Coordination effects on electron distributions for rhodium complexes of the redox-active bis(3,5-di- tert -butyl-2-phenolate)amide ligand

Article abstract of DOI:10.1021/ic2026076

New rhodium complexes of bis(3,5-di-tert-butyl-2-phenol)amine ([ONO ^cat]H₃) were synthesized, and their electronic properties were investigated. These compounds were prepared by combining [ONO^q]K and [(cod)Rh(μ-Cl)]₂ in the presence of an auxiliary donor ligand to yield complexes of the type [ONO]RhL_n (n = 3, L = py (1); n = 2, L = PMe₃ (2a), L = PMe₂Ph (2b), PMePh₂ (2c), PPh₃ (2d)). Single-crystal X-ray diffraction studies on [ONO]Rh(py)₃ (1) revealed a six-coordinate, octahedral rhodium complex. In the case of [ONO]Rh(PMe₃)₂ (2a), X-ray diffraction showed a five-coordinate, distorted square-pyramidal coordination environment around the rhodium center. While 1 is static on the NMR time scale, complexes 2a-d are fluxional, displaying both rapid isomerization of the square-pyramidal structure and exchange of coordinated and free phosphine ligands. UV-vis spectroscopy shows stark electronic differences between 1 and 2a-d. Whereas 1 displays a strong absorbance at 380 nm with a much weaker band at 585 nm in the absorption spectrum, complexes 2a-d display an intense (ε > 10⁴ M^-1 cm^-1), low-energy absorption band in the region 580-640 nm; however, in the cases of 2a and 2b, the addition of excess phosphine resulted in changes to the UV-vis spectrum indicating the formation of six-coordinate adducts [ONO]Rh(PMe₃)₃ (3a) and [ONO]Rh(PMe₂Ph)₃ (3b), respectively. The experimental and DFT computational data for the six-coordinate complexes 1, 3a, and 3b are consistent with their formulation as classical, d⁶, pseudo-octahedral, coordination complexes. In the five-coordinate complexes 2a-2d, φ-bonding between the rhodium center and the [ONO] ligand leads to a high degree of covalency and metal-ligand electron distributions that are not accurately described by formal oxidation state assignments.

Full text of DOI:10.1021/ic2026076

Article
pubs.acs.org/IC
Coordination Effects on Electron Distributions for Rhodium
Complexes of the Redox-Active Bis(3,5-di-tert-butyl-2-phenolate)
amide Ligand
Geza Szigethy, David W. Shaffer, and Alan F. Heyduk*
́
Department of Chemistry, University of California, Irvine, California 92697, United States
S
* Supporting Information
ABSTRACT: New rhodium complexes of bis(3,5-di-tert-butyl-
2-phenol)amine ([ONOcat]H3) were synthesized, and their
electronic properties were investigated. These compounds were
prepared by combining [ONOq]K and [(cod)Rh(μ-Cl)]2 in the
presence of an auxiliary donor ligand to yield complexes of the
type [ONO]RhLn (n = 3, L = py (1); n = 2, L = PMe3 (2a), L = PMe2Ph (2b), PMePh2 (2c), PPh3 (2d)). Single-crystal X-ray
diffraction studies on [ONO]Rh(py)3 (1) revealed a six-coordinate, octahedral rhodium complex. In the case of [ONO]Rh(PMe3)2
(2a), X-ray diffraction showed a five-coordinate, distorted square-pyramidal coordination environment around the rhodium center.
While 1 is static on the NMR time scale, complexes 2a−d are fluxional, displaying both rapid isomerization of the square-pyramidal
structure and exchange of coordinated and free phosphine ligands. UV−vis spectroscopy shows stark electronic differences between 1
and 2a−d. Whereas 1 displays a strong absorbance at 380 nm with a much weaker band at 585 nm in the absorption spectrum,
complexes 2a−d display an intense (ε > 104 M−1 cm−1), low-energy absorption band in the region 580−640 nm; however, in the cases
of 2a and 2b, the addition of excess phosphine resulted in changes to the UV−vis spectrum indicating the formation of six-coordinate
adducts [ONO]Rh(PMe3)3 (3a) and [ONO]Rh(PMe2Ph)3 (3b), respectively. The experimental and DFT computational data for the
six-coordinate complexes 1, 3a, and 3b are consistent with their formulation as classical, d6, pseudo-octahedral, coordination complexes.
In the five-coordinate complexes 2a−2d, π-bonding between the rhodium center and the [ONO] ligand leads to a high degree of
covalency and metal−ligand electron distributions that are not accurately described by formal oxidation state assignments.
results in a meridional coordination mode. The fully deprotonated
form of the [ONO] ligand can adopt one of three formal
oxidation states: the reduced catecholate form, [ONOcat]3−, the
one-electron oxidized semiquinonate form, [ONOsq•]2−, or the
two-electron oxidized quinonate form, [ONOq]−, as shown in
Scheme 1. Most studies of the [ONO] platform have focused on
INTRODUCTION
■
Ligands capable of adopting more than one stable oxidation
state have earned the designation as redox-active or non-
innocent because of their propensity to complicate the electronic
properties and reactivity of the metal center to which they are
bound.1−3 Installation of these ligands onto independently
redox-active metals further hinders the characterization of
valence-electron distribution in the resultant complexes, while
simultaneously providing access to electronic structures uncom-
mon in more classical metal−ligand complexes.4−8 Perhaps the
most common redox-active ligands are the catecholates and their
ortho-aminophenolate and ortho-diamide analogues, whose metal
complexes are known to exhibit rich redox chemistry independent
of the coordinated metal ion.8−10 Expansion of the catecholate
platform into polydentate, poly-aromatic geometries makes ligand
redox potentials more accessible than in less conjugated sys-
tems while also decreasing the lability of the ligand in its oxidized
form.11−14 These properties, combined with the potential for
ligand modification, make expanded catecholate-type ligands
attractive for a wide range of reaction chemistry.
Scheme 1
the structural and electronic characteristics of coordinatively
saturated M[ONO]2 complexes,15,16,19−24 but more recently,
investigations have turned toward the synthesis and reactivity of
heteroleptic M[ONO]Ln complexes,18,25 which includes alkyl-
migration on tin,26 copper-catalyzed oxidation,27 and stoichiometric
One redox-active ligand whose reactivity and electronic char-
acteristics have been extensively studied on both transition metals
and main group elements is the tridentate [ONO] ligand
([ONOcat]H3 = bis(3,5-di-tert-butyl-2-phenol)amine).15 While
capable of binding metals in a doubly deprotonated, facially
coordinating form {([ONO]H)2−},16−18 the [ONO] ligand is
most commonly studied in its fully deprotonated form, which
Received: December 3, 2011
Published: April 6, 2012
© 2012 American Chemical Society
12606
dx.doi.org/10.1021/ic2026076 | Inorg. Chem. 2012, 51, 12606−12618

Inorganic Chemistry
Article
benzene. Electrospray Ionization (ESI) mass spectrometry was
performed on a Waters LCT Premier mass spectrometer.
oxidative addition/reductive elimination transformations on
tantalum.28−30
[ONOq]K. Just-thawed tetrahydrofuran (THF, 35 mL) was added to
a combination of [ONO]H3 (2.00 g, 4.70 mmol, 1 equiv) and KH
(565 mg, 1.41 mmol, 3 equiv) in a 100 mL round-bottom flask. The
mixture was stirred at room temperature overnight while H2 gas was
released. The resulting yellow, turbid solution was frozen, and
PhI(OAc)2 (1.51 g, 4.70 mmol, 1 equiv) was added. The solution
turned purple/black with further evolution of H2 gas. The mixture was
stirred at room temperature for 6 h. Solid KOAc was removed by
filtration and washed with THF until washes were nearly colorless.
The solvent was removed from the combined filtrates under vacuum,
coevaporated with toluene, then twice with pentane, resulting in
1.99 g of product as a black powder (92%). Anal. Calcd. (Found) for
C28H40NO2K (%): C, 72.82 (72.56); H, 8.73 (9.10); N, 3.03 (2.89).
1H NMR (500 MHz) δ/ppm: 1.27 (s, CH3, 18H), 1.48 (s, CH3, 18H),
7.22 (s, br, aryl-H, 2H), 7.31 (s, br, aryl-H, 2H). 13C NMR (125 MHz)
δ/ppm: 30.65 (CH3), 34.58 (CCH3), 35.63 (CCH3), 118.53 (aryl-C),
130.88 (aryl-C), 139.95 (aryl-C), 144.21 (aryl-C), 145.27 (aryl-C),
179.45 (aryl-C). UV−vis (C6H6) λmax/nm (ε/M−1 cm−1): 346 (5,330),
537 (1,930), 810 (16,100). MS (ESI−) m/z: 422.2 (M−), 883.5
(M2K−).
Research in our group explores the unique electronic
interactions and chemical transformations accessible to metal
complexes with redox-active ligands, especially in cases where
the reactivity can be attributed explicitly to the redox properties
of the ligand.11,13,14,28,29,31−36 Our prior investigations of
[ONO] complexes have focused on early transition metal
ions with d0 electron configurations,28−30 but the possibility of
complementary redox reactivity between the [ONO] ligand
and electron-rich late transition metals prompted an inves-
tigation of the coordination chemistry of the [ONO] ligand
with rhodium. Rhodium complexes display a rich reaction
chemistry and are capable of a wide variety of stoichiometric
and catalytic bond-activation reactions,37,38 including the
activation of classically unreactive small molecules such as H2,
CO2,39 N2,40 and even CH4.41 The coordination chemistry of
rhodium with redox-active ligands remains relatively unex-
plored, although our studies on the rhodium complex
[nacnac]Rh(phdi) (nacnac− = β-diketiminate, phdi = phenan-
threnediimine) indicated a close energetic match between
rhodium and ligand valence orbitals, resulting in significant
noninnocent ligand behavior.40,42−44
This investigation details the first installation of the [ONO]
ligand onto rhodium and the physical and electronic character-
ization of the resultant complexes. Synthesis of [ONO]Rh
complexes was facilitated through the use of the potassium salt
of the oxidized [ONO] ligand, [ONOq]K, which is an as-yet
unreported starting material for [ONO] complexes. Through a
combination of UV−visible spectroscopy, solution titrations,
and density functional theory (DFT) calculations it was deter-
mined that [ONO]Rh complexes exhibit noninnocent electro-
nic behavior that is strongly influenced by the coordination of
ancillary ligands on the rhodium center. In coordinatively
saturated six-coordinate complexes, well-defined [ONOcat]-
RhIIIL3 complexes were obtained (L = pyridine, PR3). In five-
coordinate [ONO]RhL2 complexes (L = PR3), a high-degree of
covalency in metal−ligand π-bonding results in valence electron
distributions that are intermediate between [ONOcat]RhIII and
[ONOsq]RhII electronic configurations.
[ONOcat]Rh(py)3, 1. A solution of [ONOq]K (187 mg, 0.406
mmol, 1 equiv) in 18 mL of Et2O was frozen, and then [(cod)Rh
(μ-Cl)]2 (100 mg, 0.203 mmol, 0.5 equiv) and pyridine (131 μL,
1.62 mmol, 4 equiv) were added. The mixture was stirred for 4 h at room
temperature. Solvent was removed under vacuum from the brown
suspension, and the residue was extracted with 35 mL of toluene and
filtered. Removal of solvent from the filtrate under vacuum and
coevaporation with pentane yielded 306 mg (99%) of the product as a
fine brown solid. X-ray quality crystals were grown by diffusion of
pentane into a THF solution at room temperature. Anal. Calcd.
(Found) for C43H55N4O2Rh (%): C, 67.70 (67.47); H, 7.27 (7.17); N,
7.34 (7.12). 1H NMR (500 MHz): δ 1.54 (s, CH3, 18H), 2.13 (s, CH3,
18H), 6.03 (t, py-H, J = 8.5 Hz, 4H), 6.28−6.31 (m, py-H, 4H), 6.53−
6.57 (m, py-H, 1H), 7.05 (s, aryl-H, 2H), 8.16 (s, aryl-H, 2H), 8.76 (d,
py-H, J = 6.5 Hz, 4H), 9.16 (d, py-H, J = 6.0 Hz, 2H). The low
solubility of 1 precluded the acquisition of a 13C{1H} NMR spectrum.
UV−vis (C6H6) λmax/nm (ε/M−1 cm−1): 380 (23,900), 585 (1,300).
MS (ESI+) m/z: 762.6 (M+), 763.6 (MH+).
General Synthesis of [ONO]Rh(PR3)2, 2a−d. A solution of
[ONOq]K (0.20−0.25 mmol, 1 equiv) in 15 mL of Et2O was frozen,
and [(cod)Rh(μ-Cl)]2 (0.5 equiv) and phosphine (2 equiv) were
added. The mixture was stirred at room temperature for at least 4 h.
The teal/blue solution was filtered, and solvent was removed under
vacuum. The residue was in all cases clean enough for subsequent reac-
tion with near-quantitative yields (greater than 95%). Analytical
samples were purified further by recrystallization as described below.
[ONO]Rh(PMe3)2, 2a. Blue/black crystalline solid isolated in
several crops from pentane solutions at −35 °C, 61%. Anal. Calcd.
(Found) for C341H58NO2P2Rh (%): C, 60.26 (60.03); H, 8.63 (8.43);
N, 2.07 (1.98). H NMR (500 MHz): δ 0.90 (vt, P−CH3, Japp = 6.0
Hz, 18H), 1.62 (s, CH3, 18H), 1.80 (s, CH3, 18H), 7.49 (d, aryl-H, J =
1.5 Hz, 2H), 8.70 (s, aryl-H, 2H). 13C{1H} NMR (125 MHz): δ 16.37
(vt, P-CH3, Japp = 13.8 Hz), 30.61(CH3), 32.58 (CH3), 34.88 (CCH3),
35.76 (CCH3), 111.74 (aryl-C), 116.26 (vt, aryl-C, Japp = 2.4 Hz),
137.45 (aryl-C), 137.90 (aryl-C), 144.83 (vt, aryl-C, Japp = 6.6 Hz),
166.23 (vt, aryl-C, Japp = 5.5 Hz). 31P{1H} NMR (162 MHz): δ 18.80
EXPERIMENTAL SECTION
■
General Considerations. Compounds described herein are
oxygen and moisture sensitive, so manipulations were carried out
under a nitrogen atmosphere using standard Schlenk or glovebox tech-
niques. Solvents were purified by sparging with argon followed by
sequential passage through activated Q5 and alumina columns to remove
oxygen and water, respectively. Reagents were purchased from com-
mercial sources and used as received with the exception of [ONOcat]H329
and [(cod)Rh(μ-Cl)]2,45 which were synthesized following literature
procedures.
Spectroscopic Methods. NMR spectra were collected on a
Bruker DRX 500 MHz (1H 500 MHz, 13C 128 MHz) or a Bruker
DRX400 (1H 400 MHz, 31P 162 MHz) spectrometer at 298 K in C6D6
unless otherwise noted. The NMR solvent was degassed by several
freeze−pump−thaw cycles, dried over sodium benzophenone ketyl
radical, and vacuum-distilled before use. All 1H, 13C{1H}, and 31P{1H}
chemical shifts are reported using the standard δ-scale in ppm. The 1H
and 13C{1H} NMR spectra were referenced to TMS using the proteo-
and natural abundance 13C impurities in C6D6 (7.15 and 128.02 ppm).
31P{1H} NMR spectra were referenced to H3PO4 using the Ξ (Xi)
scale and the residual proton impurities C6D6.46 FTIR spectra were
recorded with a Perkin−Elmer Spectrum One spectrophotometer as
KBr pellets. UV−visible spectra were recorded with a Perkin-Elmer
Lambda 800 spectrometer in 1-cm quartz cuvettes in dry, degassed
1
(d, PMe3, JRh−P = 137.9 Hz). UV−vis (C6H6) λmax/nm (ε/M−1
cm−1): 321 (17,600), 396 (2,970), 587 (10,300). MS (ESI+) m/z:
677.0 (M+), 678.0 (MH+).
[ONO]Rh(PMe2Ph)2, 2b. Blue/black crystalline solid from satu-
rated pentane at −35 °C, 57%. Anal. Calcd. (Found) for
C44H62NO2P2Rh (%): C, 65.91 (66.19); H, 7.79 (8.17); N, 1.75
(1.89). 1H NMR (500 MHz): δ 1.10 (vt, P−CH3, Japp = 6.0 Hz, 12H),
1.63 (s, CH3, 18), 1.86 (s, CH3, 18H), 6.92−6.99 (m, aryl-H, 6H),
7.17−7.19 (m, aryl-H, 4H), 7.56 (d, aryl-H, J = 1.5 Hz, 2H), 8.63 (s,
aryl-H, 2H). 13C{1H} NMR (125 MHz): δ 14.35 (vt, P-CH3, Japp
=
14.1 Hz), 30.62 (CH3), 32.57 (CH3), 34.89 (CCH3), 35.78 (CCH3),
112.38 (aryl-C), 116.96 (aryl-C), 129.67 (aryl-C), 130.32 (vt, aryl-C,
12607
dx.doi.org/10.1021/ic2026076 | Inorg. Chem. 2012, 51, 12606−12618

Inorganic Chemistry
Article
Table 1. X-ray Data Collection and Refinement Parameters for [ONOcat]Rh(py)3 (1) and [ONO]Rh(PMe3)2 (2a)
[ONOcat]Rh(py)3·4 THF·1/2 C5H12 (1·4 THF·1/2 C5H12)
104.5H148N8O8Rh2
[ONO]Rh(PMe3)2·1/2 C5H12 (2a·1/2 C5H12)
empirical formula
C
C36.5H64NO2P2Rh
713.74
formula weight
1850.13
orthorhombic
Pbca
crystal system
monoclinic
P21/c
space group
T, K
83(2)
143(2)
a, Å
22.4517(10)
28.7074(12)
31.8954(14)
90°
13.8702(12)
23.903(2)
13.4550(12)
90°
b, Å
c, Å
α
β
90°
90°
20557.5(15)
8
118.0260(10)°
90°
3937.7(6)
4
γ
V, Å3
Z
reflections collected
217984
42307
data/restraints/parameters
21056/20/1102
0.0475
8074/0/419
0.0450
a
R1 [I > 2σ(I)]
a
wR2 (all data)
0.1322
0.0830
a
GOF
1.042
1.020
a
2
2
R1 = ∑||Fo| − |Fc||/∑|Fo|; wR2 = [∑w(Fo − Fc2)2/∑w(Fo )2]1/2; GOF = [∑w(|Fo| − |Fc|)2/(n − m)]1/2
.
1
3
Japp = 4.6 Hz), 136.2−136.6 (vm, aryl-C), 137.76 (aryl-C), 138.45
(aryl-C), 144.97 (vt, aryl-C, Japp = 7.1 Hz), 166.59 (vt, aryl-C, Japp = 5.6
(162 MHz): δ 13.70 (dd, PMe3, JRh−P = 119.4 Hz, JP−P = 65.3 Hz),
1
3
29.66 (dd, PMe2Ph, JRh−P = 155.8 Hz, JP−P = 65.3 Hz). MS (ESI+)
m/z: 739.20 (M+).
1
Hz). 31P{1H} NMR (162 MHz): δ 24.44 (d, PMe2Ph, JRh−P = 137.7
Hz). UV−vis (C6H6) λmax/nm (ε/M−1 cm−1): 323 (19,400), 416
(3,080), 625 (12,300). MS (ESI+) m/z: 800.9 (M+), 802.0 (MH+).
[ONO]Rh(PMePh2)2, 2c. Teal, amorphous powder from saturated
pentane at −35 °C, 88%. Anal. Calcd. (Found) for C54H66NO2P2Rh
Electrochemical Methods. Electrochemical experiments were
performed on a Gamry Series G 300 Potentiostat/Galvanostat/ZRA
(Gamry Instruments, Warminster, PA, U.S.A.) using a 3.0 mm glassy
carbon working electrode, a platinum wire auxiliary electrode, and a
silver wire reference electrode. Electrochemical experiments were
performed at ambient temperature (20−24 °C), in a nitrogen-filled
glovebox. Tetrabutylammonium hexafluorophosphate (Acros) was
recrystallized from ethanol three times and dried under vacuum before
use. Commercially available [CoCp2][PF6] was dried under vacuum
for 1 day before use. Analyte concentrations were 1.0 mM in THF
with 0.10 M [NBu4][PF6] as the supporting electrolyte. All potentials
are referenced to [Cp2Fe]+/0, using [Cp2Co][PF6] as an internal
standard (−1.34 V vs [Cp2Fe]+/0).47
X-ray Diffraction. X-ray diffraction data were collected on crystals
mounted on glass fibers using a Bruker CCD platform diffractometer
equipped with a CCD detector. Measurements were carried out using
Mo Kα (λ = 0.71073 Å) radiation, which was wavelength selected with
a single-crystal graphite monochromator. A full sphere of data was
collected for each crystal structure. The APEX2 program package was
used to determine unit-cell parameters and to collect data.48 The raw
frame data were processed sing SAINT49 and SADABS50 to yield the
reflection data files. Subsequent calculations were carried out using the
SHELXTL51 program suite. Structures were solved by direct methods
and refined on F2 by full-matrix least-squares techniques to con-
vergence. Analytical scattering factors for neutral atoms were used
throughout the analyses.52 Hydrogen atoms, though visible in the
difference Fourier map, were generated at calculated positions and
their positions refined using the standard riding models. ORTEP dia-
grams were generated using ORTEP-3 for Windows,53 and all thermal
ellipsoids are drawn at the 50% probability level. Diffraction collection
and refinement data are shown in Table 1.
1
(%): C, 70.04 (69.80); H, 7.18 (7.33); N, 1.51 (1.50). H NMR (500
MHz): δ 1.57 (s, CH3, 18H), 1.62 (s, br, P−CH3, 6H), 1.76 (s, CH3,
18H), 6.83−6.89 (m, aryl-H, 12H), 7.27 (m, br, aryl-H, 8H), 7.53 (s,
aryl-H, 2H), 8.37 (s, aryl-H, 2H). 13C{1H} NMR (125 MHz): δ 12.56
(P-CH3), 30.80 (CH3), 32.50 (CH3), 34.70 (CCH3), 35.67 (CCH3),
113.42 (aryl-C), 118.09 (aryl-C), 129.72 (aryl-C), 132.27 (aryl-C),
133.92 (vt, aryl-C, Japp = 5.6 Hz), 135.13 (vt, aryl-C, Japp = 21.0 Hz),
137.62 (aryl-C), 138.77 (aryl-C), 145.39 (aryl-C), 167.86 (aryl-C).
1
31P{1H} NMR (162 MHz): δ 38.35 (d, PMePh2, JRh−P = 137.7 Hz).
UV−vis (C6H6) λmax/nm (ε/M−1 cm−1): 319 (18,100), 626 (11,000).
MS (ESI+) m/z: 924.9 (M+), 925.9 (MH+).
[ONO]Rh(PPh3)2, 2d. Blue/black, microcrystalline hairs of the
hemietherate isolated in several crops by layering a saturated Et2O
solution under pentane at −35 °C, 85%. Anal. Calcd. (Found) for
C64H70NO2P2Rh·1/2(C4H10O) (%): C, 72.92 (71.33); H, 6.95 (6.76);
1
N, 1.29 (1.15). H NMR (500 MHz): δ 1.11 (t, Et2O−CH3, J = 7.0
Hz, 3H), 1.53 (s, CH3, 36 H), 3.26 (q, Et2O−CH2, J = 7.0 Hz, 2H),
6.81 (t, aryl-H, J = 7.5 Hz, 12H), 6.89 (t, aryl-H, J = 7.5 Hz, 6H), 7.36
(q, aryl-H, J = 7.0 Hz, 12H), 7.51 (d, aryl-H, J = 2.0 Hz, 2H), 8.55 (d,
aryl-H, J = 2.0 Hz, 2H). 13C{1H} NMR (125 MHz,): δ 15.58 (Et2O-
CH3), 30.95 (CH3), 32.34 (CH3), 34.74 (CCH3), 35.49 (CCH3),
65.90 (Et2O-CH2), 114.13 (aryl-C), 119.80 (vt, aryl-C, Japp = 4.6 Hz),
128.09 (aryl-C), 129.70 (aryl-C), 134.04 (vt, aryl-C, Japp = 20.3 Hz),
134.76 (vt, aryl-C, Japp = 4.8 Hz), 137.66 (aryl-C), 139.69 (vt, aryl-C,
Japp = 4.5 Hz), 146.67 (aryl-C), 168.98 (vd, aryl-C, Japp = 7.0 Hz).
1
31P{1H} NMR (162 MHz): δ 57.47 (d, PPh3, JRh−P = 139.2 Hz).
UV−vis (C6H6) λmax/nm (ε/M−1 cm−1): 315 (18,100), 446 (4,600),
639 (11,800). MS (ESI+) m/z: 1049.0 (M+), 1050.0 (MH+).
Comproportionation of 2a and 2b. Recrystallized 2a (4.8 mg,
0.0071 mmol) and 2b (5.6 mg, 0.0070 mmol) were dissolved in 0.75
mL of C6D6 and transferred to an NMR tube. NMR analysis indicated
the presence of minor amounts of 2a and 2b with the primary product
being [ONO]Rh(PMe3)(PMe2Ph), the characteristic peaks of which
are listed below. 1H NMR (400 MHz): δ 0.53 (d, P−CH3, Japp = 12.8
Hz, 9H), 1.47 (d, P−CH3, J = 10.4 Hz, 6H), 1.62 (s, CH3, 18H), 1.85
(s, CH3, 18H), 7.05−7.10 (m, arom. H, 2H), 7.53 (s, arom. H, 2H),
7.77−7.81 (m, arom. H, 2H), 8.74 (s, arom. H, 2H). 31P{1H} NMR
Titrations. Stock solutions of [ONO]Rh(PR3)2 were prepared by
dissolving a known mass of the complex accurate to 0.1 mg in benzene
in a volumetric flask. A 0.10 M stock solution of PMe3 in toluene was
prepared by diluting 1.00 mL of commercially available 1.0 M PMe3
solution in toluene to 10 mL with toluene in a volumetric flask. These
stock solutions were stored at −35° when not in use. Titrations were
conducted by two different methods, yielding the analogous results.
Method 1. Appropriate aliquots of the [ONO]Rh(PR3)2 and PR3
stock solutions were added to a series of 5 or 10 mL volumentric flasks
to generate solutions that were 40−50 μM in metal complex and with
12608
dx.doi.org/10.1021/ic2026076 | Inorg. Chem. 2012, 51, 12606−12618

Inorganic Chemistry
Article
the desired concentration of free phosphine upon dilution to the
volume mark. The samples were allowed to equilibrate for at least 30
min before acquisition of the UV−vis spectra.
Method 2. A 50 mL solution of [ONO]Rh(PR3)2 in benzene was
made in a 50 mL volumetric flask and transferred to an Erlenmeyer
flask. Successive 5 mL aliquots of this solution were removed after
each addition of PR3. Samples for both titration methods were allowed
to equilibrate at least 30 min before measurement, although color
change was immediate upon addition of PR3. Titrations were repeated
three times to verify reproducibility.
Titration Data Treatment. Spectrophotometric titration data
were analyzed using the program Hyperquad200354,55 utilizing non-
linear least-squares regression to determine formation constants. The
absorption spectrum of the starting {[ONO]Rh} complex was deter-
mined by the absorption spectrum of the first sample in the titration,
to which no titrant was added. Wavelengths between 290−750 nm
were typically used for data refinement.
Figure 1. UV−vis absorption spectra of [ONOcat]H3 and [ONOq]K
acquired in C6H6 at 298 K.
Density Functional Theory Calculations. All calculations were
performed using the TURBOMOLE program package.56 The
geometric and electronic structures were optimized without symmetry
constraints using the B3LYP functional57−60 and a polarized triple-ζ
basis set (TZVP).61 All molecules were treated in the gas phase without
accounting for solvent. The energy was converged to 10−7 Hartree.
Natural population analyses62 and plots were also obtained at the B3LYP/
TZVP level; the contour values were 0.045 for the molecular orbital plots.
Scheme 2
RESULTS
■
[ONOq]K Synthesis and Metalation. To leverage the
wealth of available rhodium(I) starting materials, a reliable
[ONOq]− synthon was targeted. The proteo species [ONOq]H
has been reported by the condensation of 3,5-di-tert-butyl-o-
quinone and 2-hydroxy-3,5-di-tert-butylaniline,18 but in our
hands this reaction was unreliable and typically gave cyclized
products.63 The Zn[ONOq]2 complex has also been used as a
source for the [ONOq]− ligand,26 but to avoid reliance on me-
tathesis of a tridentate ligand, a scalable synthesis of the potassium
salt [ONOq]K was developed. According to eq 1, deprotonation
of [ONOcat]H3 with KH followed by two-electron oxidation with
PhI(OAc)2, yielded [ONOq]K in yields of greater than 90%.
Three equivalents of KH are added at the beginning of the
reaction; however, removal of the third, nitrogen-bound proton
does not seem to occur until addition of the oxidant in the second
step. The [ONOq]K salt is purple/black, significantly soluble in
most organic solvents, and stable under a nitrogen atmosphere at
PMe3, 2a; PR3 = PMe2Ph, 2b; PR3 = PMePh2, 2c; PR3 = PPh3,
2d) in quantitative yields.
Crystallographic Characterization. To identify the
rhodium coordination environment in complexes 1 and 2,
single-crystal X-ray diffraction studies were performed. With
the exception of 2c, all rhodium complexes could be isolated as
crystalline solids; however, the crystals typically suffered from
rapid desolvation, allowing only 1 and 2a to be structurally
characterized by single-crystal X-ray diffraction. Selected
metrical parameters for 1 and 2a are listed in Table 2, and
the structures of the rhodium complexes are illustrated in
Figure 2. In both structures the [ONO] ligand is coordinated in
the expected meridional fashion. The [ONO]−Rh metal−
ligand bond lengths are similar in the two structures, and these
distances are consistent with Rh−O/N bond lengths reported
for rhodium(III) complexes with polydentate ligands.64−66
The crystal structures in Figure 2 illustrate two different
coordination geometries: 1 adopts a six-coordinate octahedral
geometry while 2a adopts a distorted square-pyramidal geometry.
The rhodium coordination environment and Rh−N(py) bond
distances in 1 are similar to those in mer-RhCl3(py)3,67 with the
largest deviations from regular octahedral geometry deriving from
the acute (84°) bite angles inherent to the [ONO] ligand. In the
case of 2a, a more relaxed geometry about the metal center is
observed. The rhodium in 2a sits 0.28 Å above the mean-squared
n
room temperature. While BuLi could be substituted for KH in
the above synthesis, the resultant [ONOq]Li salt exhibits a much
1
broader H NMR signal than the potassium salt, making purity
more difficult to verify. While [ONOcat]H3 is colorless, the absorp-
tion spectrum of [ONOq]K is characterized by a strong, broad
band at 810 nm (Figure 1). This feature is very similar to those
reported for purple [ONOq]H as well as for M[ONOq]2 com-
plexes of transition metal and main group elements.20
Successful transmetalation of [ONOq]K with rhodium(I) start-
ing materials required the inclusion of neutral donor ligands to
saturate the rhodium coordination sphere in the product. As
shown in Scheme 2, the reaction of 2 equiv of [ONOq]K with
[(cod)Rh(μ-Cl)]2 in the presence of pyridine resulted in
quantitative formation of [ONOcat]Rh(py)3 (1) as a brown
solid. Reactions using monodentate phosphine ligands afforded
the blue-green phosphine adducts [ONO]Rh(PR3)2 (PR3 =
12609
dx.doi.org/10.1021/ic2026076 | Inorg. Chem. 2012, 51, 12606−12618

Inorganic Chemistry
Article
Table 2. Selected Crystallographic/Computational Bond Distances (Å) for [ONOcat]Rh(py)3 (1), [ONO]Rh(PMe3)2 (2a),
[ONO]Rh(PPh3)2 (2d), and [ONOcat]Rh(PMe3)3 (3a)
a
Crystals of 1 had two crystallographically unique but structurally similar molecules in the asymmetric unit. The bond distances shown here are for
b
c
one molecule; the second molecule shows similar distances, see Supporting Information. Rh(1)−L = Rh−N(py). Rh(1)−L = Rh−PR3.
Figure 2. ORTEP diagrams of (a) [ONOcat]Rh(py)3 (1) and (b) [ONO]Rh(PMe3)2 (2a). Hydrogen atoms, solvent molecules, disorder, and a
crystallographically unique but structurally analogous molecule of 1 have been omitted for clarity.
plane defined by the coordinating atoms of the basal plane.
Complex 2a manifests a short Rh−P bond for the apical
phosphine ligand relative to the Rh−P bond for the phosphine
in the basal plane. The Rh−P bonds in 2a are generally shorter
12610
dx.doi.org/10.1021/ic2026076 | Inorg. Chem. 2012, 51, 12606−12618

Inorganic Chemistry
Article
than those in octahedral RhIII complexes containing PMe3
ligands,68−70 but longer than the RhI−P bonds in the square
planar complex (PEt3)3Rh[O-(p-tolyl)].71
Rh(PR3)3 (PR3 = PMe3, 3a; PR3 = PMe2Ph, 3b). The 1H NMR
spectrum of 2a in toluene-d8 in the presence of excess PMe3
showed an upfield shift of the aryl [ONO] resonances,
bringing them to within 0.2 ppm of those observed for 1. At
Intraligand distances within the [ONO] platform are often
used as a diagnostic tool for determining the ligand oxidation
state since the π bond order differs slightly for each of the
ligand oxidation states shown in Scheme 1. Both 1 and 2a
exhibited C−O and C−N bond distances 0.03−0.10 Å longer
than in the previously characterized M[ONOq]2 complexes
(M = Zn, Ni, Pb) in which the [ONO] platform is fully
oxidized,21−23 and are more consistent with bond distances in
formally reduced {[ONOcat]MV} complexes.18,28,29 The C(1)−
C(2) and C(15)−C(16) bonds are often the longest aryl C−C
bonds in [ONO] complexes regardless of coordinated metal
and ligand oxidation state;21,22,29 in both 1 and 2a these bonds
measure 1.42−1.44 Å, intermediate between C−C bonds in
fully reduced aryl rings and nonaromatic sp2−sp2 carbon
atoms.72 The other aryl C−C bond distances in 1 and 2a fall in
a narrow range of 1.38−1.40 Å, which is more consistent with
the [ONOcat]3− ligand oxidation state than the oxidized
1
298 K, the H NMR spectrum of 2a plus added phosphine
showed a single PMe3 resonance, indicating rapid exchange
of free and coordinated phosphine; however, cooling the
sample to 228 K slowed the exchange process so separate 1H
NMR resonances could be resolved, indicating the presence
of two different coordinated phosphine ligands as well as free
PMe3. A similar phenomenon was observed in the 31P{1H}
NMR spectra of 2a and added PMe3. At 298 K the
phosphorus resonances for 2a and added PMe3 in toluene-
d8 are broadened into the baseline, but at 213 K, sharp
signals for free PMe3 and two different rhodium-bound
phosphine ligand environments were resolved. Both
coordinated phosphine ligands showed Rh−P and P−P
couplings with a doublet of doublets signal at −10.34 ppm
corresponding to two PMe3 ligands cis to the [ONO]
nitrogen and a doublet of triplets signal at 4.29 ppm
corresponding to a single PMe3 ligand trans to the [ONO]
nitrogen. This NMR data is indicative of the formation of
[ONOcat]Rh(PMe3)3 (3a) from association of a third PMe3
ligand to the rhodium center in 2a. Rhodium phosphine
complex 2b behaved similarly to 2a: addition of excess PMe2Ph
to 2a afforded [ONOcat]Rh(PMe2Ph)3 (3b). Derivatives 2c and
2d showed no NMR evidence for the formation of six-coordinate
adducts upon addition of excess PMePh2 and PPh3, respectively.
Electronic Spectroscopy. The UV−visible spectra for 1
and 2a−d showed distinctly different electronic structures for
the six-coordinate pyridine adduct and the five-coordinate
phosphine adducts. Figure 3 shows the electronic absorption
[ONOsq• 2− or [ONOq]− forms.21,22
]
NMR Spectroscopy. The NMR spectrum of 1 is
consistent with the solid-state structure described above.
Complex 1 exhibited a symmetric [ONO] ligand environ-
ment with two aromatic and two tert-butyl resonances in the
1
1H NMR spectrum. The H NMR spectrum of 1 shows two
unique pyridine ligands in a 2:1 ratio, consistent with the
inequivalent pyridine environments in the X-ray structure.
The protons of the pyridine trans to the [ONO] nitrogen are
shifted downfield compared to the pyridine molecules cis to
the [ONO] nitrogen. The pyridine resonances for 1 did not
coalesce upon heating to 60 °C; however, the addition of
excess d5-pyridine to the NMR sample resulted in the
immediate (τ < 30 s) loss of the resonances for coordinated
pyridine and the appearance of the resonances associated
with 3 equiv of free h5-pyridine. Attempts to use selective
inversion-pulse NMR experiments showed that the pyridine
exchange rate is slower at 298 K than the relaxation time of
the pyridine protons (about 1 s), precluding exact measure-
ment of the pyridine exchange rate.
1
In contrast to 1, the H NMR spectra of 2a−d indicate a
fluxional coordination environment on the NMR time scale.
1
The [ONO] ligands of 2a−d are symmetric by H NMR
spectroscopy showing two aryl and two tert-butyl resonances
in all cases except 2d, for which there is coincidental overlap
of the [ONO] tert-butyl resonances. The 1H NMR spectra of
2a−d also show equivalent phosphine environments even
upon cooling the samples to −90 °C. These features suggest
a C2v molecular geometry on the NMR time scale, which can
be explained by fast exchange between apical and basal
phosphine positions in solution. Accordingly, the 31P{1H}
NMR spectra of 2a−d all showed a single doublet resonance
with a P−Rh coupling constant of about 138 Hz that was not
affected by temperature. The 31P{1H} resonances for 2a−d
are shifted to higher frequency than those in square-planar
(PR3)3RhI(OAr) complexes, suggesting a more electron-
poor rhodium center.71,73,74 Across the series 2a−d, the
31P{1H} resonance shifts downfield with the increasing
number of electron-withdrawing aryl substituents on the
phosphorus atom.
Figure 3. UV−vis absorption spectra of [ONOcat]Rh(py)3 (1),
[ONO]Rh(PMe3)2 (2a), [ONO]Rh(PMe2Ph)2 (2b), [ONO]Rh-
(PMePh2)2 (2c), and [ONO]Rh(PPh3)2 (2d), acquired in C6H6 at
298 K.
spectra of 1 and 2a−d collected at 298 K in benzene. The
spectrum of 1 is dominated by a strong absorption band in
the near-UV region at 380 nm along with a relatively weak,
broad feature in the visible range. In comparison, the UV
transitions in 2a−d are blue-shifted by 60 nm compared to
the UV transition in 1, and a broad, intense, visible-region
absorption band dominates the spectra between 530 and 630
nm. These visible absorption bands are responsible for the
blue-green colors of 2a−d, and are blue-shifted significantly
from the ligand-based π→π* transition of [ONOq]K
discussed above.15,20,22 Spectral features in the near-IR
The addition of excess phosphine to solutions of 2a or 2b
resulted in changes to the NMR spectra that are indicative of
the formation of six-coordinate phosphine adducts [ONOcat]-
12611
dx.doi.org/10.1021/ic2026076 | Inorg. Chem. 2012, 51, 12606−12618

Inorganic Chemistry
Article
portion of the UV−vis spectrum (λmax > 900) are often
observed in metal complexes containing the semiquinonate
form of the ligand, [ONOsq]2−, but were not observed in the
spectra of 1 or 2a−d.16,21,22 The broad visible transitions in
2a−d are similar to the metal-to-ligand charge-transfer
(MLCT) transitions observed in rhodium complexes
containing π-acceptor ligands such as bipyridine, phenan-
throline, and other α-diimines,42,66,75,76 in which the charge
transfer character stems from interactions between filled
metal d orbitals and antibonding ligand π* orbitals.
321 and 587 nm with concomitant growth of a band at 393 nm.
At concentrations of 40−50 μM of 2a in benzene,
approximately 160 equiv of PMe3 were required to drive the
formation of 3a to completion. Similar results were obtained
for titrations of 2b with PMe2Ph as shown in Figure 4b, though
solutions with a concentration of 40−50 μM in 2b required
approximately 1600 equiv of PMe2Ph to reach the end point.
Analysis of the spectral data shown in Figure 4 using
Hyperquad200354,55 gave log Keq = 3.77(3) for the addi-
tion of PMe3 to 2a (ΔG298K = −5.14(4) kcal mol−1) and log
Keq = 2.01(2) for the addition of PMe2Ph to 2b (ΔG298K
=
Phosphine Binding Studies. The conversion of 2a and
2b to 3a and 3b, respectively, by the addition of phosphine is
readily evident in the electronic absorption spectra of these
complexes, providing a reliable method for studying the
thermodynamics associated with coordination of the third
phosphine ligand. The addition of excess PMe3 to 2a or
PMe2Ph to 2b resulted in color changes from blue to brown/
black. The brown/black color of 3a and 3b reverted to the
blue color of 2a and 2b, respectively, upon dilution.
Similarly, concentrating solutions of 3a under vacuum
resulted in reversion to the blue color of 2a, attributable to
the removal of volatile PMe3. These results indicated
coordination of a third phosphine ligand to 2a and 2b is
reversible and concentration dependent. In the case of 2c
and 2d, the addition of excess PMePh2 and PPh3,
respectively, to solutions of the complexes did not result in
−2.74(2) kcal mol−1).
Deconvolution of the end-point spectra for the titrations of
2a and 2b with phosphine using Hyperquad200354,55 provided
calculated electronic absorption spectra for the tris(phosphine)
adducts 3a and 3b. Figure 5 shows the calculated absorption
1
an observable color change, consistent with the H NMR
behavior reported above.
Equilibrium binding constants for the addition of phosphine
to 2a and 2b were obtained from spectrophotometric titrations.
As shown in Figure 4a, the addition of PMe3 to a benzene
solution of 2a resulted in bleaching of the absorption bands at
Figure 5. Comparison of the UV−vis absorption spectrum of
[ONOcat]Rh(py)3 (1) with the calculated spectra for [ONOcat]Rh-
(PMe3)3 (3a) and [ONOcat]Rh(PMe2Ph)3 (3b). Spectra are in C6H6
at 298 K.
spectra for 3a and 3b, which are strikingly similar to the
absorption spectrum of 1. Specifically, they exhibit a weak,
broad absorbance across most of the visible region, a strong UV
absorbance at about 390 nm, which is only 10 nm red-shifted
from the similar feature in 1. The similarity between the absorp-
tion spectra of 1, 3a, and 3b and the strong differences between
the spectra for these complexes and the spectra for 2a−d indicates
that the electronic structure of these [ONO]Rh complexes is very
sensitive to the number of ligands coordinated to the metal center,
though not necessarily to the type of ligand coordinated to the
metal center.
Ligand Displacement Studies. The relative energies of
pyridine adduct 1 and phosphine adducts 2a−d were probed
by titration experiments. The conversion of solutions of
[ONOcat]Rh(py)3 (1) to solutions of [ONO]Rh(PPh3)2
(2d) could be effected by the addition of PPh3. Similar
conversions were observed for the other phosphines. To
avoid the complications associated with a second equilibrium
between the bis- and tris(phosphine) derivatives, only the
conversion between 1 and 2d was studied in detail since
there is no evidence for the formation of a [ONOcat]Rh-
(PPh3)3 complex. Addition of at least 2 equiv of PPh3 to a
brown sample of 1 in C6D6 resulted in a blue/green solution
1
whose H NMR spectrum was consistent with 2d and free
Figure 4. UV−visible absorption spectra for the titrations of (a)
[ONO]Rh(PMe3)2 (2a) with 160 equiv of PMe3 and (b) [ONO]-
Rh(PMe2Ph)2 (2b) with 1,600 equiv of PMe2Ph. Titrations were
conducted in C6H6 at 298 K.
pyridine. When [ONOcat]Rh(py)3 (1) was titrated with PPh3
using short equilibration times (2−4 h), a green solution was
obtained for the addition of up to 1 equiv of phosphine.
12612
dx.doi.org/10.1021/ic2026076 | Inorg. Chem. 2012, 51, 12606−12618

Inorganic Chemistry
Article
As the amount of phosphine increased above 1 equiv, the
solution color became bluer. Figure 6a shows this result
disproportionation of eq 4 must be slow relative to the ligand
displacement reactions of eq 2 and 3.
The ligand exchange reactions of eqs 2 and 3 are reversible as
the triphenylphosphine ligands of 2d could be replaced by
pyridine to afford 1, provided a large excess of pyridine is
employed. The addition of excess pyridine-d5 (ca. 5000 equiv)
to a solution of 2d in C6D6 gave a 1H NMR spectrum
consistent with the formation of 1 and liberated PPh3. When
this reaction was carried out as a spectrophotometric titration
(low concentrations are required for phosphine dissociation to
go to completion), the UV−vis spectra shown in Figure 6c
were obtained, showing clean conversion of 2d to 1 with no
evidence of a long-lived intermediate. An overall equilibrium
constant, Kdisp = 4.6(7) × 10−4, was extracted from the plot in
Figure 6c corresponding to a free energy change ΔG298 = 4.55
0.10 kcal mol−1 for eq 5.
Phosphine Exchange Studies. In addition to the
association of phosphine ligands to five-coordinate complexes
2a and 2b, these complexes undergo a pseudo self-exchange or
comproportionation of the coordinated phosphine ligands.
Complexes 2a and 2b in C6D6 solution at room temperature
reacted to afford the mixed phosphine complex [ONO]Rh-
(PMe3)(PMe2Ph) (4). Complex 4 was identified by a new set
1
of [ONO] ligand resonances in the H NMR spectrum along
with two new phosphine resonances in the 31P{1H} NMR
spectrum (see Supporting Information). ESI mass spectrometry
confirmed the formation of 4 with the parent ion (M+) peak
observed at 739.2 amu. The formation of 4 is an equilibrium
exchange process that proceeds without an observable color
change, but that can be monitored by 31P{1H} NMR
spectroscopy. By NMR integration measurement, the
equilibrium constant (Kcom) at 298 K for eq 6 was determined
to be 12.5(3), corresponding to a free energy change of −1.50(2)
kcal mol−1.
Figure 6. UV−visible absorption spectra for the titration of (a) 1 (ca.
50 μM) with PPh3 (∼2 equiv at end point) using a 2 h equilibration
time, (b) 1 (ca. 50 μM) with PPh3 (∼2 equiv at end point) using a 4
day equilibration time, and (c) 2d (ca. 50 μM) with pyridine (∼2300
equiv at end point) using a 36 h equilibration time. All titrations were
carried out in benzene at 298 K.
spectrophotometrically: for samples with up to 1 equiv of
phosphine an absorption band grows in at 665 nm, but as the
amount of phosphine increases above 1 equiv, this band
blue-shifts to 640 nm. This result suggests the formation of
an intermediate species, presumably [ONO]Rh(py)(PPh3)
(eq 2), which then converts to 2d upon addition of more
PPh3 (eq 3).
The putative, green, phosphine-pyridine intermediate,
[ONO]Rh(py)(PPh3), is only metastable and dispropor-
tionates to 1 and 2d at longer equilibration times. Figure 6b
shows the UV−vis absorption spectra for a titration of
[ONOcat]Rh(py)3 (1) with PPh3 using a long equilibration
time for each sample (24 h). No intermediate green species is
observed when employing a long equilibration time, indicating
that the putative [ONO]Rh(py)(PPh3) intermediate dispropor-
tionates to 1 and 2d in the presence of pyridine (eq 4). To
observe the intermediate species evident in Figure 6a, the
Electrochemistry. Cyclic voltammetric measurements were
conducted on 1 and 2a−d to probe their redox behavior. Figure 7
shows the CV traces for each complex; redox potentials are
summarized in Table 3. All potentials are referenced to the
[Cp2Fe]+/0 couple. The CVs of 2a−d were insensitive to the
presence of excess phosphine ligand. Each complex shows two
reversible, one-electron oxidation processes. The first oxidation
process is strongly dependent on the nature of the ancillary
ligand. The first oxidation of tris(pyridine) complex 1,
{[ONO]Rh(py)3}+/0, is observed at −0.91 V whereas the first
oxidation of bis(triphenylphosphine) complex 2d is shifted 640
mV more positive to −0.27 V. The first oxidation events for
2a−c fall in the region −0.60 to −0.69 V, approximately
12613
dx.doi.org/10.1021/ic2026076 | Inorg. Chem. 2012, 51, 12606−12618

Inorganic Chemistry
Article
DFT Calculations. The electronic structural properties of 1,
2a, 2d, and 3a were examined by DFT computations.
Calculations were carried out without symmetry constraints
at the B3LYP/TZVP level of theory. For complexes 1 and 2a,
the coordinates from the single-crystal X-ray diffraction
structures were used as a starting point for the geometry
optimizations. The starting structure for the optimization of
[ONOcat]Rh(PMe3)3 (3a) was generated by replacing the three
pyridine ligands of 1 with three PMe3 ligands. The ground state
solution for all four complexes proved to be the closed-shell
singlet state (S = 0), consistent with the diamagnetic character
of 1, 2a, 2d, and 3a. Attempts to locate an open-shell singlet
(i.e., a broken-symmetry singlet) resulted in collapse of the
calculation to the closed-shell solution. The triplet (S = 1 spin
system) energies for 1, 2a, and 3a were found to lie several kcal
mol−1 above the S = 0 state. The optimized geometries of both
1 and 2a were similar to the crystal structures, and pertinent
bond distances are listed in Table 2. The largest discrepancies
in the bond lengths between crystallographic and calculated
structures is the 0.06 Å difference in the Rh(1)−P(2) distance,
but most bonds varied by less than 0.03 Å. The structure of 3a
minimized to a distorted octahedral coordination geometry
similar to that of 1; selected bond distances for this solution
are shown in Table 2. The calculated structure of 2d minimized
to a coordination geometry of a slightly distorted trigonal
bipyramid (see Supporting Information). This assignment is
borne out by the calculated equatorial bond angles in the
calculated structure of 2d, in which NONO−Rh−P angles are
138° and 119°, with a P−Rh−P angle of 103°; in comparison,
measured NONO−Rh−P angles in 2a were 155° and 106° with a
P−Rh−P bond angle of 96°. The relaxation of 2d toward
trigonal bipyramidal geometry presumably is driven by the
increased steric bulk of the PPh3 ligands. Notably, the aryl C−C
bond distances within the [ONO] ligands (excepting C(1)−
C(2) and C(15)−C(16)) of 1, 2a, 2d, and 3a, were all cal-
culated to be 1.40 0.02 Å, strongly indicative of a [ONOcat]3−
oxidation state of the redox-active ligand across all four
complexes.
Figure 7. Cyclic voltammograms of 1 and 2a−2d. Spectra were
recorded at 1 mM analyte concentration in 0.1 M [NBu4][PF6] in dry,
degassed THF under a nitrogen atmosphere using a 3 mm glassy
carbon working electrode, Pt counter electrode, and silver wire
reference electrode, at 300 K with a 200 mV/s scan rate.
Table 3. Reduction Potentials vs. [FeCp2]+/0 of [ONO]RhLn
a
Complexes
E°′/V vs [Cp2Fe]1+/0
{[ONO]RhLn}2+/1+
{[ONO]RhLn}1+/0
{[ONO]RhLn}0/2−
b
1
0.12(1) [0.98]
0.10(1) [1.01]
0.10(1) [1.01]
0.07(3) [1.02]
−0.91(1) [1.03]
−0.69(1) [1.06]
−0.60(1) [1.03]
−0.61(3) [1.09]
−0.27(1) [1.06]
−2.87(2)
−2.40(3)
−2.32(4)
−2.34(3)
An analysis of the frontier molecular orbitals of 1, 2a, and 3a
provide insight into the electronic differences between the five-
and six-coordinate [ONO]RhLn complexes. Figure 8 shows
calculated frontier molecular orbital diagrams for each complex.
Orbitals of principally d character are labeled accordingly, with
the [ONO] ligand oriented in the xy plane. The relative orbital
energies and overall appearance of the frontier molecular
orbitals in 1 and 3a are similar, and follow the expected pattern
for a d6 octahedral complex. The highest occupied molecular
orbital (HOMO) and HOMO−1 correspond to the Rh(dyz)−
NONO(pz) and Rh(dxz)−OONO(pz) π* orbitals, respectively, and
are localized mainly on the [ONO] ligand. The t2g-like metal
orbitals dxy, dxz, and dyz are lower in energy and filled. As
2
b
2a
2b
2c
2d
b
b
c
b
c
0.19(2) [1.06]
−2.11(3) [1.90]
a
Conditions are detailed in Figure 7. Bracketed values are ipa/ipc ratios at
b
c
200 mV sec−1. Irreversible, 2e− process (Epc). ipc/ipa at 200 mV sec−1.
halfway between those of 1 and 2d. The second oxidation
process, {[ONO]RhLn}2+/+ is insensitive both to the rhodium
coordination geometry and to the ancillary ligands with the
potentials for 1 and 2a−d falling in the narrow window +0.07
to +0.19 V. Reduction of 1 and 2a−d occurred in an
irreversible two-electron process (as determined by peak
integration) that again was dependent on the ancillary ligand.
Triphenylphosphine derivative 2d was reduced most easily at
−2.11 V while tris(pyridine) complex 1 required −2.87 V to
drive the reduction. As with the first oxidation potential, the
reduction potentials of complexes 2a−c fell in between those
for 1 and 2d. The two-electron reductions of 1 and 2a−c
appear completely irreversible at 200 mV sec−1, and while the
reduction of 2d appears to be at least partially reversible, the
ipc/ipa ratio for this reduction was found to be independent of
scan rate, indicating a partially irreversible process.
2
2
expected for an octahedral complex, the metal dz and dx −y
orbitals are σ-antibonding with respect to the ligand set. In the
2
case of 3a the dz orbital is the LUMO of the complex and the
2
2
dx −y is the LUMO+1. In the case of 1, the π* orbitals of the
2
2
2
pyridine ligands fall below the metal dz and dx −y orbitals;
however, the metal orbitals maintain the same energy ordering
as in 3a.
The relative energies and appearance of the frontier orbitals
in 2a (Figure 8) and 2d (see Supporting Information) are
significantly different from the orbitals of 1 and 3a. Notably, the
HOMO and lowest unoccupied molecular orbital (LUMO) of
2
2a both contain significant rhodium dz character. The distorted
12614
dx.doi.org/10.1021/ic2026076 | Inorg. Chem. 2012, 51, 12606−12618

Inorganic Chemistry
Article
Figure 8. Calculated frontier molecular orbital diagrams displaying Kohn−Sham orbitals for 1 (left), 2a (center), and 3a (right). For each
compound, the HOMO is designated with an electron pair. For clarity, only the HOMO, HOMO-1, and primarily d-based orbitals are illustrated.
square pyramidal geometry of 2a places the rhodium atom 0.28
Å above the mean-squared plane defined by the coordinating
atoms of the [ONO] ligand and P(1). Because of this dis-
Experimental evidence showed that complexes 1−3 exist as S =
0 complexes with no unpaired electrons. In the case of
[ONOcat]RhIII and [ONOq]RhI electron distributions, closed-
shell electron configurations on both the ligand and the metal
would give rise to an overall S = 0 spin state. For a [ONOsq]RhII
electron configuration, two qualitative descriptions are possible.
The first possibility is an open-shell biradical configuration with
antiferromagnetic coupling between the unpaired metal and ligand
electrons to give an S = 0 spin state as has been observed for other
rhodium complexes with redox-active ligands.38 DFT computa-
tions on 1−3 do not support this open-shell, singlet-diradical
electron configuration. Consistent with these results, all attempts
to calculate a “broken-symmetry” singlet electronic configuration
resulted in collapse of the calculation to a closed-shell, S = 0
solution; furthermore, the absence of low-energy (λmax > 900 nm),
intraligand π→π* transitions typical of the semiquinonate
[ONOsq]2− ligand19−22 further refute the existence of open-shell,
singlet-diradical character in 1−3. The second qualitative
description that would give rise to a [ONOsq]RhII electron
configuration would be to pair up the odd [ONOsq]2− electron
and the odd RhII electron in a shared molecular orbital such as a π
bond to afford an S = 0 spin state, consistent with the
experimental evidence for 1−3.
2
tortion, the rhodium dz has a nonzero overlap with the highest-
energy filled π orbital of the [ONO] ligand, creating a filled
Rh−N π orbital and an empty Rh−N π* orbital. A similar π
bonding interaction exists in 2d, with the caveat that the
d-orbital hybridization changes slightly owing to the change
between square-pyramidal and trigonal bipyramidal geometries.
Natural population analysis was performed on the calculated
structures of 2a, 2d, and 3a to better quantify the localization of
the HOMO and LUMO in these representative structures. The
results of these calculations are listed in Table 4, with a more
Table 4. Natural Population Analysis Orbital Contributions
in [ONO]RhLn Complexes in Number of Electrons
(percent)
2a
HOMO LUMO HOMO LUMO HOMO LUMO
0.53 0.79 0.54 0.66 0.06 1.10
2d
3a
Rh
(27%) (40%) (27%) (33%) (3%) (55%)
[ONO] 1.28 0.76 1.20 0.86 1.80 0.31
(64%) (38%) (60%) (43%) (90%) (16%)
Structural metrics and spectroscopic properties of the six-
coordinate rhodium complexes are consistent with their
assignment as [ONOcat]RhIII complexes. The six-coordinate,
pseudo-octahedral structure of 1 is completely consistent with
its assignment as a rhodium(III) complex. Bond metrics within
the [ONO] ligand support this contention, specifically with
C−C ring distances in the range of 1.39−1.44 Å and single-
bond C−N and C−O distances. Spectroscopically, complex 1
shows only weak absorptions in the visible region accompanied
by a strong charge-transfer absorption at 380 nm. The 1H NMR
spectrum of 1 shows two distinct pyridine environments, con-
sistent with a meridional arrangement of three pyridine ligands
and slow ligand exchange as expected for a d6 octahedral com-
plex of rhodium. These data indicate that [ONOcat]RhIII(py)3 is
both the best formal and spectroscopic oxidation state assign-
ment for 1, an assignment that is supported by DFT
detailed table in the Supporting Information. As the molecular
orbitals in Figure 8 suggest, 90% of the HOMO of the octahedral
3a resides on the [ONO] ligand, with only 3% rhodium character
in the HOMO, supporting a formal [ONOcat]RhIIIL3 electronic
configuration. In 2a and 2d, however, the [ONO] contribution to
the HOMO decreases to approximately 60% with the rhodium
contribution to the HOMO increasing to 27%, indicating a more
reduced rhodium metal center in 2a and 2d compared to 3a.
DISCUSSION
■
Electronic Structure. Both the [ONO] ligand and
rhodium metal can exist in multiple oxidation states, allowing
for three plausible electron distributions for [ONO]Rh com-
plexes 1−3: [ONOcat]RhIII, [ONOsq•]RhII, and [ONOq]RhI.
12615
dx.doi.org/10.1021/ic2026076 | Inorg. Chem. 2012, 51, 12606−12618

Inorganic Chemistry
Article
2
calculations. The strong correlation between the UV−vis
absorption spectra of 1 and 3a−b support the assignment of 3a
and 3b as [ONOcat]RhIII(PMe3)3 and [ONOcat]RhIII(PMe2Ph)3,
respectively, a conclusion that is also supported by DFT results.
The striking UV−vis absorption spectra of 2a−d suggested a
different electronic configuration for these five-coordinate com-
plexes relative to the six-coordinate complexes 1 and 3a−b.
Isolated phosphine adducts of the [ONO]Rh core in the solid
state contained only two phosphine ligands to give five-
coordinate [ONO]RhL2 complexes. There is no direct evidence
for four-coordinate, [ONO]RhL derivatives, which is surprising
given the proclivity for rhodium to adopt a square-planar
geometry and a d8 electron configuration; even with very bulky
phosphines, a stable four-coordinate [ONO]Rh(PR3) complex
was never successfully isolated. Complexes 2a−d show strong
UV−vis absorptions in the visible region of the spectrum with
maxima in the range of 530−630 nm. Such absorptions are
reminiscent of the absorption features of rhodium(I) complexes
such as the [Rh(bpy)2]+ cation,77,78 which has an MLCT
absorbance at 556 nm, suggesting a possible oxidation state
assignment of [ONOq]RhIL2 for complexes 2a−d. Conversely,
the intraligand bond metrics in the X-ray structure of 2a are not
consistent with an [ONOq]− ligand oxidation state. Specifically,
the C−C ring distances vary by only 0.03 Å, whereas [ONOq]−
ligands typically have C−C bond distances that vary by 0.10
Å,21−23 indicative of a cyclohexadiene ring. Thus, the
experimental data does not support a [ONOq]RhI electronic
configuration in 2a (and by analogy in 2b−2d).
The electronic configuration of 2a−d is further muddled
upon consideration of the electrochemical data for complexes 1
and 2a−d. All five complexes show the same number of redox
processes in the electrochemical window between −3 V and
+1 V. Complex 1 is the easiest to oxidize and the hardest
complex to reduce, while complex 2d, with the bulky PPh3
ligand, is the hardest to oxidize and the easiest to reduce. Both
the oxidations and the reductions of complexes 2a−c lie in
between the extremes represented by 1 and 2d. The consistent
trend in redox properties among the five complexes argues
against an abrupt change in electronic structure between 1 and
2a−d. Furthermore, since the electrochemical data acquired for
2a−d is the same in the absence and in the presence of excess
phosphine, the formation of six-coordinate adducts 3a and 3b
do not perturb the overall electronic structure of these
complexes significantly
HOMO and LUMO of 2a have significant metal dz parentage.
The distorted square-pyramidal geometry of 2a results in an
obtuse angle (100°) between the apical phosphine donor, the
rhodium center, and the [ONO] nitrogen and bestows nonzero
2
overlap between the rhodium dz orbital and the highest energy
[ONO]-ligand π* orbital. These two orbitals interact as
illustrated in Figure 8 to form Rh−[ONO] π and π* com-
binations that constitute the HOMO and LUMO, respectively,
of the complex. Thus, the strong, low-energy absorption
observed in the UV−vis spectrum of 2a can be assigned as a
HOMO−LUMO, π to π* transition. DFT calculations of the
calculated structure for 2d revealed similar Rh−[ONO] π and
π* interactions in the HOMO and LUMO, respectively.
Five-coordinate rhodium complexes 2a−d are best described
as having experimental oxidation states that lie between
[ONOsq]RhII(PR3)2 and [ONOcat]RhIII(PR3)2. As shown in
Figure 8 and in the Supporting Information, the HOMO elec-
tron pair in 2a and 2d (and by analogy, 2b and 2c) is shared
between the rhodium center and the [ONO] ligand in a
π-bonding orbital. The sharing of this electron pair between the
rhodium center and the [ONO] ligand leads to the distinct
electronic character of complexes 2a−d, while the polarization
of the π bond toward the [ONO] ligand suggests an electron
distribution between [ONOsq]RhII(PR3)2 and [ONOcat]-
RhIII(PR3)2. The proclivity for 2a and 2b to add reversibly a
third phosphine ligand is consistent with an experimental
oxidation state between [ONOsq]RhII(PR3)2 and [ONOcat]-
RhIII(PR3)2, since a five-coordinate rhodium(III) complex
would be expected to be Lewis acidic. Incidentally, population
analysis of 2d revealed similar contributions from the [ONO]
ligand and the rhodium center to the HOMO and LUMO,
supporting the argument that association of a third phosphine
ligand ligand to 2c and 2d is inhibited by steric rather than
electronic considerations.
Ligand Exchange Reactions. Ligand exchange studies of
the [ONO]RhLn complexes reveal a complicated energetic
landscape with stable coordination geometries that are
influenced by both the steric and the electronic properties of
the auxiliary ligand. In the case of 1, with pyridine donor ligands,
only the six-coordinate geometry was observed; however, the
pyridine ligands of 1 are labile and can be readily substituted in
solution. Considering that 1 is a d6, pseudo-octahedral complex,
ligand exchange likely occurs through a dissociative pathway; this
suggests that a five-coordinate intermediate, [ONO]Rh(py)2, is
energetically accessible under mild conditions. Consistent with this
proposal, degenerate pyridine exchange occurs on a time scale of
seconds and replacement of the pyridine ligands of 1 with stronger
σ-donor phosphines such as PMe3 or PMe2Ph, rapidly affords
both five and six-coordinate phosphine adducts. Titration experi-
ments indicated an overall thermodynamic stability in which the
tris(phosphine) complexes 3a and 3b are favored over bis-
(phosphine) complexes 2a−d, which are favored over tris-
(pyridine) complex 1a; however, the presence of a large excess of
phosphine is required to push the formation of 3a and 3b to
completion at low concentrations. The favorability of the six-
coordinate phosphine complexes is likely driven by electronic
considerations, with coordination of the sixth ligand quenching the
The electronic difference between five- and six-coordinate
[ONO]Rh complexes was best illustrated in the DFT cal-
culations. In the case of pseudo-octahedral complexes 1 and 3a,
DFT calculations afforded a frontier orbital picture consistent
with the expected d6 RhIII configuration. The rhodium dxy, dxz,
2
and dyz orbitals are fully occupied (Figure 8) while the dz and
2
2
dx −y orbitals are empty and displaced to higher energy owing
to their σ* parentage with respect to the ligand field. The
calculated HOMOs for 1 and 3a are [ONO]-ligand-centered
π* orbitals. In the case of 1, the LUMO is based on the
coordinated pyridine ligands; for 3a, the LUMO is the rhodium
2
dz orbital. In the cases of 1 and 3a−b, the regular octahedral
geometry of these complexes confers classical coordination
complex electronic behavior and these complexes are well
described as [ONOcat]RhIIIL3 species.
2
Lewis acidic behavior of the rhodium dz orbital to form a tradi-
tional Werner-type coordination complex. The electronic pre-
ference for a pseudo-octahedral geometry is slight, though, since
five-coordinate complexes are only observed for 2c and 2d,
which employ the sterically demanding phosphine ligands PMePh2
or PPh3, respectively. In these two complexes, the energetic
The distorted five-coordinate geometry of 2a results in a
different frontier orbital configuration. While the rhodium dxy,
2
2
dxz, and dyz orbitals are fully occupied and the dx −y orbital is an
empty σ* orbital, Figure 8 shows that both the calculated
12616
dx.doi.org/10.1021/ic2026076 | Inorg. Chem. 2012, 51, 12606−12618

Inorganic Chemistry
Article
penalty for coordinating a third, sterically demanding phosphine
ligand is not fully compensated by the quenching of the Lewis
REFERENCES
■
(1) Kaim, W.; Schwederski, B. Coord. Chem. Rev. 2010, 254, 1580−
1588.
(2) Kaim, W.; Schwederski, B. Pure Appl. Chem. 2004, 76, 351−364.
(3) Kaim, W.; Lahiri, G. K. Angew. Chem., Int. Ed. 2007, 46, 1778−
1796.
2
acidic rhodium dz orbital and formation of the third Rh−P bond.
2
In the five-coordinate complexes 2a−d the rhodium dz orbital is
partially stabilized through π bonding with the nitrogen donor of
the [ONOcat]3− ligand.
(4) Bart, S. C.; Chlopek, K.; Bill, E.; Bouwkamp, M. W.; Lobkovsky,
E.; Neese, F.; Wieghardt, K.; Chirik, P. J. J. Am. Chem. Soc. 2006, 128,
13901−13912.
CONCLUDING REMARKS
■
Care must be taken when describing the electronic structure of
coordination complexes with redox-active or noninnocent
ligands. In the case of complexes 2a−d, a rigid view of formal
metal and ligand oxidation states de-emphasizes the importance
of covalency in the bonding between the rhodium and [ONO]
ligand. This fact is readily apparent in complexes 2a−d, in
which both the [ONO] ligand and the rhodium metal center
make significant contributions to the π-bonding HOMO and
π-antibonding LUMO giving rise to unique spectroscopic
properties. In these complexes, the formal oxidation state is
[ONOcat]RhIII(PR3)2; however, it is better to consider
[ONOcat]RhIII(PR3)2 and [ONOsq]RhII(PR3)2 as limiting
resonance structures for 2a−d to emphasize that the HOMO
electron pair is shared by both metal and ligand centers. It is
important to note that the [ONOsq]RhII(PR3)2 electron con-
figuration proposed here, which arises from a covalent π-bonding
interaction between the [ONO] ligand and the rhodium metal
center, is distinct from an open-shell, singlet-biradical electron
configuration.
(5) Masui, H.; Lever, A. B. P.; Auburn, P. R. Inorg. Chem. 1991, 30,
2402−2410.
(6) Adams, D. M.; Noodleman, L.; Hendrickson, D. N. Inorg. Chem.
1997, 36, 3966−3984.
(7) Pierpont, C. G. Coord. Chem. Rev. 2001, 219−221, 415−433.
(8) Pierpont, C. G.; Buchanan, R. M. Coord. Chem. Rev. 1981, 38,
45−87.
(9) Chaudhuri, P.; Verani, C. N.; Bill, E.; Bothe, E.; Weyhermuller,
̈
T.; Wieghardt, K. J. Am. Chem. Soc. 2001, 123, 2213−2223.
(10) Zanello, P.; Corsini, M. Coord. Chem. Rev. 2006, 250, 2000−
2022.
(11) Blackmore, K. J.; Lal, N.; Ziller, J. W.; Heyduk, A. F. J. Am.
Chem. Soc. 2008, 130, 2728−2729.
(12) Vechorkin, O.; Csok, Z.; Scopelliti, R.; Hu, X. Chem.Eur. J.
2009, 15, 3889−3899.
(13) Nguyen, A. I.; Blackmore, K. J.; Carter, S. M.; Zarkesh, R. A.;
Heyduk, A. F. J. Am. Chem. Soc. 2009, 131, 3307−3316.
(14) Nguyen, A. I.; Zarkesh, R. A.; Lacy, D. C.; Thorson, M. K.;
Heyduk, A. F. Chem. Sci. 2011, 2, 166−169.
(15) Girgis, A. Y.; Balch, A. L. Inorg. Chem. 1975, 14, 2724−2727.
(16) Camacho-Camacho, C.; Merino, G.; Martinez-Martinez, F. J.;
The reversible coordination of a sixth ligand to 2a and 2b is
intriguing because it results in a dramatic change to the
spectroscopic properties of the rhodium complexes. That these
spectroscopic changes can be contributed to a change in the
Noth, H.; Contreras, R. Eur. J. Inorg. Chem. 1999, 1021−1027.
̈
(17) Ohkata, K.; Yano, T.; Kuwaki, T.; Akiba, K. Chem. Lett. 1990,
19, 1721−1724.
(18) Poddel’sky, A. I.; Vavilina, N. N.; Somov, N. V.; Cherkasov,
V. K.; Abakumov, G. A. J. Organomet. Chem. 2009, 694, 3462−3469.
2
rhodium dz orbital energy and a transfer of electron density
(19) Evangelio, E.; Bonnet, M.-L.; Cabanas, M.; Nakano, M.; Sutter,
̃
from rhodium to the [ONO] ligand is reminiscent of similar
coordination-induced electronic rearrangements in four-coor-
dinate nickel salen and germanium porphyrin complexes.79,80 In
these examples, the four coordinate metal centers readily add
two pyridine ligands to afford six-coordinate complexes which
leads to the transfer of electron density away from the metal
center and to the salen or porphyrin ligand.
J.-P.; Dei, A.; Robert, V.; Ruiz-Molina, D. Chem.Eur. J. 2010, 16,
6666−6677.
(20) Caneschi, A.; Corina, A.; Dei, A. Inorg. Chem. 1998, 37, 3419−
3421.
(21) Bruni, S.; Caneschi, A.; Cariati, F.; Delfs, C.; Dei, A.; Gatteschi,
D. J. Am. Chem. Soc. 1994, 116, 1388−1394.
(22) Chaudhuri, P.; Hess, M.; Hildenbrand, K.; Bill, E.;
Weyhermuller, T.; Wieghardt, K. Inorg. Chem. 1999, 38, 2781−2790.
̈
(23) McGarvey, B. R.; Ozarowski, A.; Tian, Z.; Tuck, D. G. Can. J.
Chem. 1995, 73, 1213−1222.
ASSOCIATED CONTENT
■
S
* Supporting Information
Crystallographic data refinement details and crystallographic
(24) Simpson, C. L.; Boone, S. R.; Pierpont, C. G. Inorg. Chem. 1989,
28, 4379−4385.
(25) Camacho-Camacho, C.; Tlahuext, H.; Noth, H.; Contreras, R.
̈
information files for reported X-ray structures, supporting
titration spectra, MO diagrams for 2d, and H, 13C{1H}, and
1
Heteroat. Chem. 1998, 9, 321−326.
(26) Camacho-Camacho, C.; Mijangos, E.; Castillo-Ramos, M. E.;
31P{1H} NMR spectra for all new complexes. This material is
available free of charge via the Internet at http://pubs.acs.org.
́
Esparza-Ruiz, A.; Vasquez-Badillo, A.; Noth, H.; Flores-Parra, A.;
̈
Contreras, R. J. Organomet. Chem. 2010, 695, 833−840.
(27) Chaudhuri, P.; Hess, M.; Weyhermuller, T.; Wieghardt, K.
̈
Angew. Chem., Int. Ed. 1999, 38, 1095−1098.
AUTHOR INFORMATION
■
(28) Zarkesh, R. A.; Heyduk, A. F. Organometallics 2009, 28, 6629−
6631.
Corresponding Author
*E-mail: aheyduk@uci.edu.
(29) Zarkesh, R. A.; Ziller, J. W.; Heyduk, A. F. Angew. Chem., Int. Ed.
2008, 47, 4715−4718.
(30) Zarkesh, R. A.; Heyduk, A. F. Organometallics 2011, 30, 4890−
4898.
Notes
The authors declare no competing financial interest.
(31) Szigethy, G.; Heyduk, A. F. Inorg. Chem. 2011, 50, 125−135.
(32) Blackmore, K. J.; Lal, N.; Ziller, J. W.; Heyduk, A. F. Eur. J. Inorg.
Chem. 2009, 735−743.
(33) Blackmore, K. J.; Sly, M. B.; Haneline, M. R.; Ziller, J. W.;
Heyduk, A. F. Inorg. Chem. 2008, 47, 10522−10532.
(34) Ketterer, N. A.; Fan, H.; Blackmore, K. J.; Yang, X.; Ziller, J. W.;
Baik, M.-H.; Heyduk, A. F. J. Am. Chem. Soc. 2008, 130, 4364−4374.
ACKNOWLEDGMENTS
■
The authors thank the UCI Physical Science Center for Solar
Energy for supporting this work. They also thank Dr. Philip
Dennison for help with the NMR measurements. A.F.H. is a
Camille Dreyfus Teacher-Scholar.
12617
dx.doi.org/10.1021/ic2026076 | Inorg. Chem. 2012, 51, 12606−12618

Inorganic Chemistry
Article
(35) Haneline, M. R.; Heyduk, A. F. J. Am. Chem. Soc. 2006, 128,
8410−8411.
(71) Nishihara, Y.; Nara, K.; Nishide, Y.; Osakada, K. J. Chem. Soc.,
Dalton Trans. 2004, 1366−1375.
(72) Allen, F. H.; Kennard, O.; Watson, D. G.; Brammer, L.; Orpen,
A. G.; Taylor, R. J. Chem. Soc., Dalton Trans. II 1987, S1−S19.
(73) Osakada, K.; Ishii, H. Inorg. Chim. Acta 2004, 357, 3007−3013.
(74) Kuznetsov, V. F.; Yap, G. P. A; Bensimon, C.; Alper, H. Inorg.
Chim. Acta 1998, 280, 172−182.
(36) Blackmore, K. J.; Ziller, J. W.; Heyduk, A. F. Inorg. Chem. 2005,
44, 5559−5561.
(37) Colby, D. A.; Bergman, R. G.; Ellman, J. A. Chem. Rev. 2010,
110, 624−655.
(38) DeWit, D. G. Coord. Chem. Rev. 1996, 147, 209−246.
(39) Huang, K.-W.; Han, J. H.; Musgrave, C. B.; Fujita, M.
Organometallics 2007, 26, 508−513.
(40) Schwartzburd, L.; Iron, M. A.; Konstantinovski, L.; Ben-Ari, E.;
Milstein, D. Organometallics 2011, 30, 2721−2729.
(41) Albert, G.; Berg, C.; Beyer, M.; Achatz, U.; Joos, S.; Niedner-
Schatteburg, G.; Bondybey, V. E. Chem. Phys. Lett. 1997, 268, 235−
241.
(75) de Pater, B. C.; Fruhous, H.-W.; Vrieze, K.; de Gelder, R.;
̈
Baerends, E. J.; McCormack, D.; Lutz, M.; Spek, A. L.; Hartl, F. Eur. J.
Inorg. Chem. 2004, 1675−1686.
(76) Youssef, T. E. Polyhedron 2010, 29, 1776−1873.
(77) Chou, M.; Creutz, C.; Mahajan, D.; Sutin, N.; Zipp, A. P. Inorg.
Chem. 1982, 21, 3989−3997.
(78) Martin, B.; McWhinnie, W. R.; Waind, G. M. J. Inorg. Nucl.
Chem. 1961, 23, 207−223.
(42) Shaffer, D. W.; Ryken, S. A.; Zarkesh, R. A.; Heyduk, A. F. Inorg.
(79) Cissell, J. A.; Vaid, T. P.; Yap, G. P. A. J. Am. Chem. Soc. 2007,
129, 7841−7847.
Chem. 2011, 50, 13−21.
(43) Allgeier, A. M.; Mirkin, C. A. Angew. Chem., Int. Ed. 1998, 37,
895−908.
(80) Storr, T.; Wasinger, E. C.; C., P. R.; Stack, D. P. Angew. Chem.,
Int. Ed. 2007, 46, 5198−5201.
(44) van der Vlugt, J. I.; Reek, J. N. H. Angew. Chem., Int. Ed. 2009,
48, 8832−8846.
(45) Giordano, G.; Crabtree, R. H.; Heintz, R. M.; Forster, D.;
Morris, D. E. Inorg. Synth. 1990, 28, 88−90.
(46) Harris, R. K.; Becker, E. D.; Cabral De Menezes, S. M.; Granger,
P.; Hoffman, R. E.; Zilm, K. W. Pure Appl. Chem. 2008, 80, 59−84.
(47) Connelly, N. G.; Geiger, W. E. Chem. Rev. 1996, 96, 877−910.
(48) APEX2, Version 2008.3-0, or (Version 2.2-0); Bruker AXS, Inc.;
Madison, WI, 2007.
(49) SAINT, Version 7.53a (or 7.46a); Bruker AXS, Inc.: Madison,
̀
WI, 2007.
(50) Sheldrick, G. M. SADABS, Version 2007/4; Bruker AXS, Inc.:
Madison, WI, 2007.
(51) Sheldrick, G. M. SHELXTL, Version 6.12; Bruker AXS, Inc.:
Madison, WI, 2001.
(52) International Tables for X-Ray Crystallography; Kluwer Academic
Publishers: Dordrecht, The Netherlands, 1992; Vol. C.
(53) Farrugia, L. J. J. Appl. Crystallogr. 1997, 30, 565.
(54) Gans, P.; Sabatini, A.; Vacca, A. HYPERQUAD2003; University
of Leed and University of Florence: Leeds, U.K., and Florence, Italy,
2003.
(55) Gans, P.; Sabatini, A.; Vacca, A. Talanta 1996, 43, 1739−1753.
(56) TURBOMOLE GmbH: Karlsruhe, Germany.
(57) Becke, A. D. J. Chem. Phys. 1993, 98, 1372−1377.
(58) Lee, C.; Yang, W.; Parr, R. G. Phys. Rev. B. 1988, 37, 785−789.
(59) Vosko, S. H.; Wilk, L.; Nusair, M. Can. J. Phys. 1980, 58, 1200−
1211.
(60) Stephens, P. J.; Devlin, F. J.; Cabalowski, C. F.; Frisch, M. F.
J. Phys. Chem. 1994, 98, 11623−11627.
(61) Weigend, F.; Ahlrichs, R. Phys. Chem. Chem. Phys. 2005, 18,
3297.
(62) Reed, A. E.; Weinstock, R. B.; Weinhold, F. J. Chem. Phys. 1985,
83, 735−746.
(63) Klein, R. F. X.; Bargas, L. M.; Horak, V. J. Org. Chem. 1988, 53,
5994−5998.
(64) Dutta, S.; Basuli, F.; Peng, S.-M.; Lee, G.-H.; Bhattacharya, S.
New J. Chem. 2002, 26, 1607−1612.
(65) Baksi, S.; Acharyya, R.; Dutta, S.; Blake, A. J.; B., D. M. G.;
Bhattacharya, S. J. Organomet. Chem. 2007, 692, 1025−1032.
(66) Blacker, A. J.; Clot, E.; Duckett, S. B.; Eisenstein, O.; Grace, J.;
Nova, A.; Perutz, R. N.; Taylor, D. J.; Whitwood, A. C. Chem.
Commun. 2009, 6801−6803.
(67) Acharya, K. R.; Tavale, S. S.; Guru Row, T. N. Acta Crystallogr.
1984, C40, 1327−1328.
(68) Hughes, R. P.; Lindner, D. C.; Liable-Sands, L. M.; Rheingold,
A. L. Organometallics 2001, 20, 363−366.
(69) Egan, J. W. Jr.; Hughes, R. P.; Rheingold, A. L. Organometallics
1987, 6, 1578−1581.
(70) Choi, J.-C.; Sakakura, T. J. Am. Chem. Soc. 2003, 125, 77762−
77763.
12618
dx.doi.org/10.1021/ic2026076 | Inorg. Chem. 2012, 51, 12606−12618