Ancillary ligand effects upon dithiolene redox noninnocence in tungsten Bis(dithiolene) complexes

Article abstract of DOI:10.1021/ic4009174

An expanded set of compounds of the type [W(S₂C ₂Me₂)₂L₁L₂]ⁿ (n = 0: L₁ = L₂ = CO, 1; L₁ = L₂ = CN^tBu, 2; L₁ = CO, L₂ = carbene, 3; L ₁ = CO, L₂ = phosphine, 4; L₁ = L₂ = phosphine, 5. n = 2-: L₁ = L₂ = CN^-, [6] ^2-) has been synthesized and characterized. Despite isoelectronic formulations, the compound set reveals gradations in the dithiolene ligand redox level as revealed by intraligand bond lengths, υ_CCchelate, and rising edge energies in the sulfur K-edge X-ray absorption spectra (XAS). Differences among the terminal series members, 1 and [6]^2-, are comparable to differences seen in homoleptic dithiolene complexes related by full electron transfer to/from a dithiolene-based MO. The key feature governing these differences is the favorable energy of the CO π* orbitals, which are suitably positioned to overlap with tungsten d orbitals and exert an oxidizing effect on both metal and dithiolene ligand via π-backbonding. The CN^- π* orbitals are too high in energy to mix effectively with tungsten and thus leave the filled dithiolene π* orbitals unperturbed. This work shows how, and the degree to which, the redox level of a noninnocent ligand can be modulated by the choice of ancillary ligands(s).

Full text of DOI:10.1021/ic4009174

Article
pubs.acs.org/IC
Ancillary Ligand Effects upon Dithiolene Redox Noninnocence in
Tungsten Bis(dithiolene) Complexes
Yong Yan,†,# Christopher Keating,† Perumalreddy Chandrasekaran,⊥ Upul Jayarathne,†,○ Joel T. Mague,†
Serena DeBeer,§,‡ Kyle M. Lancaster,‡ Stephen Sproules,∥ Igor V. Rubtsov,† and James P. Donahue*,†
†Department of Chemistry, Tulane University, 6400 Freret Street, New Orleans, Louisiana 70118-5698, United States
⊥Department of Chemistry and Biochemistry, Lamar University, Beaumont, Texas 77710, United States
§Max-Planck Institute for Chemical Energy Conversion, Stiftstrasse 34-36, D-45470, Mulheim an der Ruhr, Germany
̈
‡Department of Chemistry and Chemical Biology, Baker Laboratory, Cornell University, Ithaca, New York 14853, United States
∥WestCHEM, School of Chemistry, University of Glasgow, Glasgow, G12 8QQ, United Kingdom
S
* Supporting Information
ABSTRACT: An expanded set of compounds of the type
[W(S2C2Me2)2L1L2]n (n = 0: L1 = L2 = CO, 1; L1 = L2 =
CNtBu, 2; L1 = CO, L2 = carbene, 3; L1 = CO, L2 = phosphine,
4; L1 = L2 = phosphine, 5. n = 2−: L1 = L2 = CN−, [6]2−) has
been synthesized and characterized. Despite isoelectronic
formulations, the compound set reveals gradations in the
dithiolene ligand redox level as revealed by intraligand bond
lengths, υCCchelate, and rising edge energies in the sulfur K-edge
X-ray absorption spectra (XAS). Differences among the
terminal series members, 1 and [6]2−, are comparable to
differences seen in homoleptic dithiolene complexes related by
full electron transfer to/from a dithiolene-based MO. The key
feature governing these differences is the favorable energy of the CO π* orbitals, which are suitably positioned to overlap with
tungsten d orbitals and exert an oxidizing effect on both metal and dithiolene ligand via π-backbonding. The CN− π* orbitals are
too high in energy to mix effectively with tungsten and thus leave the filled dithiolene π* orbitals unperturbed. This work shows
how, and the degree to which, the redox level of a noninnocent ligand can be modulated by the choice of ancillary ligands(s).
a
Scheme 1. (a−c) Dithiolene Ligand Redox States
INTRODUCTION
■
Redox “noninnocent” behavior by a ligand can present
complications in describing the electronic structures of its
transition metal complexes but, to an ever increasing degree, is
now appreciated as an aspect of coordination chemistry that
can be definitively clarified with the combined insights afforded
by spectroscopic data, X-ray crystallography, and computational
analysis. A relatively recent trend is the deliberate incorporation
of redox noninnocence into ligand design such that a
supporting ligand itself serves as a source or sink of reducing
equivalents,1−4 thereby enabling reactivity with complexes of
base metals that has traditionally been associated with noble
metals.5 The dithiolene ligand (Scheme 1) is a prototypical
redox noninnocent ligand. For example, it was recognized in
the 1970s that ligand reduction, rather than metal reduction,
governs the series [V(S2C2Ph2)3]n (n = 0, 1−, 2−) on the basis
that EPR spectra of the n = 0 and 2− members were essentially
identical and consistent with a d1 count at vanadium.8 More
recently, the power of sulfur K-edge X-ray absorption
spectroscopy (XAS) to discern different levels of oxidation at
a
Bond lengths are taken from the structures of [Ni(S2C2Me2)2]n (n =
2−, 0, for (a) and (b), respectively)6 and [Ni(Me2pipdt)2]2+
(Me2pipdt = 1,4-dimethylpiperazine-2,3-dithione).
7
dithiolate(2-)),11 [Mo(S2C2Me2)3]0,1−,2−
,
and [Ni-
12
13
(S2C2Me2)2]0,1−,2− are all attributable to dithiolene ligand.
In the course of surveying general reactivity patterns of
midvalent tungsten dithiolene complexes,14 we have synthe-
sized an expanded set of heteroleptic compounds of the general
type [W(S2C2Me2)2L2]n (L = CO, CNR, PR3, carbene, CN−; n
= 0 or 2−). Although they exist ostensibly with the same formal
sulfur has been used to show decisively that the redox processes
9
involved in the series [V(S2C2R2)3]0,1−,2−
,
[Re-
Received: April 13, 2013
10
(S2C2Ph2)3]1+,0,1−
,
[Mo(bdt)3]0,1− (bdt = benzene-1,2-
© XXXX American Chemical Society
A
dx.doi.org/10.1021/ic4009174 | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
9,14 and [Et4N][W(S2C2Me2)3].19 All other reagents were purchased
from commercial sources and used as received. All solvents were
purchased from either Pharmco-AAPER or Sigma-Aldrich. Solvents
either were dried with a system of drying columns from the Glass
Contour Company (CH2Cl2, THF, Et2O, n-pentane, hexanes) or were
freshly distilled prior to use according to standard literature
procedures20 and stored in a glovebox (tBuOMe, MeCN, MeOH,
1,2-dichloroethane). Silica columns were run in the open air using 60-
230 μm silica (Dynamic Adsorbents). All reactions were conducted
under an atmosphere of either N2 or Ar.
charge at tungsten, these complexes manifest variations in color
that suggest fundamental differences in electronic structure.
The study reported herein has revealed that changes in the
dithiolene ligand redox level underpin these patent differences
in electronic absorption spectra. Not only are these differences
quantifiable by crystallography, but they are readily correlated
to sulfur K-edge XAS data and to ligand CCchelate stretching
frequencies in the resonance Raman spectra. At the extremes in
this series of isoelectronic complexes, the differences in
apparent dithiolene reduction are equivalent to changes
observed upon full electron transfer in homoleptic dithiolene
complexes in which the redox active MO is known to be
dithiolene-based. Since the only parameter varied in this series
is the nature and identity of the ancillary ligands L, this effect is
referred to here as an “ancillary ligand effect” on dithiolene
redox noninnocence.
That ancillary ligands, while “innocent,” may play an
important role in deciding the electronic structure of a redox-
active ligand(s) coordinated to a transition metal is an aspect
on the topic of noninnocent systems that is eliciting increasing
notice. Switching between FeII and FeIII attended by offsetting
dithiolene redox chemistry has been noted in 5-coordinate
Fe(dithiolene)2L complexes as a function of L.15 Other
examples include five-coordinate bis(imino)pyridine complexes
of Mn that may be either MnII or MnI depending on whether
the remaining two ligands are THF or CO, respectively,16 and
the observation of substantial changes in Ir-corrole covalency as
a function of the natures of the axially bound pyridyl ligands.17
In conjunction with a fundamental interest in the interplay
between metal and ligand in the electronic structure of these
compounds, the practical consideration that ancillary ligands
may exert a substantial tuning effect upon the redox behavior of
a designed redox noninnocent ligand, and consequently the
reactivity of its complex with a metal, motivates this research.
[W(S2C2Me2)2(CNtBu)2], 2. To a solution of 1 (0.20 g, 0.42 mmol)
in CH2Cl2 (20 mL) was slowly added CNtBu (0.25 mL, 0.18 g, 2.21
mmol) via syringe at ambient temperature. During the course of this
addition, the evolution of gaseous CO was observed, and the color of
the reaction mixture turned from magenta to a dark orange hue.
Stirring was continued at room temperature for 4 h, whereupon the
solvent was removed under reduced pressure to afford a dark orange
solid residue, which was washed with n-pentane (2 × 5 mL) and dried
overnight. Yield: 0.19 g, 0.32 mmol, 76%. 1H NMR (δ, ppm in
CDCl3): 1.54 (s, 18H, C(CH3)3), 2.68 (s, 12H, dithiolene CH3). IR
(KBr, cm−1): 2125 (vs, CN), 1520 (vs, CC). Absorption spectrum
(CH2Cl2) λmax nm (εM): ∼272 (sh, 29200), 367 (14400), 459
(15900), ∼510 (sh, 11700). [W(mdt)2(CNtBu)2] + e− → [W-
(mdt)2(CNtBu)2]1−, −1.25 V; [W(mdt)2(CNtBu)2]1− + e−
→
[W(mdt)2(CNtBu)2]2−, −1.59 V. MALDI MS: m/e C18H30N2S4W:
586.081, Observed: 586.022. Anal. Calcd for C18H30N2S4W: C, 36.86;
H, 5.16; N, 4.78. Found: C, 37.09; H, 5.12; N, 4.81.
[W(S2C2Me2)2(CO)(IMes)], 3. A CH2Cl2 solution of 1,3-bis(2,4,6-
trimethylphenyl)imidazol-2-ylidene (IMes) (0.13 g, 0.42 mmol) was
transferred via cannula to a solution of 1 (0.20 g, 0.42 mmol) in
CH2Cl2 (10 mL) at ambient temperature. The resulting dark orange
reaction mixture was stirred for 12 h at ambient temperature,
whereupon the solvent was removed under reduced pressure, and the
remaining solid was washed with hexanes (3 × 5 mL) and dried under
a vacuum. Yield: 0.22 g, 0.29 mmol, 69%. 1H NMR (δ, ppm in
CD2Cl2): 2.03 (s, 12H, mesityl CH3), 2.25 (s, 6H, mesityl CH3), 2.40
(broad singlet, 12H, dithiolene CH3), 6.71 (s, 4H, aromatic C−H),
6.92 (s, 2H, CHCH). IR (KBr, cm−1): 1929 (vs, CO), 1936 (vs,
CO), 1519 (CC, dithiolene). Absorption spectrum (CH2Cl2) λmax
nm (εM): 370 (15900), 465 (18900), 532 (sh, 8210). [W(mdt)2(CO)-
(IMes)] − e− → [W(mdt)2(CO)(IMes)]1+, 0.67 V (qr); [W-
(mdt)2(CO)(IMes)] + e− → [W(mdt)2(CO)(IMes)]1−, −1.43 V.
(MALDI MS: m/e for C30H36N2OS4W: 752.122, observed: 724.156
(M-CO). Anal. Calcd for C30H36N2OS4W: C, 47.87; H, 4.82; N, 3.72.
Found: C, 47.72; H, 4.82; N, 3.83.
EXPERIMENTAL SECTION
■
Syntheses. The numbering system by which all compounds are
hereafter identified is defined in Chart 1 and Scheme 2. The starting
materials [W(CO)3(MeCN)3]18 and [Ni(mdt)2]14 were synthesized
according to literature procedures, as were compounds 1,14 4a,14 5a,14
[W(S2C2Me2)2(CO)(PPhMe2)], 4b. A 50 mL flask with magnetic stir
bar was charged with 1 (0.406 g, 0.853 mmol) under an Ar
atmosphere. Toluene (20 mL) was added, and the mixture was stirred
for 10 min, whereupon PMe2Ph (0.117 g, 0.853 mmol) in toluene (5
mL) was added via syringe. The resulting reaction mixture was stirred
for 4 h at ambient temperature. The solvent was removed under
reduced pressure, and the residue was washed with 3 × 5 mL of
pentane and dried. Yield: 0.324 g, 0.552 mmol, 65%. 1H NMR (δ, ppm
in CDCl3): 7.22−7.20 (m, 3H, Ph), 6.83−6.78 (m, 2H, Ph), 2.74 (s,
12H, Me), 2.11 (d, JPH = 9.2 Hz, 6H, PMe2Ph). 13C NMR (δ, ppm in
CDCl3): δ 231.53 (s, CO), 155.20 (br s, CC), 136.81 (s, Ph),
136.31 (s, Ph), 131.65 (s, Ph), 130.37 (s, Ph), 128.7 (s, Ph), 24.085 (s,
Me), 20.27 (d, JPC = 35.8 Hz, PMe2Ph). 31P NMR (δ, ppm in CDCl3):
δ 11.22 (s, JPW = 220.7 Hz). IR (KBr, cm−1): 2905 (w), 1936 (vs, CO),
1435 (m), 1290 (m), 1085 (m), 901 (vs), 736 (s), 485 (s). Anal. Calcd
for C17H23OPS4W: C, 34.81; H, 3.95; P, 5.28. Found: C, 34.87; H,
3.91; P, 5.40.
Chart 1. Compound Number Scheme
[W(S2C2Me2)2(CO)(PMePh2)], 4c. Solutions of 1 (0.345 g, 7.2
mmol) and PMePh2 (0.269 mL, 0.290 g, 14.4 mmol) were separately
prepared in toluene (30 and 10 mL, respectively). The phosphine
solution was transferred to the stirring solution of the 1 via cannula,
which induced an immediate color change from magenta to red. A
reflux condenser was affixed to the reaction vessel, and the mixture was
refluxed for 3 h, during which time the color assumed more of a
reddish-orange aspect. The solvent was evaporated under reduced
a
IMes = 1,3-bis(2,4,6-trimethylphenyl)imidazol-2-ylidene.
B
dx.doi.org/10.1021/ic4009174 | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
a
Scheme 2. Syntheses and Numbering System for Compounds
a
IMes = 1,3-bis(2,4,6-trimethylphenyl)imidazol-2-ylidene.
pressure, and a reddish colored solid was obtained. This solid residue
was purified on a silica column eluted with a mixture of CH2Cl2/
hexanes (2:3). Complex 4c was readily crystallized from the eluant as
dark reddish-orange crystals. Yield: 0.292 g, 0.45 mmol, 62%. 31P
NMR (δ, ppm in C6D6): 26.17 (s, JPW = 221 Hz). IR (KBr, υ in cm−1):
1938 υCO. UV−vis (CH2Cl2), λmax (εM): 482 (51500), 366 (27200).
Anal. Calc. for C22H25OPS4W: C, 40.75; H, 3.89; P, 4.78; S, 19.78.
Found: C, 40.32; H, 3.78; P, 3.04; S, 19.90.
[W(S2C2Me2)2(dppe)], 5c. A 100 mL Schlenk flask with stir bar was
charged with 1 (0.200 g, 0.42 mmol) and toluene (30 mL), while a
solution of dppe (0.167 g, 0.42 mmol) in 10 mL of toluene was
prepared in a separate flask. The dppe solution was transferred to that
of 1, which induced an immediate color change from magenta to red.
A reflux condenser was fitted to this reaction flask, and the mixture was
refluxed for 3 h, during which time the color turned to dark red. The
solvent was evaporated under reduced pressure, and a reddish-colored
solid was obtained. This solid residue was purified on a silica column
eluted with a mixture of CH2Cl2/hexanes (1:4). Complex 5c was
recrystallized by diffusion of pentane vapor into a concentrated
solution in benzene. Yield: 0.180 g, 0.220 mmol, 52%. MALDI MS: m/
e for C34H36P2S4W: 818.6972, observed: 818.856 (M+). Anal. Calcd for
[W(S2C2Me2)2(dppe)]·5/6C6H6, C39H41P2S4W: C, 53.00; H, 4.67; P,
7.01. Found: C, 53.02; H, 4.82; P, 6.73.
[W(S2C2Me2)2(CO)(PtBuPh2)], 4d. The procedure described for the
synthesis and purification of 4c was employed on a scale employing
0.200 g (0.42 mmol) of 1 and 0.101 g (0.42 mmol) of PtBuPh2.
Complex 4d readily crystallized from the chromatography eluant as
dark red column-shaped crystals. Yield: 0.185 g, 0.254 mmol, 61%. IR
(KBr, υ in cm−1): 1936 υCO. MALDI MS: m/e for C25H31OPS4W:
690.5983; observed: 661.838 (M-CO)+, 419.604 (M-CO-PtBuPh2)+.
Anal. Calc. for C25H31OPS4W: C, 43.48; H, 4.52; P, 4.49. Found: C,
43.79; H, 4.51; P, 4.35.
[Et4N]2[W(S2C2Me2)2(CN)2], [Et4N]2[6]. A 50 mL Schlenk flask under
N2 was charged with 1 (0.100 g, 0.21 mmol) and 20 mL of dry
dichloromethane added via syringe. Under outward nitrogen flow,
[Et4N][CN] (0.066 g, 0.42 mmol) was added in one portion to the
solution in the flask. Vigorous evolution of bubbles and a color change
from reddish purple to orange were observed immediately. The
resulting reaction mixture was stirred for another 2 h, and the solvent
was then removed under reduced pressure. The residue was washed
with diethyl ether (3 × 10 mL). Orange, block-shaped crystals that
were suitable for X-ray diffraction were grown by diffusion of tert-butyl
methyl ether into an acetonitrile solution. Yield: 0.098 g, 0.120 mmol,
57%. 1H NMR (δ, ppm in CD2Cl2): 1.16 (t, 24H, NCH2CH3), 2.51 (s,
12H, dithiolene CH3), 3.05 (q, 16H, NCH2CH3). IR (KBr, cm−1):
2081 (vs, CN), 1556 (weak, dithiolene CC symmetric), 1480
(strong, CC, asymmetric), 1458 (strong, CC, asymmetric).
Absorption spectrum (CH2Cl2) λmax nm (εM): 383 (34800), ∼404 (sh,
[W(S2C2Me2)2(CO)(PPh3)], 4e. A Schlenk flask with stir bar was
charged with portions of 1 (0.106 g, 0.223 mmol) and PPh3 (0.117 g,
0.449 mmol) under a N2 atmosphere. Freshly distilled benzene (15
mL) was added, and the mixture was heated to 85 °C with stirring for
12 h. After being cooled to room temperature, the solvent was
removed from the reaction mixture under reduced pressure. The
resulting crude product was purified on a silica chromatography
column eluted first with hexanes to remove [W(S2C2Me2)3].
Continued elution with Et2O enabled movement and collection of
4e from the column. After evaporation of the eluant under reduced
pressure and drying under a vacuum, 4e was obtained as a dark orange
product. Crystalline samples were prepared by the diffusion of n-
pentane vapor into a solution in Et2O. Yield: 0.129 g, 0.182 mmol,
1
81%. H NMR (δ, ppm in CDCl3): 7.28 (br s, 12H, Ph), 6.51 (br s,
3H, Ph), 2.72 (s, 12H, Me). 13C NMR (δ, ppm in CDCl3, 25 °C):
230.62 (CO), 156.18 (CC), 135.21 (s, Ph), 135.12 (s, Ph), 130.60
(s, Ph), 128.53 (s, Ph), 128.43 (s, Ph), 24.15 (s, CH3). 31P NMR (δ,
ppm in CD2Cl2): 39.58 (s, JPW = 222.4 Hz). IR (KBr, cm−1): 2909
(w), 1949 (vs, CO), 1432 (m), 1090 (s), 927 (w), 743 (m). Anal.
Calcd for C27H27OPS4W: C, 45.64; H, 3.83. Found: C, 45.80; H, 3.79.
[W(S2C2Me2)2(CO)(P(NMe2)3)], 4f. The procedure executed for the
synthesis of 4f was similar to that described for 4b. The reaction scale
involved 0.100 g (0.21 mmol) of 1 and excess P(NMe2)3 (0.120 mL,
0.66 mmol). Compound 4f was obtained as red plate crystals by the
diffusion of n-pentane vapor into a CH2Cl2-Et2O solution. Yield: 0.015
g, 0.025 mmol, 12%. 1H NMR (δ, ppm in CD2Cl2): 2.26 (very broad,
18H, (P(N(CH3)2)3), 2.65 (s, 12H, dithiolene CH3) Absorption
spectrum (CH2Cl2) λmax nm (εM): 356 (14700), 481 (16200). MALDI
MS: m/e for C15H30N3OPS4W: 611.052, observed: 567.220 (M-CO-
Me).
33100), 494 (17200). [W(mdt)2(CN)2]2−
+ → [W-
e−
(mdt)2(CN)2]3−, −1.25 V; [W(mdt)2(CN)2]3− + e− → [W-
(mdt)2(CN)2]4−, −1.85 V. Anal. Calcd for C30H58N6S4W: C, 44.22;
H, 7.17; N, 10.31. Found: C, 44.13; H, 6.96; N, 10.16.
[W(S2C2Me2)(CO)2(PPhMe2)2], 7b. A 50 mL Schlenk flask with
magnetic stir bar was charged with 9 (0.052 g, 0.127 mmol), PMe2Ph
(0.035 g, 0.255 mmol), and 10 mL of toluene. The reaction mixture
was refluxed for 4 h, whereupon the solvent was removed under
reduced pressure. The dark orange solid residue was washed with n-
pentane and dried under a vacuum for 24 h. Yield: 0.058 g, 0.091
1
mmol, 73%. H NMR (δ, ppm in CDCl3): δ 7.24−7.22 (m, 6H, Ph),
7.14−7.09 (m, 4H, Ph), 2.53 (s, 6H, Me), 2.12 (d, JPH = 9.6 Hz, 12H,
PMe2Ph). 31P NMR (δ, ppm in CDCl3): δ 6.41 (s, JPW = 183.66 Hz).
IR (KBr, cm−1): 2965 (w), 1924 (vs, CO), 1826 (vs, CO), 1433 (m),
1256 (m), 1086 (s), 1021 (s), 909 (s), 792 (m), 698 (m).
C
dx.doi.org/10.1021/ic4009174 | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Table 1. Unit Cell and Refinement Data for Compounds 2, 3, 4cdef, 5bc, [Et4N]2[6], 7bcd, 8, [Et4N][W(S2C2Me2)3]
compound
2
3
4c
4d
4e
solvent
formula
fw
none
none
none
none
none
C18H30N2S4W
586.53
C30H36N2OS4W
752.70
C22H25OPS4W
648.48
C25H31OPS4W
690.56
C27H27OPS4W
710.55
xtl system
space grp
color, habit
a, Å
monoclinic
P21/n
monoclinic
P21/c
monoclinic
P21
monoclinic
P21/n
triclinic
P1
̅
red slat
15.7588(16)
8.8549(9)
17.5588(18)
90
red column
12.991(3)
14.996(3)
15.949(4)
90
orange plate
10.323(3)
9.456(3)
13.370(4)
90
red plate
9.4584(7)
16.915(1)
16.942(1)
90
brown block
9.878(3)
10.203(3)
14.082(4)
79.445(4)
81.466(4)
78.219(3)
1356.7(7)
100
b, Å
c, Å
α, deg
β, deg
γ, deg
V, Å3
106.355(1)
90
98.188(3)
90
110.902(4)
90
99.637(1)
90
2351.1(4)
100
3075.4(12)
100
1219.2(6)
100
2672.2(4)
100
T, K
Z
4
4
2
4
2
R1, wR2
GOF
0.0302, 0.0729
1.026
0.0212, 0.0494
1.035
0.0167, 0.0413
0.807
0.0211, 0.0431
1.072
0.0251, 0.0653
1.060
compound
4f
5b
5c
[Et4N]2[6]
7b
solvent
formula
fw
none
none
1.5C6H6
C43H45P2S4W
935.82
2MeCN
none
C15H30N3OPS4W
611.48
C24H34P2S4W
696.54
C30H58N6S4W
814.91
C22H28O2P2S2W
634.35
monoclinic
P21/n
xtl system
space grp
color, habit
a, Å
monoclinic
C2/c
monoclinic
P21/c
monoclinic
P21/n
triclinic
P1
̅
red plate
27.292(6)
9.699(2)
17.754(4)
90
yellow plate
19.413(4)
9.0400(17)
15.947(3)
90
red plate
14.3288(7)
16.5017(8)
17.5022(9)
90
orange block
9.431(3)
13.622(5)
15.748(6)
99.509(5)
92.373(5)
109.995(5)
1864.6(11)
100
orange plate
9.516(4)
15.328(6)
16.853(7)
90
b, Å
c, Å
α, deg
β, deg
γ, deg
V, Å3
101.431(3)
90
104.240(2)
90
99.2550(10)
90
97.849(7)
90
4606.4(16)
100
2712.7(9)
100
4084.5(4)
150
2435.0(16)
100
T, K
Z
8
4
4
2
4
R1, wR2
GOF
0.0250, 0.0491
1.061
0.0260, 0.0491
1.026
0.0329, 0.0712
1.071
0.0503, 0.1316
1.048
0.0182, 0.0436
1.041
compound
7c
7d
8
[Et4N][W(S2C2Me2)3]
solvent
formula
fw
none
none
none
1/2C6H6
C23H41NS6W
707.78
C32H32O2P2S2W
C42H36O2P2S2W
882.62
C28H30N2O3S2W
690.51
758.49
xtl system
space grp
color, habit
a, Å
monoclinic
P21/n
monoclinic
P21
triclinic
monoclinic
P21/n
P1
̅
orange parallelpiped
14.6817(7)
11.2860(5)
19.6091(9)
90
orange plate
18.5335(13)
9.9638(10)
21.866(3)
90
brown slab
8.1260(15)
9.2813(17)
18.779(3)
81.487(2)
83.283(2)
81.982(2)
1380.3(4)
100
blue plate
9.5750(11)
24.536(3)
12.7760(15)
90
b, Å
c, Å
α, deg
β, deg
γ, deg
V, Å3
107.800(1)
90
114.207(9)
90
106.042(1)
90
3093.6(2)
100
3682.9(6)
100
2884.6(6)
100
T, K
Z
4
4
2
4
ab
,
bc
,
R1, wR2
0.0282, 0.0546
1.191
0.0377, 0.0911
0.0209, 0.0511
0.0434, 0.0773
1.046
d
GOF
1.062
1.070
a
b
c
2
2
R1 = Σ∥Fo| − |Fc∥/Σ|Fo|. R indices for data cut off at I > 2σ(I). wR2 = {[Σw(Fo2 − Fc2)2/Σw(Fo )2}1/2; w = 1/[σ2(Fo ) + (xP)2 + yP], where P =
d
(Fo + 2Fc2)/3. GOF = {Σ[w(Fo − Fc2)2]/(n − p)}1/2, where n = number of reflections and p is the total number of parameters refined.
2
2
[W(S2C2Me2)(CO)3(IMes)], 8. The procedure employed was
analogous to that described for the synthesis of compound 3. The
scale employed for the reaction involved 0.050 g (0.12 mmol) of 9 and
excess IMes (0.120 g, 0.39 mmol). Crystallization of 8 as brown, slab-
shaped crystals was effected by diffusion of hexanes into a 1,2-
1
dichloroethane solution. Yield: 0.041 g, 0.059 mmol, 49%. H NMR
(δ, ppm in CD2Cl2): 1.95 (s, 12H, mesityl CH3), 2.25 (s, 6H, mesityl
CH3), 2.31 (s, 6H, dithiolene CH3), 6.80 (s, 2H carbene CHCH),
D
dx.doi.org/10.1021/ic4009174 | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
7.00 (s, 4H, aromatic C−H). IR (KBr, cm−1): 1992 (vs, CO), 1907
(vs, CO), 1884 (vs, CO), 1859 (s, CO), 1562 (m), 1481 (m), 612
(m). Absorption spectrum (CH2Cl2) λmax nm (εM): ∼362 (sh, 15500),
summarized in Table 1. All the bis(dithiolene) [W-
(S2C2Me2)2L2]n compounds are trigonal prismatic with non-
dithiolene ligands arranged is cis configuration along one of the
long edges of a rectangular face. Gross structural features, such
as Bailar twist angles and dihedral angles between dithiolene
ligands, are typical of such complexes (Table S2). Thermal
ellipsoid plots of 3 and [6]2− suffice to represent all the
bis(dithiolene) complexes (Figure 1). What is noteworthy
433 (16800), ∼597 (sh, 4530). [W(mdt)(CO)3(IMes)] − e−
→
[W(mdt)(CO)3(IMes)]1+, −0.74 V (ir); [W(mdt)(CO)3(IMes)] + e−
→ [W(mdt)(CO)3(IMes)]1−, −1.16 V (qr). MALDI MS: m/e
C28H30O3N2S2W: 690.121, observed: 635.372 (M-2CO), 606.097
(M-3CO).
Physical Methods. All Raman spectra were acquired with a Nexus
FT-Raman module with a 1064 nm excitation wavelength. All samples
were prepared by smearing a solid sample of the complex onto paper,
which then was mounted onto a magnetic plate placed in the focus of
the laser. To reduce the absorbance background and avoid
photobleaching, spectra were collected at a laser power ranging from
25 mW to 300 mW. The spectra presented are an average of 3000
scans, which subsequently improved the signal-to-noise ratio.
Electronic absorption spectra were obtained at ambient temperature
with a Hewlett-Packard 8452A diode array spectrophotometer, while
IR spectra were taken as pressed KBr pellets with a Thermo Nicolet
Nexus 670 FTIR instrument in absorption mode. All NMR spectra
were recorded at 25 °C with a Varian Unity Inova spectrometer
operating at 400, 100.5, or 161.8 MHz for 1H, 13C, and 31P,
respectively, and referenced to the solvent residual. Electrochemical
measurements were obtained with a CHI620C electroanalyzer
workstation using a Ag/AgCl reference electrode, a platinum disk
working electrode, Pt wire as auxiliary electrode, and [nBu4N][PF6] as
the supporting electrolyte. Under these conditions, the [Cp2Fe]+/
Cp2Fe couple consistently occurred at +540 mV. X-ray absorption
spectra were measured at the Stanford Synchrotron Radiation
Lightsource (SSRL) under ring conditions of 3.0 GeV and 60−100
mA. Sulfur K-edge data were obtained using the 20-pole wiggler
beamline 4-3 and 54-pole wiggler beamline 6-2. Tungsten L-edge
spectra were measured on the 20-pole wiggler beamline 7-3, as
previously described.21 Elemental analyses were performed by
Midwest Microlab, LLC of Indianapolis, IN, or by Canadian
Microanalytical of Delta, British Columbia. Descriptions of the
procedures for collection of X-ray diffraction data, the solution and
refinement of crystal structures, and all computational details are
deferred to Supporting Information.
Figure 1. Thermal ellipsoid plots at the 50% probability level for 3
(top) and [6]2− (bottom). Hydrogen atoms are omitted for clarity.
RESULTS AND DISCUSSION
■
The synthesis of [W(S2C2Me2)2L2]n (L = tBuNC, n = 0, 2; L =
CN−, n = 2−, [6]2−) readily proceeds by straightforward ligand
substitution from [W(S2C2Me2)2(CO)2] (1), the preparation
of which was initially described by Schrauzer22 but materially
improved by Holm and Goddard23 (Scheme 2). The analogous
bis(phosphine) complexes, except that with dppe, require
elevated temperatures and the use of Me3NO as facilitating
oxidant for the liberation of CO as CO2. An enhanced W−CO
backbonding interaction following substitution with a single
phosphine ligand apparently necessitates forcing conditions for
displacement of the second CO ligand. The π-acidic nature of
both CNtBu and CN−, in contrast, appears to render the CO
ligand quite labile toward substitution in the presumed
monocarbonyl intermediates, [W(S2C2Me2)2(CO)(CNtBu)]
and [W(S2C2Me2)2(CO)(CN)]1−, which were not isolable.
The monocarbonyl compounds [W(S2C2Me2)2(CO)L] (L =
PR3, carbene) form immediately with 1 equiv of L, excess L
alone being ineffective in producing further substitution. All
compounds occur nominally at the same total redox level and
differ only in the character of the non-dithiolene ligands. A
more limited set of compounds of the type [W(S2C2Me2)-
(CO)2LL′] (L = L′ = phosphine, 7; L = carbene, L′ = CO, 8)
has been prepared from [W(S2C2Me2)(CO)4] and is included.
With the exception of 4b, all the compounds in Chart 1 have
been structurally identified in this work or previously.14,24,25
Unit cell, refinement data, and other crystallographic data are
about the bis(dithiolene) compounds, when considered as a set,
are dithiolene SC and CCchelate interatomic distances (Table 2)
that increase and decrease, respectively, as the table is traversed
from top toward bottom. Although juxtaposed pairs of
compounds in the table do not have SC and CCchelate distances
with uncertainties that meet the 3σ criterion that is indicative of
a statistically significant difference,26 1 and [6]2−, which define
the two extremes in this set of bis(dithiolene) compounds, do
have meaningfully different SC bond lengths. These distances
inversely correlate with the CCchelate bond lengths in a linear
fashion (Figure S1) and are indicative of the redox level of the
dithiolene ligand. Scheme 1 summarizes the canonical forms of
the dithiolene ligand with corresponding intraligand bond
lengths. With reference to the values in Scheme 1, the data in
Table 2 imply radical monoanionic ligands in 1 and fully
reduced ene-1,2-dithiolate dianions in [6]2−. The conclusion
that emerges from the metric parameters in Table 2 is that the
variable non-dithiolene ligands, referred to hereafter as the
“ancillary” ligands, exert an important effect upon the redox
state of the dithiolene ligand, and possibly the tungsten atom as
well, even though all compounds exist with a metallodithiolene
fragment that may be described with a common charge
formalism. A similar situation likely pertains to the related
mono(dithiolene) compounds 7−9, but a suitably broad set of
compounds with discernible structural differences is unavail-
able.
E
dx.doi.org/10.1021/ic4009174 | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Table 2. Dithiolene Redox Level in W Bis(dithiolene) Compounds Gauged by Crystallographic and Spectroscopic Markers
a
bond lengths, Å
sulfur K-edge XAS
υCC chelate, cm−1
b
c
d
d
SC
CCchelate
pre-edge (eV)
α2, LUMO
edge (eV)
Raman
infrared
1
1.722[2]
1.732[2]
1.733[1]
1.733[1]
1.743[1]
1.753[3]
1.74[1]
1.363[3]
1.360[4]
1.354[2]
1.354[2]
1.346[2]
1.339[6]
1.37[1]
2470.8, 2472.4
2471.1, 2472.4
67
57
2474.2
2473.9
1491.5 (18.1), 1476.0 (4.0)
1470.6
1520.5
1522.1
1511.5
2
1523.7
3
1523.1
4a14
2471.0, 2472.3
2471.2, 2472.2
2471.3, 2472.3
54
46
48
2474.0
2473.8
2473.4
2473.9
2473.4
1515.5
5a14
1555.7 (10.4), 1539.0 (4.8)
1556.8
1553.3 (0.4), 1538.9 (1.1)
1557.0
[6]2−
[W(S2C2Me2)3]14
1487.0 (6.0), 1458.9 (25.0)
[W(S2C2Me2)3]1−
1.743[3]
1.763[1]
1.754[4]
1.736[3]
1.744[4]
1.354[5]
1.337[2]
1.349(3)
1.348(4)
1.351(8)
20
[W(S2C2Me2)3]2−
7a14
8
914,21
2473.4
a
Uncertainties in averaged bond lengths were determined using the general formula for uncertainty in a function of multiple variables as detailed by
b
Taylor, J. R. An Introduction to Error Analysis; University Science Books: Sausalito, California, 1997; pp 73−77. α2 = covalency expressed as %.
c
d
Defined as the inflection point in the rising edge, determined from the second derivative of the intensity vs energy plot. Two peaks visible owing to
symmetric and asymmetric stretches; parenthetical values are relative intensities in arbitrary units.
The varying levels of dithiolene ligand reduction indicated by
the structural data in Table 2 have been probed by sulfur K-
edge X-ray absorption spectroscopy (XAS). This method
involves transitions of core 1s electrons of sulfur to acceptor
MOs with a degree of sulfur p character and, at high enough
energy, their promotion into the continuum. The energy at
which the latter excitation occurs directly correlates with Zeff at
sulfur and thereby can distinguish between reduced thiolate-
type sulfur and an oxidized radical-type sulfur27 (Scheme 1,
panels a and b, respectively). The sulfur K-edge XAS for 1 and
[6]2− are overlaid in Figure 2, and their rising edge energies as
well as those of other representative compounds from the set
are presented in Table 2. The rising edge energy for 1 is 0.8 eV
higher in energy than that for [6]2−, which accords with a
higher Zeff arising from its diminished thiolate character. This
difference is comparable to the rising edge energy change (∼0.5
eV) observed in reducing [W(S2C2Me2)3] to its corresponding
monoanion (Table 2). By analogy to the related molybdenum
system,12 the reduction of [W(S2C2Me2)3] to [W-
(S2C2Me2)3]1− involves the addition of an electron to an MO
comprised equally of the three dithiolene ligands such that the
remaining “hole” is distributed equally among six sulfur atoms.
The magnitude of the reduction seen at the dithiolene sulfur
atoms in going from 1 to [6]2− is surprising in view of the ease
with which both CO ligands in 1 are displaced by CN−, which
would seem to suggest negligible difference between the two
ligands in the nature and degree of their interaction with
tungsten.
The pre-edge features in the sulfur K-edge XAS of 1 and
[6]2− are also informative of the fundamental differences
between them. The lowest energy absorption is assigned as a
sulfur 1s → LUMO transition in these complexes as well as in
2, 4a, and 5a. The intensity of these features correlates to the
degree of sulfur p character in the composition of the LUMO, a
greater contribution by the sulfur p orbitals being indicative of a
higher degree of covalency (mixing) between metal and ligand.
Deconvolution of the lowest energy features in the sulfur K-
edge XAS from overlapping features (Figure S2), followed by
integration of the absorption intensity (Do), permits a
quantitative assessment of the covalency (α2) to the acceptor
orbital via the relationship Do = (α2hIs)/3n (h = # holes in
acceptor orbitals; Is = radial transition dipole integral =14.22; n
= # absorbing S atoms).12 Expressed as percentages, covalencies
for the LUMOs for 1, 2, 4a, 5a, and [6]2− range from 67 to
46% (Tables 2 and S3). These values comport with the rising
edge energies in showing that the fundamental difference
between 1 and [6]2− is greater ionic character (less covalency)
to the latter, which is equivalent to saying that the dithiolene
sulfur atoms in [6]2− are more thiolate-like in nature.
The bond order changes implied by the variations in CS and
CCchelate distances can be separately gauged by vibrational
spectroscopy. The effectiveness of this technique in systems of
this kind has been demonstrated by its use in identifying
Figure 2. Sulfur K-edge (top), W L1-edge XAS for 1, [6]2− (bottom).
F
dx.doi.org/10.1021/ic4009174 | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Figure 3. Comparison of the frontier MOs for 1 (left) and [6]2− (right). The pseudo C3 axis of the trigonal prism is taken as the z axis. Except for
the HOMO of [6]2−, all orbital images (shown at the 0.05 contour level) are presented in the same relative orientation of the complexes.
dithiolene ligation to molybdenum in R. sphaeroides DMSO
reductase.28 The direct relationship between bond stretching
frequency and interatomic force constant, and therefore
interatomic bond order, provides independent but comple-
menting insight into the ligand redox level. Here, symmetric
CC bond stretching frequencies observed by Raman spectros-
copy have been more accurately identified than their
asymmetric modes in the IR, as the latter technique is
complicated by the occurrence of strong interfering modes.
Table 2 summarizes the SC and CC vibrational frequencies for
compounds 1-[6]2−, while Figure S3 presents a plot of υCC
against the crystallographically determined CCchelate distances.
The relationship is monotonic and, allowing for the
uncertainties in the measurements, can be fit by a linear
function.
As part of a DFT analysis of compounds 1−8, Mulliken
charges associated with the dithiolene ligand have been assessed
and plotted against both the symmetric CC stretching
frequencies and the rising edge energies in the sulfur K-edge
X-ray absorption spectra (Figure S4). The linearity seen in
these plots further affirms that these dithiolene ligand variations
are real. With only one of the following values in hand −
crystallographic CCchelate bond length, υCC from resonance
Raman, rising edge energy in the sulfur K-edge XAS or
computed Mulliken charge, the plots in Figures S3 and S4
provide a basis for estimating any of the remaining three.
A Mulliken population analysis of the frontier MOs in the
optimized structures reveals changes in their composition that
underpin the structural and spectroscopic data. Table S4
summarizes changes in the composition of the HOMO and
LUMO, while Figure 3 illustrates these changes with orbital
images for 1 and [6]2−, which are the terminal members of this
series continuum and the two complexes that offer the clearest
contrast. As noted by Holm and co-workers,19 the frontier MOs
for 1 are highly mixed between metal and ligands in a way that
does not provide a straightforward d electron count at tungsten.
Its HOMO is comprised of dithiolene π* orbital and tungsten d
orbital in a ∼64:21 ratio. The HOMO of [6]2− differs sharply in
2
being principally the dz orbital (68%). The lower lying
HOMO-1 and HOMO-2 of [6]2− are similar to one another
(Figure 3) and are primarily filled dithiolene π* orbitals with
modest metal character and therefore support the description
of [6]2− as a WIV d2 species. The LUMO for [6]2− contrasts
with that for 1 in having visibly less dithiolene π character
(∼31% vs 62%) and proportionately more tungsten d character
(49% vs 18%), a difference which is manifested in the lesser
intensity of its first pre-edge feature in the sulfur K-edge XAS as
compared to that of 1.
Figure 3 reveals that the essential difference between 1 and
[6]2− is the suitable energy match, and therefore effective
mixing, between the CO π* orbitals and the tungsten d orbital
manifold and π system of the dithiolene ligand, as compared to
the CN− π* orbitals, which are too high in energy. The poor
overlap between the CN− π* and tungsten d orbitals in the MO
images for [6]2− in Figure 3 is emphasized. In contrast to [6]2−,
1 has two filled MOs with significant W → CO π* backbonding
character. One of these is the HOMO-2, which effectively
overlaps with the tungsten dxz, draws electron density away
from tungsten, and raises Zeff at tungsten. Even more important
to the understanding of 1 is the HOMO-1, which is comprised
of filled dithiolene π* orbitals and empty CO π* orbitals
interacting with the tungsten dxy orbital. The consequence of
this particular makeup is effective transfer of charge density
from dithiolene to CO via tungsten, thereby oxidizing the
dithiolene ligands in 1 as compared to [6]2−, as has been
evidenced by several physical methods. A natural inclination is
to presume that, if the dithiolene ligands are appreciably more
oxidized in 1 than in [6]2−, then tungsten is commensurately
G
dx.doi.org/10.1021/ic4009174 | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
more reduced. Importantly, Figure 3 shows that the relatively
oxidized condition of the dithiolene ligands occurs not at the
expense of tungsten but rather is mediated by tungsten d orbitals
and driven by the π-acidic nature of the CO ligands. In support of
this interpretation, Figure 2 presents an overlay of the tungsten
L1-edge X-ray absorption spectra for 1 and [6]2− and shows a
rising edge energy for 1 that is comparable to, even slightly
higher in energy, than that for [6]2−. These values correlate
directly with Zeff at metal and indicate that, at least as this
method can assess, the spectroscopic oxidation state of
tungsten in both complexes is closer to WIV than otherwise.
AUTHOR INFORMATION
■
Corresponding Author
*E-mail: donahue@tulane.edu.
Present Addresses
#(Y.Y.) Department of Chemistry, Princeton University,
Princeton, New Jersey 08544.
○(U.J.) Department of Chemistry, University of Illinois at
Chicago, 845 West Taylor Street, MC 111, Chicago, Illinois,
60607.
Notes
The authors declare no competing financial interest.
CONCLUSIONS
■
ACKNOWLEDGMENTS
■
A principal conclusion arising from this study of isostructual,
isoelectronic tungsten bis(dithiolene) complexes is that the
ancillary ligands completing the coordination sphere can exert a
significant effect upon the dithiolene redox level. This influence
likely reflects the combination of ancillary ligand σ-donor
strength with π-acidity, although the former is not as easily
visualized as the latter in the orbital images of Figure 3. The
variations seen in dithiolene redox level, which have been
assessed structurally, spectroscopically, and computationally,
are comparable in magnitude to the changes seen upon full
electron transfer in homoleptic bis- and tris(dithiolene)
complexes where the redox active orbital is dithiolene-based.
Furthermore, the correlation between the data obtained by
different methods provides a basis for gauging dithiolene CS or
CCchelate bond distances, dithiolene υCC, sulfur K-edge XAS
rising edge energy, or Mulliken charge for the dithiolene ligand
if only one of the values is available. Our results demonstrate
that the behavior of a redox noninnocent ligand engineered for
a particular reactivity is subject, at least potentially, to
considerable modulation by choice of ancillary ligand(s).
More immediate to the present work is the observation that
the M(dithiolene)2 (M = Mo, W) fragment is common to a
broad family of molybdenum and tungsten enzymes. Vibra-
tional spectroscopy studies of R. sphaeroides DMSO reductase
suggest differences in the redox level of the two pterin-
dithiolene ligands in the enzyme’s oxidized form, implying a
degree of π-delocalization to one of them.29 While the enzyme’s
redox chemistry is primarily Mo-based, partially oxidized
dithiolene ligand may play a critical role in lowering the
transition state energy during oxygen atom transfer from
substrate to metal. Our results show that ancillary ligands with
π orbitals of the right symmetry and energy (e.g., CO, possibly
NO+ or a suitably functionalized alkyne) have the capacity to
effectively engage the π system of the dithiolene ligand(s) and
render redox noninnocence operative.
Support from NSF (Grant CHE-0845829 to J.P.D., Grant
CHE-1012371 to I.V.R.), from DOE and NIH for support of
the Stanford Synchrotron Radiation Lightsource at which
portions of this research were conducted, from the Max Planck
Society (SD), Cornell University (SD), the Sloan Foundation
(SD), the Louisiana Board of Regents (Grant LEQSF-(2002-
03)-ENH-TR-67) for the Tulane crystallography facility, and
from IBM (fellowship support to Y.Y., C.K.) is gratefully
acknowledged. Support from the EPSRC National UK EPR
Facility and Service is gratefully acknowledged (S.S.).
REFERENCES
■
(1) Chirik, P. J. Inorg. Chem. 2011, 50, 9737−9740.
(2) Heyduk, A. F.; Zarkesh, R. A.; Nguyen, A. Inorg. Chem. 2011, 50,
9849−9863.
(3) Ouch, K.; Mashuta, M. S.; Grapperhaus, C. A. Inorg. Chem. 2011,
50, 9904−9914.
(4) Lippert, C. A.; Riener, K.; Soper, J. D. Eur. J. Inorg. Chem. 2012,
554−561.
(5) Chirik, P. J.; Wieghardt, K. Science 2010, 327, 794−795.
(6) Lim, B. S.; Fomitchev, D. V.; Holm, R. H. Inorg. Chem. 2001, 40,
4257−4262.
(7) Bigoli, F.; Chen, C.-T.; Wu, W.-C.; Deplano, P.; Mercuri, M. L.;
Pellinghelli, M. A.; Pilia, L.; Pintus, G.; Serpe, A.; Trogu, E. F. Chem.
Commun. 2001, 2246−2247.
(8) Kwik, W.-L.; Stiefel, E. I. Inorg. Chem. 1973, 12, 2337−2342.
(9) Sproules, S.; Weyhermuller, T.; DeBeer, S.; Wieghardt, K. Inorg.
̈
Chem. 2010, 49, 5241−5261.
(10) Sproules, S.; Weyhermuller, T.; Goddard, R.; Wieghardt, K.
̈
Inorg. Chem. 2011, 50, 12623−12631.
(11) Kapre, R. R.; Bothe, E.; Weyhermuller, T.; DeBeer George, S.;
̈
Wieghardt, K. Inorg. Chem. 2007, 46, 5642−5650.
(12) Tenderholt, A.; Szilagyi, R. K.; Holm, R. H.; Hodgson, K. O.;
Hedman, B.; Solomon, E. I. Inorg. Chem. 2008, 47, 6382−6392.
(13) Szilagyi, R. K.; Lim, B. S.; Glaser, T.; Holm, R. H.; Hedman, B.;
Hodgson, K. O.; Solomon, E. I. J. Am. Chem. Soc. 2003, 125, 9158−
9169.
́
(14) Chandrasekaran, P.; Arumugam, K.; Jayarathne, U.; Perez, L.
M.; Mague, J. T.; Donahue, J. P. Inorg. Chem. 2009, 48, 2103−2113.
(15) Sproules, S.; Wieghardt, K. Coord. Chem. Rev. 2010, 254, 1358−
1382.
ASSOCIATED CONTENT
■
S
* Supporting Information
(16) Russell, S. K.; Bowman, A. C.; Lobkovsky, E.; Wieghardt, K.;
Chirik, P. J. Eur. J. Inorg. Chem. 2012, 535−545.
(17) Palmer, J. H.; Lancaster, K. M. Inorg. Chem. 2012, 51, 12473−
12482.
Full description of all procedures for crystal growth, diffraction
data collection and processing, and structure solution and
refinement; complete crystallographic data for all new
structures in CIF format; thermal ellipsoid plots with complete
atom labeling; S K-edge X-ray absorption spectra with first and
second derivatives; description of basis sets employed and
computational methods; xyz coordinates for optimized
structures; supplementary figures as noted in text. This material
is available free of charge via the Internet at http://pubs.acs.org.
(18) Tate, D. P.; Knipple, W. R.; Augl, J. M. Inorg. Chem. 1962, 1,
433−434.
(19) Fomitchev, D. V.; Lim, B. S.; Holm, R. H. Inorg. Chem. 2001, 40,
645−654.
(20) Armarego, W. L. F.; Perrin, D. D. Purification of Laboratory
Chemicals, 4th ed.; Butterworth-Heinemann: Oxford, Great Britain,
2000.
H
dx.doi.org/10.1021/ic4009174 | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
(21) Sproules, S.; Benedito, F. L.; Bill, E.; Weyhermuller, T.; DeBeer
̈
George, S.; Wieghardt, K. Inorg. Chem. 2009, 48, 10926−10941.
(22) Schrauzer, G. N.; Mayveg, V. P.; Heinrich, W. J. Am. Chem. Soc.
1966, 88, 5174−5179.
(23) Goddard, C. A.; Holm, R. H. Inorg. Chem. 1999, 38, 5389−
5398.
(24) Sproules, S.; Banerjee, P.; Weyhermuller, T.; Yan, Y.; Donahue,
̈
J. P.; Wieghardt, K. Inorg. Chem. 2011, 50, 7106−7122.
(25) Yan, Y.; Chandrasekaran, P.; Mague, J. T.; DeBeer, S.; Sproules,
S.; Donahue, J. P. Inorg. Chem. 2012, 51, 346−361.
(26) Stout, G. H.; Jensen, L. H. X-Ray Structure Determination: A
Practical Guide; Wiley & Sons: New York, 1989; Chapter 18.
(27) Glaser, T.; Hedman, B.; Hodgson, K. O.; Solomon, E. I. Acc.
Chem. Res. 2000, 33, 859−868.
(28) Gruber, S.; Kilpatrick, L.; Bastian, N. R.; Rajagopalan, K. V.;
Spiro, T. G. J. Am. Chem. Soc. 1990, 112, 8179−8180.
(29) Johnson, M. K. Prog. Inorg. Chem. 2004, 52, 213−266.
I
dx.doi.org/10.1021/ic4009174 | Inorg. Chem. XXXX, XXX, XXX−XXX