Synthesis, characterization, and photophysics of oxadiazole- and diphenylaniline-substituted Re(I) and Cu(I) complexes

Article abstract of DOI:10.1021/ic3018387

Transition-metal complexes of the types [Re(CO)₃Cl(NN)], [Re(CO)₃py(NN)]⁺, and [Cu(PPh₃) ₂(NN)]⁺, where NN = 4,4′-bis(5-phenyl-1,3,4- oxadiazol-2-yl)-2,2′-bipyridine (OX) and 4,4′-bis(N,N-diphenyl-4- [ethen-1-yl]-aniline)-2,2′-bipyridine (DPA), have been synthesized and characterized. Crystal structures for [Re(CO)₃Cl(DPA)] and [Cu(PPh₃)₂(OX)]BF₄ are presented. The crystal structure of the rhenium complex shows a trans arrangement of the ethylene groups, in agreement with density functional theory calculations. The structure of the copper complex displays the planar aromatic nature of the bpy-oxadiazole ligand. Density functional theory modeling of the complexes was supported by comparison of calculated and experimental normalized Raman spectra; the mean absolute deviations of the complexes were <10 cm^-1. The Franck-Condon state was investigated using UV-vis and resonance Raman spectroscopic as well as density functional theory computational techniques. It was shown that the lowest energy absorption peaks are metal to ligand charge transfer and ligand-centered charge transfer for the oxadiazole- and diphenylaniline-substituted bipyridine ligands, respectively. The lowest energy excited states were characterized using transient emission and absorption spectroscopic techniques in conjunction with density functional theory calculations. These showed that the DPA complexes had ligand-centered nonemissive dark states with lifetimes ranging from 300 to 2000 ns.

Full text of DOI:10.1021/ic3018387

Article
pubs.acs.org/IC
Synthesis, Characterization, and Photophysics of Oxadiazole- and
Diphenylaniline-Substituted Re(I) and Cu(I) Complexes
Raphael Horvath,† Michael G. Fraser,† Scott A. Cameron,† Allan G. Blackman,† Pawel Wagner,‡
David L. Officer,‡ and Keith C. Gordon*,†
†MacDiarmid Institute for Advanced Materials and Nanotechnology, Department of Chemistry, University of Otago, Dunedin, New
Zealand
‡ARC Centre of Excellence for Electromaterials Science, Intelligent Polymer Research Institute, University of Wollongong, Australia
S
* Supporting Information
ABSTRACT: Transition-metal complexes of the types [Re(CO)3Cl(NN)],
[Re(CO)3py(NN)]+, and [Cu(PPh3)2(NN)]+, where NN = 4,4′-bis(5-phenyl-
1,3,4-oxadiazol-2-yl)-2,2′-bipyridine (OX) and 4,4′-bis(N,N-diphenyl-4-[ethen-1-
yl]-aniline)-2,2′-bipyridine (DPA), have been synthesized and characterized.
Crystal structures for [Re(CO)3Cl(DPA)] and [Cu(PPh3)2(OX)]BF4 are
presented. The crystal structure of the rhenium complex shows a trans
arrangement of the ethylene groups, in agreement with density functional theory
calculations. The structure of the copper complex displays the planar aromatic
nature of the bpy−oxadiazole ligand. Density functional theory modeling of the
complexes was supported by comparison of calculated and experimental
normalized Raman spectra; the mean absolute deviations of the complexes
were <10 cm−1. The Franck−Condon state was investigated using UV−vis and
resonance Raman spectroscopic as well as density functional theory computa-
tional techniques. It was shown that the lowest energy absorption peaks are metal to ligand charge transfer and ligand-centered
charge transfer for the oxadiazole- and diphenylaniline-substituted bipyridine ligands, respectively. The lowest energy excited
states were characterized using transient emission and absorption spectroscopic techniques in conjunction with density functional
theory calculations. These showed that the DPA complexes had ligand-centered nonemissive “dark” states with lifetimes ranging
from 300 to 2000 ns.
photophysical properties that these substituents may induce is
crucial in designing useful functional materials. While the
{Re(CO)3Cl} moiety has been shown to be electron
withdrawing,21 in [Cu(PPh3)2(bpy)]+ and related compounds,
the level of communication between the copper(I) and the
ligand is relatively low. For MLCT transitions, the radical anion
of the ligand can in theory be observed with little interference
from the metal. Furthermore, in contrast to Ru(II)-based
complexes, little interference from ancillary ligands is expected.
This makes both Re(I) and Cu(I) complexes excellent
reference compounds to ascertain the effects of the
aforementioned substituents on the ground and excited-state
electronic properties of metal−polypyridyl complexes.
This work examines the effect of substitution of 2,2′-
bipyridine ligands at the 4,4′ positions by electron-withdrawing
oxadiazole and electron-donating diphenylaniline substituents
to give the 4,4′-bis(5-phenyl-1,3,4-oxadiazol-2-yl)-2,2′-bipyri-
dine (OX) and 4,4′-bis(N,N-diphenyl-4-[ethen-1-yl]-aniline)-
2,2-bipyridine (DPA) ligands, which are shown in Figure 1.
Complexes investigated were [Re(CO)3Cl(NN)], [Re-
(CO)3py(NN)]+, and [Cu(PPh3)2(NN)]+ (where NN = bpy,
INTRODUCTION
■
Metal polypyridyl complexes, such as [Ru(bpy)3]2+, [Re-
(CO)3Cl(bpy)], and [Cu(PPh3)2(bpy)]+ (bpy = 2,2′-bipyr-
idine) and their derivatives, have been studied for applications
such as solar energy conversion or display technologies1−8 due
3
to the presence of low-energy MLCT (metal-to-ligand charge
transfer) states. Consequently, the ground and excited-state
electronic structures, dipoles, and associated vibrational modes
of [Re(CO)3Cl(bpy)], [Cu(PPh3)2(bpy)]+, and related com-
pounds are relatively well studied and known.1,9−12 By
appropriate functionalization, these complexes can be used to
effect charge separation by photolysis or act as emissive centers
under application of a potential. To this end, electron-
withdrawing and -donating groups such as oxadiazoles and
diphenylanilines are often appended to metal complexes to
induce electron- and hole-transporting properties, respectively.
Oxadiazole-substituted complexes3,13,14 and complex poly-
mers15 have been investigated for their use in electro-
luminescent and sensitizing applications. A number of groups
have substituted ruthenium-based solar cell dyes with
diphenylanilines, attached via ethenyl linkers, in order to
improve the solar absorption profiles.16−20 In every case this
induced strong absorption bands at ca. 420−500 nm; however,
these remain largely unassigned. Correct understanding of the
Received: August 22, 2012
© XXXX American Chemical Society
A
dx.doi.org/10.1021/ic3018387 | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Figure 1. OX and DPA ligands used to synthesize complexes [Re(CO)3Cl(NN)], [Re(CO)3py(NN)]+, and [Cu(PPh3)2(NN)]+.
Hz, phenyl), 8.681 (m, 4H, bpy and py overlapping), 9.63 (s, 2H,
bpy), 9.80 (d, 2H, J = 6, bpy). Anal. Calcd for C34H21N7O5RePF6: C,
43.50; H, 2.26; N, 10.45. Found: C, 43.32; H, 2.43; N, 10.27. ESI
POS: m/z 793.74 {(Re(CO)3py(OX))+}.
OX, or DPA and py = pyridine). Resonance Raman
spectroscopy, transient emission, and ground- and excited-
state absorption spectroscopic techniques were used. Density
functional theory (DFT) optimization and frequency calcu-
lations were carried out at the 6-31G(d)/LANL2DZ level to aid
assignment of vibrational modes in the resonance Raman
spectra; their accuracies were verified using FT-IR and FT-
Raman spectroscopic techniques. Furthermore, time-dependent
(TD) DFT calculations in a solvent field have been used to
quantify electronic transitions using both the B3LYP and the
CAM-B3LYP functionals.
[Cu(PPh3)2(OX)]BF4. Dry diethyl ether (20 mL) was purged with
nitrogen for 10 min. [Cu(PPh3)4] BF4 (0.363 g, 0.337 mmol) and OX
(0.150 g, 0.337 mmol) were added, and the suspension was stirred
under N2 for 48 h. The yellow precipitate was filtered and washed with
diethyl ether and recrystallized from a minimum amount of methanol.
Yield: 0.200 g (0.178 mmol, 53%). Crystals of suitable quality for X-
ray crystallography were grown by slow diffusion of diethyl ether into a
concentrated dichloromethane solution. 1H NMR (CDCl3, 400 MHz):
δ 7.10−7.75 (m, 34H), 8.26 (dd, 2H, J1 = 5 Hz, J2 = 1 Hz), 8.36 (m,
6H), 8.66 (m, 2H), 9.14 (s, 2H). Anal. Calcd for
C62H46N6O2P2CuBF4: C, 65.47; H, 4.25; N, 7.39. Found: C, 65.23;
H, 4.00; N, 7.22. ESI POS: m/z 769.15 {(Cu(PPh3)(OX)+)}.
[Re(CO)3Cl(DPA)]. A 0.100 g amount of DPA (0.144 mmol) and
0.052 g of [Re(CO)5Cl] (0.144 mmol) were refluxed for 7 h in 30 mL
of ethanol under N2. The resulting red solution was refrigerated
overnight. The red solid was filtered to give analytically pure solid.
EXPERIMENTAL SECTION
■
Synthesis. All commercially available reagents were used as
received. The ligands OX (4,4′-bis(5-phenyl-1,3,4-oxadiazole)-2,2′-
bipyridine) and DPA (4,4′-bis(2-(4-diphenylaminophenyl))-2,2′-bi-
pyridine) were prepared by literature methods.13,15,22
All metal complexes were synthesized by established literature
techniques with minor variations.12,23 The precursor [Cu(PPh3)4]BF4
was synthesized by the method of Rader et al.24
1
Yield: 0.126 g (0.127 mmol, 88%). H NMR: (CDCl3, 400 MHz) δ
6.72 (d, 2H, J = 16 Hz, ethenyl), 7.01−7.19 (m, 20H), 7.30−7.35 (m,
8H), 7.53 (d, 2H, J = 16 Hz, ethenyl), 7.58 (d, 2H, J = 6 Hz), 8.28 (s,
2H, bpy), 8.58 (d, 2H, J = 6 Hz, bpy). Anal. Calcd for
C53H38N4O3ClRe: C, 63.62; H, 3.83; N, 5.60. Found: C, 63.51; H,
4.05; N, 5.63. ESI POS: m/z 983.25 {(Re(CO)3(DPA) + H2O)+},
965.24 {Re(CO)3(DPA)+}.
[Re(CO)3Cl(OX)]. A 0.100 g amount of OX (0.225 mmol) and 0.082
g of Re(CO)5Cl (0.225 mmol) were refluxed for 7 h in 30 mL of
ethanol under N2. The resulting orange solution was refrigerated
overnight. The orange solid was filtered to give the analytically pure
solid. Yield: 0.157 g (0.209 mmol, 93%). 1H NMR (CDCl3, 400
MHz): δ 7.60−7.69 (m, 6H, phenyl), 8.21−8.28 (m, 6H, phenyl and
bpy overlapping), 9.09 (s, 2H, bpy), 9.30 (d, 2H, J = 6 Hz, bpy). Anal.
Calcd for C29H16N6O5ClRe: C, 46.43; H, 2.15; N, 11.21. Found: C,
46.51; H, 2.24; N, 11.08. ESI POS: m/z 773 {([Re(CO)3Cl(OX)] +
Na+)}.
[Re(CO)3py(DPA)]PF6. A 0.1 g of amount of [Re(CO)3Cl(DPA)]
(0.1 mmol) and 0.025 g of AgPF6 (0.20 mmol) were refluxed in 50 mL
of acetonitrile under N2 in the dark for 18 h. The resulting red solution
was filtered through Celite to remove the AgCl precipitate. A 10 mL
amount of saturated aqueous NH4PF6 was added to the filtrate.
Acetonitrile was then removed under reduced pressure. The resulting
red solid (0.1 g) was filtered and washed with water. The solid,
assumed to be [Re(CO)3(CH3CN)DPA]PF6, was reacted to form the
pyridine adduct without characterization. A 0.1 g amount of
[Re(CO)3(CH3CN)(DPA)]PF6 (0.09 mmol) and 0.200 g of pyridine
(2.61 mmol) were refluxed in 30 mL of methanol for 4 h. The solution
was reduced in volume to 25 mL under reduced pressure and
subsequently refrigerated overnight. The red solid was filtered and
washed with water. A 10 mL amount of NH4PF6(aq) was added to the
filtrate to precipitate further product. Yield: 0.082 g (0.068 mmol, 68%
[Re(CO)3py(OX)]PF6. A 0.15 g amount of [Re(CO)3Cl(OX)] (0.2
mmol) and 0.051 g of AgPF6 (0.2 mmol) were refluxed in 50 mL of
acetonitrile under N2 in the dark for 18 h. The resulting yellow
solution was filtered through Celite to remove the AgCl precipitate. A
10 mL amount of saturated aqueous NH4PF6 was added to the filtrate.
Acetonitrile was then removed under reduced pressure. The yellow
solid (0.16 g) was filtered and washed with water. The solid, assumed
to be [Re(CO)3(CH3CN)OX]PF6, was reacted to form the pyridine
adduct without characterization. A 0.16 g amount of [Re-
(CO)3(CH3CN)OX]PF6 (0.18 mmol) and 0.445 g of pyridine (5.63
mmol) were refluxed in 50 mL of methanol for 4 h. The resulting
solution was reduced in volume to 25 mL under reduced pressure and
refrigerated overnight. The resulting yellow solid was filtered and
recrystallized from methanol. Yield: 0.14 g (0.149 mmol, 75% overall).
1H NMR ((CD3)2CO, 400 MHz): δ 7.5 (m, 2H, py), 7.7 (m, 6H, bpy
1
overall). H NMR ((CD3)2CO, 400 MHz): δ 7.01 (d, 4H, J = 8 Hz,
phenyl), 7.16 (m, 12H, phenyl), 7.30 (d, 2H, J = 16 Hz, ethenyl), 7.37
(m, 8H, phenyl), 7.52 (m, 2H, py), 7.6 (d, 4H, J = 8 Hz, phenyl), 7.87
(d, 2H, J = 16 Hz, ethenyl), 8.03 (m, 3H, bpy and py overlaid), 8.60
(m, 2H), 8.87 (s, 2H, bpy), 9.28 (d, 2H, J = 6 Hz, bpy). Anal. Calcd
1
and py overlapping), 8.03 (t, H, py), 8.28 (dd, 4H, J1 = 7 Hz, J2 = 1
B
dx.doi.org/10.1021/ic3018387 | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
for C58H43N5O3RePF6: C, 58.58; H, 3.65; N, 5.89. Found: C, 58.25;
H, 3.48; N, 5.93. ESI POS: m/z 1044.15 {(Re(CO)3py(DPA))+}.
[Cu(PPh3)2(DPA)]BF4 MeOH. Dry diethyl ether (20 mL) was purged
with nitrogen for 10 min. [Cu(PPh3)4]BF4 (0.231 g, 0.215 mmol) and
DPA (0.150 g, 0.215 mmol) were added, and the suspension was
stirred under N2 for 48 h. The orange precipitate was filtered and
washed with diethyl ether and recrystallized from a minimum amount
least-squares procedures. All non-hydrogen atoms were refined with
anisotropic displacement parameters. Hydrogen atoms were included
in calculated positions with isotropic displacement parameters 1.2
times the isotropic equivalent of their carrier carbon atoms. Included
in the structure of [Cu(PPh3)2(OX)]BF4 are one molecule of CHCl3
−
solvent and a single BF4 counteranion per cation. The completeness
of the [Cu(PPh3)2(OX)]BF4 structure is 94% due to data collection in
a monoclinic unit cell rather than the correct symmetry, triclinic.
However, the data are of high quality. FT-IR spectra were acquired in
solid-state KBr disks as 64 coadded scans using a Perkin-Elmer BX FT-
IR spectrometer. FT-Raman data were acquired using a Bruker
Equinox 55 interferometer coupled with a FRA-106 Raman module
and a D418T liquid-nitrogen-cooled germanium detector, controlled
by the Bruker OPUS v5.5 software package. Spectra were acquired as
1024−2048 coadded scans using 1064 nm radiation of a Nd:YAG laser
operating at 450 mW. The resolution for both FT-IR and FT-Raman
spectra is 4 cm−1. For absorption, emission, and lifetime measure-
ments, ca. 10−5 mol L−1 CH2Cl2 solutions were used. These were
argon purged for 20 min before spectral acquisition to remove any
dissolved gases. UV−vis absorption spectra were acquired with a Jasco
V-550 spectrophotometer. Steady-state emission spectra were acquired
with a Perkin-Elmer LS-50 B luminescence spectrometer using an
excitation wavelength of 350 nm. Excited-state emission and
absorption transients were acquired using a LP920K transient
absorption (TA) system (Edinburgh Instruments). Excitation was
carried out using pulsed third-harmonic radiation (355 nm) from a
Brilliant (Quantel) Nd:YAG laser operating at 5 Hz, and in TA mode a
Xe900 450W xenon arc lamp was used as the probe source. Photons
were dispersed using a TMS300-A monochromator with a 1800
grooves/mm grating, recorded on a R928 (Hamamatsu) photo-
multiplier and transcribed on a TDS3012C (Tektronix) oscilloscope.25
Fluorescence correction was applied for the transient absorption
spectra as part of the optical density (ΔOD) calculation. An
exponential decay curve was fitted to the output to obtain the
lifetime. Samples were in CH2Cl2 solution and thoroughly degassed
using argon prior to the experiment. Temperature was controlled using
a TLC 50/E (Quantum Northwest Inc.) sample holder. Resonance
Raman measurements were performed using the setup described in
detail by Howell et al.26,27 A continuous-wave Innova I-302 krypton-
ion laser (Coherent, Inc.) was used for excitation wavelengths λex of
350.7, 406.7, and 413.1 nm, a diode-pumped laser (CrystaLaser) for
λex = 444.3 nm, and a Melles Griot OmNichrome model 543-MAP
argon laser at λex = 488.0 and 514.5 nm. The sample and collection
lens were arranged in a 135° backscattering geometry, and the Raman
photons were focused on the entrance slit of an Acton Research
SpectraPro500i spectrograph with a 1200 grooves mm−1 grating.
Signal was recorded using a liquid-nitrogen-cooled Spec-10:100B
CCD controlled by WinSpec/32 software (Roper Scientific), operating
at −100 °C. A similar system but using a 300 grooves mm−1 grating
and a temperature-controlled Flash 300 (Quantum Northwest, Inc.)
sample holder was used to collect temperature-dependent emission
data.
1
of methanol. Yield: 0.194 g (0.142 mmol, 66%). H NMR (CDCl3,
400 MHz): δ 7.04−7.17 (m, 30H), 7.19−7.24 (m, 10H), 7.28−7.37
(m, 16H), 7.57−7.64 (m, 8H), 8.11 (d, 2H, J = 5 Hz), 8.73 (s, 2H).
Anal. Calcd for C86H68N4P2CuBF4: C, 74.54; H, 5.18; N, 4.00. Found:
C, 74.42; H, 4.96; N, 3.99. ESI POS: m/z 1019.32 {(Cu(PPh3)-
(DPA)+)}.
Measurements. Crystal structures of [Cu(PPh3)2(OX)]+ and
[Re(CO)3Cl(DPA)] were acquired with a Bruker diffractometer
equipped with an Apex II charge-coupled device (CCD) area detector
using graphite-monochromated Mo Kα (λ = 0.71073 Å) radiation.
Data collection parameters are listed in Table 1. Structures were solved
by direct methods and refined on F2 by use of all data in full-matrix
Table 1. Crystal Data and Structure Refinement for
[Cu(PPh3)2(OX)]BF4 and [Re(CO)3Cl(DPA)]
[Cu(PPh3)2(OX)]BF4
C63H47BCl13CuF4N6O2P2
1238.71
[Re(CO)3Cl(DPA)]
C53.17H38.33Cl1.33N4O3Re
1014.68
empirical formula
fw
temp.
89(2) K
89(2) K
wavelength
cryst syst
0.71069 Å
0.71073 Å
triclinic
trigonal
space group
unit cell dimens
P1
R3
̅
̅
a = 11.6160(10) Å
b = 14.3910(14) Å
c = 19.589(3) Å
α = 103.990(5)°
β = 96.298(5)°
γ = 108.619(3)°
2948.3(6) Å3
2
a = 40.2046(13) Å
b = 40.2046(13) Å
c = 14.0107(6) Å
α = 90°
β = 90°
γ = 120°
19 612.9(12) Å3
volume
Z
18
density (calcd)
abs coeff
F(000)
1.395 Mg/m3
0.623 mm−1
1.546 Mg/m3
2.920 mm−1
9216
0.40 × 0.20 × 0.20 mm3
2.12−25.35°
1268
cryst size
0.38 × 0.24 × 0.15 mm3
1.09−25.50°
θ range for data
collection
index ranges
−11 ≤ h ≤ 13
−17 ≤ k ≤ 16
−23 ≤ l ≤ 22
62 816
−48 ≤ h ≤ 48
−48 ≤ k ≤ 48
−16 ≤ l ≤ 16
107 101
no. of reflns
collected
Computational Modeling. Frequency and time-dependent (TD)
calculations were performed on the optimized ground-state structures.
All calculations were performed using the B3LYP method, except the
TD calculations, which were additionally performed using CAM-
B3LYP. The latter were carried out using the Gaussian 09 package
(Gaussian, Inc.), while all B3LYP calculations were carried out using
Gaussian 03 (Gaussian, Inc.). All calculations used the 6-31G(d) basis
set for the ligands, while heavy metal centers were approximated using
a LANL2DZ effective core potential.25,28,29 TD calculations were
performed under CH2Cl2 solvent fields implemented with the IEF
polarizable continuum (PCM) model, while other calculations were
performed without a solvent model. The effective peak center for
calculated electronic transitions was determined as an average
weighted by constituent oscillator strengths. For [Cu(PPh3)2(DPA)]+
the PPh3 groups were approximated with PH3 in order to achieve a
reasonable computational time. The output frequencies were scaled to
minimize the mean actual deviations (MADs) between calculated and
experimental peaks by a factor of approximately 0.975.30 IR and
no. of independent 9936 [R(int) = 0.0449]
reflns
7974 [R(int) = 0.0963]
(25.35°) 99.9%
semiempirical from
equivalents
completeness to θ (25.00°) 93.50%
abs corr
semiempirical from
equivalents
max and min
transmission
1.000000 and 0.833427
0.5928 and 0.3880
refinement method full-matrix least-squares on full-matrix least-squares
F2
on F2
data/restraints/
params
9936/0/739
7974/92/599
goodness-of-fit on
1.053
1.071
F2
final R indices [I > R1 = 0.0432, wR2 = 0.1037 R1 = 0.0885, wR2 = 0.1778
2σ(I)]
R indices (all data) R1 = 0.0593, wR2 = 0.1169 R1 = 0.1256, wR2 = 0.1998
largest diff. peak
and hole
0.729 and −0.544 e·Å−3
3.641 and −2.119 e·Å−3
C
dx.doi.org/10.1021/ic3018387 | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Raman spectra were generated using a Lorentzian line shape with a
half-width at half-maximum (HWHM) of 3 cm−1. Vibrational modes
and molecular orbitals (MOs) were visualized using Molden31 and
GaussView v4.1 (Gaussian, Inc.), respectively, after extraction from the
output files using GaussSum.32
carbonyl equatorial plane. The N1−Re1−Cl1 angle of 81.78° is
also consistent with a distorted octahedral geometry. This is not
uncommon for complexes of the type [Re(CO)3(L)Cl], where
L = polypyridyl. A search of the Cambridge structural database
of fac-[Re(CO)3Cl(bpy)]-based complexes gives 33 results
with a mean N−Re−Cl bond angle of 84.463°. Similarly, the
same analysis carried out on 16 fac-[Re(CO)3Cl(phen)]-based
(phen =1,10-phenanthroline) complexes gives a mean N−Re−
Cl angle of 84.413°. This is partly due to the size of the
rhenium ion which forces it to sit out of the plane of the bpy or
phen ligands.
RESULTS AND DISCUSSION
■
Crystal Structures. Crystals of [Re(CO)3Cl(DPA)] of
suitable quality for X-ray crystallography were grown from
vapor diffusion of diisopropyl ether into a dichloromethane
solution of the complex. The structure is shown in Figure 2
The crystal structure displays a trans,trans arrangement of
the ethylene groups. This is consistent with 1H NMR
spectroscopy, where J = 16 Hz is observed for the ethylene
protons (vide supra) and is consistent with the lowest energy
geometry calculated by DFT. Two rotamers are possible for the
trans ethylene groups with respect to their orientation to the
bpy moiety. Rotation of the C3−C6−C7 angle 180° around the
C3−C6 bond is possible and likewise rotation of the C28−
C31−C32 angle 180° around the C31−C32 bond. The
structure in Figure 3 shows an “inward−inward” (rotamer A)
orientation of the ethylene groups, i.e., C7 and C32 are eclipsed
with C4 and C29, respectively. Other possibilities are that both
ethylene groups are pointing “outward” (rotamer C) or one
ethylene group is pointing in and the other out (rotamer B).
Structures of the related complexes, {4,4′-bis[p-diethylamino)-
α-styryl]-2,2′-bipyridine}tricarbonylchlororhenium(I)33 and
(4,4′-distyryl-2,2′-bipyridine)tricarbonyl acetonitrilerhenium(I)
tetrafluoroborate34 both show a “inward−outward” orientation
of the ethylene groups (rotamer B). It should be noted that
rotamer C has been calculated as the lowest energy form and
thus likely corresponds to the species predominantly present in
solution.
Figure 2. POVRAY perspective view of [Re(CO)3Cl(DPA)].
Disordered solvent molecules and hydrogen atoms are omitted for
clarity.
with the solvent and counterion removed for clarity, and
selected bond lengths and angles are presented in Table 2. The
Table 2. Bond Lengths (Angstroms) and Angles (degrees)
for the Crystal Structures Described
[Re(CO)3Cl(DPA)]
The outer phenyl ring (C45−C50) is disordered over two
sites. Atoms labeled C45−C50 in Figure 2 have 60%
occupancy. Rotation of the phenyl ring around the N3−C45
bond of 29.88° results in the remaining 40% occupancy (these
atoms are not shown in Figure 2). C37 and C38 of the styryl
phenyl ring are also disordered over two sites with a 60/40
occupancy. The two halves of the DPA ligand are twisted
slightly from planarity. Relevant distortion angles between the
pyridyl ring and the phenyl rings range from 170.1° to 178.43°,
indicating slight nonplanarity, enough to retain conjugation
through the bpy ligand. Within the crystal, molecules of
[Re(CO)3Cl(DPA)] form dimers; a depiction of this can be
found in Figure S1 (Supporting Information). The shortest of
the intermolecular H−Cl contacts is 2.726°. A similar dimer
formation is observed in the related complex {4,4′-bis[p-
d i e t h y l a m i n o ) - α - s t y r y l ] - 2 , 2 ′ - b i p y r i d i n e } -
tricarbonylchlororhenium(I).33 This intermolecular interaction
may be facilitating formation of rotamer A in the crystal.
Crystals of [Cu(PPh3)2(OX)]BF4 were grown from vapor
diffusion of diethyl ether into a chloroform solution. A diagram
of the complex cation is shown in Figure 4, and selected bond
lengths and angles are presented in Table 2. Crystallization
occurred with a noncoordinating tetrafluoroborate anion and a
Re1−C62
Re1−C60
Re1−N1
N2−Re1−N1
C62−Re1−C60
1.8562(7)
1.9122(8)
2.178(9)
75.4(4)
Re1−C61
Re1−N2
Re1−Cl1
C62−Re1−C61
C61−Re1−C60
1.9151(4)
2.1311(3)
2.480(3)
85.3(6)
92.2(6)
88.5(6)
[Cu(PPh3)2(OX)]+
Cu1−N1
Cu1−N2
N1−Cu1−N2
N1−Cu1−P1
N2−Cu1−P1
2.063(2)
2.071(2)
Cu1−P1
Cu1−P2
N1−Cu1−P2
N2−Cu1−P2
P1−Cu1−P2
2.2295(8)
2.2520(8)
101.74(6)
107.51(6)
129.00(3)
80.16(8)
116.62(6)
110.94(6)
a
For four-coordinate complexes, τ4 can be used to establish a
quantitative measure of how close a geometry is to tetrahedral (τ4 =
1), trigonal pyramid (τ4 = 0:85), square planar (τ4 = 0). τ4 = {360° −
(Θ + θ)/141° (Θ and θ are the two largest bond angles of the
tetrahedron).
complex crystallized in the trigonal space group R3. Eighteen
̅
complex molecules comprise the unit cell, as well as three
highly disordered dichloromethane solvent molecules, which
are located on symmetry sites. This was modeled with each
dichloromethane molecule disordered over six sites. The
SQUEEZE function identifies a 40-electron hole, consistent
with this. The {Re(CO)3Cl} fragment displays bond lengths
and angles consistent with a distorted octahedral geometry,
with the N1−Re1−N2 angle of 75.4°. A mean plane through
the atoms of the bpy moiety and a mean plane through the Re
and the equatorial carbonyl carbons (C61 and C62) lie at an
angle of 16.14°. This shows the bpy ligand bent out of the Re−
chloroform solvent molecule in the triclinic space group P1. As
̅
found for many [Cu(I)(PPh3)2(NN)]+ complexes (NN =
diimine ligand), the Cu(I) metal center displays a distorted
tetrahedral geometry and the P−Cu−P angle of 129° and the
N−P−N angle of 80.16° exemplify this distortion. A
substructure search of the Cambridge structural database of
[Cu(I)(PPh3)2(bpy)]+ compounds result in 16 complexes, with
D
dx.doi.org/10.1021/ic3018387 | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Figure 3. Possible rotamers of [Re(CO)3Cl(DPA)].
consistently higher than those of [Re(CO)3Cl(DPA)]. This is
due to decreased back-bonding from the ligand π*L orbitals to
the CO antibonding orbitals via the metal dπ system and
thus consistent with the electron-withdrawing properties of the
oxadiazole unit. This trend is well represented in computational
spectra; however, the agreement between calculated and
observed peak positions is less than optimal (vide infra). The
wavenumber of the high-frequency CO stretch of [Re-
(CO)3Cl(bpy)] lies between that of the OX and DPA
compounds; however, the low-frequency CO stretches are
at significantly lower wavenumbers. Other significant features in
the FT-IR spectra are due to phenyl- or bipyridine-related
vibrations.
FT-Raman spectra of the OX- and DPA-containing
complexes are shown in Figure 5. For reference, spectra of
[Re(CO)3Cl(bpy)] and [Cu(PPh3)2(bpy)]+ are shown in
Figure S2 (Supporting Information). All complexes feature
spectra that are dominated by delocalized vibrations,
corresponding to aromatic ring deformations. The vibrational
modes of the OX- and DPA-substituted compounds can be
identified both from comparison to the bipyridine complexes
and calculated vibrational spectra. Marked peaks correspond to
modes that are significant in resonance Raman spectra and will
be further discussed below.
IR and Raman vibrational stretching frequencies and
intensities can be calculated with reasonable accuracy for
transition-metal complexes using DFT at the 6-31G(d) level for
ligands and LANL2DZ level for heavy atoms.28 As IR and
Raman spectroscopic techniques are concerned with the
electronic dipole and polarizability, respectively, good spectral
correlation gives confidence to the calculated values of these
and related quantities, such as structure, vibrational modes, and
energetics.25,30 Mean absolute deviations (MADs) were
determined as they provide a way to quantitatively monitor
the accuracy of the peak wavenumbers in a calculated
spectrum.36 These data were calculated by averaging the
absolute difference in wavenumbers of the most prominent
experimental peaks to the corresponding calculated peaks in the
spectral range from 400 to 1800 cm−1 and are shown in Table
S1 (Supporting Information). Each spectrum was scaled to
minimize the MADs. The average of the MAD values obtained
is <7 cm−1, and the assignment of vibrational peaks is definitely
possible from the calculated spectra. Carbonyl stretching peaks
Figure 4. POVRAY perspective view of the [Cu(PPh3)2(OX)]+ cation.
BF4 counteranion and hydrogen atoms are omitted for clarity.
−
a mean P−Cu−P angle of 120.7° and a mean N−Cu−N angle
of 78.4°. The geometry index τ4* = 0.81, consistent with a
trigonally distorted tetrahedron.35 The OX ligand can be
treated as two halves, each half consisting of a pyridyl,
oxadiazole, and phenyl ring (py−Ox−Ph). Each half of the
ligand displays only small deviations from planarity. The planes
between the py and the Ox sections are 7.3° and 5.9° to one
another for the sides of the ligand bonded to the Cu through
N1 and N2, respectively. The pyridyl rings of the OX ligand are
twisted by 10.67° to each other.
Ground-State Vibrational Spectra. The CO stretching
frequencies (Table 3) of [Re(CO)3Cl(OX)] are subtly but
Table 3. Carbonyl Stretching Frequencies of the Rhenium
a
Complexes
ν
̃
(CO)/cm−1
symmetry
a′(2)
a″
a′(1)
[Re(CO)3Cl(bpy)]
[Re(CO)3Cl(OX)]
[Re(CO)3Cl(DPA)]
1883
1923
1919
1950
1938
1922
2019
2026
2016
2040
2034
2030
1902
1897
b,c
[Re(CO)3py(bpy)]+
[Re(CO)3py(OX)]+
[Re(CO)3py(DPA)]+
a
b
c
Symmetry assignments are from ref 11. From ref 12. In CH2Cl2.
E
dx.doi.org/10.1021/ic3018387 | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Figure 5. FT-Raman spectra acquired in a KBr disk. (a) Complexes containing the OX ligand; (b) complexes containing the DPA ligand.
Figure 6. Electronic absorption spectra acquired in CH2Cl2. (a) Complexes containing the OX ligand; (b) complexes containing the DPA ligand.
were not included in the calculations, as they require different
scale factors;37 however, as they are spectrally isolated from
other features, their vibrational modes are straightforward to
assign nevertheless.
an increase in the energies of the ligand πL orbitals, causing
them to become the new HOMOs, as shown in Figure 7. The
πL* energy appears unchanged with respect to [Re(CO)3Cl-
(bpy)]. This gives rise to an electron shift from diphenylaniline
to bpy that is best described as a ligand-centered charge transfer
(LCCT). As visible from the MO pictures in Table 5, this
transition also contains significant π → π* character due to
overlap of the donor and acceptor orbitals. Two sets of
transitions are calculated for this peak in B3LYP (see Table 4),
which differ in whether the transitions terminate on the LUMO
or LUMO+1 orbitals. For the former, the transition energy is
significantly underestimated, which is not unexpected for
charge transfer transitions calculated using this functional.38
Using CAM-B3LYP as the functional leads to a single set of
transitions but at a slightly overestimated energy. This method
has been designed to compensate for the shortcomings in the
long-range modeling of charge transfer transitions of
B3LYP.38,39 The higher energy π → π* transitions on the
other hand are better approximated with the B3LYP functional.
Compared to [Re(CO)3Cl(OX)], the lowest energy
absorption peak of [Re(CO)3py(OX)]+ is blue shifted by ca.
40 nm while [Re(CO)3py(DPA)]+ shows a red shift of ca. 20
nm. These effects can both be explained by the relative
electron-withdrawing influence of the py ligand compared to
Cl−. The most significant effect of this is a stabilization of the
metal dπ set of orbitals, which is of significantly greater extent
than that of bpy-based πL and πL* orbitals, leading to an increase
in the HOMO−LUMO gap of the material (see Figure 7).
Electronic Absorption Spectroscopy. Electronic absorp-
tion spectra of the OX- and DPA-based complexes in CH2Cl2
are shown in Figure 6a and 6b, respectively, and the observed
peaks, along with the electronic transitions calculated using the
B3LYP and CAM-B3LYP functionals, are summarized in Table
4. The energies and shapes of the highest occupied and lowest
unoccupied molecular orbitals (HOMOs and LUMOs,
respectively) are shown in Figure 7 and presented in Table
5; additional MOs can be found in Table S2 (Supporting
Information).
The lowest energy transitions of [Re(CO)3Cl(OX)] and
[Re(CO)3Cl(DPA)] are both calculated as HOMO−1/
HOMO → LUMO, where LUMO signifies bpy-based orbitals
in the complexes (see Table 5); however, their origins vary
significantly. The oxadiazole-substituted complex is calculated
to behave much like unsubstituted [Re(CO)3Cl(bpy)] in that
the lowest energy transition is MLCT in nature, followed by π
→ π* transitions at higher energies. These transitions are red
shifted compared to the unsubstituted complex, which can be
attributed to a stabilization of the π*L acceptor orbital, as shown
in Figure 7. Metal-based dπ orbitals appear relatively unaffected
in comparison. The absorption spectrum of [Re(CO)3Cl-
(DPA)] shows a strong band at 460 nm. For this complex the
electron-rich nature of the diphenylaniline substituents effects
F
dx.doi.org/10.1021/ic3018387 | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Table 4. UV−Vis Absorption Data Correlated to TD-DFT-Calculated Electronic Transitions Using the B3LYP and CAM-
a
B3LYP Functionals in a CH2Cl2 Solvent Field
expt.
B3LYP
CAM-B3LYP
transition
a
a
λ/nm (ε/M−1 cm−1
)
λ/nm (f)
transition
λ/nm (f)
a
[Re(CO)3Cl(bpy)]
[Re(CO)3Cl(OX)]
386 (3.7)
405 (0.071)
473 (0.183)
378 (1.054)
321 (0.753)
554 (1.174)
442 (1.293)
351 (0.563)
H → L
H−1, H → L
H−4 to H → L, L+1
H−9, H−7 to H−2 → L, L+1
H−1, H → L
336 (0.099)
376 (0.265)
307 (1.492)
298 (0.600)
399 (3.083)
H−1, H → L
H−1, H → L
H−4, H−3, H−1 → L, L+1
H−4, H → L, L+1
437 (9.0)
338sh (44.2)
324 (54.5)
460 (57.9)
[Re(CO)3Cl(DPA)]
H−3, H−1, H → L, L+1
H−1, H → L+1
H−5 to H−2 → L, L+1
376 (28.3)
303 (49.7)
350 (0.113)
H−3 to H → L, L+1
[Re(CO)3py(bpy)]+
[Re(CO)3py(OX)]+
366 (24.0)
365 (0.103)
416 (0.986)
373 (0.036)
353 (0.737)
321 (0.385)
594 (1.289)
456 (1.543)
394 (0.470)
382 (0.089)
463 (0.208)
368 (0.146)
515 (1.354)
428 (1.348)
373 (0.227)
H−2, H−1 →
H−3, H → L
334 (0.100)
363 (0.144)
344 (0.513)
316 (1.373)
299 (0.441)
429 (3.284)
H−2, H−1 → L
H−2, H → L
H−3, H−1 → L
H−2, H → L, L+1
H−4, H−3, H → L, L+1
H−1, H → L, L+1
b
393 (14.2)
347sh (39.0)
333 (43.5)
306 (36.6)
479 (60.0)
H−4, H−3 → L
H−1, H → L+1
H−7, H−4 to H−1 → L to L+2
H−1, H → L
H−1, H → L+1
H−4, H → L, L+3
H−2 to H → L
H → L
[Re(CO)3py(DPA)]+
389 (32.5)
267 (0.587)
292 (0.129)
333 (0.360)
H−6, H → L, L+2
H−1, H → L
H−4, H → L
[Cu(PPh3)2(bpy)]+
[Cu(PPh3)2(OX)]+
356
c
415 (7.9)
333sh (46.7)
446 (65.7)
H → L+1
[Cu(PPh3)2(DPA)]+
H−1, H → L
H−1, H → L+1
H−1, H → L+2
390 (3.353)
H−1, H → L, L+1
368 (26.9)
294 (0.238)
H−2 to H → L to L+2
a
a
H and L refer to highest occupied molecular orbitals and lowest unoccupied molecular orbitals respectively. The dominant orbital contributions
are shown.
complexes, that is, the ionization energies are expected to be
similar. [Cu(PPh3)2(OX)]+ shows MLCT transitions as a peak
at 425 nm and a shoulder at 333 nm on a strong PPh3
absorption to the blue edge of the window. The ligand-based
charge transfer transition of [Cu(PPh3)2(DPA)]+ is blue shifted
to 446 nm compared to the rhenium complexes; the involved
acceptor orbitals appear in analogous locations to those of the
rhenium complexes. The difference in energy can be attributed
to an increase in the energies of the virtual orbitals, reflecting a
lack of stabilization by the comparatively innocent copper
metal. Again, B3LYP and CAM-B3LYP functionals under- and
overestimate the MLCT transition energies, respectively, while
the higher energy π → π* transitions are adequately described
by B3LYP.
Resonance Raman (RR) Spectroscopy. If Raman spec-
troscopy is carried out with incident light (λex) of similar energy
to an electronic transition of the analyte, resonance enhance-
ment is observed of specific peak intensities, namely, those
corresponding to normal modes that mimic the change in
equilibrium bond lengths (ΔQ) upon electronic excitation.40
This enables the activity in a particular part of the molecule for
a given λex to be determined and thus correlated to an
electronic transition.36 Although the change in intensity means
that direct comparison of calculated spectra to RR spectra is
difficult, comparison is possible via FT-Raman spectra as they
show identical peak positions. Resonance Raman spectra of the
rhenium and copper complexes are shown in Figures 8 and 10,
respectively. In order to obtain a satisfactory signal-to-noise
ratio at long wavelengths, spectra of [Re(CO)3Cl(OX)] and
[Re(CO)3Cl(DPA)] were acquired in DMSO, where higher
concentrations could be achieved compared to CH2Cl2, which
was used for the remaining complexes.
Figure 7. Energies of the frontier MOs of all complexes, as calculated
by B3LYP. Red shades correspond to metal-based occupied MOs
while blue shades correspond to ligand-based occupied MOs. All
virtual orbitals shown are π*L . Complexes with bpy ligands are included
for comparison.
Diphenylamine-based πL and π*L orbitals on the other hand are
too distant to experience significant inductive effects from py,
which, coupled with the energy decrease of bpy π*L orbitals,
decreases the energy of the charge transfer transition taking
place. As with the [Re(CO)3Cl(NN)] complexes, CAM-B3LYP
underestimates transition energies; generally predictions are
somewhat more accurate than those made using B3LYP. Both
methods appear to calculate enhanced mixing of dπ and πL
orbitals, which is most likely related to the aforementioned
decrease in dπ orbital energy. This sometimes makes it difficult
to classify MOs as “ligand based” or “metal based”.
The copper complexes exhibit a number of similar
transitions, albeit at different energies. As shown in Figure 7,
the occupied orbitals have been calculated to be of similar
energies as those of the respective [Re(CO)3Cl(NN)]
G
dx.doi.org/10.1021/ic3018387 | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
a
Table 5. MO Diagrams of the [Re(CO)3Cl(NN)] Complexes
a
The orbitals correspond to calculations carried out in a CH2Cl2 solvent field using B3LYP.
1
At long wavelengths, [Re(CO)3Cl(OX)] shows resonance
enhancement for the peaks at 1028, 1283, 1486, and 1628
cm−1. The peak at 1028 cm−1 has been assigned as
corresponding to a vibration delocalized over the entire ligand,
while the other peaks correspond to vibrations localized on the
bpy portion of the ligand. Most significantly, however, the fully
symmetric carbonyl stretch at 2022 cm−1 is visible in all spectra;
it is usually the most intense peak, with the exception of the λex
= 350.7 nm spectrum. This supports the assignment of the
evidence that the nature of the MLCT excited state is not
significantly perturbed by substitution of the pyridine ligand.
The resonance Raman spectra of [Re(CO)3Cl(DPA)] show
no CO enhancement at any probed wavelength, which
indicates negligible contributions from MLCT transitions.
Spectra acquired at λex = 444.3 nm and longer wavelengths
appear very similar, corresponding to the lowest energy
absorption peak found in this region. Resonance enhancement
for these spectra is observed most prominently for peaks at
1031, 1179, 1192, 1540, and 1586 cm−1. While the peaks at
1031 and 1540 cm−1 are due to vibrations of the bipyridine
moiety, the peaks at 1179 and 1192 cm−1 involve vibrations of
the diphenylaniline and CC moieties. The example
presented in Figure 9 shows activity in similar regions to the
donor MOs, as discussed above (see Table 5). These data are
consistent with an LCCT transition, best described as a transfer
of an electron from the diphenylaniline to the bpy section of
the ligand. The main contribution from the metal lies in
conformational constraints as well as a slight electron-
withdrawing effect, which leads to a red shift of this transition
in the absorption spectrum when compared to [Cu-
(PPh3)2(DPA)]+. At λex = 350.7 nm, resonance enhancement
occurs for the peak at 1612 cm−1, corresponding to a vibration
that is delocalized over the entire ligand (see Figure 9). It
should be noted that this peak is distinct from peaks observed
1
lowest energy electronic absorption peak as MLCT, as such a
transition would cause a significant change in the equilibrium
CO bond length. At λex = 350.7 nm some of these peaks
diminish in intensity, although they are still visible. Peaks at
1548 and 1608 cm−1 have been assigned as delocalized and
oxadiazole based and display an increase in intensity in the
resonance Raman spectra. This is consistent with resonance
with a π → π* transition; the peak is centered at 324 nm in the
absorption spectrum.
Similar spectra are observed for the complex [Re(CO)3py-
(OX)]+ (Figure 8c). Due to strong fluorescence, spectra with
λex > 457.9 nm show a low signal-to-noise ratio and are
therefore omitted. The most obvious difference with respect to
the [Re(CO)3Cl(NN)] complex is a red shift of the CO
stretch to 2040 cm−1, as discussed for the FT-Raman spectra.
The similarity of the spectra to those of [Re(CO)3Cl(OX)] is
H
dx.doi.org/10.1021/ic3018387 | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Figure 8. Resonance Raman spectra at a number of excitation wavelengths: (a) [Re(CO)3Cl(OX)] acquired in DMSO, (b) [Re(CO)3Cl(DPA)]
acquired in DMSO, (c) [Re(CO)3py(OX)]+ acquired in CH2Cl2, (d) [Re(CO)3py(DPA)]+ acquired in CH2Cl2.
Figure 9. Examples of normal modes, with their corresponding calculated wavenumbers, corresponding to bpy-centered, diphenylaniline-centered,
and delocalized vibrations seen in the resonance Raman spectra of [Re(CO)3Cl(DPA)].
at all other wavelengths and thus highly characteristic of the
transition in this wavelength range. The peak at 376 nm in the
absorption spectrum has thus been assigned as corresponding
to a π → π* transition.
457.9 nm spectrum. This is consistent with the lowering in
energy of the bpy-based πL* orbital that acts as an acceptor
during the LCCT transition, effected by the relative electron-
withdrawing effect of the py ligand, as discussed above. Due to
the relative distance of the diphenylaniline-based π*L orbital, it is
expected to be less affected by the stabilizing influence of the py
ligand compared to the bpy-based acceptor orbital, which is
manifested in a decreased LCCT energy. Conversely, the π →
π* transition remains relatively unaffected. This is also evident
from comparison of the MO energies presented in Figure 7.
Spectra of [Re(CO)3py(DPA)]+ are similar to those of
[Re(CO)3Cl(DPA)], showing a similar pattern of peak
enhancements. The peaks characteristic of the LCCT transition
give way to the π → π* marker bands at somewhat longer
wavelengths compared to the [Re(CO)3Cl(NN)] complex,
evidenced by the shoulder at 1613 cm−1 visible in the λex
=
I
dx.doi.org/10.1021/ic3018387 | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Figure 10. Resonance Raman spectra at various excitation wavelengths, acquired in CH2Cl2: (a) [Cu(PPh3)2(OX)]+ and (b) [Cu(PPh3)2(DPA)]+.
Solvent subtraction has been performed. Residuals that appear are marked with asterisks.
a
Table 6. Summary of Emission and Transient Absorption Data
emission
transient absorption
b
c
b
c
solvent
CH2Cl2
λmax /nm
τ /ns
λmax /nm
τ /ns
assignment
[Re(CO)3Cl(OX)]
[Re(CO)3Cl(DPA)]
440
690
<10
17
1(π → π*)bpy
3MLCTbpy
3MLCTbpy
1(π → π*)
380, 520
370, 525
21
13
CH3CN
CH2Cl2
520
<10
418, 555, 670
600
LCCTDPA →
bpy
CH3CN
CH2Cl2
CH3CN
CH2Cl2
514
603
622
510
<10
300
90
1(π → π*)
3MLCTbpy
3MLCTbpy
1(π → π*)
[Re(CO)3py(OX)]+
[Re(CO)3py(DPA)]+
375, 506, 760
370, 500, 740
300
86
<10
430, 640
410, 640
1700
1000
LCCTDPA →
bpy
bpy
CH3CN
510
<10
<10
1(π → π*)
LCCTDPA
PPh3
→
[Cu(PPh3)2OX]+
CH2Cl2
CH2Cl2
460
407
1(π → π*)
3MLCTbpy
3MLCTbpy
1(π → π*)
d
370, 620
16
20
e
<350, 510
[Cu(PPh3)2(DPA)]+
CH2Cl2
CH3CN
495
620
513
<10
<10
<10
LCCTDPA →
bpy
1(π → π*)
370, 490
580, 650
260
LCCTDPA→bpy
a
b
c
d
All measurements are at 293 K in deaerated solvents with excitation at 350.0−354.7 nm. Errors are 20 nm. Errors are 10%. Excess PPh3
e
added. No PPh3 added.
At λex = 514.5 nm, the RR spectrum of [Cu(PPh3)2(OX)]+
appears very similar to the FT-Raman spectrum and indeed the
FT-Raman spectrum of [Cu(PPh3)2(bpy)]+ (see Figure S2,
Supporting Information). Some resonance enhancement is
observed for peaks at 1001, 1285, and 1471 cm−1, which are
assigned as bipyridine vibrations from the frequency calculation.
This is very similar up to excitation energies of 413.1 nm. At λex
= 350.7 nm the peaks at 1013 and 1550 cm−1 are significantly
enhanced, and these are both identified as delocalized
vibrations involving the entire ligand. Identification is some-
what ambiguous for the latter peak, as there is significant
activity in that part of the calculated spectrum; however, most
of this activity is due to various delocalized modes, giving some
confidence to the assignment. Furthermore, a similar peak is
absent in the FT-Raman spectrum of [Cu(PPh3)2(bpy)]+. This
is evidence for a MLCT transition at 437 nm in the absorption
spectrum and a π → π* transition at short wavelengths, similar
to [Re(CO)3Cl(bpy)].
Due to significant fluorescence, the signal-to-noise ratio for
complex [Cu(PPh3)2(DPA)]+ becomes prohibitively small at
λex > 488.0 nm, and therefore, spectra acquired at higher
wavelengths are omitted from analysis. Visible in all spectra but
diminishing for λex = 350.7 nm are vinyl stretching peaks at
1633 cm−1. Also observed at longer wavelengths are peaks at
J
dx.doi.org/10.1021/ic3018387 | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Figure 11. (a) Transient emission and (b) transient absorption spectra of [Re(CO)3Cl(OX)] in deaerated CH2Cl2 at room temperature.
Figure 12. Transient absorption spectra of (a) [Re(CO)3Cl(DPA)] in deaerated CH2Cl2 and (b) [Re(CO)3py(DPA)]+ in deaerated CH3CN at
room temperature.
1179 and 1287 cm−1, which are assigned as diphenylaniline-
based and bpy-based, respectively. This is evidence for an
LCCT transition, as concluded for the rhenium complexes and
predicted by TD-DFT. At λex = 350.7 nm, the spectrum is
dominated by peaks at 1589 and 1609 cm−1.
using different samples, and changes were monitored by
carrying out electronic absorption measurements before and
after pulse laser irradiation. All samples were thoroughly
degassed with dry argon to remove dissolved oxygen, and the
temperature was held constant at 293 K. For the [Re(CO)3Cl-
(NN]) complexes, a prevalent mode of breakdown was thought
to be via loss of Cl−. A potential monitor for this is to carry out
experiments in CH3CN, which is a much more strongly
coordinating solvent than CH2Cl2 and should thus exacerbate
any such effects. Conversely, as py is expected to be much more
strongly coordinating than Cl−, the [Re(CO)3py(NN)]+
complexes represent a step toward increased stability.
In summary, the resonance Raman data show that the lowest
energy transitions of complexes with the OX ligand are MLCT
in nature, similar to the bpy complexes, albeit at lower energies
due to the electron-withdrawing nature of the oxadiazole
groups and increase in conjugation length. The electron-
donating nature of the diphenylaniline group effects com-
pounds with the DPA ligand to undergo ligand-centered charge
transfer as their lowest energy electronic transition. The regions
of activity predicted by calculated MOs match those observed
in the resonance Raman experiments. MLCT transitions could
not be detected for these compounds, as evidenced by the lack
of carbonyl peaks for rhenium DPA complexes; they are
potentially masked by the stronger intraligand transitions. This
is consistent with electronic data and TD-DFT calculations
presented above.
Excited-State Electronic Spectroscopy. Transient ab-
sorption and emission data are summarized in Table 6, and
representative examples of acquired data are shown in Figures
11 and 12.
Due to the high photon flux of the pump laser, photo-
chemical effects such as conformational changes were observed
for some samples. A number of steps were taken to prevent or
elucidate these effects. Acquisitions were repeated several times
As shown in Figure 11, [Re(CO)3Cl(OX)] shows two
emission transients. A short-lived band (τ < 10 ns) at 440 nm is
1
assigned to a radiative decay of the intraligand (π → π*)bpy
state. No corresponding transient absorption peaks were
observed. A lower energy band is found at 690 nm, the
corresponding state of which possesses a lifetime of 17 ns. This
3
has been assigned to the radiative decay of a MLCTbpy state.
However, some mixing with a Cl− → bpy intraligand charge
transfer (ILCT) state is expected.41 In comparison, unsub-
stituted [Re(CO)3Cl(bpy)] shows an emission band at 622 nm
corresponding to a state with a lifetime of 51 ns (in methylene
chloride).12 The observed red shift of this band of ca. 70 nm
represents a decrease in the energy of the thermally equilibrated
excited (THEXI) state, effected by the electron-withdrawing
oxadiazole substituent. Working under the assumption that ΔE
≈ Eem, where ΔE is the energy gap between ground and excited
K
dx.doi.org/10.1021/ic3018387 | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
states and Eem is the emission energy, the decreased emission
lifetime compared to unsubstituted [Re(CO)3Cl(bpy)]1,10,12 is
consistent with the energy gap law.1,12,42 In other words, a
smaller ΔE leads to increased overlap between ground and
excited states and thus an increased rate of nonradiative decay.
It should be pointed out that the Stokes shift of this complex is
actually decreased with respect to [Re(CO)3Cl(bpy)] (8400
compared to 9800 cm−1) (measured as the difference between
the peak of the absorption band to the peak of the emission
band), indicating that ΔQ, the change in equilibrium bond
lengths, is smaller for [Re(CO)3Cl(OX)]. This makes sense
when considering that the excited electron would be
delocalized over a larger and more electrophilic ligand
compared to the bpy complex. Emission decay from the
lower energy state appears at a time slightly delayed with
respect to the high-energy band, indicating some (π → π*)bpy
addition to these peaks, a broad low-energy peak is also
observed at 760 nm (∼13 000 cm−1). Ground-state calculations
suggest an energy difference between the LUMO and the
lowest unoccupied py orbital of ∼12 000 cm−1. Assuming the
MO energies show little relative variation during excitation, an
assignment that is energetically reasonable would be a
3MLCTbpy → 3MLCTpy or 3MLCTbpy → (π → π*)py transition,
that is, a transition involving an electron shift from bpy to py.
Such processes have been shown to occur spontaneously for
[Re(CO)3L(bpy)]+, where L is 1,2-bis(4-pyridyl)ethylene or
trans-4-styrylpyridine, using time-resolved infrared (TRIR)
measurements.46,47 Due to the increased reduction potential
of py compared to these ligands, this transition would be
optically driven in [Re(CO)3py(OX)]+ instead. An electron
transfer from the Cl− ligand to the Re metal is unlikely, as it has
been observed for [Re(CO)3Cl(bpy)] at 510 nm,43 and given
the relative electron withdrawing nature of py, this transition is
not expected to shift to lower energies. As with [Re(CO)3Cl-
(OX)], a decreased transient lifetime is observed in CH3CN.
[Cu(PPh3)2(OX)]+ shows very low signal-to-noise ratios for
both transient absorption and emission measurements.
However, it has been possible to observe some transient states.
A short-lived and weak emission peak was seen at 460 nm.
Using the high-sensitivity resonance Raman detector a variable-
temperature emission experiment was carried out. At 253 K, a
peak at 407 nm was observed, which decreased in intensity as
the temperature was increased, diminishing sufficiently at 305 K
to show a prominent shoulder at 470 nm. No appreciable shift
in the peak positions was observed. Addition of excess PPh3
significantly enhanced the intensity of the low-energy peak.
Luminescence of PPh3 (or impurities therein) has previously
been observed at 450 nm48 (in ethanol), and we therefore
assign the short-lived peak at 470 nm to this species. The peak
→
3MLCTbpy conversion. Transient absorption peaks of
[Re(CO)3Cl(OX)] are observed at 380 and 520 nm and
show similar lifetimes to the emissive 3MLCT state. Both bands
are assigned as π → π* transitions of the radical anion of bpy,
analogous to [Re(CO)3Cl(bpy)]43 and consistent with
absorption peaks observed at 380 and 440 nm in chemically
reduced bpy.44 The lower energy band appears affected by the
oxadiazole substituent, as indicated by a red shift compared to
[Re(CO)3Cl(bpy)]. In CH3CN, the emission is quenched and
the lifetime of the transient absorption peaks is reduced to close
to the duration of the laser pulse; however, the observed peak
positions remain the same.
Changing the Cl− ligand to py changes the excited-state
kinetics but not the nature of the lowest energy excited state
3
accessed, which is again assigned as MLCTbpy. In CH2Cl2 an
emission peak is observed, centered at 603 nm, with a lifetime
of 300 ns. Blue-shifted emission compared to [Re(CO)3Cl-
(OX)] is consistent with an increased energy gap between the
dπ and the πL orbitals, as discussed above (see Figure 7). The
Stokes shift increases slightly, to 8700 cm−1. The increase in
excited-state lifetime is comparable to that observed for
[Re(CO)3Cl(bpy)] compared to [Re(CO)3py(bpy)]+ (51 ns
compared to 669 ns in methylene chloride)12 and is consistent
with the change in ΔE. The mechanism for the large increase in
lifetime has also been rationalized by decreased ILCT mixing
expected for this state, due to an increased energy gap between
dπ and πL orbitals compared to the [Re(CO)3Cl(NN)]
complex.41,45 Unlike [Re(CO)3Cl(OX)], [Re(CO)3py(OX)]+
shows no obvious high-energy (π → π*)bpy emission. However,
it should be noted that the initial phase of the observed
emission is best modeled by including a second transient with a
lifetime of 26 ns (cf. the low-energy band in Figure 11a), which
is possibly due to such a state. An observed blue shift in
emission is consistent with a lowering in energy of the metal dπ
orbitals with substitution of the more electronegative py ligand.
This shift is on the order of 2100 cm−1, compared with 2500
cm−1 for the lowest energy absorption peaks of these
complexes, indicating similar levels of communication of the
metal with the FC and THEXI states of the complex. In
CH3CN the transient lifetime is decreased to 90 ns and the
emission red shifted to 622 nm; this solvatochromic shift is a
consequence of increased stabilization of the THEXI state by
the higher polarity of the solvent. Again, the relationship
between ΔE and τ is qualitatively consistent with the energy
gap law. The transient absorption peaks at 375 and 506 nm,
which are comparable to those of [Re(CO)3Cl(OX)], are
assigned as π → π* transitions of the radical anion of bpy. In
1
at 407 nm is tentatively assigned as radiative decay of a (π →
π*)bpy state, analogous to the high-energy decay observed for
[Re(CO)3Cl(OX)]. Weak transient absorption bands at 370
and 620 nm are observed (τ ≈ 16 ns) when excess PPh3 is
added to the solution and at <350 and 510 nm (τ ≈ 20 ns) with
no PPh3 present. The compound was found to break down
immediately when dissolved in CH3CN, as evidenced by
complete discoloration and changes in the absorption
spectrum. No attempt to acquire excited-state data was
therefore made in this solvent.
[Re(CO)3Cl(DPA)] shows an emission peak, with a lifetime
shorter than the laser pulse (∼7 ns), centered at 520 nm. As
mentioned below, emission from the free ligand occurs at 495
nm (in CH2Cl2); thus, this is consistent with prompt
1
fluorescence from a (π → π*) state. Transient absorption
data show a number of peaks belonging to a dark state, with a
comparatively long lifetime of 600 ns. Two peaks are visible, at
ca. 555 and 670 nm, and part of a broad transient that stretches
to the red edge of the window (Figure 12a). Unlike the OX-
based complexes, depletions of the ground state, corresponding
to the presence of absorption peaks in the ground-state
absorption spectrum, are also observed with matching lifetimes.
The relatively long excited-state lifetime is inconsistent with
3MLCT states of complexes of the type Re(CO)3Cl(L) (where
L = polypyridyl). Furthermore, there is no evidence for a strong
absorption at ca. 380 nm that has been used as evidence for the
presence of the radical anion of bpy.49,50 The nature of the
3
THEXI state can therefore be concluded to be either (π →
π*) or LCCT. MO pictures from triplet-state calculations (see
Figure 13 for an example involving [Re(CO)3Cl(DPA)]+)
L
dx.doi.org/10.1021/ic3018387 | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Figure 13. MO pictures of the highest occupied, partially occupied, and lowest unoccupied molecular orbitals of the calculated triplet state (using
B3LYP) of [Re(CO)3py(DPA)]+.
show singly occupied molecular orbitals (SOMOs) that are
diphenylaniline and bpy based, which is consistent with a
THEXI state that is LCCT in nature. One should note that the
LCCT state is not charge separated; both the donor and the
emission at 620 nm can still be observed if the solution is
excited at 450 nm. It is known that the ligand absorbs at 395
nm and emits strongly at 475 nm (on a glass substrate),22
which is consistent with the observed high-energy emission
peak. Thus, if addition of excess PPh3 causes the following
equilibrium24 to shift to the right
3
acceptor orbitals have overlap and thus appreciable (π → π*)
character. A number of good examples of charge-separated
states have been studied, and they are characterized with short-
lived excited states which show electron tunneling through
spacer groups51,52 which are not present in [Re(CO)3Cl-
(DPA)]+
[Cu(DPA)(PPh3)2]+ + 2PPh3 ⇌ [Cu(PPh3)4]+ + DPA
it is likely that the observed high-energy emission stems from
free ligand in solution. Any remaining ligand bound to copper
will absorb 450 nm light and emit at 620 nm, and thus, the low-
energy emission band is assigned as LCCT based. A solution of
[Cu(PPh3)2(DPA)]+ in CH3CN is highly emissive, with a peak
at 513 nm. This is probably due to emission of free ligand in
CH3CN; use of a coordinating solvent causes more ligand to be
liberated. No transient absorption signals could be obtained
from [Cu(PPh3)2(DPA)]+ in CH2Cl2. However, ground-state
depletion is observed at 430 nm when the experiment is carried
out in CH3CN, along with grow-in of a broad band with two
peaks at ca. 580 and 650 nm.
In summary, the OX complexes possess a THEXI state that is
3MLCTbpy in nature. This state is relatively short lived for
[Re(CO)3Cl(OX)] and [Cu(PPh3)2(OX)]+ and, where ob-
servable, decreases in lifetime in CH3CN, a higher polarity
solvent. [Re(CO)3Cl(OX)] shows a lifetime of 300 ns, which
decreases to ca. 90 ns in CH3CN. Complexes based on the
DPA ligand show dark excited states that are considerably
longer lived compared to either unsubstituted or oxadiazole-
substituted [Re(CO)3Cl(bpy)]. Due to the important role of
the {Re(CO)3Cl} unit in nonradiative excited-state deactiva-
tion, this suggests an excited state that is mostly ligand based,
an assignment that is supported by triplet-state DFT
calculations.
While the SOMOs of the Cl− and py complexes appear
almost identical, the HOMO of the former is metal based (not
shown), which is consistent with a smaller energy gap between
the dπ and the πL orbitals. As can be seen from Figure 12a, the
transients of the peaks at 418 and 555 nm persist for longer
than depletion of the ground state. This is not inconsistent with
a cis/trans isomerization of one of the CC groups. While
none of the cis product has been isolated, an electronic
transition of this species has been calculated using TD-DFT at
ca. 540 nm, which points to a retained ΔOD in this spectral
region. It appears that the trans isomer reforms over a period of
minutes as no permanent photoproduct can be observed. In
CH3CN this complex undergoes photochemical reactions upon
excitation, thus precluding analysis of the transient absorption
data. However, short-lived emission has been observed at 514
nm, which will be further discussed in reference to the copper
complex below.
The py complexes are stable in both CH2Cl2 and CH3CN; an
example of transient absorption spectra in the latter solvent is
presented in Figure 12b. The low-energy band is somewhat
broader than for [Re(CO)3Cl(DPA)], and individual peaks can
no longer be assigned. However, the spectra bear many
similarities otherwise. Most importantly, the excited-state
lifetime of the dark state increases to 1700 and 1000 ns in
CH2Cl2 and CH3CN, respectively. This is most likely an effect
of communication with a 3MLCT state of similar energy in the
[Re(CO)3Cl(NN)] complex,45 facilitating deactivation via the
a′(1) vibrational mode of the {Re(CO)3Cl} moiety. This mode
has been identified as a major energy acceptor in the
nonradiative deactivation of [Re(CO)3Cl(bpy)] complexes.12,53
The level of communication and associated excited-state
deactivation would be expected to be decreased for [Re-
(CO)3py(DPA)]+ due to the aforementioned increase in
separation between the dπ and the πL orbitals. The calculated
separation between the highest energy occupied dπ orbital and
the singly occupied πL orbital in the triplet state is ca. 13 300
and 7900 cm−1 for [Re(CO)3py(DPA)] and [Re(CO)3Cl-
(DPA)], respectively. MO pictures of the calculated triplet state
clearly show the LCCT nature of this state (Figure 13).
Two short-lived emission peaks are observed for [Cu-
(PPh3)2(DPA)]+ in CH2Cl2 at 495 and 620 nm. The former
dominates when excess PPh3 is added; however, strong
CONCLUSION
■
We synthesized rhenium and copper complexes of ligands bpy,
OX, and DPA and investigated their photophysical properties.
The crystal structure of [Re(CO)3Cl(DPA)] confirms the trans
arrangement of the ethylene bonds. The [Cu(PPh3)2(OX)]+
structure shows that the oxadiazole ligand adopts a near planar
arrangement. Electronic absorption and resonance Raman
suggest that the oxadiazole-substituted complexes undergo
MLCT excitation at slightly longer wavelengths than the
unsubstituted complexes, which is consistent with previous
observations. Conversely, the DPA complexes show a strong
absorption peak in the visible region, which has been assigned,
using resonance Raman to a LCCT transition involving the
diphenylaniline and bpy as donor and acceptor units,
respectively. A lack of carbonyl stretches suggests this transition
is entirely located on the DPA ligand. Relevant calculated MOs
are consistent with these data. Excited states of these complexes
M
dx.doi.org/10.1021/ic3018387 | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
(20) Myahkostupov, M.; Castellano, F. N. Inorg. Chem. 2011, 50,
9714−9727.
were probed using transient absorption and emission
1
spectroscopic techniques. Short-lived (π → π*) transitions
(21) Fraser, M. G.; Clark, C. A.; Horvath, R.; Lind, S. J.; Blackman, A.
G.; Sun, X.-Z.; George, M. W.; Gordon, K. C. Inorg. Chem. 2011, 50,
6093−6106.
were found for a number of complexes. The THEXI states of
3
the OX complexes have been assigned as MLCT, while the
DPA compounds showed relatively long-lived dark states. For
the latter, attachment of a metal has a detrimental effect on the
luminescence. Consistent with triplet-state DFT calculations,
the dark states have been assigned as LCCT in nature; their
lifetimes are consistent with excited states that experience
relatively little influence from the attatched metals, which serve
essentially as conformational constraints.
(22) Kay, K. Y.; Cho, S. Y.; Park, H. C.; Park, J. W. Synth. Met. 2003,
137, 1045−1046.
(23) Lundin, N. J.; Walsh, P. J.; Howell, S. L.; McGarvey, J. J.;
Blackman, A. G.; Gordon, K. C. Inorg. Chem. 2005, 44, 3551−3560.
(24) Rader, R. A.; McMillin, D. R.; Buckner, M. T.; Matthews, T. G.;
Casadonte, D. J.; Lengel, R. K.; Whittaker, S. B.; Darmon, L. M.; Lytle,
F. E. J. Am. Chem. Soc. 1981, 103, 5906−5912.
(25) Kilpin, K. J.; Horvath, R.; Jameson, G. B.; Telfer, S. G.; Gordon,
K. C.; Crowley, J. D. Organometallics 2010, 29, 6186−6195.
(26) Howell, S. L.; Matthewson, B. J.; Polson, M. I. J.; Burrell, A. K.;
Gordon, K. C. Inorg. Chem. 2004, 43, 2876−2887.
(27) Howell, S. L.; Gordon, K. C. J. Phys. Chem. A 2004, 108, 2536−
2544.
ASSOCIATED CONTENT
■
S
* Supporting Information
Additional crystal structures, MADs, and scale factors of
calculated vibrations, FT-Raman spectra of bpy complexes,
transient absorption data and emission spectra of [Re(CO)3py-
(OX)]+, and additional MO diagrams. This material is available
free of charge via the Internet at http://pubs.acs.org.
(28) Howell, S. L.; Gordon, K. C.; McGarvey, J. J. J. Phys. Chem. A
2005, 109, 2948−2956.
(29) Fraser, M. G.; Blackman, A. G.; Irwin, G. I. S.; Easton, C. P.;
Gordon, K. C. Inorg. Chem. 2010, 49, 5180−5189.
(30) Walsh, P. J.; Gordon, K. C.; Lundin, N. J.; Blackman, A. G. J.
Phys. Chem. A 2005, 109, 5933−5942.
AUTHOR INFORMATION
■
(31) Schaftenaar, G.; Noordik, J. H. J. Comput.-Aided Mol. Des. 2000,
14, 123−134.
Corresponding Author
*E-mail: keith.gordon@otago.ac.nz.
(32) O’Boyle, N. M.; Tenderholt, A. L.; Langner, K. M. J. Comput.
Chem. 2008, 29, 839−845.
Notes
The authors declare no competing financial interest.
(33) Gerlits, O. O.; Coppens, P. Acta Crystallogr. E 2001, 57, m164−
m166.
(34) Wenger, O. S.; Henling, L. M.; Day, M. W.; Winkler, J. R.; Gray,
REFERENCES
■
H. B. Polyhedron 2004, 23, 2955−2958.
(35) Yang, L.; Powell, D. R.; Houser, R. P. Dalton Trans. 2007, 955−
964.
(36) Horvath, R.; Gordon, K. C. Coord. Chem. Rev. 2010, 254, 2505−
2518.
(37) Howell, S. L.; Gordon, K. C. J. Phys. Chem. A 2006, 110, 4880−
4887.
(1) Worl, L. A.; Duesing, R.; Chen, P.; Della Ciana, L.; Meyer, T. J. J.
Chem. Soc., Dalton Trans. 1991, 849−858.
(2) Shirota, Y.; Kageyama, H. Chem. Rev. 2007, 107, 953−1010.
(3) Lundin, N. J.; Blackman, A. G.; Gordon, K. C.; Officer, D. L.
Angew. Chem., Int. Ed. 2006, 45, 2582−2584.
(4) Easun, T. L.; Alsindi, W. Z.; Towrie, M.; Ronayne, K. L.; Sun, X.-
Z.; Ward, M. D.; George, M. W. Inorg. Chem. 2008, 47, 5071−5078.
(5) Chen, J.-L.; Cao, X.-F.; Gu, W.; Su, B.-T.; Zhang, F.; Wen, H.-R.;
Hong, R. Inorg. Chem. Commun. 2012, 15, 65−68.
(6) Elliott, A. B. S.; Horvath, R.; Gordon, K. C. Chem. Soc. Rev. 2012,
41, 1929−1946.
̌ ́ ̌
(38) Vlcek, A., Jr.; Zalis, S. Coord. Chem. Rev. 2007, 251, 258−287.
(39) Yanai, T.; Tew, D. P.; Handy, N. C. Chem. Phys. Lett. 2004, 393,
51−57.
(40) Hirakawa, A. Y.; Tsuboi, M. Science 1975, 188, 359−361.
̌
(41) Rossenaar, B. D.; Stufkens, D. J.; Vlcek, A., Jr. Inorg. Chem.
(7) Zhang, Q.; Zhou, Q.; Cheng, Y.; Wang, L.; Ma, D.; Jing, X.;
Wang, F. Adv. Mater. 2004, 16, 432−436.
1996, 35, 2902−2900.
(42) Caspar, J. V.; Kober, E. M.; Sullivan, B. P.; Meyer, T. J. J. Am.
Chem. Soc. 1982, 104, 630−632.
(8) Zhang, Q.; Ding, J.; Cheng, Y.; Wang, L.; Xie, Z.; Jing, X.; Wang,
F. Adv. Funct. Mater. 2007, 17, 2983−2990.
̌
(43) Liard, D. J.; Busby, M.; Matousek, P.; Towrie, M.; Vlcek, A., Jr. J.
(9) Smothers, W. K.; Wrighton, M. S. J. Am. Chem. Soc. 1983, 105,
1067−1069.
Phys. Chem. A 2004, 108, 2363−2369.
(44) Noble, B. C.; Peacock, R. D. Spectrochim. Acta A 1990, 46, 407−
412.
(45) Vlce
(10) Walters, K. A.; Kim, Y.-J.; Hupp, J. T. Inorg. Chem. 2002, 41,
2909−2919.
̌
k Jr., A. Topics in Organometallic Chemistry; Springer:
(11) Gamelin, D. R.; George, M. W.; Glyn, P.; Grevels, F.-W.;
Johnson, F. P. A.; Klotzbuecher, W.; Morrison, S. L.; Russell, G.;
Schaffner, K.; Turner, J. J. Inorg. Chem. 1994, 33, 3246−3250.
(12) Caspar, J. V.; Meyer, T. J. J. Phys. Chem. 1983, 87, 952−957.
(13) Kim, Y.; Vanhelmont, F. W. M.; Stern, C. L.; Hupp, J. T. Inorg.
Chim. Acta 2001, 318, 53−60.
Berlin/Heidelberg, 2010; Vol. 29, pp 115−1588.
̌
(46) Busby, M.; Matousek, P.; Towrie, M.; Vlcek, A., Jr. J. Phys.
Chem. A 2005, 109, 3000−3008.
(47) Polo, A. S.; Itokazu, M. K.; Frin, K. M.; Patrocinio, A. O. d. T.;
Murakami Iha, N. Y. Coord. Chem. Rev. 2006, 250, 1669−1680.
(48) Buckner, M. T.; McMillin, D. R. J. Chem. Soc., Chem. Commun.
1978, 759−761.
(49) Creutz, C.; Chou, M.; Netzel, T. L.; Okumura, M.; Sutin, N. J.
Am. Chem. Soc. 1980, 102, 1309−1319.
(50) Wolcan, E.; Torchia, G.; Tocho, J.; Piro, O. E.; Juliarena, P.;
Ruiz, G.; Feliz, M. R. J. Chem. Soc., Dalton Trans. 2002, 2194−2202.
(51) Walther, M. E.; Grilj, J.; Hanss, D.; Vauthey, E.; Wenger, O. S.
Eur. J. Inorg. Chem. 2010, 4843−4850.
(52) Hanss, D.; Wenger, O. S. Inorg. Chem. 2009, 48, 671.
(53) Baiano, J. A.; Murphy, J.; Rorer, W. Inorg. Chem. 1991, 30,
4594−4598.
(14) Lundin, N. J.; Walsh, P. J.; Howell, S. L.; Blackman, A. G.;
Gordon, K. C. Chem.−Eur. J. 2008, 14, 11573−11583.
(15) Ng, P. K.; Gong, X.; Chan, S. H.; Lam, L. S. M.; Chan, W. K.
Chem.−Eur. J. 2001, 7, 4358−4367.
(16) Karthikeyan, C. S.; Wietasch, H.; Thelakkat, M. Adv. Mater.
2007, 19, 1091−1095.
(17) Giribabu, L.; Bessho, T.; Srinivasu, M.; Vijaykumar, C.;
Soujanya, Y.; Reddy, V. G.; Reddy, P. Y.; Yum, J.-H.; Gratzel, M.;
Nazeeruddin, M. K. Dalton Trans. 2011, 40, 4497−4504.
(18) Bandara, J.; Willinger, K.; Thelakkat, M. Phys. Chem. Chem. Phys.
2011, 13, 12906−12911.
(19) Willinger, K.; Fischer, K.; Kisselev, R.; Thelakkat, M. J. Mater.
Chem. 2009, 19, 5364−5376.
N
dx.doi.org/10.1021/ic3018387 | Inorg. Chem. XXXX, XXX, XXX−XXX