any of the 961.4, 689.1
and 664.6 cm−1 bands, which supports the assignment to the
Nb–O, antisymmetric and symmetric F–Nb–F stretches of Nb
(O)(S)F2.
To further confirm the assignments, DFT frequency calcu-
lations were carried out on these new products. Four infrared
Fig. 5 Infrared spectra from the reactions of laser-ablated niobium
atoms and isotopically labelled SOF2 in solid argon at 4 K. (a) Nb + 0.5%
SOF2; (b) Nb + 0.5% 34SOF2; and (c) Nb + 0.5% S18OF2. Spectra were
taken after sample annealing at 30 K following λ > 220 nm UV-vis
irradiation. P, 34P and 18P denote the Nb(O)(S)F2, Nb(O)(34S)F2 and Nb
(
18O)(S)F2 absorptions.
absorptions above 400 cm−1 were predicted for V(O)(S)F2
(Table 1). The V–O stretching vibrational mode at 1079.4 cm−1
was found to shift to 1033.9 cm−1 upon 18O substitution with
an 16O/18O frequency ratio of 1.0440. Both the band position
and 16O/18O isotopic frequency ratio fit the experimental
values of 999.4 cm−1 and 1.0418 well. The antisymmetric and
symmetric F–V–F stretching vibrations were predicted at 782.8
and 723.6 cm−1, in excellent agreement with the experimental
values of 770.1 and 704.0 cm−1. The observed small 18O shifts
of 0.1 and 1.5 cm−1 also fit the calculated values of 0.1 and
1.1 cm−1. For the last band predicted at 455.6 cm−1 due to the
V–S stretch, its infrared intensity is only 1/4 that of the V–O
Fig. 3 Infrared spectra in the product regions from the reaction of
laser-ablated niobium atoms and 0.5% SOF2 in argon matrixes at 4 K (a)
after codeposition for 60 min; (b) after annealing at 25 K; (c) after λ >
220 nm UV-vis irradiation for 20 min; and (d) after annealing at 30 K.
11302 | Dalton Trans., 2021, 50, 11300–11306
This journal is © The Royal Society of Chemistry 2021

View Article Online
Dalton Transactions
Paper
Fig. 7 Optimized structures of M(O)(S)F2 (M = V, Nb and Ta) (bond
lengths in angstroms and bond angles in degrees) with Cs symmetry and
2A’ ground state at the B3LYP/6-311+G(3df)/SDD level of theory.
only 0.1 cm−1 red shift was predicted for the 979.6 cm−1 band
of Nb(O)(S)F2 and no shift was found for the 696.9 and
669.8 cm−1 bands, which are in line with the experimental
results. We also performed additional calculations using the
BP86 and M06 functionals,32–34 and the calculation results are
listed in Table S2.† The vibrational frequencies and infrared
intensities of V(O)(S)F2, Nb(O)(S)F2 and Ta(O)(S)F2 are almost
the same as those predicted at the B3LYP level.
Fig. 7 shows the optimized structures of V(O)(S)F2, Nb(O)(S)
F2 and Ta(O)(S)F2, all of which possess non-planar Cs sym-
metries and doublet ground states. The metal–fluorine and
metal–oxygen bond lengths are quite close to the values of the
same bonds in other molecules containing group
5 metals.30,31,35,36 Different from the relatively short terminal
oxo bonds, the metal–sulfur bond lengths were calculated to
be 2.194, 2.350 and 2.351 Å for V(O)(S)F2, Nb(O)(S)F2 and Ta
(O)(S)F2, which are about 0.2 Å longer in comparison with the
distances of group 5 metal complexes with terminal sulfido
ligands.37 Such a difference in the M–S bond length is consist-
Fig. 6 Infrared spectra from the reactions of laser-ablated tantalum
atoms and isotopically labelled SOF2 in solid argon at 4 K. (a) Ta + 0.5%
SOF2 and (b) Ta + 0.5% S18OF2. The spectra were recorded after anneal-
ing at 30 K following λ > 220 nm UV-vis irradiation. P and 18P denote the
Nb(O)(S)F2 and Nb(18O)(S)F2 absorptions.
stretch, which makes it difficult to detect in a region approach-
ing the detection limit of the infrared spectrometer (Fig. S2,
ESI†). Similar agreements were found between the observed
and calculated vibrational frequencies and isotopic frequency
ratios for the niobium and tantalum products as shown in
Table 1. The Nb–S and Ta–S stretches are even lower in posi-
tion than the V–S stretch, and it is unlikely for them to be
detected (Fig. S3 and S4, ESI†). In the case of 34S substitution,
Table 1 Comparison between the experimental and calculated (in italic) vibrational frequencies (cm−1) and 16O/18O frequency ratios of the M(O)(S)
F2 (M = V, Nb and Ta) complexes
Mode
Frequencya
16O/18
O
V(O)(S)F2
Nb(O)(S)F2
Ta(O)(S)F2
V–O stretch
999.4
1079.4(133)
1.0418
1.0440
antisymmetric F–V–F stretch
symmetric F–V–F stretch
770.1
782.8(174)
1.0001
1.0001
704.0
723.6(167)
b455.6(32)
1.0021
1.0015
c1.0020
V–S stretch
Nb–O stretch
961.4
979.6(114)
1.0513
1.0508
antisymmetric F–Nb–F stretch
symmetric F–Nb–F stretch
689.1
696.9(164)
1.0001
1.0001
664.6
669.8(140)
b405.2(35)
1.0002
1.0003
c1.0012
Nb–S stretch
Ta–O stretch
955.2
946.8(81)
1.0547
1.0552
antisymmetric F–Ta–F stretch
symmetric F–Ta–F stretch
Ta–S stretch
674.4
671.3(129)
1.0001
1.0001
672.8
664.9(112)
b381.9(22)
1.0000
1.0000
c1.0010
a Intensity (km mol−1) in parentheses. b Not observed. c Not available.
This journal is © The Royal Society of Chemistry 2021
Dalton Trans., 2021, 50, 11300–11306 | 11303

View Article Online
Paper
Dalton Transactions
Fig. 8 Natural bond orbitals showing the V–S interactions of V(O)(S)F2.
Fig. 9 Potential energy profile for the isomerization reaction [MF2(η2-
SO) ↔ M(O)(S)F2] (M = V, Nb and Ta) at the B3LYP level of theory.
ent with the difference in the M–S stretching vibrational fre-
quency. As listed in Table 1, the calculated M–S stretching
vibrational frequencies range from 455.6 cm−1 for V to
381.9 cm−1 for Ta, which are lower than the frequencies of
inal sulfur ligand. For the NbF2(η2-SO) and TaF2(η2-SO) com-
plexes, the S–O stretches were predicted at 770.7 and
752.0 cm−1 which would be observable between the two
intense bands of SOF2 if they were formed. The absence of any
metal dependent band in this region demonstrates the domi-
nance of Nb(O)(S)F2 and Ta(O)(S)F2 as the reaction products of
SOF2 and metal atoms. Similarly, V(O)(S)F2 is the only
observed product in the vanadium case since there would be a
S–O stretching band between 800 and 820 cm−1 if VF2(η2-SO)
were formed. The conversion from MF2(η2-SO) to the more
stable M(O)(S)F2 isomer is mediated by a transition state where
the S–O bond is elongated and the M–O bond is shortened
(Fig. 9). The energy barriers for this process are 16.4 (V), 8.3
(Nb) and 4.4 kcal mol−1 (Ta), respectively. Since the formation
of VF2(η2-SO), NbF2(η2-SO) and TaF2(η2-SO) from ground state
vanadium (4F), niobium (6D) and tantalum(4F) atoms and
SOF2 (singlet) is highly exothermic by 208.0, 229.1 and
248.3 kcal mol−1, the energy barrier for breaking the S–O bond
is too low to prevent the occurrence of M(O)(S)F2. This is con-
sistent with the oxophilic character of group 5 metals41 and
the energy required to break the S–O bond is compensated by
forming a terminal oxo bond leaving the sulfur atom as a
radical.
group
5
metal
sulfides
and
hydrido
sulfides
(500–600 cm−1).37,38 Given that the highest oxidation state of
the metal center is V, the terminal sulfur atom in M(O)(S)F2 (M
= V, Nb and Ta) should be a radical that forms a single bond
with the metal center. Such radical character is confirmed by
natural bond orbital (NBO) analysis. It can be found that the
metal–sulfur bond is composed of sulfur 3p and metal s/d
hybrid orbitals with about 25% contribution from the metal
(Table S3†). The other two sulfur 3p orbitals remain nonbond-
ing with two valence electrons occupying one of them as lone
pair electrons, leaving the last 3p orbital singly occupied
(Fig. 8). Consistent with this notion, the unpaired electron is
mainly located on sulfur with spin densities of 1.01, 0.95 and
0.93 for the vanadium, niobium and tantalum complexes,
respectively.
In addition to M(O)(S)F2 observed in the experiments, DFT
calculations also predict the existence of another isomer
MF2(η2-SO) that is higher in energy by 5.4 (V), 28.0 (Nb) and
35.0 kcal mol−1 (Ta) (Fig. S5, ESI†). The major difference
between the two isomers is the presence of a side-on SO ligand
in MF2(η2-SO) with S–O distances of 1.684, 1.709 and 1.730 Å
for vanadium, niobium and tantalum, respectively, while there
is no S–O bond in M(O)(S)F2. These S–O bond lengths are
about the same as those of other sulfur monoxide complexes
such as OMoF2(η2-SO)/OWF2(η2-SO) and OV(η2-SO)/ONb(η2-
SO)/OTa(η2-SO),39,40 but are longer than the distances for the
OVF2(η2-SO), ONbF2(η2-SO) and OTaF2(η2-SO) complexes,30
Conclusions
suggesting that the SO ligand in MF2(η2-SO) should be peroxo- Vanadium, niobium and tantalum radical complexes in the
like instead of superoxo-like in character. This is consistent form of M(O)(S)F2 were prepared via the reactions of laser-
with the fact that the S–O stretching frequencies of MF2(η2-SO) ablated metal atoms and SOF2 in cryogenic matrixes. M(O)(S)
are closer to those of OMoF2(η2-SO) and OWF2(η2-SO), and the F2 is characterized to possess two M–F bonds, one terminal
spin densities are mainly located on the metal center (V: 1.27; oxo bond and one M–S radical bond with a doublet ground
Nb: 1.09; and Ta: 1.06). Compared with MF2(η2-SO), the M–O state and non-planar geometry. The metal–sulfur bond is com-
bond of M(O)(S)F2 is shortened by 0.2 Å while the M–S bond is posed of sulfur 3p and metal s/d hybrid orbitals with the
slightly changed by less than 0.03 Å. This indicates that the M– unpaired electron residing in one of the two sulfur 3p orbitals
S single bond remains unchanged from MF2(η2-SO) to M(O)(S) that are nonbonding in character. Although the MF2(η2-SO)
F2, which further confirms the radical character of the term- complex with a side-on SO ligand is also predicted to be stable
11304 | Dalton Trans., 2021, 50, 11300–11306
This journal is © The Royal Society of Chemistry 2021

View Article Online
Dalton Transactions
Paper
by DFT calculations, it is higher in energy by 5.4 (V), 28.0 (Nb) 11 Z. K. Sweeney, J. L. Polse, R. A. Andersen, R. G. Bergman
and 35.0 kcal mol−1 (Ta). The highly exothermic character of
and M. G. Kubinec, J. Am. Chem. Soc., 1997, 119, 4543.
the formation of this SO complex from metal atoms and SOF2 12 W. A. Howard and G. Parkin, J. Am. Chem. Soc., 1994, 116, 606.
[208.0 (V), 229.1 (Nb) and 248.3 kcal mol−1 (Ta)], makes the 13 M. W. Rosenzweig, A. Scheurer, C. A. Lamsfus,
energy barrier towards the occurrence of M(O)(S)F2 [16.4 (V),
8.3 (Nb) and 4.4 kcal mol−1 (Ta)] easily overcome. The obser-
F. W. Heinemann, L. Maron, J. Andrez, M. Mazzanti and
K. Meyer, Chem. Sci., 2016, 7, 5857.
vation of the sulfur radical complex instead of the sulfur mon- 14 D. E. Smiles, G. Wu, N. Kaltsoyannis and T. W. Hayton,
oxide complex is a result of the oxophilic character of Chem. Sci., 2015, 6, 3891.
vanadium, niobium and tantalum, and the energy required to 15 J. J. Wang, S. Groysman, S. C. Lee and R. H. Holm, J. Am.
break the S–O bond is compensated by forming a terminal oxo Chem. Soc., 2007, 129, 7512.
bond. The successful preparation of M(O)(S)F2 via the oxi- 16 S. Groysman, J. J. Wang, R. Tagore, S. C. Lee and
dation of metal atoms by SOF2 provides a new route towards R. H. Holm, J. Am. Chem. Soc., 2008, 130, 12794.
transition metal complexes with sulfur radical ligands. This 17 R. Wei, X. Chen and Y. Gong, Chem. Commun., 2020, 56,
could be extended to the synthesis of similar compounds in 6782.
solution and solid state under appropriate conditions for a 18 L. Andrews and H. G. Cho, Organometallics, 2006, 25, 4040.
comprehensive understanding of the structure and bonding of 19 L. Andrews, Chem. Soc. Rev., 2004, 33, 123.
metal stabilized sulfur radicals.
20 M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria,
M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone,
B. Mennucci, G. A. Petersson, H. Nakatsuji, M. Caricato,
X. Li, H. P. Hratchian, A. F. Izmaylov, J. Bloino, G. Zheng,
J. L. Sonnenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda,
J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao,
H. Nakai, T. Vreven, J. A. Montgomery, Jr., J. E. Peralta,
F. Ogliaro, M. Bearpark, J. J. Heyd, E. Brothers,
K. N. Kudin, V. N. Staroverov, R. Kobayashi, J. Normand,
K. Raghavachari, A. Rendell, J. C. Burant, S. S. Iyengar,
J. Tomasi, M. Cossi, N. Rega, J. M. Millam, M. Klene,
J. E. Knox, J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo,
R. Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin,
R. Cammi, C. Pomelli, J. W. Ochterski, R. L. Martin,
K. Morokuma, V. G. Zakrzewski, G. A. Voth, P. Salvador,
J. J. Dannenberg, S. Dapprich, A. D. Daniels, O. Farkas,
J. B. Foresman, J. V. Ortiz, J. Cioslowski and D. J. Fox,
Gaussian 09, revision A.01, Gaussian, Inc., Wallingford, CT,
2009.
Conflicts of interest
There are no conflicts to declare.
Acknowledgements
This work was supported by the Natural Science Foundation of
Shanghai (20ZR1474300), the Strategic Priority Research
Program, and the Frontier Science Key Program of the Chinese
Academy
of
Sciences
(Grants
XDA02030000
and
QYZDY-SSW-JSC016).
Notes and references
1 M. Drosou, F. Kamatsos and C. A. Mitsopoulou, Inorg.
Chem. Front., 2020, 7, 37.
21 A. D. Becke, J. Chem. Phys., 1993, 98, 5648.
2 K. Ouch, M. S. Mashuta and C. A. Grapperhaus, Inorg. 22 C. Lee, W. Yang and R. G. Parr, Phys. Rev. B: Condens.
Chem., 2011, 50, 9904. Matter Mater. Phys., 1988, 37, 785.
3 J. A. Yan, Z. K. Yang, Y. S. Chen, Y. H. Chang, C. L. Lyu, 23 A. D. McLean and G. S. Chandler, J. Chem. Phys., 1980, 72,
C. G. Luo, M. J. Cheng and H. F. Hsu, Chem. – Eur. J., 2018,
24, 15190.
4 K. T. Chu, Y. C. Liu, Y. L. Huang, G. H. Lee, M. C. Tseng
and M. H. Chiang, Chem. – Eur. J., 2015, 21, 6852.
5639.
24 R. Krishnan, J. S. Binkley, R. Seeger and J. A. Pople,
J. Chem. Phys., 1980, 72, 650.
25 A. J. H. Wachters, J. Chem. Phys., 1970, 52, 1033.
5 U. Das, T. Ghorui, B. Adhikari, S. Roy, S. Pramanik and 26 P. J. Hay, J. Chem. Phys., 1977, 66, 4377.
K. Pramanik, Dalton Trans., 2015, 44, 8625.
6 H. Tang, E. N. Brothers and M. B. Hall, Inorg. Chem., 2017,
56, 583.
7 H. Tang, J. Guan and M. B. Hall, J. Am. Chem. Soc., 2015,
137, 15616.
27 D. Andrae, U. Haeussermann, M. Dolg, H. Stoll and
H. Preuss, Theor. Chem. Acc., 1990, 77, 123.
28 E. D. Glendening, J. K. Badenhoop, A. E. Reed,
J. E. Carpenter, J. A. Bohmann, C. M. Morales, C. R. Landis
and F. Weinhold, Natural Bond Order 6.0; Theoretical
Chemistry Institute, University of Wisconsin, Madison, WI,
2013; http://nbo6.chem.wisc.edu/.
8 R. S. Glass, Top. Curr. Chem., 2018, 376, 22.
9 E. Y. Sneeden, M. J. Hackett, J. J. H. Cotelesage,
R. C. Prince, M. Barney, K. Goto, E. Block, I. J. Pickering 29 G. V. Chertihin, W. D. Bare and L. Andrews, J. Phys. Chem.
and G. N. George, J. Am. Chem. Soc., 2017, 139, 11519. A, 1997, 101, 5090.
10 J. Xu, Z. Wu, H. Wan, G. Deng, B. Lu, A. K. Eckhardt, 30 R. Wei, X. Chen and Y. Gong, Inorg. Chem., 2019, 58, 3807.
P. R. Schreiner, T. Trabelsi, J. S. Francisco and X. Zeng, 31 R. Wei, Z. Fang, M. Vasiliu, D. A. Dixon, L. Andrews and
J. Am. Chem. Soc., 2018, 140, 9972.
Y. Gong, Inorg. Chem., 2019, 58, 9796.
This journal is © The Royal Society of Chemistry 2021
Dalton Trans., 2021, 50, 11300–11306 | 11305

View Article Online
Paper
Dalton Transactions
32 A. D. Becke, Phys. Rev. A, 1988, 38, 3098.
33 J. P. Perdew, Phys. Rev. B: Condens. Matter Mater. Phys.,
1986, 33, 8822.
37 J. Zhao, B. Xu, W. Yu and X. Wang, Chin. J. Chem. Phys.,
2016, 29, 10.
38 B. Liang and L. Andrews, J. Phys. Chem. A, 2002, 106, 3738.
34 Y. Zhao and D. G. Truhlar, Theor. Chem. Acc., 2008, 120, 215. 39 R. Wei, X. Chen and Y. Gong, Inorg. Chem., 2019, 58, 15652.
35 M. Zhou, J. Zhuang, G. Wang and M. Chen, J. Phys. Chem. 40 X. Liu, X. Wang, Q. Wang and L. Andrews, Phys. Chem.
A, 2011, 115, 2238.
Chem. Phys., 2013, 15, 9823.
36 J. T. Lyon, H. G. Cho and L. Andrews, J. Phys. Chem. A, 41 R. Wesendrup, C. A. Schalley, D. Schroder and H. Schwarz,
2015, 119, 12742.
Chem. – Eur. J., 1995, 1, 608.
11306 | Dalton Trans., 2021, 50, 11300–11306
This journal is © The Royal Society of Chemistry 2021