Heteroleptic Complexes of Cyclometalated Platinum with Triarylformazanate Ligands

Article abstract of DOI:10.1021/acs.inorgchem.5b02595

Formazanates are a ligand class featuring a 1,2,4,5-tetraazapentadienyl core, with variable substitution at the 1, 3, and 5 positions. Here we describe a set of four heteroleptic cylcometalated platinum complexes containing triarylformazanate ligands. The

Full text of DOI:10.1021/acs.inorgchem.5b02595

Article
pubs.acs.org/IC
Heteroleptic Complexes of Cyclometalated Platinum with
Triarylformazanate Ligands
Evanta Kabir, Chia-Hua Wu, Judy I-Chia Wu, and Thomas S. Teets*
Department of Chemistry, University of Houston, 3585 Cullen Boulevard Room 112, Houston, Texas 77204-5003, United States
S Supporting Information
*
ABSTRACT: Formazanates are a ligand class featuring a
,2,4,5-tetraazapentadienyl core, with variable substitution at
1
the 1, 3, and 5 positions. Here we describe a set of four
heteroleptic cylcometalated platinum complexes containing
triarylformazanate ligands. The complexes are prepared by
metathesis reactions of chloro-bridged dimers [Pt(C∧N)(μ-
Cl)]2 (C∧N = 2-phenylpyridine or 2-(2,4-difluorophenyl)-
pyridine) with triarylformazans in the presence of base. X-ray
diffraction studies reveal the molecular structures of three such
complexes. Cyclic voltammograms and UV−vis absorption
spectra of the complexes show features characteristic of both
the cyclometalated platinum fragment and the formazanate,
with the latter giving rise to two reversible one-electron reductions in the CV and an intense visible π → π* absorption which is
red-shifted by >100 nm relative to the free formazan. The electronic structures and redox properties of the complexes were
further investigated by UV−vis spectroelectrochemistry and density functional theory calculations. All of the experimental and
theoretical work points to a frontier molecular orbital manifold where the formazanate π and π* orbitals are substantially mixed
with d-orbitals derived from the platinum center.
INTRODUCTION
articles detail the coordination chemistry of formazanate
derivatives with mercury(II); these previous examples all report
formazanates with additional o-phenoxy or o-benzoate groups
■
Formazans have been known for over a century, with most of the
early work on these compounds devoted to synthetic procedures,
structural elucidation, and evaluation of their biomedical
21,22
which are also coordinated.
The only examples of platinum
1
formazanate complexes likewise consist of o-phenoxy-substituted
applications. Featuring a 1,2,4,5-tetraazapentadienyl core, the
ligands which result in dianionic formazanates coordinated in a
chelating ability of these compounds has long been recognized,
with sporadic accounts of their transition-metal coordination
chemistry appearing over the past several decades.
desirable optical properties of formazans, namely, intense visible
absorption, coupled with their accessible reduction potentials,
make them attractive ligands for a number of applications, and in
recent years, the coordination chemistry of formazans has
experienced a renaissance. The latest accounts have expanded the
23,24
tridentate fashion.
These previous studies described a
2
−7
number of spectroscopic and reactivity properties of the
complexes, but the electrochemical properties and electronic
structures of these platinum complexes were not elaborated. A
partnership between formazans and heavy transition metals
offers the possibility of designing complexes where the intense
visible absorption and redox activity of formazans are coupled
with unique bond activation chemistry and/or triplet excited-
state processes engendered by the metal center, motivating
further work on such compounds.
The
8
chemistry of formazans with first-row transition metals and
9
palladium, demonstrated their utility as supporting ligands for
10,11
copper-mediated oxygen activation,
and paired formazans
Here we describe examples of complexes of platinum(II) with
triarylformazan ligands and provide insight into their electronic
structures. Reasoning that C∧N cyclometalated ligands would
provide a robust and tunable platform to support platinum
formazanate complexes, we have prepared heteroleptic bis-
chelated complexes with both a formazanate and a C∧N
cyclometalated ligand. Furthermore, this approach permits
electronic tuning of both the C∧N ligand and formazanate as a
means of gauging the extent of electronic communication
with other redox-active ligands in heteroleptic cobalt com-
1
2
plexes. Furthermore, zinc complexes of formazanates have
13,14
been used to showcase their redox noninnocence,
and
luminescent boron chelates have emerged which further
expound the redox and optical properties of these com-
1
5−19
pounds.
In spite of these numerous previous accounts, there are only
sparse examples of formazan ligands coordinated to third-row
transition metals. There are a number of third-row transition
metal complexes of dithizonates, which contain a formazan-like
core, but in these examples, coordination of the sulfur is
Received: November 10, 2015
2
0
proposed, leaving one uncoordinated nitrogen atom. Several
©
XXXX American Chemical Society
A
DOI: 10.1021/acs.inorgchem.5b02595
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
between the platinum center and the formazanate π system. A
general synthetic procedure is described for four such complexes,
three of which are structurally characterized by single-crystal X-
ray diffraction. The electronic structures were interrogated
experimentally using cyclic voltammetry, electronic absorption
spectroscopy, and UV−vis-NIR spectroelectrochemistry.
Whereas formazanate-based reductions are observed, the
potentials do shift upon fluorination of the phenylpyridine
ligand, consistent with orbital mixing between the platinum
center and the formazanate ligand, a contention that is supported
by DFT calculations. The optical absorption spectra, with
features attributable to both the formazanate and the Pt−C∧N
fragment, also evince mixing between the formazanate π orbitals
and platinum-centered atomic orbitals. Solvent-dependent
electronic absorption spectra reveal some charge-transfer
character in the HOMO → LUMO transition, which is
confirmed by TD-DFT analysis. All told, this work introduces
a new class of robust coordination complexes and demonstrates
how binding the formazanate to cyclometalated platinum
perturbs the formazanate’s electronic structure and gives rise to
pronounced changes in the electrochemical and optical proper-
ties.
concert with elemental analysis, establish bulk purity and are
consistent with the proposed structures.
Complexes 1−3 were characterized by single crystal X-ray
diffraction, with the structures depicted in Figure 1. A summary
of crystallographic data for 1−3 (Table S1) appears in the
Supporting Information. Relevant bond lengths and angles are
provided in Table 1. The coordination environments of all
complexes are planar, as anticipated for Pt(II); in each case, the
Pt atom deviates <0.075 Å from the mean plane of the ligating
atoms, and the sum of the bond angles about the Pt center is
exactly 360° within experimental error. Bond angles of 80(±2)°
are enforced by the six-membered C∧N and formazanate
chelates, such that the geometries are not rigorously square
planar. The Pt−N bond lengths to the formazanate are in the
same range as the distance between the platinum and the pyridyl
nitrogen, and they do not appear to depend systematically on the
peripheral substitution of the formazanate ligand. A side view of
the crystal structures, shown in Figure 2 for complex 3 (1 and 2
shown in Figures S13 and S14), reveals a “dragonfly” geometry
for the formazanate ligand, whereby the three noncoordinated
atoms of the formazanate core and the N-aryl substituents are
canted out of the coordination plane in opposite directions. A
similar coordination mode for formazans has been observed in
4
,8,12
+
16,17
RESULTS
■
other transition metal
or BF
2
chelates.
Synthesis of Complexes. Scheme 1 depicts the general
synthetic procedure for preparing platinum formazanate
Electrochemistry. Cyclic voltammograms of complexes 1−
4 are shown in Figure 3. Table 2 summarizes the electrochemical
and optical properties of 1−4. The first reduction wave is
assigned to a formazanate-based reduction; the nearly equal peak
current values and the ca. 60 mV separation between the cathodic
and anodic peak potentials are consistent with a reversible one-
electron process. The E1/2 values are similar for the four
complexes. Referenced to ferrocene, this first E1/2 value is −1.49
V for complex 1. Replacing the para methyl groups at the 1- and
Scheme 1. Synthesis of Platinum Complexes 1−4
5
-positions of the formazanate with methoxy groups induces a
small cathodic shift in the reduction potential, which is observed
to be −1.55 V in 2. This shift is identical in magnitude for 3 (E
1
/2
=
−1.44 V) and 4 (E = −1.50 V), and we note that fluorination
1/2
of the phenylpyridine ligand induces a 50 mV anodic shift, as
determined by comparing the potentials for complexes 1 and 3
and those of 2 and 4.
Two additional reductive waves are observed, an irreversible
wave ranging between −2.09 and −2.23 V and another reversible
feature between −2.57 and −2.66 V. The potentials of these
Me,OMe
waves also shift cathodically when Fz
Fz
is replaced with
OMe,OMe
, with the third reduction wave showing only minimal
sensitivity to the formazanate’s substitution pattern, shifting by
only 10−20 mV when the flanking methyl groups are substituted
with methoxy groups.
Complexes 1−4 are also oxidized at mild potentials, although
in acetonitrile, the oxidation waves in the CVs are complex and
typically irreversible (Figure S15). The oxidation waves become
reversible in dichloromethane, as shown in Figure S16,
suggesting that the irreversibility in acetonitrile results from
solvent binding to the oxidized species. Measured in CH Cl vs
2
2
+
Fc /Fc, the first oxidation wave occurs at 0.14 V for 1, 0.04 for 2,
.21 V for 3, and 0.10 V for 4, again depending on both the
0
complexes 1−4. Two different cyclometalating ligands, 2-
phenylpyridine (ppy) and 2-(2,4-difluorophenyl)pyridine
substitution of the formazan and the ppy ligand. A second
reversible oxidation is observed in complexes 2−4, occurring at
0.70 V in 2, 0.89 V in 3, and 0.66 V in 4. In complex 1, this second
oxidation is chemically irreversible, with an anodic peak potential
of 0.90 V and two return waves observed.
Electronic Absorption and Emission Spectra. Electronic
absorption spectra of complexes 1−4 were recorded and
(
(
F ppy), are combined with two different triarylformazan ligands
2
Me,OMe
OMe,OMe
Fz
and Fz
) to furnish a set of four complexes.
Reactions in refluxing methanol, with Na CO acting as a base,
2
3
1
gave good isolated yields (70−94%) of all complexes. H,
C{ H}, and F (for 3 and 4) NMR (Figures S1−S12), in
1
3
1
19
B
DOI: 10.1021/acs.inorgchem.5b02595
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Figure 1. X-ray crystal structures of complexes 1−3. For complex 1, only one crystallographically independent molecule is depicted, and for all
structures, hydrogen atoms and solvate molecules are omitted. Data was collected at 213(2) K (1) or 123(2) K (2 and 3).
Table 1. Summary of Crystallographic Bond Lengths and
Angles for Complexes 1−3
a
1
2
3
d(Pt−N) (Å) (formazanate, trans
2.041(4)
2.021(4)
2.003(2)
to N)
2
.055(4)
2.065(4)
.059(4)
2.022(4)
.055(4)
2.006(4)
.011(4)
78.67(16)
8.30(15)
80.14(19)
0.30(19)
359.9(7)
59.8(7)
d(Pt−N) (Å) (formazanate, trans
2.058(3)
2.020(4)
2.015(4)
80.84(14)
80.60(17)
360.0(6)
2.084(2)
2.036(2)
1.998(3)
78.92(9)
80.31(10)
359.7(4)
to C)
2
d(Pt−N) (Å) (C∧N)
2
d(Pt−C) (Å) (C∧N)
2
∠N−Pt−N (deg) (formazanate)
7
∠
N−Pt−C (deg) (C∧N)
Σ bond angles (deg)
8
3
a
Two crystallographically independent molecules.
Figure 3. Cyclic voltammograms of complexes 1−4, showing the
reduction waves. CVs were recorded in acetonitrile with 0.1 M NBu PF
4
6
supporting electrolyte, using a glassy carbon working electrode and a
scan rate of 0.1 V/s. The arrows indicate the scan direction. Currents are
normalized to bring the individual traces onto the same scale.
Table 2. Summary of Electrochemical Data and Electronic
a
Absorption Spectra for 1−4
1
2
3
4
+
E1/2 (V vs Fc /Fc)
−1.49,
−1.55
−2.23
−2.66
−1.44
−2.09
−2.57
−1.50
−2.14
−2.59
b
b
b
b
−
2.12
2.65
252 (39)
80 (41)
−
abs. in CH Cl λ, nm
256 (36)
277 (33)
353 (18)
251 (35)
269 (32)
294 (25)
254 (43)
2
2
3
(
ε/10 )
c
2
292 (27)
c
3
39 (22)
44 (7.9)
64 (12)
353 (21)
462 (5.5)
662 (11)
Figure 2. Side view of the crystal structure of complex 3.
c
c
4
449 (4.9)
338 (17)
6
665 (8.3)
449 (5.5)
compared to the free formazan ligands. Figure 4 shows the
absorption spectra of 1−4, overlaid with their respective free
formazan ligand. The data is summarized in Table 2. The broad,
6
53 (9.6)
c
a
b
CVs recorded in MeCN solution. Irreversible. Shoulder.
Me,OMe
OMe,OMe
intense visible absorption bands in Fz
and Fz
occur
−1
−1
−1
at 524 nm (ε = 15 000 M cm ) and 545 nm (ε = 14 000 M
absorption is noted in platinum complexes 1−4, with little
dependence of λ on the formazanate’s substituents and molar
−
1
cm ), respectively. A substantial red shift in this visible
max
C
DOI: 10.1021/acs.inorgchem.5b02595
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
electron reduction chemistry by UV−vis spectroelectrochemis-
+
try. For complex 1, electrolysis at −1.8 V vs Fc /Fc, which elicits
the first reduction event, gave rise to the spectral changes shown
in the top panel of Figure 5. The band at 440 nm was minimally
Figure 4. Electronic absorption spectra of 1−4, overlaid with their
respective free formazan ligand. Spectra were recorded in dichloro-
methane solution at room temperature.
4
absorptivities that are ∼10 in each case. Complexes 1 and 3,
Me,OMe
ligated by Fz
1
, show visible absorption bands at 664 (ε =
Figure 5. UV−vis-NIR spectroelectrochemistry of complex 1 in MeCN
2 000 M− cm ) and 653 nm (ε = 9600 M cm ),
1
−1
−1
−1
+
with 0.1 M NBu PF . The applied potential (vs Fc /Fc) was −1.8 V for
4
6
OMe,OMe
respectively, which are only minimally different in Fz
the first reduction and −2.3 V for the second reduction.
−1
−1
complexes 2 (665 nm, ε = 8300 M cm ) and 4 (662 nm, ε =
1
−1
1
1 000 M− cm ). This low-energy band shows some sensitivity
affected by reduction. The low-energy absorption band,
occurring at 657 nm in this solvent (acetonitrile), disappears,
and new bands at 559 and 950 nm grow in. Reduction of complex
to solvent polarity, as judged by comparing the absorption
spectra recorded in toluene, dichloromethane, and methanol
+
(
Figures S17−S20, Table S2). The band shifts to shorter
1 at −2.3 V vs Fc /Fc induces a second reduction of the complex.
wavelengths as the solvent polarity is increased, but the difference
in absorption maxima is <10 nm when comparing toluene
solutions to methanol solutions.
This second reduction results in less substantial spectral changes,
depicted in the bottom panel of Figure 5. The low-energy
•
−
transitions attributed to 1 diminish slightly in intensity during
the second reduction, whereas the band that occurs at 425 nm in
1 shifts to 435 nm and intensifies significantly in the two-electron
In addition to these broad, intense absorption features in the
red, complexes 1−4 all show a band between 440−460 nm (ε ∼
000−8000), which is absent in the free formazanate ligand. This
absorption band is much more sensitive to the solvent
environment than the low-energy band, also shifting to higher
energy as the solvent polarity is increased. As a representative
example, this band for complex 1 is observed at 460 nm in
toluene, 444 nm in dichloromethane, and 413 nm in methanol.
Complexes 1−4 are nonemissive in the visible region at room
temperature and 77 K.
2
−
5
reduced complex 1 . Spectroelectrochemistry was also
performed on complex 4, as shown in Figure S21, with a nearly
identical outcome to that of complex 1.
DFT Computations. DFT-optimized geometries are good
matches for the crystal structures, and the “dragonfly” shape of
the formazanate is reproduced. Time-dependent density func-
tional theory (TD-DFT) computations characterized the key
H,H
electronic transitions of two 1,3,5-triphenylformazanate (Fz
)
H,H
Spectroelectrochemistry. To further investigate the nature
of the electrochemical features, we investigated the one- and two-
models: an unfluorinated complex 5 Pt(ppy) (Fz ) and a
fluorinated derivative 6 Pt(F ppy) (Fz ). Figure S22 shows the
H,H
2
D
DOI: 10.1021/acs.inorgchem.5b02595
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
simulated absorption spectra generated from the TD-DFT
results. A HOMO → LUMO transition dominates the one-
electron, low-energy excitation of both 5 (86% weight) and 6
involved formazan ligands FzMe,OMe and FzOMe,OMe, where the
methyl and methoxy groups provide convenient NMR handles
and allow us to probe the effect of electron-donating groups on
the electronic properties of the complexes. The synthetic route
described in Scheme 1 proved to be general for the four
complexes we describe here, and we envision it will be applicable
to a wide range of C∧N and formazanate ligands with diverse
electronic structures. Once prepared, complexes 1−4 have
proven to be stable under ambient conditions and can be stored
and handled on the benchtop with no evidence for
decomposition over a period of several months.
(87% weight), and the computed wavelengths (5: 677 nm, 6: 658
nm) agree well with the experimental optical spectra of 4 (cf.
Table 2). This HOMO → LUMO transition is best described as a
formazanate-centered π→π* excitation with some Pt charge
transfer character mixed in. As shown in Figure 6, the HOMO
Three of the new complexes were structurally characterized by
single-crystal X-ray diffraction. There is a slight asymmetry in the
Pt−N bond distances to the chelated formazanate; the distances
trans to the phenyl are longer, consistent with a stronger trans
influence of phenyl as compared to pyridine. Also of note, the
Pt−N bond distances involving the C∧N ligand and the
formazanate are quite similar; on average the former are only
marginally shorter than the latter. The Pt−N (C∧N) bond
distances observed here, which average 2.033(4) Å for the three
complexes, are considerably longer than the corresponding
25
distance in Pt(ppy) (acac) (1.979(6) Å). That said, the bond
distances we observe here between the platinum and the C∧N
ligand are markedly similar to those reported for cyclometalated
Figure 6. Computed HOMO and LUMO contour plots for the model
complex (5) Pt(ppy) (Fz ) (contour level = 0.3 au).
H,H
2
6
platinum complexes with 5-aryldipyrrinato ligands. Although
there is some variability among the complexes reported here, the
average Pt−N and Pt−C distance for 1−3 are within 0.01 Å of
the corresponding distances in the dipyrrinato complexes. In
addition, the average formazanate Pt−N distances we observe are
quite similar to the pyrrinato Pt−N distances in these previously
reported complexes, which represent rare examples of
cyclometalated platinum with monoanionic diaza ligands.
There are few literature compounds which are directly
comparable to the complexes we describe here, but we do not
believe that the structural metrics observed 1−3 are remarkable
in any way.
and LUMO orbitals of 5 show substantial mixing of the platinum
orbitals with the formazanate π-system. Population analyses
show that the HOMO for both 5 and 6 exhibit a ca. 15%
2
contribution from the Pt-centered orbitals (primarily d ), while
z
the LUMO displays only ca. 2% Pt character. Direct comparisons
of the molecular electrostatic potential (MEP) maps of 5 and 6
2
6
(see Figure 7) reveal that fluorination modestly reduces the
Complexes 1−4 are dark green in color, owing to a low-energy
4
−1
−1
transition centered at ∼660 nm with ε ∼ 10 M cm . We
assign this transition as a formazanate-centered π−π* transition,
which is significantly perturbed from that of the free ligand,
which has a low energy λmax of 520−540 nm. Consistent with this
assignment, this low-energy band shows minimal dependence on
solvent polarity, decreasing by no more than 10 nm when the
solvent is changed from toluene to methanol, with dichloro-
methane solutions giving maxima between these extremes.
Calculations confirm the nature of this transition. TD-DFT
reveals an intense low-energy transition which is principally
HOMO→LUMO, consisting mostly of formazanate π→π*
character. The calculations also reveal some Pt(d)→ formazanate
charge transfer character in the excited state, consistent with the
slight solvent dependence of the this absorption maximum. The
perturbation of the formazanate’s π→π* absorption in these
Figure 7. Computed MEP plots for the model complexes, (5) Pt(ppy)
H,H
H,H
(
Fz ) and (6) Pt(F ppy) (Fz ), employing the Molekel program
2
(
isosurface values were set to 0.1 for illustrative purposes).
electrostatic potential of the Pt center (5: V(Pt) = −399.12 V, 6:
V(Pt) = −398.93 V) as well as the nitrogen atoms at the 2- (5:
V(N2) = −499.69 V, 6: V(N2) = −499.53 V) and 4- (5: V(N4) =
−
499.61 V, 6: V(N4) = −499.54 V) positions of the 1,2,4,5-
17,19
tetraazapentadienyl core.
platinum complexes is larger than that observed in boron
or
14
8
zinc formazanate complexes, though in group 10 Ni(II)
complexes, similarly large bathochromic shifts are observed. The
participation of Pt atomic orbitals in the frontier molecular
orbitals is responsible for the large perturbation of the HOMO→
LUMO gap when the formazanate is coordinated to platinum. A
second absorption band is observed at higher energy, consistent
with a Pt(d)→C∧N(π*) MLCT transition, as seen in other
DISCUSSION
■
The coordination chemistry of formazanate ligands with third-
row transition metals remains underexplored, and in particular,
the electrochemical and optical properties of such complexes
have received little attention. Seeking a versatile platform to
support platinum formazanate coordination compounds, this
initial work explored the design and synthesis of heteroleptic
complexes, where the formazanate is partnered with a C∧N
cyclometalated phenylpyridine derivative. Our early efforts
27
cyclometalated platinum complexes. The solvent dependence
of this absorption band clearly points to substantial charge
transfer character, with the absorption maximum shifting by ∼30
E
DOI: 10.1021/acs.inorgchem.5b02595
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
nm or more when the solvent is changed from toluene to
methanol.
Two subsequent reduction waves are observed in complexes
1−4, and these are likewise sensitive to electronic modification of
both the formazanate and C∧N ligand. The first of these, which is
irreversible, is assigned to a Pt−C∧N based reduction, and the
most cathodically shifted reversible reduction is assigned to a
second formazanate-based reduction. The assignment of these
waves is supported by literature precedent. The two reversible
waves are separated by ca. 1.1 V, which is similar to the separation
The electrochemical behavior of complexes 1−4 offers
additional insight into their electronic structure. Although the
potential of the first reversible process depends on both the
formazanate and the C∧N ligand, we assign it primarily to a
formazanate-based reduction. Related Pt(II) cyclometalates,
supported by redox-inactive acac ligands, are reduced at much
more negative potentials, −2.41 V for Pt(ppy) (acac), and −2.29
between the successive reduction potentials of boron for-
2
7
17,19
V for Pt(F ppy) (acac). Our observed potentials of ca. −1.5 V
mazanates,
leading us to favor an assignment of the wave at
2
are much more in line with the ligand-based reductions observed
in other triarylformazanate complexes. Several boron formaza+-
nates have been electrochemically characterized, and a BF2
ca. −2.6 V to a second formazanate-based reduction.
Furthermore, this second formazanate-based wave is consid-
erably less sensitive to addition of a methoxy group to the 1- and
5-aryl positions, which is also true for previously studied boron
chelates when making analogous substitutions. The spectroelec-
trochemistry for the second reduction is inconclusive but is
consistent with the assignment of the wave as a platinum-
centered reduction. With this reduction (Figure 5, bottom), only
the Pt(d)→ppy MLCT transition is substantially perturbed,
whereas the bands assigned the formazanate(2−) ligand
diminish in intensity with no new low-energy features emerging.
This suggests that the second electron enters an orbital that is
primarily Pt−C∧N-centered and does not further reduce the
formazanate. Ultimately, a more definitive assignment of the
second and third reduction waves will come from investigation of
a larger suite of complexes, with more electronically diverse C∧N
and formazanate ligands, which are expected to perturb the
reduction potentials more substantially and will clarify the
assignment of these features.
Me,OMe
17
chelate of Fz
shows a first reduction potential of −0.90 V.
Metal complexes of triarylformazans are rarer and tend to be
more difficult to reduce; a series of homoleptic zinc
triarylformazanates are reduced at potentials in the range of
1
4
−
1.2 to −1.8 V, similar to the values we observe for Pt(II)
complexes 1−4. Moreover, DFT calculations on model
H,H
H,H
complexes Pt(ppy) (Fz ) (5) and Pt(F ppy) (Fz ) (6)
2
reveal LUMOs that are almost exclusively localized on the
formazanate ligand, also suggesting the formazanate as the site of
the first reduction event. The assignment of the first reduction
wave is further supported by UV−vis spectroelectrochemistry.
Reduction of 1 by one electron results in the disappearance of the
low-energy HOMO→LUMO band (657 nm in MeCN) and the
growth of two new low-energy bands, one at 559 nm and one at
950 nm. These new features are entirely consistent with a one-
electron reduced formazanate and are qualitatively similar to the
absorption bands in other compounds featuring formazanate-
derived radical ligands. As described previously, one-electron
CONCLUSION
■
2
8
15,29
reduction of zinc and boron
formazanate complexes gives
In this work, we outlined a general synthetic strategy for
coordinating formazanate ligands to cyclometalated platinum
complexes. The desirable properties of the formazanate, namely,
the facile ligand-based redox chemistry and intense visible
absorption, are retained in these complexes, and these features
motivate an extension to other transition metals and supporting
ligand platforms where differential mingling of metal-based and
formazanate-based orbitals could lead to unique properties.
Efforts are underway to prepare an expanded library of
complexes with more structurally and electronically diverse
C∧N and formazanate ligands, with the goal of gaining a more
thorough understanding of the electronic structure and ensuant
properties of this novel class of complexes.
rise to two new absorption bands, which resemble the absorption
bands of verdazyl radicals. The two new bands flank the π→π*
13
band of the closed-shell, monoanionic formazanate, with one
growing at higher energy, and one appearing at lower energy.
·−
·−
The absorption bands we observe for 1 and 4 , which contain a
one-electron reduced verdazyl-type ligand, are substantially red-
shifted from those of zinc or boron formazanates and verdazyl
radicals, much in the same manner as the HOMO→LUMO
absorption in neutral 1−4 undergoes a significant red-shift upon
coordination to platinum.
The behavior of the first formazanate-based reduction
provides further evidence for the mixing of platinum- and
formazanate-centered orbitals in the frontier MO’s. Addition of
electron-donating methoxy groups to the formazanate predict-
ably perturbs the reduction potential to a more negative value;
EXPERIMENTAL SECTION
■
Materials. Reactions were carried out in a nitrogen atmosphere
using standard Schlenk techniques. Solvents, starting materials, and
reagents were of commercial origin and used without further
purification, unless stated otherwise below. Dichloromethane and
toluene for UV−vis spectroscopy and acetonitrile for electrochemical
replacing p-CH groups on the 1- and 5-aryl substituents with p-
3
OCH induces a 60 mV cathodic shift in this potential. By
3
replacing ppy (complexes 1 and 2) with F ppy (3 and 4), the first
2
reduction wave is anodically shifted by 50 mV. This is consistent
with an electronic structure where the Pt is donating electron
density to the redox-active formazanate. The electron donation is
attenuated by fluorination of the phenylpyridine ligand, which
results in a positive shift in the formazanate-based reduction
potential. This supposition is supported by computational
studies of the model complexes. The molecular electrostatic
potential map shows that the addition of electronegative fluorine
substituents to the ppy ligand pulls electron density from the
opposing formazanate, which is consistent with the observation
30
measurements were dried by the method of Grubbs, passing through
dual alumina columns on a commercial solvent purification system
(SPS). The acetonitrile was further dried by storage over 3A molecular
sieves. Tetrabutylammonium hexafluorophosphate (TBAPF ) was
6
recrystallized from hot ethanol and ferrocene was sublimed at ambient
pressure before use in electrochemical experiments. CDCl for NMR
3
spectroscopy was stored over potassium carbonate and molecular sieves
to remove acidic impurities and moisture. The ligand 3-p-methox-
Me,OMe
yphenyl-1,5-di-p-tolylformazan (Fz
) was prepared by the method
28
of Hicks et al. The ligand 1,3,5-tri-p-methoxyphenylformazan
that the F ppy complexes are easier to reduce by 50 mV, even
OMe,OMe
2
(Fz
) was prepared by the same method; details of its synthesis
though the reduction event is primarily localized on the
formazanate.
have not been previously reported and are given below. The platinum
precursors [Pt(ppy)(μ-Cl)] (ppy =2-phenylpyridine) and [Pt(F ppy)-
2
2
F
DOI: 10.1021/acs.inorgchem.5b02595
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
(
μ-Cl)] (F ppy =2-(2,4-difluorophenyl)pyridine) were prepared as
by TLC. The methanol was removed via rotary evaporation, and the
product was taken up in dichloromethane. The mixture was filtered
through a short silica column, eluting with dichloromethane until all of
the green color had passed through. The solution was taken to dryness in
vacuo. A spectroscopically pure green microcrystalline solid was
obtained by adding pentane to a concentrated THF solution. Yield
133.6 mg (70%). Material satisfactory for elemental analysis required
recrystallization from CH Cl /pentane. The crystalline material was
2
2
previously described, using conventional heating and a 1:1 mol ratio of
31
K PtCl and the C∧N ligand. (Note: Many other references describe
2
4
the syntheses of these dimers, using >2 equiv of C∧N with K PtCl . We
2
4
have found, as reported in the referenced synthetic procedure, that using
excess C∧N ligand leads to the formation of substantial amounts of
monomeric Pt(κ -C∧N)(κ −N-C∧N) (Cl) as a side product.)
2
1
Physical Methods. NMR spectra were recorded at room temper-
ature using a JEOL ECA-600 NMR spectrometer. UV−vis absorption
spectra were recorded in CH Cl solutions in screw-capped quartz
2
2
spectroscopically and electrochemically indistinguishable from the
1
2
2
initially isolated powder. H NMR (400 MHz, CDCl ) δ: 8.19 (d, J =
3
cuvettes using an Agilent Cary 60 UV−vis spectrophotometer. Cyclic
voltammetry (CV) measurements were performed with a CH
Instruments 602E potentiostat interfaced with a nitrogen glovebox via
wire feedthroughs. Samples were dissolved in acetonitrile with 0.1 M
8.7 Hz, 2H, ArH), 8.14 (d, J = 8.2 Hz, 3H, ArH), 8.00 (d, J = 6.8 Hz, 2H,
ArH), 7.58−7.67 (m, 2H, ArH), 7.43 (d, J = 6.8 Hz, 1H, ArH), 7.15 (d, J
=
8.4 Hz, 2H, ArH), 7.10 (d, J = 8.0 Hz, 2H, ArH), 6.93−6.99 (m, 3H,
ArH), 6.80 (t, J = 7.2 Hz, 1H, ArH), 6.72 (d, J = 6.8 Hz, 1H, ArH), 6.68
(t, J = 6.8 Hz, 1H, ArH), 3.87 (s, 3H, OCH ), 2.38 (s, 6H, CH ).
TBAPF as a supporting electrolyte. A 3 mm diameter glassy carbon
6
3
3
working electrode, a platinum wire counter electrode, and a silver wire
pseudoreference electrode were used. Potentials were referenced to an
internal standard of ferrocene. Spectroelectrochemistry measurements
were executed in thin-layer quartz cuvettes, using a patterned
13
1
C{ H} NMR: δ 168.5, 159.9, 153.8, 151.1, 150.9, 150.6, 148.0, 145.9,
1
1
38.4, 136.86, 136.84, 136.6, 130.1,129.8, 129.2, 129.1, 126.4, 125.8,
24.7, 123.6, 123.0, 121.0, 118.5, 113.7, 55.5, 21.3, 21.2. IR (solid): 3020
(
1
m), 2951 (m), 2924 (m), 2908 (m), 2831 (m), 1607 (s), 1584 (m),
510 (s), 1494 (s), 1482 (s), 1439 (m), 1427 (m), 1389 (m) cm . Anal.
“honeycomb” electrode from Pine Research Instrumentation consisting
−1
of a gold working electrode and a platinum counter electrode, in
combination with a separate silver wire pseudoreference. Solutions were
thoroughly sparged with argon prior to measurement, and spectra were
recorded on a Cary 8354 diode array spectrophotometer. IR spectra
were recorded on solid samples using a Nicolet iS10 Spectrometer with
an ATR (attenuated total reflectance) accessory. Elemental analyses
were performed by Midwest Microlab, LLC.
Computational Details. Geometry optimizations and time-
dependent density functional theory (TD-DFT) computations for 5
and 6 were performed at the B97/Def2-TZVPP level employing
Gaussian09. For TD-DFT, only the lowest-energy excited state was
computed at the B97/Def2-TZVPP level, whereas the full simulated
spectrum shown in Figure S22 was computed at B97D/Def2-SVPP//
B97D/Def2-TZVPP. Molecular electrostatic potential (MEP) maps
were computed at B97/Def2-SVPP//B97/Def2-TZVPP and plotted
with the Molekel program. The optimized geometries of both 5 and 6
exhibit a characteristic “dragonfly” shape (see Cartesian coordinates
included in the Supporting Information) and match well with the crystal
structures of 1−4. TD-DFT computations quantified the HOMO →
LUMO transition oscillator strengths and composition of the one-
Calcd for C H N OPt: C, 56.09; H, 4.14, N, 9.91. Found: C, 56.32, H,
33
29
5
4
.41, N, 9.63.
Pt(ppy) (FzOMe,OMe) (2). The title compound was prepared by the
general method described above for complex 1, using 100 mg of
OMe,OMe
1
[
Pt(ppy)(μ-Cl)] and 101 mg of Fz
. Yield: 189 mg (94.0%). H
2
NMR (400 MHz, CDCl ) δ: 8.14−8.25 (m, 5H, ArH), 7.96−8.00 (m,
3
2
6
3
1
1
5
9
2
1
9
H, ArH), 7.58−7.67 (m, 2H, ArH), 7.43 (d, J = 7.6 Hz, 1H, ArH),
.79−6.98 (m, 8H, ArH), 6.66−6.74 (m, 2H, ArH), 3.87 (s, 3H, OCH ),
3
13
1
.84 (s, 3H, OCH ), 3.83 (s, 3H, OCH ). C{ H} NMR: 168.6, 159.9,
3
3
58.6, 158.4, 153.8, 151.0, 148.1, 147.1, 146.5, 146.0, 138.5, 136.9, 132.4,
30.3, 129.4, 127.0, 126.4, 125.9, 123.7, 123.0, 121.2, 118.6, 114.5, 113.7,
5.79, 55.76, 55.6. Anal. Calcd for C H N O Pt: C, 53.66; H, 3.96, N,
33
29
5
3
.48. IR (solid): 3045 (m), 3000 (m), 2951 (m), 2928 (m), 2907 (m),
834 (m), 1597 (s), 1582 (s), 1509 (s), 1495 (s), 1483 (s), 1459 (s),
−
1
438 (s) cm . Anal. Calcd for C H N O Pt: C, 53.66; H, 3.96, N,
33
29
5
3
.48. Found: C, 53.68, H, 3.91, N, 9.39.
Me,OMe
Pt(F ppy) (Fz
) (3). The title compound was prepared by the
2
general method described above, using 100 mg of [Pt(F ppy)(μ-Cl)]
2
2
Me,OMe
and 85 mg of Fz
. Yield: 151 mg (85.7%). Recrystallization from
1
CH Cl /pentane gave the product as 3·0.5CH Cl , evident from H
2
2
2
2
electron excitation transitions.
1
Preparation of FzOMe,OMe. 4-Methoxyphenylhydrazine hydro-
NMR spectra of the crystals and combustion analysis data. H NMR
400 MHz, CDCl ) δ: 8.15−8.19 (m, 3H, ArH), 8.06 (d, J = 8.2 Hz, 2H,
(
3
chloride (5.76g, 33.0 mmol) was combined with triethylamine (8.34
mL, 59.7 mmol) and ethanol (50 mL). After the mixture was stirred for
ArH), 7.96−8.01 (m, 3H, ArH), 7.69 (t, J = 7.8 Hz, 1H, ArH), 7.11−7.16
m, 4H, ArH), 6.97−7.01 (m, 2H, ArH), 6.66−6.69 (m, 1H, ArH),
.37−6.43 (m, 1H, ArH), 6.20 (dd, J = 9.6, 2.4 Hz, 1H, ArH), 3.87 (s,
(
3
0 min, 4-methoxybenzaldehyde (4.01 mL, 33.0 mmol) was added, and
6
3
1
3
1
the mixture was allowed to stir for an additional 2 h, at which time a dark
orange precipitate had formed. The reaction mixture was treated with
sodium carbonate hydrate (11.87 g, 111.5 mmol), tetrabutyl ammonium
bromide (1.07 g, 3.34 mmol), water (100 mL) and dichloromethane
1
3
1
H, OCH ), 2.40 (s, 3H, CH ), 2.38 (s, 3H, CH ). C{ H} NMR:
3
3
3
65.1 (d, JCF = 7.4 Hz), 162.1 (dd, JCF = 343, 12 Hz), 160.4 (dd, JCF
46, 12 Hz), 160.0, 153.9, 152.8 (d, JCF = 5.9 Hz), 151.0, 150.4, 150.2,
38.9, 137.3, 136.8, 129.9, 129.7, 129.6, 129.2, 126.4, 125.7, 124.7, 122.2
=
(
100 mL) and stirred at 0 °C for another 1 h. A solution of a diazonium
(d, JCF = 21 Hz), 120.9, 118.7 (d, JCF = 16 Hz), 113.8, 99.6 (t, JCF = 27
salt made from stirring 4-methoxy aniline (4.68 g, 38.0 mmol), sodium
nitrite (3.17 g, 46.6 mmol), water (10 mL), and hydrochloric acid (10
mL) for 1 h at 0 °C was then added dropwise to the hydrazone mixture.
After addition, the organic phase in the biphasic reaction turned blood
red. After stirring for 2 h at room temperature, the organic layer was
collected and washed with water (500 mL) in a separatory funnel. The
aqueous layer was extracted with an additional 20 mL of dichloro-
Hz), 55.5, 21.24, 21.21. IR (solid): 3077 (m), 3023 (m), 3004 (m), 2950
(
(
m), 2922 (m), 2855 (m), 2838 (m), 1601 (s), 1571 (s), 1560 (s), 1513
s), 1497 (s), 1480 (s), 1427 (s), 1407 (s) cm . Anal. Calcd for
−1
C H F N OPt·0.5CH Cl : C, 51.25; H, 3.59, N, 8.92. Found: C,
33 27
2
5
2
2
5
1.44, H, 3.55, N, 8.58.
Pt(F ppy) (FzOMe,OMe) (4). The title compound was prepared by the
2
general method above, using 100 mg of [Pt(F ppy)(μ-Cl)] and 93 mg
methane. MgSO (5 g) was added to the combined organic layer to
2
2
4
OMe,OMe
1
of Fz
. Yield: 148 mg (80.4%). H NMR (400 MHz, CDCl ) δ:
remove the extra water, and the solution was taken to dryness on a rotary
3
evaporator after filtration. The solid was dissolved in boiling methanol
8.19−8.22 (m, 3H, ArH), 8.14 (d, J = 9.2 Hz, 2H, ArH), 7.97−7.99 (m,
3H, ArH), 7.66−7.70 (m, 1H, ArH), 6.97−7.00 (m, 2H, ArH), 6.85−
6.91 (m, 4H, ArH), 6.68−6.72 (m, 1H, ArH), 6.38−6.44 (m, 1H, ArH),
followed by cooling to furnish a dark purple microcrystalline solid. Yield:
1
6
8
(
.20 g (48.2%). H NMR (400 MHz, CDCl ) δ: 15.38 (br, s, 1H, NH),
3
.05 (m, 2H, ArH), 7.62 (m, 4H, ArH), 6.95−6.99 (m, 6H, ArH), 3.870
6.23 (dd, J = 9.2, 2.2 Hz, 1H, ArH), 3.87 (s, 3H, OCH
3
), 3.83 (s, 6H,
). C{ H} NMR: 165.2 (d, JCF = 5.9 Hz), 162.2 (dd, JCF = 347, 13
Hz), 160.5 (dd, JCF = 352, 12 Hz), 160.0, 158.9, 158.5, 153.8, 152.9 (d,
JCF = 5.9 Hz), 150.9, 146.3, 146.1, 139.0, 129.8, 129.6, 126.9, 126.4,
13
1
s, 3H, OCH ), 3.867 (s, 6H, OCH ).
OCH
3
3
3
Me,OMe
Pt(ppy) (Fz
) (1). [Pt(ppy)(μ-Cl)] (100 mg, 0.130 mmol),
2
Me,OMe
Fz
(93 mg, 0.26 mmol) and sodium carbonate hydrate (110 mg,
1
.04 mmol) were combined in MeOH (12 mL), and the mixture was
deoxygenated by bubbling with nitrogen. The mixture was refluxed at 65
C, and a color change was observed from deep purple to bright green
over the course of refluxing for 16 h. Reaction completion was confirmed
125.8, 122.3 (d, JCF = 19 Hz), 121.0, 118.7 (d, JCF = 18 Hz), 114.5,
113.83, 113.75, 99.6 (t, JCF = 27 Hz), 55.74, 55.71, 55.5. IR (solid):
3070(m), 3043 (m), 3006 (m), 2938 (m), 2911 (m), 2836 (m), 1598
(s), 1574 (s), 1514 (s), 1494 (s), 1481 (s), 1464 (s), 1451 (s), 1434 (s),
°
G
DOI: 10.1021/acs.inorgchem.5b02595
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
−
1
1
9
405 (s) cm . Anal. Calcd for C H F N O Pt: C, 51.16; H, 3.51, N,
(9) Gilroy, J. B.; Ferguson, M. J.; McDonald, R.; Hicks, R. G. Inorg.
Chim. Acta 2008, 361, 3388−3393.
33
27
2
5
3
.04. Found: C, 50.98, H, 3.53, N, 8.90.
X-ray Crystallography Details. Single crystals of 1−3 were grown
by layering concentrated CH Cl solutions with pentane. Crystals were
mounted on a Bruker Apex II three-circle diffractometer using Mo Kα
radiation (λ = 0.71073 Å). The data was collected at 213(2) K (1) or
(10) Hong, S.; Hill, L. M. R.; Gupta, A. K.; Naab, B. D.; Gilroy, J. B.;
Hicks, R. G.; Cramer, C. J.; Tolman, W. B. Inorg. Chem. 2009, 48, 4514−
4523.
2
2
(11) Hong, S.; Gupta, A. K.; Tolman, W. B. Inorg. Chem. 2009, 48,
6323−6325.
1
23(2) K (2 and 3) and was processed and refined within the APEXII
software. Structures were solved by direct methods in SHELXS and
refined by standard difference Fourier techniques in the program
SHELXL. Hydrogen atoms were placed in calculated positions using the
standard riding model and refined isotropically; all non-hydrogen atoms
were refined anisotropically. The structure of 3 contained a heavily
disordered solvent molecule, likely dichlormethane, which was removed
using the SQUEEZE function within Platon. The crystal of 1 was found
to be partially desolvated; one dichloromethane was modeled as half-
occupied, and a second was modeled as a two-part disorder about a
special position with total one-half occupancy. Distance restraints
(12) Protasenko, N. A.; Poddel’sky, A. I.; Bogomyakov, A. S.; Fukin, G.
K.; Cherkasov, V. K. Inorg. Chem. 2015, 54, 6078−6080.
(13) Chang, M.-C.; Dann, T.; Day, D. P.; Lutz, M.; Wildgoose, G. G.;
Otten, E. Angew. Chem., Int. Ed. 2014, 53, 4118−4122.
(14) Chang, M.-C.; Roewen, P.; Travieso-Puente, R.; Lutz, M.; Otten,
E. Inorg. Chem. 2015, 54, 379−388.
(15) Chang, M.-C.; Otten, E. Chem. Commun. 2014, 50, 7431−7433.
(16) Barbon, S. M.; Reinkeluers, P. A.; Price, J. T.; Staroverov, V. N.;
Gilroy, J. B. Chem. - Eur. J. 2014, 20, 11340−11344.
(
17) Barbon, S. M.; Price, J. T.; Reinkeluers, P. A.; Gilroy, J. B. Inorg.
Chem. 2014, 53, 10585−10593.
18) Hesari, M.; Barbon, S. M.; Staroverov, V. N.; Ding, Z.; Gilroy, J. B.
Chem. Commun. 2015, 51, 3766−3769.
19) Barbon, S. M.; Staroverov, V. N.; Gilroy, J. B. J. Org. Chem. 2015,
(
SADI) were used for all 1,2 and 1,3 distances within the disordered
solvent molecules, and rigid bond restraints SIMU and DELU were
employed for the thermal displacement parameters. Crystallographic
details are summarized in Table S1.
(
(
8
0, 5226−5235.
ASSOCIATED CONTENT
Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.inorg-
chem.5b02595.
(20) Meriwether, L. S.; Breitner, E. C.; Sloan, C. L. J. Am. Chem. Soc.
1
(
1
■
965, 87, 4441−4448.
21) Petersen, R. L.; McFarland, J. T.; Watters, K. L. Bioinorg. Chem.
978, 9, 355−367.
22) Zhang, H.-Y.; Sun, W.-X.; Wu, Q.-A.; Zhangb, H.-Q.; Chen, Y.-Y.
*
S
(
Synth. React. Inorg. Met.-Org. Chem. 2000, 30, 571−582.
(23) Bait, S.; Meuldijk, J.; Renkema, W. E. Transition Met. Chem. 1980,
5
(
Crystallographic data tables, NMR spectra for new
compounds, alternative views of the crystal structures of
, 357−361.
24) Balt, S.; Meuldijk, J. Inorg. Chim. Acta 1981, 47, 217−223.
1
and 2, additional cyclic voltammograms, solvent-
(
25) Bossi, A.; Rausch, A. F.; Leitl, M. J.; Czerwieniec, R.; Whited, M.
dependent electronic absorption spectra, UV−vis-NIR
spectroelectrochemistry of complex 4, simulated absorp-
tion spectra for 5 and 6, and optimized Cartesian
coordinates for 5 and 6 (PDF)
T.; Djurovich, P. I.; Yersin, H.; Thompson, M. E. Inorg. Chem. 2013, 52,
2403−12415.
26) Bronner, C.; Baudron, S. A.; Hosseini, M. W.; Strassert, C. A.;
1
(
Guenet, A.; De Cola, L. Dalton Trans. 2010, 39, 180−184.
(27) Brooks, J.; Babayan, Y.; Lamansky, S.; Djurovich, P. I.; Tsyba, I.;
Bau, R.; Thompson, M. E. Inorg. Chem. 2002, 41, 3055−3066.
X-ray data (CIF)
(
28) Gilroy, J. B.; McKinnon, S. D. J.; Koivisto, B. D.; Hicks, R. G. Org.
AUTHOR INFORMATION
Corresponding Author
E-mail: tteets@uh.edu.
■
*
Lett. 2007, 9, 4837−4840.
29) Gilroy, J. B.; Ferguson, M. J.; McDonald, R.; Patrick, B. O.; Hicks,
R. G. Chem. Commun. 2007, 126−128.
30) Pangborn, A. B.; Giardello, M. A.; Grubbs, R. H.; Rosen, R. K.;
Timmers, F. J. Organometallics 1996, 15, 1518−1520.
31) Godbert, N.; Pugliese, T.; Aiello, I.; Bellusci, A.; Crispini, A.;
Ghedini, M. Eur. J. Inorg. Chem. 2007, 2007, 5105−5111.
(
(
Author Contributions
The manuscript was written through contributions of all authors.
All authors have given approval to the final version of the
manuscript.
(
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This work was supported by the University of Houston and by a
grant from the Welch Foundation (Grant No. E-1887). We
thank Prof. Loi Do for allowing use of the UV−vis
spectrophotometer, and Prof. Karl Kadish and Dr. Yuanyuan
Fang for helpful discussions regarding spectroelectrochemistry.
REFERENCES
1) Nineham, A. W. Chem. Rev. 1955, 55, 355−483.
■
(
(
(
(
(
(
(
2) Hunter, L.; Roberts, C. B. J. Chem. Soc. 1941, 823−826.
3) Irving, H.; Gill, J. B.; Cross, W. R. J. Chem. Soc. 1960, 2087−2095.
4) Dale, D. J. Chem. Soc. A 1967, 278−287.
5) Budesinsky, B.; Svecova, J. Inorg. Chem. 1971, 10, 313−317.
6) Siedle, A. R.; Pignolet, L. H. Inorg. Chem. 1980, 19, 2052−2056.
7) Kawamura, Y.; Yamauchi, J.; Ohya-Nishiguchi, H. Bull. Chem. Soc.
Jpn. 1993, 66, 3593−3599.
8) Gilroy, J. B.; Patrick, B. O.; McDonald, R.; Hicks, R. G. Inorg. Chem.
008, 47, 1287−1294.
(
2
H
DOI: 10.1021/acs.inorgchem.5b02595
Inorg. Chem. XXXX, XXX, XXX−XXX