Fluorescence in rhoda- and iridacyclopentadienes neglecting the spin-orbit coupling of the heavy atom: The ligand dominates

Article abstract of DOI:10.1021/ic501115k

We present a detailed photophysical study and theoretical analysis of 2,5-bis(arylethynyl)rhodacyclopenta-2,4-dienes (1a-c and 2a-c) and a 2,5-bis(arylethynyl)iridacyclopenta-2,4-diene (3). Despite the presence of heavy atoms, these systems display unusually intense fluorescence from the S ₁ excited state and no phosphorescence from T₁. The S ₁ → T₁ intersystem crossing (ISC) is remarkably slow with a rate constant of 10⁸ s^-1 (i.e., on the nanosecond time scale). Traditionally, for organometallic systems bearing 4d or 5d metals, ISC is 2-3 orders of magnitude faster. Emission lifetime measurements suggest that the title compounds undergo S₁ → T₁ interconversion mainly via a thermally activated ISC channel above 233 K. The associated experimental activation energy is found to be ΔH _ISC^? = 28 kJ mol^-1 (2340 cm ^-1) for 1a, which is supported by density functional theory (DFT) and time-dependent DFT calculations [ΔH_ISC ^?(calc.) = 11 kJ mol^-1 (920 cm^-1) for 1a-H]. However, below 233 K a second, temperature-independent ISC process via spin-orbit coupling occurs. The calculated lifetime for this S₁ → T₁ ISC process is 1.1 s, indicating that although this is the main path for triplet state formation upon photoexcitation in common organometallic luminophores, it plays a minor role in our Rh compounds. Thus, the organic π-chromophore ligand seems to neglect the presence of the heavy rhodium or iridium atom, winning control over the excited-state photophysical behavior. This is attributed to a large energy separation of the ligand-centered highest occupied molecular orbital (HOMO) and lowest unoccupied MO (LUMO) from the metal-centered orbitals. The lowest excited states S₁ and T ₁ arise exclusively from a HOMO-to-LUMO transition. The weak metal participation and the cumulenic distortion of the T₁ state associated with a large S₁-T₁ energy separation favor an organic-like photophysical behavior.

Full text of DOI:10.1021/ic501115k

Article
pubs.acs.org/IC
Fluorescence in Rhoda- and Iridacyclopentadienes Neglecting the
Spin−Orbit Coupling of the Heavy Atom: The Ligand Dominates
Andreas Steffen,*,† Karine Costuas,*,‡ Abdou Boucekkine,‡ Marie-Helene Thibault,§ Andrew Beeby,§
́
̀
Andrei S. Batsanov,§ Azzam Charaf-Eddin,∥ Denis Jacquemin,∥,⊥ Jean-Francois Halet,*,‡
̧
and Todd B. Marder*,†,§
†Institut fur Anorganische Chemie, Universitat Wurzburg, Am Hubland, 97074 Wurzburg, Germany
̈
̈
̈
̈
‡Institut des Sciences Chimiques de Rennes, UMR 6226 CNRS - Universite
́
de Rennes 1, 35042 Rennes Cedex, France
§Department of Chemistry, Durham University, South Road, Durham DH1 3LE, U.K.
∥CEISAM, UMR CNRS 6230, Universite
́ ̀
de Nantes, 2 rue de la Houssiniere, 44322 Nantes Cedex 3, France
⊥Institut Universitaire de France, 103 bd Saint-Michel, 75005 Paris Cedex 05, France
S
* Supporting Information
ABSTRACT: We present a detailed photophysical study and
theoretical analysis of 2,5-bis(arylethynyl)rhodacyclopenta-2,4-
dienes (1a−c and 2a−c) and a 2,5-bis(arylethynyl)-
iridacyclopenta-2,4-diene (3). Despite the presence of heavy
atoms, these systems display unusually intense fluorescence
from the S1 excited state and no phosphorescence from T1.
The S1 → T1 intersystem crossing (ISC) is remarkably slow
with a rate constant of 108 s−1 (i.e., on the nanosecond time
scale). Traditionally, for organometallic systems bearing 4d or
5d metals, ISC is 2−3 orders of magnitude faster. Emission
lifetime measurements suggest that the title compounds undergo S1 → T1 interconversion mainly via a thermally activated ISC
channel above 233 K. The associated experimental activation energy is found to be ΔH⧧ISC = 28 kJ mol−1 (2340 cm−1) for 1a,
which is supported by density functional theory (DFT) and time-dependent DFT calculations [ΔH⧧ISC(calc.) = 11 kJ mol−1 (920
cm−1) for 1a-H]. However, below 233 K a second, temperature-independent ISC process via spin−orbit coupling occurs. The
calculated lifetime for this S1 → T1 ISC process is 1.1 s, indicating that although this is the main path for triplet state formation
upon photoexcitation in common organometallic luminophores, it plays a minor role in our Rh compounds. Thus, the organic π-
chromophore ligand seems to neglect the presence of the heavy rhodium or iridium atom, winning control over the excited-state
photophysical behavior. This is attributed to a large energy separation of the ligand-centered highest occupied molecular orbital
(HOMO) and lowest unoccupied MO (LUMO) from the metal-centered orbitals. The lowest excited states S1 and T1 arise
exclusively from a HOMO-to-LUMO transition. The weak metal participation and the cumulenic distortion of the T1 state
associated with a large S1−T1 energy separation favor an “organic-like” photophysical behavior.
However, efficient formation of the excited triplet states Tn and
emission from them can be achieved by incorporation of a
heavy atom into a rigid organic framework. Because of the non-
negligible SOC effects, this leads to the partial breakdown of
the spin selection rules that determine transitions between
states. The rate of intersystem crossing (ISC) processes tends
to scale with the spin−orbit coupling constant (ξ) of the heavy
atom incorporated into the molecule, as observed, for example,
for halogen-substituted naphthalenes.21
Organometallic complexes incorporating late transition
metals have gained special attention for triplet state formation
in photocatalysis and as phosphorescent materials. It soon
became clear from the early works of Balzani and co-workers on
the prototypical compound [Ru(bpy)3]2+ and its derivatives
INTRODUCTION
■
Major developments have been achieved in recent years in
applications of materials exhibiting long-lived excited triplet
states, such as in organic light-emitting diodes (OLEDs),1−6
light-harvesting antennas for solar cells,7−12 and photo-
catalysis.13−18 Research in these areas has mainly been
stimulated by the fact that excited triplet states (Tn) exhibit
different photophysical and photochemical properties com-
pared with excited singlet states (Sn), such as longer lifetimes τ,
lower-energy emission wavelength λem, different redox
potentials Eox/Ered, diradical character, etc.19 T1 → S0 emission
(phosphorescence) is normally spin-forbidden if the spin−orbit
coupling (SOC) between the two states is weak. This is
generally the case in ordinary organic luminophores.4,20 Thus,
the S1 state is formed according to Kasha’s rule and tends to
release its energy by S1 → S0 emission (fluorescence), whereas
the T1 state, if generated at all, relaxes by the release of heat.
Received: May 16, 2014
Published: June 12, 2014
© 2014 American Chemical Society
7055
dx.doi.org/10.1021/ic501115k | Inorg. Chem. 2014, 53, 7055−7069

Inorganic Chemistry
Article
that the combination of heavy metals exhibiting exceptionally
high intrinsic SOC with a variety of different ligands allows
facile tuning of the excited states.22,23 As a consequence, fast
ISC occurs, quenching the fluorescence from S1, and highly
efficient radiative decay from the T1 state (phosphorescence)
Unusually slow ISC leading to fluorescence (Φf = 0.22) has
been observed in [(TEE){AuPCy3}4] (TEE = tetraethynyl-
ethene) despite the presence of four gold atoms, while
[{TEB}{AuPCy3}3] (TEB = 1,3,5-triethynylbenzene) with
only three Au(I) ions displays phosphorescence (Φp
=
with quantum yields (Φp) of up to unity can be observed.19,24
A
0.46).41−45 On the basis of time-dependent density functional
theory (TD-DFT) calculations, Che and co-workers explained
the behavior of these linear-coordinated Au(I) systems by the
mutual cancellation of the SOC matrix elements of each Au
atom in the Au4 compound.43
large number of phosphorescent transition-metal compounds
mainly based on Ru (ξ = 1259 cm−1), Re (ξ = 2200 cm−1), Os
(ξ = 3500 cm−1), Ir (ξ = 3909 cm−1), or Pt (ξ = 4000 cm−1)
with bipyridine, 2-phenylpyridine, phenyltriazolate, or phenyl-
tetrazolate ligands have been synthesized.1,2,10,17,23,25−29 Their
excited states are usually described as metal-to-ligand charge
transfer (MLCT), ligand-to-ligand charge transfer (LLCT), or
intraligand charge transfer (ILCT) states or mixtures thereof.
Only a few of those compounds show residual fluorescence
(fluorescence quantum yield Φf < 1%), the low efficiency of
which is attributed to the fast ISC.25,30−33 In fact, ultrafast
conversion of the S1 state to the Tn states on the time scale of
vibrations (i.e., a few femtoseconds) was found by the groups of
Chergui and McCusker.31,34−36 They carried out picosecond
flash spectroscopy on bipyridyl complexes of Fe, Ru, and Re in
order to elucidate the time dependence of the ISC processes on
the nature of the transition metal present. Surprisingly, the iron
and ruthenium bipyridyl complexes exhibited the fastest
intersystem crossing (ca. 15 fs), although one would expect
an opposite trend according to the SOC constants of the metal
atoms (Re, ξ = 2200; Ru, ξ = 1259; Fe, ξ = 382 cm−1). The
authors explained this behavior by the geometric similarity of
We recently reported the synthesis and photophysical
properties of 2,5-bis(arylethynyl)rhodacyclopenta-2,4-dienes,
which exhibit fluorescence with quantum yields of up to Φf =
0.69 and no observed phosphorescence even at 77 K in a glass
matrix.46 A picosecond time-resolved IR spectroscopic
investigation confirmed the slow formation of the T1 state in
nanoseconds instead of the fast ISC in femto- or picoseconds
expected for octahedral organometallic complexes. To the best
of our knowledge, this was the first reported example of such a
slow ISC process in octahedral organometallic compounds.
In a series of oligothienylpyridine cyclometalated complexes
of platinum and iridium, Kozhevnikov and co-workers observed
fluorescence as well as phosphorescence.47 Their interpretation
was that the contribution of the metal character is still sufficient
to promote phosphorescence but not large enough to induce
the ultrafast depletion of the singlet state that normally occurs
for complexes with discrete arylpyridine ligands. Similar
behavior was found in Rh-based dipyrromethane complexes
by Kirsch-De Mesmaeker and co-workers, who observed that at
room temperature fluorescence originating from the ligand
dominates, while at 77 K phosphorescence is the main
process.48
Ligand-based fluorescence instead of the expected phosphor-
escence has also been reported for various fused aromatic
systems (e.g., anthracene or pentacene) linked to a transition-
metal atom via an alkynyl group or a keto group.49−66 It has
been suggested that acceleration of the S1 → T1 ISC does not
occur in this case because of the long distance between the
metal atom and the ligand chromophore, diminishing the
MLCT character and the influence of the metal in the excited
states. An exceptional example is [{C6H3(CH3)2}NCPt-
(PEt3)2]2-5,12-diethynyltetracene dication, which exhibits a
fluorescence quantum yield of Φf = 0.97.67
1
3
the MLCT and MLCT states, which, in addition to the
already very strong coupling by spin−orbit interactions of the
metal d electrons, can couple diabatically by high-frequency
vibrations in a strongly non-Born−Oppenheimer regime.
Those recent results gave fresh impetus to the discussion of
the importance of the “heavy atom effect” in such complexes,
which questions the traditional picture of the physics of
photoexcitation, according to which all steps [excitation,
internal conversion (IC), ISC, emission] should be well-
separated in time and energy (Figure 1).20,37−40 It is of great
importance to note that in the past few years, some reports
have appeared addressing the opposite effect of unusually slow
Sn → Tn ISC in organometallic compounds despite the
presence of 4d and 5d transition-metal atoms.
All of these observations raise the question of the importance
of the heavy atom effect with respect to structural and
electronic parameters and the correlation with the formation of
the triplet excited state and its emissive properties (i.e.,
phosphorescence). Herein we provide a detailed account,
including temperature-dependent photophysical measurements
and detailed DFT/TD-DFT analysis, of some representative
examples of a series of 2,5-bis(arylethynyl)rhodacyclopenta-2,4-
dienes and a related iridacyclopenta-2,4-diene to explain their
unusual excited-state behavior, namely, exceptionally high
fluorescence quantum yields (Φf = 0.33−0.69) with slow ISC
and no observable phosphorescence at 77 K. We show that
several factors, including the specific energies and nodal
properties of the frontier orbitals in the ground states, structural
changes in the excited states, and rigidity of the ligand, have to
be taken into account to understand the role of the heavy atom
in the ISC between S1 and Tn and between T1 and S0. In the
title metallacyclopentadiene series, a combination of those
factors leads to small spin−orbit couplings regardless of the
Figure 1. Simplified Jablonski diagram illustrating the fundamental
processes and their typical rate constants in organometallic complexes
after photoexcitation (f = fluorescence, p = phosphorescence, IC =
internal conversion, ISC = intersystem crossing).
7056
dx.doi.org/10.1021/ic501115k | Inorg. Chem. 2014, 53, 7055−7069

Inorganic Chemistry
Article
Scheme 1. Synthesis of 2,5-Bis(p-R-arylethynyl)rhodacyclopenta-2,4-dienes 1a−c and 2a−c46
Figure 2. (left) Absorption spectra (dot-dashed lines) and emission spectra with excitation at the respective absorption maxima (solid lines) of 1a−c.
(right) Emission spectrum of 1a at 77 K in an isopentane/Et2O/EtOH glass matrix (λex = 450 nm).46
Table 1. Selected Photophysical Properties of Rhodacyclopenta-2,4-dienes46 1a−c and 2a−c and Iridacyclopenta-2,4-diene 3
Measured in Degassed Toluene at Room Temperature
a
b
c
d
e
ε [mol−1 cm−1 dm3]
λ
[nm]
max
Φf
τf [ns]
τ0 [ns]
ΦΔ
kf [108 s−1
]
kΔ [108 s−1
]
abs
max
em
compd
λ
[nm]
1a
1b
1c
2a
2b
2c
3
456
22000
41000
44000
24000
21000
15000
501
0.33
0.34
0.69
0.07
0.16
0.46
0.08
1.2
1.8
3.0
3.6
5.3
4.3
6.8
6.2
5.4
0.65
0.40
0.26
0.32
0.19
0.20
0.22
2.75
1.89
2.30
1.47
1.61
1.84
0.61
5.42
2.22
0.87
467
497
476
487
518
515
518
560
526
541
586
595
1.0 (13%), 0.4 (87%)
1.1 (72%), 0.7 (28%)
2.5
0.80
21000
1.9 (41%), 0.9 (59%)
16.3
a
b
c
1
Observed fluorescence lifetime. Intrinsic (pure radiative) lifetime (see eq 1). Quantum yield for O2 (Δ) formation in O2-saturated solutions.
d
e
Rate constant for fluorescence. Rate constant for Δ generation.
intrinsically large spin−orbit coupling constant of the metal
O2-saturated solution was found to coincide with the S1 → T1
ISC efficiency deduced from picosecond time-resolved
transient IR measurements on 1a, indicating that 1O2 formation
occurs only from the T1 state of the rhodium complex. This
assumption also appears reasonable because the intrinsic
lifetime of S1 is below 10 ns, the minimum lifetime for an
atom involved.
RESULTS
■
Synthesis and Photophysical and Electrochemical
Measurements. We previously described the synthesis of a
series of 2,5-bis(p-R-arylethynyl)-3,4-bis(aryl)rhodacyclopenta-
2,4-dienes by the remarkably regioselective reductive coupling
of 1,4-bis(aryl)buta-1,3-diynes at Rh(I) precursors.68,69 In-
trigued by indications of unusual excited-state behavior in these
complexes, we prepared more rigid derivatives (Scheme 1),
which appeared to be surprisingly fluorescent from their S1
states, as evidenced by their short emission lifetimes (τf = 1−3
ns) in degassed toluene solution at room temperature and by
the overlap of their excitation and emission spectra (Figure 2
1
excited state to be able to generate O2 considering a diffusion
rate constant of ca. 1010 dm3 mol−1 s−1 and an O2 concentration
of ca. 10−3 mol dm−3.
As the exceptionally high fluorescence quantum yields (Φf up
1
to 0.69) and the quantum yields for O2 formation in O2-
saturated solution (ΦΔ) of the respective compounds sum to
unity, one can conclude that the ISC also occurs on the
nanosecond time scale, with very little to no nonradiative decay
from S1, and that the efficiency of 1O2 formation from T1 must
be essentially 100%. An emission spectrum recorded at 77 K in
and Table 1).46 The quantum yield for O2 generation in an
1
7057
dx.doi.org/10.1021/ic501115k | Inorg. Chem. 2014, 53, 7055−7069

Inorganic Chemistry
Article
Scheme 2. Synthesis of 2,5-Bis(p-carbomethoxyphenylethynyl)iridacyclopentadiene 3
a glass matrix showed no additional bathochromically shifted
phosphorescence between 400 and 1000 nm (Figure 2).
In order to elucidate the influence of the transition-metal
center of the metallacyclopentadiene ring on the photophysical
properties of the complexes, we also prepared an iridium
analogue by the reaction of [IrCl(PEt3)3] with 1,12-bis(p-
carbomethoxyphenyl)dodeca-1,3,9,11-tetrayne (Scheme 2), as
iridium complexes are very well known for their triplet emission
and are the most widely employed class of organometallic
phosphorescent materials in OLEDs.1−5,27,70 The formation of
the 2,5-bis(p-MeO2C−C6H4−CC)iridacyclopenta-2,4-diene
(3) occurred at room temperature after a few days, but the
Figure 4. Absorption and emission spectra (λex = 515 nm) of 3. Inset:
emission spectrum of 3 at 77 K in a glass matrix (λex = 515 nm).
reaction rate could be increased by heating the THF solution to
reflux. Interestingly, the formation of a π complex by
coordination of the bis(diyne) occurs faster than the activation
of the THF solvent by the Ir(I) fragment, a process that we
reported to occur when PMe3 instead of PEt3 is used as a ligand
in our studies concerning iridium-catalyzed nitrile hydration.71
ISC process on the nanosecond time scale (kISC = 108 s−1). An
emission spectrum recorded at 77 K in a glass matrix (Figure 4
inset) showed a well-resolved emission band with an apparent
vibrational progression of ca. 1020 cm−1, but no phosphor-
escence in the range of 500−1000 nm was observed. Indirect
measurements of the quantum yield for ISC by 1O2
sensitization gave an estimated value of ΦISC = 0.22. However,
it must be noted that 3 seems to be unstable in oxygen-
saturated solutions under photolytic conditions.
1
Complex 3 was unambiguously characterized by means of H
and 31P NMR spectroscopy, mass spectrometry, and elemental
analysis as well as X-ray diffraction studies of single crystals
obtained from diffusion of hexane into a solution of 3 in
chloroform (Figure 3).
The fact that emission occurs from the S1 state in
iridacyclopenta-2,4-diene 3 implies that the ISC processes are
little-influenced or maybe even independent of the metal atom
incorporated in the metallacycle, as the intrinsic spin−orbit
coupling constant of Ir (ξ = 3909 cm−1) is 3 times larger than
that of Rh (1200 cm−1). To the best of our knowledge, there is
no other organometallic Ir system that exhibits fluorescence but
not phosphorescence even at 77 K and has such slow ISC on
the nanosecond time scale.
Compound 1a was used as a model for further studies in
order to rationalize the excited-state behavior of such
conjugated, rodlike metallacyclopentadienes. This choice was
driven by the fact that only two paths of deactivation were
found experimentally: triplet state formation and single-
exponential radiative decay (Φf + ΦΔ ≈ 1; vide supra, Table
1). Measurement of the emission lifetime of 1a at 77 K gave a
value of τf = 3.2 ns, which is close to the calculated pure
radiative lifetime of the first singlet excited state (τ0 = τf /Φf =
3.6 ns; Table 1). This finding leads us to conclude that the ISC
process occurs under thermal activation at room temperature
and is inhibited at low temperatures. The temperature
dependence of the conversion from the first singlet excited
state to the triplet excited state T1 was further probed by
fluorescence lifetime measurements from room temperature
down to 183 K in 5 K increments (Figure 5 and Table S1 in the
Supporting Information). It can be seen from Figure 5 and eqs
1−3,
Figure 3. Molecular structure of 3 in the solid state as determined by
single-crystal X-ray diffraction. Thermal ellipsoids are drawn at the
50% probability level, and the disorder of the ethyl groups and
cyclohexyl ring and all of the H atoms have been omitted for clarity.
Selected bond distances (Å): Ir−Cl 2.4662(16), Ir−P1 2.3616(12),
Ir−P2 2.3891(15), Ir−P3 2.3693(13), Ir−C19 2.100(4), Ir−C26
2.028(4).
Compound 3 shows a poorly resolved low-energy absorption
band in the visible region of the electromagnetic spectrum with
λ
= 515 nm (ε = 21 000 M−1 cm−1) and a broad emission
abs
max
em
with λ
= 595 nm (Figure 4 and Table 1). Surprisingly,
max
iridium complex 3 also displays fluorescence (Φf = 0.08) in
deaerated toluene solution and no visible phosphorescence, a
behavior not observed previously for luminescent iridium
complexes. This is evidenced by the overlap of the absorption
and emission bands as well as by the short luminescence
lifetime [τf = 0.9 (59%)/1.9 (41%) ns], similar to that which
was observed for the 2,5-bis(p-R-arylethynyl)rhodacyclopenta-
2,4-dienes 1a−c and 2a−c, pointing again to an unusually slow
τf
τ0 =
Φ
(1)
f
7058
dx.doi.org/10.1021/ic501115k | Inorg. Chem. 2014, 53, 7055−7069

Inorganic Chemistry
Article
Table 2. Cyclic Voltammetric and Spectroscopic Data for
1a−c and 2a
EHOMO [eV]
a
b
c
abs
max
em
compd
E1/2 [V ]
λ
[nm]
λ
[nm]
max
expt
calc
1a
1b
1c
2a
0.354
0.310
0.461
0.321
456
501
−4.56
−4.52
−4.67
−4.53
−4.43
−4.42
−4.71
−4.42
467
497
476
518
560
526
a
All potentials are reported vs [FeCp*2]/[FeCp*2]+ = 0.00 V and
were obtained from 0.1 M nBu4NPF6/MeCN solutions at ambient
temperature at a carbon working electrode. EHOMO = −e[(Eox
E1/2(Fc)) + 4.8 V]. Obtained from DFT calculations on model
b
−
c
compounds (see the text and Figure 6).
The one-electron reduction wave lies outside the measurement
window and thus was not recorded. With the assumption that
the highest occupied molecular orbital (HOMO) energy of
ferrocene lies 4.8 eV below vacuum,72 the HOMO energies of
1a−c can be estimated to be 4.56, 4.52, and 4.67 eV,
respectively. Exchanging the PMe3 and CC−SiMe3 ligands
in the equatorial plane of the metallacyclopentadiene ring in 1a
for a dithiocarbamate ligand in 2a leads to a lower oxidation
potential, indicating some metal contribution to the HOMO.
The influence of the dithiocarbamate ligand on the oxidation
potential and the HOMO energy compared with 1a appears to
be similar to the introduction of the −SMe substituent at the
para position of the phenyl rings. However, this influence is
small compared with the change in compound 1c. Unfortu-
nately, no data for the iridium complex 3 could be collected
because it decomposed during the experiment. Nevertheless,
the results are in agreement with previously collected data on
related 2,5-bis(arylethynyl)-3,4-bis(aryl)rhodacyclopenta-2,4-
dienes obtained by reductive coupling of 1,4-bis(aryl)-1,3-
butadienes,69 leading us to conclude that the HOMO is mainly
located on the conjugated ligand system with little contribution
from the metal.
DFT and TD-DFT Calculations. Ground-State Properties
and Vertical Franck−Condon Excitations. In order to gain
further insight into this exceptional excited-state behavior and
to explain the observation of intense fluorescence rather than
phosphorescence, we carried out DFT and TD-DFT studies on
simplified derivatives of 1a−c (exchange of SiMe3 for H;
1a‑H−1c-H, respectively) and 3 (PMe3 instead of PEt3; 3‑Me)
and on the original compound 2a. Details of the calculations
are given in Computational Details. The following discussion is
mostly focused on mer,cis-[tris(trimethylphosphine)(ethynyl)-
2,5-bis(phenylethynyl)cyclohexa[1′,2′:3,4]rhodacyclopenta-
2,4-diene] (1a-H) (Scheme 3), as the parent compound 1a was
chosen for the temperature-dependent lifetime measurements.
However, the general analysis applies equally to 1b-H, 1c‑H,
2a, and 3-Me (also see the Supporting Information). DFT and
TD-DFT calculations were performed to delineate the
geometric and electronic structures of its ground state (S0),
Figure 5. (top) Temperature-dependent emission lifetime measure-
ments (λex = 520 nm) in degassed toluene and (bottom) Eyring plot
for 1a.
kf
Φ =
f
kf + kISC
(2)
1 − Φ
1
τf
1
τ0
f
kISC
=
=
−
τf
(3)
which are valid under the justified assumption that only
radiative decay from S1 to S0 occurs (vide supra), that the
emission lifetime τf and the fluorescence quantum yield Φf
increase and thus the ISC rate constant (kISC) decreases
gradually with decreasing temperature until T = 233 K, where a
plateau is reached. The plateau ranging from 233 to 183 K
indicates a constant Φf value of 0.77 in solution, and thus, the
ISC does not seem to be completely shut off, although it is
reduced to ΦΔ = 0.23 (see eq 1 and Table 1). However, the
ISC is further inhibited in an EPA glass matrix at 77 K, where
an estimated Φf of 0.89 and an estimated ΦΔ of 0.11 were
observed. An Eyring plot [i.e., a plot of ln(kISC/T) vs 1/T] over
the temperature range 263−308 K, according to eqs 4 and 5,
⧧
⧧
kBT
h
kISC
=
e−ΔH /RTeΔS /R
(4)
ΔH⧧
R
1
T
ΔS⧧
R
kISC
T
kB
h
ln
= −
+ ln
+
(5)
gives an activation enthalpy of ΔH⧧ISC = 28 kJ/mol (0.29 eV)
for the ISC in toluene (Figure 5).
Cyclovoltammetry measurements were performed in aceto-
nitrile versus [FeCp*2]0/+ as the internal standard for the
rhodacyclopentadienes 1a−c, and the results are given in Table
2. All three compounds show fully reversible one-electron
oxidation, which appears to be dependent on the para
substituent of the phenyl rings, with the acceptor −CO2Me
(1c) having a larger effect than the donor −SMe (1b). A similar
trend was observed in the red shift of the absorption and
emission maxima of the respective compounds (see Figure 2).
Scheme 3. Atom Numbering in 1a-H
7059
dx.doi.org/10.1021/ic501115k | Inorg. Chem. 2014, 53, 7055−7069

Inorganic Chemistry
Article
first singlet and triplet excited states (S1 and T1, respectively),
and second triplet excited state (T2).
The ground-state (S0) geometry of 1a-H was first optimized,
and selected atomic separations are given in Table 3. A
Table 3. Selected Interatomic Distances (Å) of the Ground
and Adiabatic Excited States of 1a-H
a
bond
expt
S0
S1
T1
T2
Rh−Pe
Rh−P1a
Rh−P2a
Rh−Cα
Cα−Cβ
Rh−C1
Rh−C1′
C1−C2
C1′−C2′
C2−C2′
C1−C3
C1′−C3′
C3−C4
C3′−C4′
C4−C5
C4′−C5′
C5−C6
C5′−C6′
C6−C7
C6′-C7′
C7−C8
C7′−C8′
C8−C9
C8′−C9′
C9−C10
C9′−C10′
C10−C5
2.3606(5)
2.3160(5)
2.3153(5)
2.0479(18)
1.218(2)
2.0993(17)
2.0806(17)
1.374(2)
1.365(2)
1.447(2)
1.419(2)
1.413(2)
1.205(3)
1.206(2)
1.435(3)
1.433(2)
1.401(3)
1.403(2)
1.383(3)
1.385(3)
1.384(3)
1.378(3)
1.384(3)
1.384(3)
1.388(3)
1.387(3)
1.403(3)
1.399(3)
2.424
2.353
2.354
2.039
1.233
2.101
2.073
1.383
1.378
1.447
1.407
1.404
1.234
1.230
1.426
1.424
1.413
1.413
1.395
1.395
1.400
1.401
1.401
1.401
1.394
1.395
1.414
1.412
2.440
2.366
2.366
2.043
1.233
2.077
2.057
1.436
1.430
1.399
1.382
1.377
1.247
1.243
1.410
1.409
1.422
1.421
1.392
1.392
1.404
1.404
1.405
1.404
1.391
1.392
1.424
1.421
2.429
2.360
2.360
2.038
1.233
2.090
2.062
1.467
1.460
1.371
1.369
1.367
1.251
1.246
1.410
1.410
1.421
1.420
1.393
1.392
1.403
1.403
1.404
1.403
1.392
1.393
1.422
1.419
2.425
2.357
2.357
2.036
1.233
2.104
2.075
1.395
1.390
1.466
1.394
1.390
1.251
1.249
1.400
1.395
1.434
1.437
1.387
1.386
1.413
1.414
1.413
1.414
1.387
1.386
1.436
1.436
Figure 6. Molecular orbital diagrams of 1a-H, 1c-H, 2a, and 3-Me.
octahedral environment. Orbitals of “t2g” and “eg*” pseudosym-
metry expected for octahedral ML6 systems are found below
the HOMO and above the LUMO, respectively. Introduction
of the acceptor substituent −CO2Me (1c-H) at the para
position of the phenyl rings lowers the energy of the HOMO (a
trend nicely reproduced by the cyclic voltammetry measure-
ments), but its effect on the π* orbitals LUMO and LUMO+1
is greater, leading to a smaller HOMO−LUMO gap (Figure 6;
also see the Supporting Information). Substitution of the PMe3
and CC−SiMe3 ligands in the equatorial plane by the
electron-donating dithiocarbamate ligand changes the nature of
the lower-lying occupied orbitals but leaves the frontier orbitals
unaltered (compare 1a-H and 2a in Figure 6). Furthermore,
although the substitution of iridium for rhodium does increase
the metal d-orbital participation to ca. 8%, the frontier orbitals
are still dominated by the ligand π system, as can be seen from
a comparison of 1c-H and 3-Me.
One important question is the following: What makes the
HOMO and the LUMO so energetically isolated from the
other molecular orbitals? In comparison with a shorter and less
π-conjugated ligand system such as a simple biphenylene, in the
present case we have an extended and highly conjugated π
system, namely, the bis(arylethynyl)butadiene backbone, with a
small HOMO−LUMO gap. In addition, the strong ligand field
gives rise to a large dπ−dσ* splitting. The small HOMO−
LUMO gap of the π ligand and the large d-orbital splitting
combine to disfavor interactions of the ligand-based orbitals
with the metal d orbitals as a result of the increase in the energy
difference and the decrease in overlap.
C
10′−C5′
a
Obtained from single-crystal X-ray diffraction studies of 2,5-
bis(phenylethynyl)rhodacyclopenta-2,4-diene 1a.
comparison with the data experimentally determined by single-
crystal X-ray diffraction studies of the 2,5-bis(phenylethynyl)-
rhodacyclopenta-2,4-diene complex 1a reveals that the metal−
carbon and carbon−carbon separations are well-reproduced.
However, the CC bond lengths are slightly overestimated by
ca. 0.025 Å in the calculations, and as often seen in
organometallic complexes, the DFT-computed values of the
metal−phosphorus distances are somewhat overestimated with
respect to the experimentally obtained separations, in this case
by ca. 0.05 Å.73,74
The frontier orbital region of the molecular orbital (MO)
diagram of 1a-H is shown in Figure 6. The HOMO and lowest-
unoccupied MO (LUMO) are well-separated in energy. In
addition, they are somewhat isolated from the other occupied
and vacant MOs by 1.36 and 1.14 eV, respectively. An analysis
of their nodal properties indicates that they are π-type MOs,
and they are heavily weighted on the organic π system of the
conjugated ligand with very modest participation of the
rhodium (ca. 4% in the HOMO and <2% in the LUMO)
(Figure 7). The general trend of the relative HOMO energies
as well as the small metal contribution are in agreement with
the conclusions drawn from the cyclovoltammetry measure-
ments (Table 2). The rhodium metal center lies in a pseudo-
TD-DFT calculations were performed starting from the
optimized S0 ground-state structure in order to obtain the
lowest excitation energies. First, vertical energies were
calculated; solvent effects and geometric relaxation occurring
after excitation were not considered. The resulting excited
states thus represent the respective Franck−Condon (FC)
7060
dx.doi.org/10.1021/ic501115k | Inorg. Chem. 2014, 53, 7055−7069

Inorganic Chemistry
Article
Figure 7. Isosurface plots of selected frontier molecular orbitals of 1a-H, 2a, and 3-Me. Isocontour values: 0.035 (e/bohr3)1/2
.
states in the gas phase. The excitation energies and main
transitions (including their percentage contributions to the
excitations) of 1a-H are given in Figure 8 and Table 4. The
lowest excited state is a triplet state (FC-T1) with a calculated
excitation energy of 1.06 eV. This would lead to phosphor-
escence in the near-IR (NIR) region (vide infra). It can be
formally described as the result of an electronic transition from
the HOMO to the LUMO. The second lowest excitation, with
an oscillator strength of f = 0.86, leads to a singlet state (FC-S1)
with an associated energy of 2.50 eV. The third excited state
(FC-T2) is found at almost the same excitation energy (2.52
eV). The same situation arises by comparison of the FC-S2
state, which is a result of a symmetry-forbidden HOMO →
LUMO+2 transition (f = 0.00), and the FC-T3 state, which are
nearly degenerate at 3.11 and 3.09 eV, respectively.
The ground-state molecular orbital structure provides an
explanation for the significant energy gap between the FC-S1
and FC-T1 states. As mentioned earlier, the HOMO and
LUMO are both essentially ligand-based orbitals with π
symmetry that are well-separated in energy from lower-lying
filled orbitals with metal contribution. The first two excited
states correspond to ligand-centered electronic transitions
(π−π*) (Table 3), as is common for fluorescent organic
systems.19,20 The large majority of organometallic compounds
exhibit excited states with MLCT admixtures, which usually
decrease the S1−T1 energy gap in comparison with that for
local (π−π*) excitation.1,4,19,37,75
The energy calculated for the FC first excited singlet state
matches well with the first absorption peak found exper-
imentally in toluene (λmax = 456 nm, 2.72 eV), although solvent
effects were not taken into account. A vibrational progression
was proposed to interpret the rest of the experimental UV−vis
absorption spectrum (i.e., the envelope of this large absorption
band).15 This assumption was based on a comparison with
organic compounds incorporating aromatic rings, for which this
type of vibronic structure can be observed. Indeed, our
theoretical results on the FC states of 1a-H reveal that the
Figure 8. Jablonski diagram of the vertical electronic excitations
(Franck−Condon states) of 1a-H obtained by TD-DFT calculations.
7061
dx.doi.org/10.1021/ic501115k | Inorg. Chem. 2014, 53, 7055−7069

Inorganic Chemistry
Article
Table 4. Vertical Electronic Excitation Energies and Main Transitions Describing the First Six Franck−Condon States of 1a-H
Obtained by TD-DFT Calculations
a
b
excitation
energy [eV]
f
transition(s)
configuration
π−π*
π−π*
π−π*
π−π*
Rh-d/π−π*
π−σ*(Rh−P)
π−σ*(Rh−P)
(Rh-d/TMS-π)−π*
π−π*
classification
3IL
1IL
S0 → T1
S0 → S1
S0 → T2
1.06
2.50
2.52
−
0.8559
−
HOMO → LUMO (99%)
HOMO → LUMO (98%)
HOMO → LUMO+1 (49%)
HOMO−4 → LUMO (26%)
HOMO−2 → LUMO (8%)
HOMO → LUMO+2 (83%)
HOMO → LUMO+2 (95%)
HOMO−1 → LUMO (65%)
HOMO−5 → LUMO (20%)
3IL/3MLCT (19%)
S0 → T3
S0 → S2
S0 → S3
3.09
3.11
3.26
−
0.0000
3IL/3LMCT (23%)
1IL/1LMCT (29%)
1IL/1MLCT (14%)
0.0004
a
b
Only the major contributions are listed. The percentage contributions of MLCT/LMCT are given in parentheses.
second singlet excited state (3.11 eV), which has negligible
oscillator strength, cannot explain the shape of the absorption
spectrum. In order to locate the vibronic states involved, we
applied a computational protocol that allows simulation of the
vibrational structure of absorption spectra (see Computational
Details).76,77 The harmonic vibrational frequencies of both
states were computed on the respective optimized geometries
using a method allowing accurate determination of vibronic
couplings. The vibrationally resolved spectra were computed
using the FC classes program taking into account solvent effects
(see Computational Details).78,79
electronically excited states. In Figure 9, the main vibrational
modes contributing to both phenomena are given. Movies of
these vibrational modes are available in the Supporting
Information. For absorption (i.e., excited-state vibrations),
modes 27, 149, 196, and 203 appear at 160, 1290, 1578, and
2191 cm−1, respectively. The first mode corresponds to the
deformation of the full molecule. The second and third modes
can be mainly ascribed to the stretching of the double and
single carbon−carbon bonds of the rhodacycle, respectively.
Finally, mode 203 corresponds to stretching of the CC triple
bonds attached to the central five-membered metallacycle. For
emission, mode 27 (159 cm−1) is similar to its excited-state
counterpart, and the same holds for mode 195 (1585 cm−1),
which is localized on the central five-membered ring (as for
mode 196 described above).
The computed vibrationally resolved absorption and
emission spectra are compared with their experimental
counterparts in Figure 9 (note that the emission has been
Excited-State Properties and ISC Processes. The geometries
of the first excited singlet state S1 and the first and second
excited triplet states (T1 and T2) of 1a-H were also optimized
(Figure 10) using the same basis set and functional as for the
ground state (see Computational Details). This allows direct
comparisons with the previously discussed data. A comparison
of the structural parameters is given in Table 3. The geometries
of the optimized excited states differ significantly from that of
the ground state, mainly in the rhodacyclopentadiene core and
to a lesser extent in the 2,5-bis(phenylethynyl) moieties.
Interestingly, calculated geometrical changes found for S1 and
T1 show the same pattern. With reference to the ground-state
S0 geometry, the respective M−C, CC, and C−C distances
in the S1 and T1 states are overall up to a few of hundredths of
an angstrom shorter, longer, and shorter, respectively, whereas
the ethynyl CC distances are approximately 0.01 Å longer. In
other words, the geometries of the S1 and T1 states develop
some cumulenic character. This is in accordance with the nodal
properties of the HOMO and LUMO in the S0 state displayed
in Figure 7, which would be depopulated and populated,
respectively, in the excited states. Nevertheless, their geometries
are different, particularly within the rhodacyclopentadiene core,
where the double and single bond alternation, reversed
compared with that in the S0 state, is more marked in T1
than in S1 (1.467−1.371−1.460 vs 1.430−1.399−1.436 Å,
respectively).
Figure 9. Vibrationally resolved absorption spectrum (black solid line)
and emission spectrum (red solid line) computed at the M06-2X level.
Both convoluted and stick spectra are shown, together with the
numbering of the most important vibrational modes (given in the
Supporting Information). The experimental absorption spectrum
(black dashed line) and renormalized fluorescence spectrum (red
dashed line) are given for comparison.
renormalized; see Computational Details). The good match
with the experimental data is obvious: for absorption, a weaker
band corresponding to the excitation between the lower-lying
vibrational states (0−0) is present, followed by a second
maximum and a shoulder; for emission, a structure with two
maxima of similar intensities and a shoulder at longer
wavelengths is found. The position of the 0−0 band, which
includes vibrational zero-point energy (ZPE) corrections, is
also in line with the experimental spectrum. This clearly
confirms that the specific shapes of these bands are related to
vibronic couplings and not to two (or more) energetically close
These geometrical differences can disfavor ISC from S1 to T1
considering their difference in energy (vide supra). Instead, we
propose the thermal population of T1 on the basis of the
temperature-dependent emission lifetime for compound 1a in
toluene (ΔH⧧ISC = 0.29 eV) (vide supra). The activation barrier
for 1a-H was calculated to be ΔH⧧ISC = 0.11 eV via geometry
optimization of the transition state between S1 and T1 (see
7062
dx.doi.org/10.1021/ic501115k | Inorg. Chem. 2014, 53, 7055−7069

Inorganic Chemistry
Article
Figure 10. Calculated structures of 1a-H in (left) the excited singlet state S1 and (right) the excited triplet state T1 (right). Hydrogen atoms have
been omitted for clarity. Bond lengths are given in Å.
Computational Details), and this value is in good agreement
with the experimental result. In addition, we searched for a
transition state that could connect S1 to T2 because the FC-T2
state is very close in energy to FC-S1 (Table 4), making ISC
from S1 to T2 a possible pathway. However, the resulting
structure had the same geometry as the transition state found
between S1 and T1. Thus, it is difficult to distinguish whether
the system undergoes thermally activated ISC to T1 or to T2,
but we can extrapolate that the energy of the transition state
between S1 and T2 is at least that found for the thermal S1 →
T1 process. It is important to note that the spin density of the
T1 state is spread along the conjugated organic ligand system
with very little rhodium contribution (less than 0.1 electron;
see Figure 11), which must disfavor the emissive T1 → S0 de-
excitation (phosphorescence).
envisage the possibility that a thermal geometrical rearrange-
ment occurs rather than direct ISC to T1 (ps time scale). The
same type of calculation revealed that the probability of ISC
from T1 to S0 is also very low, with an estimated lifetime of 0.3
s. This time scale again allows nonradiative de-excitations to
occur, which are particularly facile at low energy (above 1200
nm) according to the energy-gap law. Apart from our
calculations on the T1 state (Tables 4 and 5), which suggest
a phosphorescence energy of ca. 1 eV (1240 nm), the
quantitative sensitization of singlet oxygen from the triplet
states of the rhodacyclopentadienes also leads us to expect a
potential phosphorescence to occur within the above-
1
mentioned wavelength range. This is due to the fact that O2
emits at 1270 nm (0.98 eV), and therefore, the energy of the T1
states of the rhodium compounds must be equal to or even
higher than that value. We previously measured an emission
spectrum in the range between 400 and 1000 nm. However,
with our new knowledge of the energy of the T1 state, we
remeasured the spectrum of a degassed solution of 1a in
toluene in the region between 900 and 1400 nm at room
temperature, where no phosphorescence could be detected.
This new information is consistent with the calculated lifetime
of 0.3 s for the phosphorescence, which allows for nonradiative
decay by vibrational and/or rotational modes. Because only
very little ISC occurred at 77 K in a glass matrix (ΦΔ ≈ 0.1;
vide supra), we did not expect any emission to be observable
within the instrument noise of the NIR photomultiplier tube at
this temperature.
Figure 11. Spin density distribution of the optimized triplet excited
state T1 of 1a-H. Isocontour values: 0.005 e/bohr3.
The vibrational frequencies were computed for each
optimized state. Three vibrations were found between 1900
and 2200 cm−1 for 1a-H, corresponding to vibrational modes
involving the C3C4 and C3′C4′ stretching vibrators,
coupled or not, and the CαCβ (acetylide ligand) vibrator
(Table 5). In each case, the most intense modes are different.
For S0, the C3C4 and C3′C4′ stretching vibrators are
uncoupled and almost isoenergetic (2120 and 2146 cm−1). For
S1 and T1, the symmetric mode of the C3C4/C3′C4′
vibronic coupling is active (2005 and 1933 cm−1, respectively).
The energy difference is then explained by the change in the
In order to evaluate the intrinsic rate constants for the S1 →
T1 and T1 → S0 vertical ISC processes, spin−orbit Tamm−
Dancoff approximation (SO-TDA) TD-DFT calculations were
performed (see Computational Details). The resulting
calculated S1 lifetime, for which no competitive emissive or
nonradiative deactivation pathways were, of course, considered,
was estimated to be 1.1 s. This was calculated via the Einstein
coefficient of spontaneous emission. This value leads us to
a
Table 5. Relative ZPE-Corrected Energies ΔE (eV) and Calculated Frequencies (cm−1), Intensities (km·mol−1, in
Parentheses), and Natures of the CC Vibrational Modes for the Ground State and First Relaxed Excited States of 1a-H (See
Scheme 3 for Atom Numbering)
state
ΔE
vibrational frequency (intensity) nature
S0
T1
S1
T2
0.000
0.985
2.259
1.339
1949 (40) CαCβH
2120 (513) C3C4Ph
1950 (39) CαCβH
2146 (287) C3′C4′Ph
b
b
c
c
c
1933 (445) C3C4Ph + C3′C4′Ph
1948 (25) CαCβH
1986 (2) C3C4Ph − C3′C4′Ph
2057 (7) C3C4Ph − C3′C4′Ph
2088 (3) C3C4Ph − C3′C4′Ph
b
2005 (836) C3C4Ph + C3′C4′Ph
1963 (845) C3C4Ph + C3′C4′Ph
2049 (31) CαCβH
a
b
c
A scaling factor of 0.9521 was applied to the vibrational frequencies.80 Symmetric vibration mode. Asymmetric vibration mode.
7063
dx.doi.org/10.1021/ic501115k | Inorg. Chem. 2014, 53, 7055−7069

Inorganic Chemistry
Article
C−C distances of the conjugated ligand in T1 (see Table 3).
These calculated results are in excellent agreement with the
evolution of the time-resolved infrared (TRIR) vibrational
absorption spectra, which we reported previously,46 showing
the disappearance of two bands at 2128 and 2142 cm−1
belonging to the ground state S0. Immediately after photo-
excitation, the appearance of a new band at 2008 cm−1 (S1) is
noted, and after 300 ps, a second band arises at 1942 cm−1
(T1), growing with a time constant of τ ≈ 1.6 ns. The fact that
T2 (or another higher triplet excited state) was not noticeable
in the TRIR experiments is most likely due to the expected
short lifetimes of these states according to Kasha’s rule.19,20
Similar theoretical analyses were also performed for
rhodacyclopentadienes 1b-H, 1c-H, and 2a and the new
iridium compound 3-Me (see the Supporting Information). All
of them give a similar picture with the same conclusions as for
the model compound 1a-H.
Scheme 4. Summary of the Excited-State Processes upon
Photoexcitation for 1a−c, 2a−c, and 3
DISCUSSION
■
The metallacyclopentadienes 1a−c, 2a−c, and 3 investigated
here exhibit highly intense fluorescence from the S1 state
instead of phosphorescence at both room temperature and 77
K. The temperature-dependent emission lifetime measurements
on 1a (Figure 5) suggest that the S1 → Tn ISC process is largely
inhibited at low temperatures. We therefore conclude that a
thermally activated ISC pathway exists with an activation
barrier of ΔH⧧ISC = 28 kJ mol−1 (0.29 eV) for 1a. The energetic
barriers for the ISC processes S1 → T1 and S1 → T2 in 1a-H,
which share the same transition state, were calculated by DFT
methods to be on the same order of magnitude as the
experimental value [i.e., ΔH⧧ISC(calc.) = 11 kJ mol−1 (0.11 eV)],
supporting this conclusion. In addition, for all of the
absorption process has significant MLCT contributions (and
the reverse for emission) involving d electrons.5,23,25−27,31,37
Rhodium complexes are usually less emissive because of an
important intraligand (IL) component in the excitation or
emission processes that leads to weaker SOC than, for example,
in iridium analogues, which show typical MLCT.33,81−95
In the case of the metallacyclopentadienes 1a−c, 2a−c, and
3, the “organic-like” photophysical behavior appears to be due
to the electronic structure of the frontier orbitals, in which the
HOMO and the LUMO are nearly pure ligand-type π and π*
orbitals, as confirmed by cyclic voltammetric measurements,
and are well-separated from the lower- and higher-lying MOs,
respectively (including metallic orbitals; vide supra). Con-
sequently, the incorporation of Ir (ξ = 3909 cm−1) in 3 instead
of Rh (ξ = 1200 cm−1) does not affect the composition of these
frontier orbitals much (Figures 6 and 7) and does not
significantly enhance the SOC component overlap between S1
and T1 or the interconversion rate (Table 1).
Two consequences arise from these findings that are
confirmed by TD-DFT calculations: (1) electronic transitions
originating from the frontier orbitals lead to the FC-S1 and FC-
T1 states, which can be described as purely 1,3IL(π−π*) with
presumably weak coupling, and (2) the relaxed S1 and T1 states
are well-separated from excited states originating from
transitions between lower occupied and higher unoccupied
metal-based orbitals.
We also found that structural changes occur during the
relaxation of T1, which develops a more cumulenic character of
the extended conjugated ligand in comparison to S1, including
important modifications of the bond distances in the
rhodacyclopentadiene core (Table 3 and Figure 10). This
fact, associated with more important electronic exchange
energy in triplet states than in singlet states, leads to an energy
difference of more than 1 eV between S1 and T1, hampering
direct ISC between these two states. However, the proximity in
energy and the similarity in structural arrangement of the S1
and FC-T2 states allows the possibility of an alternative ISC
pathway. In that case, the participation of the metallic electrons
in the excitation process would be small, definitely weaker than
1
compounds O2 was shown to be formed in moderate yields
in O2-saturated solutions at room temperature (Table 1). Thus,
the triplet state(s) is(are) formed on the nanosecond time scale
(kISC = 108 s−1) under these experimental conditions. The S1 →
T1 conversion occurs mainly by thermal activation from higher
vibrational levels of the S1 state or via a transition state (atomic
motion), as shown in Scheme 4. Our experimental results
reveal that the energy necessary for this structural change is not
reached at 233 K. The even higher fluorescence efficiency of Φf
= 0.89 at 77 K (i.e., the lower ISC rate) could possibly be a
result of the rigidity imposed by the glass matrix.
At first glance, this behavior is surprising, bearing in mind
that ultrafast ISC channels due to the strong SOC mediated by
the metal d electrons should be available to couple the singlet
and triplet excited states, as has been found for the vast majority
of Ru, Rh, Os, Re, Ir, and Pt complexes.5,23,25−27,31,37 It should
also be emphasized that although rhodium complexes are
usually not very emissive in comparison to their iridium
analogues, depletion of the S1 state, which is mostly of
1IL(π−π*) character, via ISC to a triplet state is typically faster
than fluorescence from S1.33,81−95
SOC is often relatively small, especially in organic systems,
and is often overlooked in quantum-chemical calculations.
Nevertheless, it affects several physical phenomena such as ISC,
for which the possibility of transitions between the involved
states would be zero without the SOC component. ISC is
emphasized, of course, when the difference in energy between
the states is small but also when photoexcitation or emission
involves strongly relativistic electrons (i.e., electrons surround-
ing heavy atoms). It is for this reason that the large majority of
Ru, Os, Re, Ir, and Pt complexes are phosphorescent, as the
7064
dx.doi.org/10.1021/ic501115k | Inorg. Chem. 2014, 53, 7055−7069

Inorganic Chemistry
Article
that usually found in rhodium systems. Most probably, a
thermal conversion to either T2 or T1 occurs (vide supra and
Scheme 4). Thermally activated conversion does not rely on
the presence of heavy atoms; indeed, it has been reported for
organic perylene and anthracene derivatives.96−99
Our SO-TDA calculations suggest that the radiative S1 → T1
ISC process occurs with a lifetime of 1.1 s, underlining the
exceptionally weak coupling between these two states.
Consequently, the emissive T1 → S0 process has a calculated
lifetime of ca. 0.3 s because of a weak SOC component in the
overlap of its wave function with that of the FC singlet state
FC-S0 (almost-nonrelativistic electrons involved in the de-
excitation process). It has to be noted also that the long
intrinsic lifetime of T1 allows for effective nonradiative decay by
atomic motions and thus for geometry relaxation.
A comparison with other fluorescent organometallic systems
incorporating heavy transition metals, such as rhodium, iridium,
or platinum, shows the uniqueness of the 2,5-bis(arylethynyl)-
metallacyclopentadienes. While mostly only residual fluores-
cence is observed,25,30−33 more efficient fluorescence (i.e., slow
ISC) is possible for chromophoric ligands to which a transition-
metal center is linked, for example, via an alkynyl chain, as
shown recently for dinuclear [X(Et3P)2Pt]2n+-5,12-diethynylte-
tracenes and -pentacenes (X = Cl, Br, I, CCC6H5, SC6H5,
SeC6H5 for n = 0; X = PPh3, PEt3, NC5H5, CN(Me)2C6H3 for n
= 2).55,67 However, the fact that the metal is linked instead of
directly attached to the chromophore reduces the influence of
the metal on the excited states. This increases the distance of
the metal from the chromophore, leading to less SOC by the
external heavy atom effect. A good, albeit very simplified,
approximation for the efficiency of the SOC is given by the
expression kISC ∝ Z8/r6, where Z is the atomic number and r is
the distance between the metal atom and the center of the
chromophore involved in the transition.100 The transition metal
in our 2,5-bis(arylethynyl)metallacyclopenta-2,4-dienes 1a−c,
2a−c, and 3 is covalently bound directly to the center of the
chromophoric ligand, but still does not participate in the
transitions, leading to pure organic 1,3IL(π−π*) states as found
for structurally related main-group 2,5-bis(arylethynyl)boroles,
-siloles, -thiophenes, and -phospholes.101−108
interact. DFT and TD-DFT studies revealed that the extended
conjugation of the peculiar ligand causes the HOMO and
LUMO to be ligand-centered π and π* MOs that lie in between
the metal-centered filled and vacant orbitals. The lowest excited
states, S1 and T1, originate from electronic transitions between
the HOMO and LUMO. This, in combination with a structural
distortion to a cumulenic geometry in the excited T1 state, leads
to a weak coupling of the different spin states and thus to
“organic-like” photophysical behavior.
Our results show that the ligand can win control over the
photophysical excited-state behavior to such an extent that even
heavy transition-metal atoms such as Rh or Ir participate in
enhancing the fluorescence compared with their main-group
analogues but basically do not influence the ISC processes. The
awareness of the importance of the peculiar structure of the π-
chromophoric ligand, such as the ones with extended
conjugation used in this work, and the careful analysis of the
spin−orbit interactions by the d electrons provide the
possibility of other applications in the area of organometallic
photophysics such as upconversion, nonlinear optics, NIR
emitters, and even the use of first-row transition-metal centers
for luminescent materials. This study sheds light on the true
role of the heavy atom in excited-state processes that cannot be
disassociated from its ligands. Further work to explore potential
applications exploiting the combination of fluorescent “organic”
1
behavior with “organometallic” triplet properties (such as O2
sensitization) and stability is in progress.
EXPERIMENTAL SECTION
■
General Considerations. Unless otherwise stated, manipulations
were performed using standard Schlenk or glovebox (Innovative
Technology Inc.) techniques under an atmosphere of dry nitrogen
(BOC). Reagent-grade solvents (Fisher Scientific and J.T. Baker) were
nitrogen-saturated, dried and deoxygenated using an Innovative
Technology Pure-Solv 400 solvent purification system, and further
deoxygenated by the freeze−pump−thaw method. C6D6 and CDCl3
were purchased from Cambridge Isotope Laboratories, dried over
sodium/benzophenone and CaH2, respectively, deoxygenated using
the freeze−pump−thaw method, and vacuum-transferred into sealed
vessels. The syntheses of the 2,5-bis(arylethynyl)rhodacyclo-
pentadienes 1a−c and 2a−c have been reported previously.46
All of the NMR spectra were recorded at ambient temperature using
CONCLUSION
a 400 MHz Varian Mercury spectrometer (1H, 400 MHz; 13C{1H},
■
1
100 MHz; 31P{1H}, 162 MHz). H NMR chemical shifts are reported
The detailed photophysical and theoretical analysis of 2,5-
bis(arylethynyl)rhodacyclopenta-2,4-dienes 1a−c and 2a−c
and 2,5-bis(arylethynyl)iridacyclopenta-2,4-diene 3 carried out
in this work provides a full picture of the excited-state behavior
and clarifies the origin of the highly unusual slow intersystem
crossing, which occurs with a rate constant of 108 s−1 (i.e., in
nanoseconds). Normally, the intersystem crossing in 4d and 5d
organometallic systems would be expected to be several orders
of magnitude faster than the S1 → S0 radiative decay. These
compounds, which exhibit fluorescence (S1 → S0 radiative
decay) and no observable phosphorescence, undergo inter-
conversion of their excited singlet state S1 to the triplet state T1
mainly via a thermally activated ISC channel above 233 K, for
relative to tetramethylsilane (TMS) and were referenced to residual
proton resonances of the corresponding deuterated solvent (C6D6,
7.16 ppm), whereas 13C NMR spectra are reported relative to TMS
using the carbon signals of the deuterated solvent (C6D6, 128.39
ppm). 31P NMR spectra were referenced to external 85% H3PO4.
Elemental analyses were obtained using an Exeter Analytical Inc. CE-
440 elemental analyzer. Unit mass resolution spectrometric determi-
nations were obtained using a MALDI ToF Applied Biosystems
Voyager-DE STR mass spectrometer.
Cyclic voltammograms were recorded at v = 100 mV s−1 from 0.1 M
(nBu)4NPF6/MeCN solutions containing ca. 1 × 10−4 M analyte using
a three-electrode cell equipped with a glassy carbon working electrode,
Pt wire counter electrode, and Pt wire pseudoreference electrode. All
of the redox potentials are reported with reference to an internal
standard of the decamethylferrocene/decamethylferrocenium couple
([FeCp*2]/[FeCp*2]+ = 0.00 V).
UV−vis absorption and emission spectra, lifetime, and quantum
yield measurements were all recorded in degassed toluene. UV−vis
absorption spectra and extinction coefficients were obtained on a
Hewlett-Packard 8453 diode array spectrophotometer using standard 1
cm path length quartz cells. Fluorescence spectra and quantum yield
measurements on dilute solutions with absorbance maxima of less than
which we have determined the activation energy to be ΔHI⧧SC
=
28 kJ mol−1 for 1a, while only spin−orbit ISC occurs at lower
temperatures. The general photophysical behavior of these
octahedral compounds does not change dramatically upon
substitution of the other ligands at the rhodium center or upon
substitution of rhodium by iridium. The organic π-chromo-
phore ligand apparently neglects the presence of the heavy
rhodium or iridium atom, although they strongly covalently
7065
dx.doi.org/10.1021/ic501115k | Inorg. Chem. 2014, 53, 7055−7069

Inorganic Chemistry
Article
0.2 were recorded on a HORIBA Jobin Yvon Fluoromax-3
spectrophotometer using the conventional 90° geometry. The
emission spectra were fully corrected for the spectral response of
the emission optical components using the manufacturer’s correction
curves. The quantum yield was measured using an integrating sphere
with a HORIBA Jobin Yvon Fluorolog 3−22 Tau-3 spectropho-
tometer following a method described in the literature.109 The
absorbance of the samples was kept below 0.12 to avoid inner filter
effects, and all of the measurements were carried out at room
temperature.
17.831(2) Å, α = 91.515(4)°, β = 95.930(4)°, γ = 107.886(4)°, V =
2302.5(8) Å3, Z = 2, Dcalc = 1.449 g cm−3, μ = 3.10 mm−1, 40 936
reflections with 2θ ≤ 60°, 13 468 unique, Rint = 0.086, R1 = 0.041 [10
430 data with I ≥ 2σ(I)], wR2(F2) = 0.095 (all data), CCDC-952768.
COMPUTATIONAL DETAILS
■
Some DFT calculations were carried out with the Gaussian 09
program,113 including determination of the geometric structures,
which were fully optimized without any symmetry constraints using
the MPW1PW91 functional114−116 with the LANL2DZ effective core
potential basis set augmented by a polarization function for all atoms
except hydrogens.117 Harmonic vibrational frequency calculations were
performed to check that the optimized geometries were energy
minima and to compute ZPE corrections. The spin density isosurface
representation was created using the GaussView 5.0 program.118 The
geometries and energies of the transition state between S1 and T1 and
that between S1 and T2 were calculated using the geometries of T1 and
T2, respectively, as the starting compounds and that of S1 as the
product by means of the synchronous transit-guided quasi-Newton
method as implemented in Gaussian 09.113,119 In both cases, two
parallel calculations were performed with and without providing an
initial geometry of the transition state (average of the main diverging
distances and angles). The maximum sizes of the optimization steps
were diminished to 0.005 Å and 0.6°. Those calculations were
performed for the triplet spin state configuration, considering that the
ISC occurs at fixed geometry. Vibrationally resolved absorption and
emission spectra were calculated following a protocol detailed recently
in the literature.76,77 The ground- and excited-state structures were
optimized using analytic DFT and TD-DFT gradients. These
calculations were performed with the Gaussian 09.C01 program113
using the M06-2X functional120 and the 6-31G(d) atomic basis set.
This choice was driven by the fact that the potential energy surfaces of
excited states are different from their ground-state counterparts and
necessitate the use of exchange−correlation functionals including a
large share of exact exchange in order to avoid qualitative breakdowns
(see refs 76 and 77 and references therein). The M06-2X/6-31G(d)
combination is known to be valuable for TD-DFT calculations of 0−0
energies and band shapes.76,77,121 To ensure numerically stable
vibronic spectra, the force minimizations were performed until the
residual mean square force was smaller than 1 × 10−5 a.u. The
vibrationally resolved spectra within the harmonic approximation were
computed using the FC Classes program.78,79 The reported spectra
were simulated using convoluted Gaussian functions presenting a full
width at half-maximum (fwhm) of 0.07 eV. Maximal numbers of 25
overtones for each mode and 20 combination bands on each pair of
modes were included in the calculations. The maximum number of
integrals to be computed for each class was set to 1 × 1010, and it was
checked that such a number provided converged FC factors (>0.9; see
the discussion in ref 76). In these calculations, the electrostatic
interactions between the molecule and the environment (toluene)
were modeled using the polarizable continuum model (PCM), which
approximates solvent effects as long as no specific solute−solvent
interactions take place.122 In Figure 9, the experimental fluorescence
spectrum measured on the wavelength scale was transformed in line
shapes by applying an intensity correction proportional to ω2, as this
correction, which allows consistent theory/experiment comparisons,
significantly affects the band shapes. DFT two-component SOC
calculations were performed with the ADF2012.01 package.123 For
these calculations, the nonlocal MPW1PW91 corrections were added
to the exchange and correlation energies, respectively. Calculations
relying on the Tamm−Dancoff approximation (TDA) of the full TD-
DFT equations were used to evaluate the lifetimes of the excited states
and the oscillator strengths.124
The quantum yields of singlet oxygen formation were determined
relative to perinaphthanone in toluene (ΦΔ = 1.0) using a method
described by Nonell and Braslavsky.110 The samples and the reference
compounds were analyzed in the same solvent because of the strong
solvent dependence of the radiative and nonradiative rate constants for
deactivation of the triplet states. The singlet oxygen emission was
detected at 1269 nm from solutions in a 1 cm path length quartz
cuvette after excitation at 355 nm using a frequency-tripled Q-switched
Nd:YAG laser (Spectra Physics, Quanta Ray GCR-150-10) with a 10
Hz repetition rate. The emission was collected at 90° to the excitation
beam using a liquid-nitrogen-cooled germanium photodiode (North
Coast E0-817P) after passing through an interference filter centered at
1270 nm. The photodiode output was amplified and AC-coupled to a
digital oscilloscope, which digitized and averaged the transients. The
averaged data were then analyzed using the Microsoft Excel package.
The fluorescence lifetimes were measured via time-correlated single-
photon counting (TCSPC) using a 396 nm pulsed laser diode. The
fluorescence emission was collected at right angles to the excitation
source; the emission wavelength was selected using a monochromator,
and the emission was detected using a single-photon avalanche diode
(SPAD). The instrument response function (IRF) was measured using
a dilute LUDOX suspension as the scattering sample with the
monochromator set at the emission wavelength of the laser, giving an
IRF of 200 or 100 ps at 396 or 300 nm, respectively. The resulting
intensity decay is a convolution of the fluorescence decay with the IRF,
and iterative reconvolution of the IRF with the decay function and
nonlinear least-squares analysis were used to analyze the convoluted
data. Low-temperature emission spectra were recorded in an EPA glass
at 77 K in a liquid-nitrogen-cooled Oxford Instruments OptistatDN
cryostat. The same cryostat was used for the temperature-dependent
lifetime measurements in degassed toluene.
Synthesis of mer,cis-[Tris(triethylphosphine)(chlorido)-2,5-
bis(p-carbomethoxyphenylethynyl)cyclohexa[1′,2′:3,4]-
iridacyclopenta-2,4-diene] (3). To a stirred solution of [IrCl-
(PEt3)3] (50 mg, 0.086 mmol) in THF (5 mL) was added 1,12-bis(p-
carbomethoxyphenyl)dodeca-1,3,9,11-tetrayne (36 mg, 0.086 mmol)
dissolved in THF (5 mL) dropwise, and the reaction mixture was
stirred overnight at room temperature. The volatiles were removed in
vacuo. An NMR spectroscopic investigation showed complete
conversion to 3. The crude product was extracted with Et2O and
recrystallized several times from CHCl3/hexane to obtain high-purity
samples for photophysical studies and single crystals suitable for X-ray
diffraction of 3. Yield following purification: 61 mg (0.061 mmol,
71%).
Spectroscopic and analytical data: 1H NMR (400 MHz, benzene-d6,
25 °C, TMS): δ 8.11−7.86 (AA′XX′, 4H; CHarom), 8.08−7.38
(AA′XX′, 4H; CHarom), 3.44 (s, 3H; CO2CH3), 3.40 (s, 3H;
CO2CH3), 3.03 (br m, 2H; CH2), 2.85 (br m, 2H; CH2), 2.28 (m,
6H; PCH2), 2.10−1.80 (m, 12H; PCH2), 1.65 (br m, 4H; CH2), 1.08
(dt, 3J(H,P) = 12 Hz, 3J(H,H) = 8 Hz, 9H; PEt3), 0.93 (vq, J = 7 Hz,
2
18H; PEt3). 31P{1H} NMR (162 MHz) δ −31.9 (d, J(P,P) = 19 Hz,
2P; PEt3), −34.7 (t, 2J(P,P) = 19 Hz, 1P; PEt3). Anal. Calcd for
C46H67ClIrO4P3: C, 55.00; H, 6.72. Found: C, 55.27; H, 6.68%. MS
MALDI m/z = 1004 [M+].
Single-crystal X-ray diffraction data were collected on a Bruker
three-circle diffractometer with a SMART 6000 CCD area detector
using graphite-monochromatized Mo Kα radiation (λ = 0.71073 Å);
computations used SHELXTL-2013/2111 and OLEX2112 software.
Crystal data: C46H67ClIrO4P3, M = 1004.56, T = 120 K, triclinic, space
ASSOCIATED CONTENT
* Supporting Information
Temperature-dependent emission lifetime data for 1a; further
details of the DFT and TD-DFT results for 1a-H, 1c-H, 2a, and
3-Me, including Cartesian coordinates; X-ray crystallographic
■
S
group P1 (No. 2), a = 10.6570(13) Å, b = 12.9436(16) Å, c =
̅
7066
dx.doi.org/10.1021/ic501115k | Inorg. Chem. 2014, 53, 7055−7069

Inorganic Chemistry
Article
Domen, K.; DuBois, D. L.; Eckert, J.; Fujita, E.; Gibson, D. H.;
Goddard, W. A.; Goodman, D. W.; Keller, J.; Kubas, G. J.; Kung, H.
H.; Lyons, J. E.; Manzer, L. E.; Marks, T. J.; Morokuma, K.; Nicholas,
K. M.; Periana, R.; Que, L.; Rostrup-Nielson, J.; Sachtler, W. M. H.;
Schmidt, L. D.; Sen, A.; Somorjai, G. A.; Stair, P. C.; Stults, B. R.;
Tumas, W. Chem. Rev. 2001, 101, 953.
data for 3 in CIF format; and movies (AVI) of the calculated
main vibrational modes involved in the vibronic couplings in
absorption and in emission of 1a-H. This material is available
free of charge via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Authors
■
(17) Balzani, V.; Juris, A.; Venturi, M.; Campagna, S.; Serroni, S.
Chem. Rev. 1996, 96, 759.
*E-mail: andreas.steffen@uni-wuerzburg.de (A.S.).
*E-mail: kcostuas@univ-rennes1.fr (K.C.).
*E-mail: halet@univ-rennes1.fr (J.-F.H.).
*E-mail: todd.marder@uni-wuerzburg.de (T.B.M.).
(18) Hawecker, J.; Lehn, J.-M.; Ziessel, R. Helv. Chim. Acta 1986, 69,
1990.
(19) Montalti, M.; Credi, A.; Prodi, L.; Gandolfi, M. T. Handbook of
Photochemistry, 3rd ed.; CRC Press: Boca Raton, FL, 2006.
(20) Turro, N. J.; Ramamurthy, V.; Scaiano, J. C. Photochem.
Photobiol. 2012, 88, 1033.
Notes
The authors declare no competing financial interest.
(21) Saigusa, H.; Azumi, T.; Sumitani, M.; Yoshihara, K. J. Chem.
ACKNOWLEDGMENTS
Phys. 1980, 72, 1713.
(22) Bomben, P. G.; Robson, K. C. D.; Koivisto, B. D.; Berlinguette,
C. P. Coord. Chem. Rev. 2012, 256, 1438.
(23) Campagna, S.; Puntoriero, F.; Nastasi, F.; Bergamini, G.;
Balzani, V. Top. Curr. Chem. 2007, 280, 117.
(24) Hedley, G. J.; Ruseckas, A.; Samuel, I. D. W. J. Phys. Chem. A
2009, 113, 2.
(25) Hsu, C. C.; Lin, C. C.; Chou, P. T.; Lai, C. H.; Hsu, C. W.; Lin,
C. H.; Chi, Y. J. Am. Chem. Soc. 2012, 134, 7715.
(26) Kirgan, R. A.; Sullivan, B. P.; Rillema, D. P. Top. Curr. Chem.
2007, 281, 45.
■
We acknowledge the very helpful comments of one of the
anonymous reviewers, which further improved our work. A.S.
thanks the DAAD and the EU (Marie-Curie FP-7) for
postdoctoral fellowships. K.C., J.-F.H., and A. Boucekkine are
grateful to the GENCI-CINES and GENCI-IDRIS for HPC
resources (Grants 2012-80649 and 2013-80649). A.C.-E.
thanks the European Research Council (ERC, Marches
278845) for his postdoctoral grant. D.J. acknowledges the
́
European Research Council (ERC) and the Region des Pays de
la Loire for financial support in the framework of a Starting
(27) Flamigni, L.; Barbieri, A.; Sabatini, C.; Ventura, B.; Barigelletti,
Grant (Marches 278845) and a Recrutement sur Poste
Strategique, respectively. The vibronic calculations used
́
F. Top. Curr. Chem. 2007, 281, 143.
(28) Kaes, C.; Katz, A.; Hosseini, M. W. Chem. Rev. 2000, 100, 3553.
(29) You, Y.; Park, S. Y. Dalton Trans. 2009, 1267.
(30) Buckner, S. W.; Fischer, M. J.; Jelliss, P. A.; Luo, R. S.; Minteer,
S. D.; Rath, N. P.; Siemiarczuk, A. Inorg. Chem. 2006, 45, 7339.
(31) Cannizzo, A.; van Mourik, F.; Gawelda, W.; Zgrablic, G.;
Bressler, C.; Chergui, M. Angew. Chem., Int. Ed. 2006, 45, 3174.
(32) Cheng, Y. M.; Yeh, Y. S.; Ho, M. L.; Chou, P. T.; Chen, P. S.;
Chi, Y. Inorg. Chem. 2005, 44, 4594.
resources of the GENCI-CINES/IDRIS (Grant
c2012085117), Centre de Calcul Intensif des Pays de Loire
(CCIPL), and a local Troy cluster. This work was supported by
the Bavarian State Ministry of Science, Research, and the Arts
for the Collaborative Research Network “Solar Technologies
Go Hybrid”. Part of this work was conducted within the scope
of the CNRS Associated European Laboratory “Molecular
Materials and Catalysis (MMC)” involving the Department of
Chemistry of Durham University and the Institut des Sciences
Chimiques de Rennes of the University of Rennes 1.
(33) Sexton, D. A.; Ford, P. C.; Magde, D. J. Phys. Chem. 1983, 87,
197.
(34) Cannizzo, A.; Blanco-Rodriguez, A. M.; El Nahhas, A.; Sebera,
J.; Zalis, S.; Vlcek, A., Jr.; Chergui, M. J. Am. Chem. Soc. 2008, 130,
8967.
(35) Gawelda, W.; Cannizzo, A.; Pham, V.-T.; van Mourik, F.;
Bressler, C.; Chergui, M. J. Am. Chem. Soc. 2007, 129, 8199.
(36) Monat, J. E.; McCusker, J. K. J. Am. Chem. Soc. 2000, 122, 4092.
(37) Chou, P. T.; Chi, Y.; Chung, M. W.; Lin, C. C. Coord. Chem.
Rev. 2011, 255, 2653.
(38) Chergui, M. Dalton Trans. 2012, 41, 13022.
(39) Abedin-Siddique, Z.; Ohno, T.; Nozaki, K. Inorg. Chem. 2004,
43, 663.
(40) Abedin Siddique, Z.; Yamamoto, Y.; Ohno, T.; Nozaki, K. Inorg.
Chem. 2003, 42, 6366.
(41) Ma, C. S.; Chan, C. T. L.; Kwok, W. M.; Che, C. M. Chem. Sci.
2012, 3, 1883.
(42) Lu, W.; Kwok, W. M.; Ma, C. S.; Chan, C. T. L.; Zhu, M. X.;
Che, C. M. J. Am. Chem. Soc. 2011, 133, 14120.
(43) Tong, G. S. M.; Chow, P. K.; Che, C. M. Angew. Chem., Int. Ed.
2010, 49, 9206.
REFERENCES
■
(1) Yersin, H.; Rausch, A. F.; Czerwieniec, R.; Hofbeck, T.; Fischer,
T. Coord. Chem. Rev. 2011, 255, 2622.
(2) Yersin, H. Highly Efficient OLEDs with Phosphorescent Materials;
Wiley-VCH: Weinheim, Germany, 2008.
(3) Holder, E.; Langeveld, B. M. W.; Schubert, U. S. Adv. Mater.
2005, 17, 1109.
(4) Yersin, H. Top. Curr. Chem. 2004, 241, 1.
(5) Lamansky, S.; Djurovich, P.; Murphy, D.; Abdel-Razzaq, F.; Lee,
H. E.; Adachi, C.; Burrows, P. E.; Forrest, S. R.; Thompson, M. E. J.
Am. Chem. Soc. 2001, 123, 4304.
(6) Kalyanasundaram, K.; Gratzel, M. Coord. Chem. Rev. 1998, 177,
̈
347.
(7) Wong, W. Y.; Ho, C. L. Acc. Chem. Res. 2010, 43, 1246.
(8) Gratzel, M. Acc. Chem. Res. 2009, 42, 1788.
̈
(9) Robertson, N. Angew. Chem., Int. Ed. 2006, 45, 2338.
(10) Polo, A. S.; Itokazu, M. K.; Iha, N. Y. M. Coord. Chem. Rev.
2004, 248, 1343.
(44) Lu, W.; Zhu, N. Y.; Che, C. M. J. Organomet. Chem. 2003, 670,
11.
(11) Gratzel, M. J. Photochem. Photobiol., C 2003, 4, 145.
̈
(45) Chao, H. Y.; Lu, W.; Li, Y. Q.; Chan, M. C. W.; Che, C. M.;
Cheung, K. K.; Zhu, N. Y. J. Am. Chem. Soc. 2002, 124, 14696.
(46) Steffen, A.; Tay, M. G.; Batsanov, A. S.; Howard, J. A. K.; Beeby,
A.; Vuong, K. Q.; Sun, X. Z.; George, M. W.; Marder, T. B. Angew.
Chem., Int. Ed. 2010, 49, 2349.
(47) Kozhevnikov, D. N.; Kozhevnikov, V. N.; Shafikov, M. Z.;
Prokhorov, A. M.; Bruce, D. W.; Williams, J. A. G. Inorg. Chem. 2011,
50, 3804.
(12) Hagfeldt, A.; Gratzel, M. Acc. Chem. Res. 2000, 33, 269.
̈
(13) Crabtree, R. H. Organometallics 2011, 30, 17.
(14) Bonnemann, H.; Khelashvili, G. Appl. Organomet. Chem. 2010,
̈
24, 257.
(15) Morris, A. J.; Meyer, G. J.; Fujita, E. Acc. Chem. Res. 2009, 42,
1983.
(16) Arakawa, H.; Aresta, M.; Armor, J. N.; Barteau, M. A.; Beckman,
E. J.; Bell, A. T.; Bercaw, J. E.; Creutz, C.; Dinjus, E.; Dixon, D. A.;
7067
dx.doi.org/10.1021/ic501115k | Inorg. Chem. 2014, 53, 7055−7069

Inorganic Chemistry
Article
(48) Ramlot, D.; Leen Volker, M. R.; Ovaere, M.; Beljonne, D.;
Dehaen, W.; Van Meervelt, L.; Moucheron, C.; Kirsch-De Mesmaeker,
A. Eur. J. Inorg. Chem. 2013, 2031.
(49) Wilson, J. J.; Lippard, S. J. Inorg. Chim. Acta 2012, 389, 77.
(50) Prusakova, V.; McCusker, C. E.; Castellano, F. N. Inorg. Chem.
2012, 51, 8589.
(51) Zhao, G. J.; Yu, F. B.; Zhang, M. X.; Northrop, B. H.; Yang, H.
B.; Han, K. L.; Stang, P. J. J. Phys. Chem. A 2011, 115, 6390.
(52) Zhang, B. G.; Li, Y. J.; Sun, W. F. Eur. J. Inorg. Chem. 2011,
4964.
(53) Wu, W. T.; Wu, W. H.; Ji, S. M.; Guo, H. M.; Zhao, J. Z. Dalton
Trans. 2011, 40, 5953.
(54) Wu, W. H.; Sun, J. F.; Ji, S. M.; Wu, W. T.; Zhao, J. Z.; Guo, H.
M. Dalton Trans. 2011, 40, 11550.
(55) Nguyen, M. H.; Yip, J. H. K. Organometallics 2011, 30, 6383.
(56) Mancilha, F. S.; Barloy, L.; Rodembusch, F. S.; Dupont, J.;
Pfeffer, M. Dalton Trans. 2011, 40, 10535.
(81) Lo, K. K. W.; Li, C. K.; Lau, K. W.; Zhu, N. Y. Dalton Trans.
2003, 4682.
(82) Ghizdavu, L.; Lentzen, O.; Schumm, S.; Brodkorb, A.;
Moucheron, C.; Kirsch-De Mesmaeker, A. Inorg. Chem. 2003, 42,
1935.
(83) Humbs, W.; Yersin, H. Inorg. Chem. 1996, 35, 2220.
(84) Calogero, G.; Giuffrida, G.; Serroni, S.; Ricevuto, V.; Campagna,
S. Inorg. Chem. 1995, 34, 541.
(85) Colombo, M. G.; Brunold, T. C.; Riedener, T.; Gudel, H. U.;
̈
Fortsch, M.; Burgi, H.-B. Inorg. Chem. 1994, 33, 545.
̈
̈
(86) Miki, H.; Shimada, M.; Azumi, T.; Brozik, J. A.; Crosby, G. A. J.
Phys. Chem. 1993, 97, 11175.
(87) Zilian, A.; Gudel, H. U. Inorg. Chem. 1992, 31, 830.
(88) Milder, S. J.; Brunschwig, B. S. J. Phys. Chem. 1992, 96, 2189.
(89) van Diemen, J. H.; Haasnoot, J. G.; Hage, R.; Muller, E.;
̈
Reedijk, J. Inorg. Chim. Acta 1991, 181, 245.
(90) Barigelletti, F.; Sandrini, D.; Maestri, M.; Balzani, V.; von
Zelewsky, A.; Chassot, L.; Jolliet, P.; Maeder, U. Inorg. Chem. 1988, 27,
3644.
(91) Maestri, M.; Sandrini, D.; Balzani, V.; Maeder, U.; von
Zelewsky, A. Inorg. Chem. 1987, 26, 1323.
(92) Bergkamp, M. A.; Watts, R. J.; Ford, P. C.; Brannon, J.; Magde,
D. Chem. Phys. Lett. 1978, 59, 125.
(93) Petersen, J. D.; Watts, R. J.; Ford, P. C. J. Am. Chem. Soc. 1976,
(57) Lentijo, S.; Miguel, J. A.; Espinet, P. Organometallics 2011, 30,
1059.
(58) Nguyen, M. H.; Yip, J. H. K. Organometallics 2010, 29, 2422.
(59) Lentijo, S.; Miguel, J. A.; Espinet, P. Inorg. Chem. 2010, 49,
9169.
(60) Hudson, Z. M.; Zhao, S. B.; Wang, R. Y.; Wang, S. N. Chem.
Eur. J. 2009, 15, 6131.
98, 3188.
(61) Lanoe, P.-H.; Fillaut, J.-L.; Toupet, L.; Williams, J. A. G.; Le
̈
(94) Halper, W.; Dearmond, M. K. Chem. Phys. Lett. 1974, 24, 114.
(95) Indelli, M. T.; Chiorboli, C.; Scandola, F. Top. Curr. Chem.
2007, 280, 215.
(96) Dreeskamp, H.; Koch, E.; Zander, M. Chem. Phys. Lett. 1975, 31,
251.
(97) Dreeskamp, H.; Koch, E.; Zander, M. Ber. Bunsen-Ges. Phys.
Chem. 1974, 78, 1328.
(98) Bennett, R. G.; McCartin, P. J. J. Chem. Phys. 1966, 44, 1969.
(99) Ware, W. R.; Baldwin, B. A. J. Chem. Phys. 1965, 43, 1194.
(100) McGlynn, S. P.; Azumi, T.; Kinoshita, M. Molecular
Spectroscopy of the Triplet State; Prentice Hall: Englewood Cliffs, NJ,
1969.
(101) Siddle, J. S.; Ward, R. M.; Collings, J. C.; Rutter, S. R.; Porres,
L.; Applegarth, L.; Beeby, A.; Batsanov, A. S.; Thompson, A. L.;
Howard, J. A. K.; Boucekkine, A.; Costuas, K.; Halet, J.-F.; Marder, T.
B. New J. Chem. 2007, 31, 841.
Bozec, H.; Guerchais, V. Chem. Commun. 2008, 4333.
(62) Khan, M. S.; Al-Mandhary, M. R. A.; Al-Suti, M. K.; Hisahm, A.
K.; Raithby, P. R.; Ahrens, B.; Mahon, M. F.; Male, L.; Marseglia, E. A.;
Tedesco, E.; Friend, R. H.; Kohler, A.; Feeder, N.; Teat, S. J. J. Chem.
̈
Soc., Dalton Trans. 2002, 1358.
(63) Wilson, J. S.; Chawdhury, N.; Al-Mandhary, M. R. A.; Younus,
M.; Khan, M. S.; Raithby, P. R.; Kohler, A.; Friend, R. H. J. Am. Chem.
̈
Soc. 2001, 123, 9412.
(64) Carano, M.; Cicogna, F.; Houben, J. L.; Ingrosso, G.; Marchetti,
F.; Mottier, L.; Paolucci, F.; Pinzino, C.; Roffia, S. Inorg. Chem. 2002,
41, 3396.
(65) Cicogna, F.; Colonna, M.; Houben, J. L.; Ingrosso, G.;
Marchetti, F. J. Organomet. Chem. 2000, 593, 251.
(66) Chen, Y. L.; Li, S. W.; Chi, Y.; Cheng, Y. M.; Pu, S. C.; Yeh, Y.
S.; Chou, P. T. ChemPhysChem 2005, 6, 2012.
(67) Nguyen, M. H.; Wong, C. Y.; Yip, J. H. K. Organometallics 2013,
32, 1620.
(102) Baumgartner, T. J. Inorg. Organomet. Polym. 2005, 15, 389.
(103) Hissler, M.; Dyer, P. W.; Rea
1.
́
u, R. Coord. Chem. Rev. 2003, 244,
(68) Rourke, J. P.; Batsanov, A. S.; Howard, J. A. K.; Marder, T. B.
Chem. Commun. 2001, 2626.
(69) Steffen, A.; Ward, R. M.; Tay, M. G.; Edkins, R. M.; Seeler, F.;
(104) Boydston, A. J.; Yin, Y.; Pagenkopf, B. L. J. Am. Chem. Soc.
2004, 126, 3724.
(105) Yamaguchi, S.; Tamao, K. J. Organomet. Chem. 2002, 653, 223.
(106) Steffen, A.; Ward, R. M.; Jones, W. D.; Marder, T. B. Coord.
Chem. Rev. 2010, 254, 1950.
van Leeuwen, M.; Palsson, L.-O.; Beeby, A.; Batsanov, A. S.; Howard,
̊
J. A. K.; Marder, T. B. Chem.Eur. J. 2014, 20, 3652.
(70) Adachi, C.; Baldo, M. A.; Thompson, M. E.; Forrest, S. R. J.
Appl. Phys. 2001, 90, 5048.
(107) Matano, Y.; Nakashima, M.; Imahori, H. Angew. Chem., Int. Ed.
(71) Crestani, M. G.; Steffen, A.; Kenwright, A. M.; Batsanov, A. S.;
Howard, J. A. K.; Marder, T. B. Organometallics 2009, 28, 2904.
(72) Pommerehne, J.; Vestweber, H.; Guss, W.; Mahrt, R. F.; Bassler,
H.; Porsch, M.; Daub, J. Adv. Mater. 1995, 7, 551.
(73) Dalton, D. M.; Rappe, A. K.; Rovis, T. Chem. Sci. 2013, 4, 2062.
(74) Barthes, C.; Lepetit, C.; Canac, Y.; Duhayon, C.; Zargarian, D.;
Chauvin, R. Inorg. Chem. 2013, 52, 48.
2009, 48, 4002.
(108) Hay, C.; Hissler, M.; Fischmeister, C.; Rault-Berthelot, J.;
Toupet, L.; Nyulas
́
zi, L.; Rea
́
u, R. Chem.Eur. J. 2001, 7, 4222.
(109) Porres, L.; Holland, A.; Palsson, L.-O.; Monkman, A. P.; Kemp,
̀
̊
C.; Beeby, A. J. Fluoresc. 2006, 16, 267.
(110) Nonell, S.; Braslavsky, S. L. Methods Enzymol. 2000, 319, 37.
(111) Sheldrick, G. M. Acta Crystallogr. 2008, A64, 112.
(112) Dolomanov, O. V.; Bourhis, L. J.; Gildea, R. J.; Howard, J. A.
K.; Puschmann, H. J. Appl. Crystallogr. 2009, 42, 339.
(75) Rausch, A. F.; Homeier, H. H. H.; Yersin, H. Top. Organomet.
Chem. 2010, 29, 193.
(76) Charaf-Eddin, A.; Planchat, A.; Mennucci, B.; Adamo, C.;
Jacquemin, D. J. Chem. Theory Comput. 2013, 9, 2749.
(77) Jacquemin, D.; Planchat, A.; Adamo, C.; Mennucci, B. J. Chem.
Theory Comput. 2012, 8, 2359.
(78) Santoro, F.; Lami, A.; Improta, R.; Barone, V. J. Chem. Phys.
2007, 126, No. 184102.
(79) Santoro, F.; Improta, R.; Lami, A.; Bloino, J.; Barone, V. J. Chem.
Phys. 2007, 126, No. 084509.
(80) Yu, L.; Srinivas, G. N.; Schwartz, M. J. Mol. Struct.:
(113) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.;
Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci,
B.; Petersson, G. A.; Nakatsuji, H.; Caricato, M.; Li, X.; Hratchian, H.
P.; Izmaylov, A. F.; Bloino, J.; Zheng, G.; Sonnenberg, J. L.; Hada, M.;
Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima,
T.; Honda, Y.; Kitao, O.; Nakai, H.; Vreven, T.; Montgomery, J. A., Jr.;
Peralta, J. E.; Ogliaro, F.; Bearpark, M.; Heyd, J. J.; Brothers, E.; Kudin,
K. N.; Staroverov, V. N.; Kobayashi, R.; Normand, J.; Raghavachari, K.;
Rendell, A.; Burant, J. C.; Iyengar, S. S.; Tomasi, J.; Cossi, M.; Rega,
N.; Millam, N. J.; Klene, M.; Knox, J. E.; Cross, J. B.; Bakken, V.;
THEOCHEM 2003, 625, 215.
7068
dx.doi.org/10.1021/ic501115k | Inorg. Chem. 2014, 53, 7055−7069

Inorganic Chemistry
Article
Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.;
Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Martin, R. L.;
Morokuma, K.; Zakrzewski, V. G.; Voth, G. A.; Salvador, P.;
̈
Dannenberg, J. J.; Dapprich, S.; Daniels, A. D.; Farkas, O.;
Foresman, J. B.; Ortiz, J. V.; Cioslowski, J.; Fox, D. J. Gaussian 09,
revisions A.02 and C.01; Gaussian, Inc.: Wallingford, CT, 2009.
(114) Adamo, C.; Barone, V. J. Chem. Phys. 1998, 108, 664.
(115) Perdew, J. P.; Burke, K.; Wang, Y. Phys. Rev. B 1996, 54, 16533.
(116) Perdew, J. P.; Burke, K.; Ernzerhof, M. Phys. Rev. Lett. 1996,
77, 3865.
(117) Hay, P. J.; Wadt, W. R. J. Chem. Phys. 1985, 82, 270.
(118) Varetto, U.; Giuffreda, M. G.; Jang, Y. Abstr. Pap.Am. Chem.
Soc. 2009, 238, COMP 151.
(119) Peng, C.; Schlegel, H. B. Isr. J. Chem. 1993, 33, 449.
(120) Zhao, Y.; Truhlar, D. G. Theor. Chem. Acc. 2008, 120, 215.
(121) Isegawa, M.; Peverati, R.; Truhlar, D. G. J. Chem. Phys. 2012,
137, No. 244104.
(122) Tomasi, J.; Mennucci, B.; Cammi, R. Chem. Rev. 2005, 105,
2999.
(123) te Velde, G.; Bickelhaupt, F. M.; Baerends, E. J.; Guerra, C. F.;
Van Gisbergen, S. J. A.; Snijders, J. G.; Ziegler, T. J. Comput. Chem.
2001, 22, 931.
(124) Hirata, S.; Head-Gordon, M. Chem. Phys. Lett. 1999, 314, 291.
7069
dx.doi.org/10.1021/ic501115k | Inorg. Chem. 2014, 53, 7055−7069