Electron Transfer Across Modular Oligo-p-phenylene Bridges in Ru(bpy)(2)(bpy-ph(n)-DQ)(4+) (n=1-5) Dyads. Unusual Effects of Bridge Elongation

Article (PDF Available)inThe Journal of Physical Chemistry A 116(1):119-31 · November 2011with 186 Reads
DOI: 10.1021/jp209858p · Source: PubMed
Abstract
A series of dyads of general formula Ru(bpy)(2)(bpy-ph(n)-DQ)(4+) (n = 1-5), based on a Ru(II) polypyridine unit as photoexcitable donor, a set of oligo-p-phenylene bridges with 1-5 modular units, and a cyclo-diquaternarized 2,2'-bipyridine (DQ(2+)) as electron acceptor unit, have been synthesized. Their spectroscopic and photophysical properties have been investigated in CH(3)CN and CH(2)Cl(2) by time-resolved emission and absorption spectroscopy in the nanosecond and picosecond time scale. The experimental study has also been complemented with a computational investigation carried out on the whole series of dyads. The absorption spectra of the dyads show new spectroscopic transitions, in addition to those characteristic of the donor, bridge, and acceptor fragments. DFT calculations suggest the assignment of such bands as bridge-to-acceptor (π ph(n)) → (π* DQ) charge-transfer transitions. This assignment is consistent with the solvatochromic and spectroelectrochemical behavior of the new bands. For all the dyads at room temperature in fluid solution, the typical (3)MLCT luminescence of the Ru(II) polypyridine unit is strongly (>90%) quenched, supporting the occurrence of an efficient intramolecular photoinduced electron transfer. The study has revealed, however, that the photophysical mechanism is actually more complex than presumed on the basis of a simple photoinduced electron-transfer scheme. For n = 1, very fast (few picoseconds) photoinduced electron transfer from the MLCT state localized on the substituted bpy ligand to the DQ unit has been observed, followed by slower interligand hopping and charge recombination. For n = 2-5, MLCT excited-state quenching takes place without transient detection of charge-separated product, indicating that charge recombination is faster than charge separation. This behavior can be rationalized in terms of the superexchange couplings expected through this type of bridges for the two processes. The kinetics of MLCT quenching in the dyads with n = 1-5 does not follow the usual exponential falloff with bridge length: after a regular decrease for n = 1-3, the rate constants become almost insensitive to bridge length for n = 3-5. The rationale of this uncommon behavior, as suggested by DFT calculations, lies in a switch in the MLCT quenching mechanism with increasing bridge length, from oxidative quenching by the DQ acceptor to reductive quenching by the bridge.
Published: November 22, 2011
r2011 American Chemical Society 119 dx.doi.org/10.1021/jp209858p |J. Phys. Chem. A 2012, 116, 119131
ARTICLE
pubs.acs.org/JPCA
Electron Transfer Across Modular Oligo-p-phenylene Bridges in
Ru(bpy)
2
(bpyph
n
DQ)
4+
(n=15) Dyads. Unusual Effects of
Bridge Elongation
Maria Teresa Indelli,
,
Michele Orlandi,
,1
Claudio Chiorboli,
§
Marcella Ravaglia,
,
||
Franco Scandola,*
,,
Frederic Lafolet,
3
Steve Welter,
4
and Luisa De Cola*
,^
Dipartimento di Chimica, Universita di Ferrara, and Centro Interuniversitario per la Conversione Chimica dellEnergia Solare
(SolarChem), sezione di Ferrara, 44100 Ferrara, Italy
INSTM, Sezione di Ferrara, 44100 Ferrara, Italy
§
ISOF-CNR, Sezione di Ferrara, 44100 Ferrara, Italy
^
Westfalische Wilhelms-Universitat Munster, Physikalisches Institut, Mendelstrasse 7, 48149 Munster, Germany, and
Center for Nanotechnology (CeNTech), Heisenbergstrasse 11, 48149 Munster, Germany
b
SSupporting Information
INTRODUCTION
Driven by potential applications in the elds of articial photo-
synthesis,
1
molecular electronics, and photonics,
2
the study of
charge transport from donor to acceptor units across molecular
bridges is a subject of continuing interest.
3
In this context, the
bridges are often designated as molecular wires, despite the fact
that their behavior is generally quite far from that of a conven-
tional ohmic-type conductor. In fact, for many experimental
cases involving saturated or partially unsaturated organic bridges,
the energy levels of the wires investigated are far apart from those
of the donor and acceptor sites, so that the electron has to tunnel
in a single coherent step from donor to acceptor, without being
localized at any time on the wire. Nevertheless, the nature of
the bridge plays an important role in determining the electron-
transfer rates.
Electron transfer in a donorbridgeacceptor (DBA)
system (Figure 1) can be described by standard nonadiabatic
electron transfer theory,
4
where the transfer probability is
Received: October 13, 2011
Revised: November 21, 2011
ABSTRACT: A series of dyads of general formula Ru(bpy)
2
(bpyph
n
DQ)
4+
(n=15), based on a Ru(II) polypyridine unit as photoexcitable
donor, a set of oligo-p-phenylene bridges with 15 modular units, and a
cyclo-diquaternarized 2,20-bipyridine (DQ
2+
) as electron acceptor unit,
have been synthesized. Their spectroscopic and photophysical properties
have been investigated in CH
3
CN and CH
2
Cl
2
by time-resolved emission
and absorption spectroscopy in the nanosecond and picosecond time scale.
The experimental study has also been complemented with a computational
investigation carried out on the whole series of dyads. The absorption
spectra of the dyads show new spectroscopic transitions, in addition to those
characteristic of the donor, bridge, and acceptor fragments. DFT calculations
suggest the assignment of such bands as bridge-to-acceptor (πph
n
)f
(π*DQ)charge-transfertransitions. This assignment is consistent with the
solvatochromic and spectroelectrochemical behavior of the new bands. For all the dyads at room temperature in uid solution, the typical
3
MLCT luminescence of the Ru(II) polypyridine unit is strongly (>90%) quenched, supporting the occurrence of an ecient intramolecular
photoinduced electron transfer. The study has revealed, however, that the photophysical mechanism is actually more complex than presumed
on the basis of a simple photoinduced electron-transfer scheme. For n= 1, very fast (few picoseconds) photoinduced electron transfer from the
MLCT state localized on the substituted bpy ligand to the DQ unit has been observed, followed by slower interligand hopping and charge
recombination. For n=25, MLCT excited-state quenching takes place without transient detection of charge-separated product, indicating
that charge recombination is faster than charge separation. This behavior can be rationalized in terms of the superexchange couplings expected
through this type of bridges for the two processes. The kinetics of MLCT quenching in the dyads with n=15 does not follow the usual
exponential fallowith bridge length: after a regular decrease for n=13,therateconstantsbecomealmostinsensitivetobridgelengthforn=
35. The rationale of this uncommon behavior, as suggested by DFT calculations, lies in a switch in the MLCT quenching mechanism with
increasing bridge length, from oxidative quenching by the DQ acceptor to reductive quenching by the bridge.
120 dx.doi.org/10.1021/jp209858p |J. Phys. Chem. A 2012, 116, 119–131
The Journal of Physical Chemistry A ARTICLE
given by
k¼4π
hHif 2FCWD ð1Þ
where H
if
is the electronic matrix element between the initial
and nal states of the process (donoracceptor electronic
coupling) and FCWD is a nuclear factor (FranckCondon
weighted density of states) accounting for the combined eects
of driving force and reorganizational energy. The role of the
bridge as mediator of the donoracceptor electronic coupling can
be conveniently described in terms of superexchange,
3,5
as second-
order interaction of the initial and nal states via high-energy
charge transfer virtualstates involving the bridge.
In this formalism, the electronic coupling H
if
between the
initial (DBA) and nal (D
+
BA
) states is given by
Hif ¼HieHfe
ΔEe
þHihHfh
ΔEh
ð2Þ
where H
ie
,H
fe
,H
ih
,H
fh
are the couplings between the initial/nal
states and the bridge-localized electron-transfer (D
+
B
A)
and hole-transfer (DB
+
A
) virtual states, and ΔE
e
and ΔE
h
are the energy dierences between the virtual states and the
initial/nal levels (taken at the transition-state nuclear geometry,
where the initial and nal state have the same energy). The same
model can be applied to modular bridges involving a number (n)
of repeating units. In such a case, each term in eq 2 takes the form
Hif ¼HiHf
ΔE
Hmn
ΔE

n1
ð3Þ
where H
i
and H
f
are the couplings between the initial/nal states
and the virtual states localized on the bridge subunits adjacent to
the donor and acceptor, while H
mn
is the coupling between
adjacent subunits of the bridge, and ΔEis the energy dierence
between the virtual states of the bridge and the initial/nal state
(at the transition-state nuclear geometry), ideally assumed to
be independent of the bridge length. This translates into an
exponential dependence on the number of modular units in the
bridge, i.e., on donoracceptor distance r
AB
Hif ¼Hif ð0Þexp β
2ðrAB r0Þ
 ð4Þ
In eq 4, r
0
and H
if
(0) represent the donoracceptor distance
and electronic coupling for a single-module bridge, and β=
(2/r
m
) ln(ΔE/H
mn
) (where r
m
is the length of an individual
module) is the attenuation factor of the electron-transfer rate
with distance. The βfactor depends on the intermodule cou-
plings within the bridge and is often considered as a parameter
that quanties the intrinsic ability of the bridge to promote long-
distance electron transfer (the lower β, the higher the conductivity
of the molecular wire). It should be noticed, however, that βis
not entirely bridge-specic, as it also depends on the energy gap
between the bridge-localized virtual states and the initial/nal
state,
3d
and thus also on the donor and acceptor units involved.
6
These superexchange models have been extensively tested with
studies of photoinduced electron transfer in donorbridgeacceptor
systems (dyads) where the process can be triggered by elec-
tronic excitation of either the donor or the acceptor. In particular,
the predicted exponential decay of the rates with distance has
been often veried in homogeneous series of dyads containing
modular organic bridges of variable length.
3
Exceptions to this general behavior have been observed with
highly conjugated organic bridges, when the bridge-localized
electron/hole-transfer states may become lower in energy than
the electronically excited state. In such cases, injection of elec-
trons (or holes) into the wire can occur and the process (charge
hoppingor 00transport00
3a
) has an inverse dependence on a
small power (12) of distance, more similar to conventional ohmic
conduction. Interesting examples of switch from superexchange
to electron injection
7
or hole injection
8
upon bridge elongation
have been reported by Wasielewski and co-workers.
Because of their modular nature and longitudinal rigidity, an
interesting class of molecular bridges is that of oligo-p-phenylenes.
Although the individual units have fully aromatic πsystems,
extensive delocalization along the oligomeric chain is prevented
by nonbonded interactions that force adjacent units in a twisted
geometry (ca. 40angle in the gas phase
9
). Onuchich and
Beratan
10
rst considered electron transfer through biphenyl
spacers from a theoretical viewpoint. In pioneering experimental
work by McLendon and co-workers
11
on bis-porphyrin systems
with oligo-p-phenylene spacers, a sevenfold attenuation in
photoinduced electron-transfer rate was obtained for each addi-
tional p-phenylene unit (β= 0.4 Å
1
).
11b
Extensive work has
recently been performed by Wasielewski and co-workers
8,12
on
oligo-p-phenylene bridged dyads where a perylenebisimide acts
as a photoexcited acceptor and a phenothiazine as the electron
donor quencher. Inorganic dyads with oligo-p-phenylene spacers
have also been investigated. The extensive series of dyads
developed by Sauvage and co-workers
13
and by De Cola and
co-workers,
14,15
involving polypyridyl complexes of Ru(II),
Os(II), and Ir(III) as donor/acceptor units, were designed for the
study of energy (rather than electron) transfer. Recently, we have
investigated bimetallic dyads involving Ru(II) polypyridine units as
photoexcited donor and a Rh(III) unit as electron acceptor.
16
Despite some limitations (small driving force, limited number of
modular spacers) the study of photoinduced electron transfer in
these dyads gave interesting information on the βattenuation
factor and on the eect of intermodule twist angle on rates. Wegner
has recently reviewed the eect of the donorbridgeacceptor
energetics on the distance dependence of electron transfer across
modular bridges, including oligo p-xylene bridges.
17
In this work, we set out to synthesize and study a series of dyads
(Scheme 1) based ona Ru(II) polypyridine unit as photoexcitable
donor, a set of oligo-p-phenylene bridges with 15 modular units,
and a cyclo-diquaternarized bipyridine (DQ
2+
) as electron accep-
tor unit. In this series, the long lifetime of the donor excited state
and the favorable driving force are appropriate, in principle, to
obtain ecient long-range photoinduced electron transfer. As it
Figure 1. Schematic representation of superexchange in a donor
bridgeacceptor system. The donoracceptor electronic coupling is
mediated by interaction of initial (DBA) and nal (D
+
BA
)
states with bridge-localized virtual states of electron- (D
+
B
A) and
hole-transfer (DB
+
A
) type.
121 dx.doi.org/10.1021/jp209858p |J. Phys. Chem. A 2012, 116, 119–131
The Journal of Physical Chemistry A ARTICLE
will be seen, the study has revealed that the behavior of this type of
bridges is actually much more complex than presumed on the basis
of the above simple superexchange mechanism. To help in the
interpretation of this intriguing behavior, a DFT computational
investigation has also been carried out on the series of dyads.
EXPERIMENTAL SECTION
Materials. For photophysical measurements spectroscopic
grade organic solvents (Merck Uvasol) were used without further
purification. Other chemicals were all of reagent grade quality.
Silica Gel (Merck) used in chromatographic purication was
200400 mesh.
Syntheses. The series of nonquaternarized Ru(bpy)
2
(bpyph
n
bpy)
2+
(n=15) complexes, used in this work as chemical precursors
and as model compounds, was available from previous work.
15,16
The syntheses of the dyads were carried out starting from the
appropriate Ru(bpy)
2
(bpyph
n
bpy)
2+
complex by cycloqua-
ternarization of the free bipyridyl group of the bridging ligand.
The same general procedure was followed for all the compounds
with minor changes concerning the volume of the reaction
mixture. A typical synthesis is described below for Ru(bpy)
2
-
(bpyph
2
DQ)
4+
. All the compounds were isolated as PF
6
salts and identied (see data below) by
1
H NMR (400 MHz,
CD
3
CN) and mass spectroscopy (electron spray ionization,
CH
3
CN matrix). The
1
H NMR spectrum of Ru(bpy)
2
(bpy
ph
2
DQ)
4+
is reported as an example in the Supporting Infor-
mation (Figure S1).
Ru(bpy)
2
(bpyph
2
DQ)
4+
.A 116 mg (0.25 mmol) amount
of [Ru(bpy)
2
(bpyph
n
bpy)](PF
6
)
2
was added to 25 mL of
hot 1,3 dibromopropane and stirred in argon atmosphere for
15 min. The reaction mixture was then heated for 72 h at ca. 110 C
under continuous stirring. After cooling at room temperature,
the solvent was evaporated and acetonitrile and NaPF
6
were
added. The hexafluorophosphate salt was precipitated by addition
of water and evaporation of acetonitrile. The crude product was
washed with water, dissolved in a minimum of acetonitrile, and
purified by column chromatography on silica gel using MeCN/
MeOH/H
2
O/saturated aqueous KNO
3
, 40/10/10/1 as eluent.
Elution was first performed to collect the unreacted [Ru(bpy)
2
-
(bpyph
2
bpy)](PF
6
)
2
. Then elution was continued until the
main fraction containing the desired product was completely
eluted, leaving a small fraction containing unidentified bypro-
ducts at the top of the column. The fraction containing the pure
product was concentrated to 2 mL, and solid NH
4
PF
6
was added.
The product was isolated by filtration, washed extensively with
water, and dried under vacuum. Purity was verified by TLC (silica
gel, MeCN/MeOH/H
2
O/saturated aqueous KNO
3
, 40/10/10/1
as eluent). (Yield 60%.)
Ru(bpy)
2
(bpyphDQ)
4+
.
1
H NMR (400 MHz, CD
3
CN)
δ9.08 (d, 1H, J= 5.76 Hz), 9.04 (m, 1H), 8.93 (t, 1H, J=
9.5 Hz), 8.87 (bs, 1H), 8.76 (d, 1H, J= 10.5 Hz), 8.67 (m, 2H),
8.56 (m, 5H), 8.43 (t, 1H, J= 9.5 Hz), 8.22 (d, 2H, J=11.4Hz),
8.28 (d, 2H, J= 14.4 Hz), 8.13 (m, 5H), 7.83 (m, 7H), 7.48
(m, 5H), 4.95 (m, 2H), 4.48 (m, 2H), 2.95 (m, 2H). MS (ESI)
m/z221.0 ([MCH
3
CN]
4+
/4 requires m/z221.4).
Ru(bpy)
2
(bpyph
2
DQ)
4+
.
1
H NMR (400 MHz, CD
3
CN)
δ9.08 (d, 1H, J= 5.76 Hz), 9.04 (m, 1H), 8.93 (t, 1H, J= 9.5 Hz),
8.86 (bs, 1H), 8.77 (d, 1H, J= 10.5 Hz), 8.67 (m, 2H), 8.56
(m, 5H), 8.43 (t, 1H, J= 9.5 Hz), 8.23 (d, 2H, J= 11.4 Hz), 8.12
(m, 11 H), 7.82 (m, 7H), 7.48 (m, 5H), 4.95 (m, 2H), 4.48
(m, 2H), 2.95 (m, 2H). MS (ESI) m/z229.9 ([M]
4+
/4 requires
m/z230.11), m/z240.2 ([MCH
3
CN]
4+
/4 requires m/z240.3).
Ru(bpy)
2
(bpyph
3
DQ)
4+
.
1
H NMR (400 MHz, CD
3
CN)
δ9.10 (d, 1H, J= 5.76 Hz), 9.03 (m, 1H), 8.93 (t, 1H, J= 9.5 Hz),
8.86 (bs, 1H), 8.78 (d, 1H, J= 10.5 Hz), 8.68 (m, 2H), 8.58
(m, 5H), 8.43 (t, 1H, J= 9.5 Hz), 8.24 (d, 2H, J=11.4Hz),8.10(m,
15H), 7.82 (m, 7H), 7.48 (m, 5H), 4.95 (m, 2H), 4.48 (m, 2H),
2.95 (m, 2H). MS (ESI) m/z248.9 ([M]
4+
/4 requires m/z249.1).
Ru(bpy)
2
(bpyph
4
DQ)
4+
.
1
H NMR(400 MHz, CD
3
CN)
δ9.10 (d, 1H, J= 5.76 Hz), 9.02 (m, 1H), 8.92 (t, 1H, J= 9.5 Hz),
8.85 (bs, 1H), 8.76 (d, 1H, J= 10.5 Hz), 8.68 (m, 2H), 8.57 (m,
5H), 8.45 (t, 1H, J= 9.5 Hz), 8.23 (m, 2H), 8.10 (m, 19H), 7.82
(m, 7H), 7.47 (m, 5H), 4.95 (m, 2H), 4.48 (m, 2H), 2.95 (m,
2H). MS (ESI) m/z267.9 ([M]
4+
/4 requires m/z268.1).
Ru(bpy)
2
(bpyph
5
DQ)
4+
.
1
H NMR(400 MHz, CD
3
CN)
δ9.08 (d, 1H, J= 5.76 Hz), 9.00 (m, 1H), 8.92 (t, 1H, J= 9.5 Hz),
8.85 (bs, 1H), 8.76 (d, 1H, J= 10.5 Hz), 8.67 (m, 2H), 8.57
(m, 5H), 8.43 (t, 1H, J= 9.5 Hz), 8.22 (d, 2H, J=11.4Hz),8.10(m,
23H), 7.82 (m, 7H), 7.47 (m, 5H), 4.95 (m, 2H), 4.48 (m, 2H),
2.95 (m, 2H). MS (ESI) m/z287.0 ([M]
4+
/4 requires m/z287.1).
Apparatus and Procedures.
1
H NMR spectra were recorded
in CD
3
Cl on a Varian Mercury spectrometer (400 MHz) with
residual nondeuterated solvent signal as reference .Electron spray
ionization (ESI) mass spectra were measured with a Micromass
ZMD2000 spectrometer.
UVvis spectra were recorded with a Perkin-Elmer LAM-
DA40 spectrophotometer. Luminescence spectra were taken on
a Spex Fluoromax-2 equipped with Hamamatsu R928 tubes.
Emission lifetimes were measured by time-correlated single-
photon counting technique, using a TCSPC apparatus (PicoQuant
Picoharp 300), equipped with subnanosecond LED sources
(280600 nm range, 300700 ps pulse width) powered with
a PicoQuant PDL 800-B variable (2.540 MHz) pulsed power
supply. The experiments were performed using 460 nm LED
(300 ps pulse width) as excitation source. The decays were anal-
yzed by means of PicoQuant FluoFit Global Fluorescence Decay
Analysis software. The time resolution of the system after the
deconvolution procedure is 250 ps, and estimated errors are 10%
for lifetime values.
Femtosecond time-resolved experiments were performed
using a pumpprobe setup
18
based on the Spectra-Physics
Hurricane Ti:sapphire laser source and the Ultrafast Systems
Helios spectrometer. The 560 nm pump pulses were generated
with a Spectra Physics 800 OPA. Probe pulses were obtained by
continuum generation on a sapphire plate (useful spectral range:
450800 nm). Eective time resolution ca. 300 fs, temporal
chirp over the white-light 450750 nm range ca. 200 fs, temporal
window of the optical delay stage 01000 ps. The time-resolved
spectral data were analyzed with the Ultrafast Systems Surface
Explorer Pro software.
Scheme 1
122 dx.doi.org/10.1021/jp209858p |J. Phys. Chem. A 2012, 116, 119–131
The Journal of Physical Chemistry A ARTICLE
Nanosecond transient absorption measurements were
performed with an Applied Photophysics laser ash photolysis
apparatus, using a frequency-doubled (532 nm, 330 mJ) or
tripled (355 nm, 160 mJ) Surelite Continuum II Nd/YAG laser
(half-width 68 ns) as excitation source. Transient detection
using photomultiplieroscilloscope combination (Hamamatsu
R928, LeCroy 9360) or gated intensied CCD-Camera
(Princeton Instruments PI-MAX II, with Acton SpectraPro
2300i triple grating at eld monochromator, RB GenII intensi-
er, ST133 controller, and a PTG pulser).
Cyclic voltammetric measurements were carried out with a PC-
interfaced Eco Chemie Autolab/Pgstat30 Potentiostat. Argon-
purged solutions in CH
2
Cl
2
and CH
3
CN (Romil, Hi-dry) con-
taining 0.1 M TBAPF
6
(Fluka, electrochemical grade; dried in an
oven) were used. A conventional three-electrode cell assembly
was used. A saturated calomel electrode (SCE, 6 mm
2
AMEL)
and a platinum wire were used as reference and counter electro-
des, respectively; a glassy carbon electrode (8 mm
2
, AMEL) was
used as a working electrode. The scan rate was 200 mV/s.
Spectroelectrochemistry measurements were performed in
CH
3
CN, using an optically transparent thin layer electrochemi-
cal (OTTLE) cell (optical path 1 mm, platinum mini-grid work-
ing electrode, platinum wire as counter electrode, silver wire as
quasi-reference electrode, 0.1 M (TBA)PF
6
electrolyte).
Computational Techniques. Ground-state and lowest triplet-
state geometries have been optimized at the DFT level of theory,
using the B3LYP hybrid functional or the unrestricted version
uB3LYP for triplet. The 6-31G* basis set has been employed for all
elements except Ru, for which LANL2DZ has been used. Energies,
molecular orbitals, and singly occupied molecular orbitals in the case
of triplet have been calculated at the same level of theory and with the
same basis set used for geometry optimization. Solvents have been
modeled with the polarizable continuum model. All calculations have
been performed with the Gaussian 2003 software package.
19
RESULTS
Absorption Spectra. The absorption spectra of the Ru(bpy)
2
-
(bpyph
n
DQ)
4+
dyads in acetonitrile solution are shown in
Figure 2a. The spectra of the Ru(bpy)
2
(bpyph
n
bpy)
2+
non-
quaternarized complexes, employed as model compounds for the
donor unit of the dyads, are also shown for comparison (Figure 2b).
The absorption spectra of the dyads were also measured in
CH
2
Cl
2
solutions. Going from CH
3
CN to CH
2
Cl
2
, signicant
spectral changes in the 350450 region are observed, especially
for the dyads with n> 2. The absorption spectrum of Ru(bpy)
2
-
(bpyph
4
DQ)
4+
in CH
3
CN and in CH
2
Cl
2
is reported in
Figure 3. The corresponding nonquaternarized model com-
pounds do not show any signicant solvatochromism.
Redox Behavior. The electrochemical behavior of the dyads
was studied by cyclic voltammetry in acetonitrile solution
(TBAPF
6
0.1 M supporting electrolyte, glassy carbon working
electrode, SCE reference electrode, platinum counter electrode).
For purposes of comparison, the electrochemical behavior of the free
bpyphDQ
2+
ligand was studied under the same experimental
conditions. All the dyads exhibit the same behavior regardless of
the number of phenylene spacers. The anodic region (0.0, + 1.5 V
vs SCE) is characterized by a reversible oxidation wave, due to
oxidation of the Ru(II) center. In the cathodic region (0.0, 1.2 V
vs SCE), two subsequent well-resolved reversible reduction waves
are observed. By comparison with the bpyphDQ
2+
ligand,
these waves are easily assigned to the reduction of the quaternar-
ized bipyridine unit and are almost unaffected by the number of
phenylene units, as exemplified by the results for n=1andn=4
reported in Table 1.
Spectroelectrochemical measurements were performed in
CH
3
CN to obtain the absorption spectrum of the reduced forms
Figure 2. Absorption spectra of the Ru(bpy)
2
(bpyph
n
DQ)
4+
dyads
(a) and of the Ru(bpy)
2
(bpyph
n
bpy)
2+
model complexes (b) in
CH
3
CN solution at room temperature.
Figure 3. Solvatochromic behavior of the Ru(bpy)
2
(bpyph
4
DQ)
4+
dyad.
Table 1. Redox Potentials of the [Ruph
n
DQ]
4+
Dyads
(n=1,4)inCH
3
CN
a
E
1/2
(V)
oxidation reduction
compound Ru(II/III) DQ
2+/+
DQ
+/0
Ru(bpy)
2
(bpyphDQ)
4+
+1.26 0.54 0.84
Ru(bpy)
2
(bpyph
4
DQ)
4+
+1.29 0.52 0.83
bpyphDQ
2+
0.61 0.93
a
CH
3
CN solution with 0.1 M TBAPF
6
; reversible waves, E
pa
E
pc
=
ca. 60 mV; potential values vs SCE.
123 dx.doi.org/10.1021/jp209858p |J. Phys. Chem. A 2012, 116, 119–131
The Journal of Physical Chemistry A ARTICLE
of the dyads. Upon reduction of the DQ
2+
unit, besides the
expected increase in absorbance at 387 nm and at λ> 490 nm,
20
the disappearance of the intense bands in the 350450 nm region,
characteristic of the dyads with n=25, was observed. Spectro-
electrochemical results for Ru(bpy)
2
(bpyphDQ)
4+
and Ru-
(bpy)
2
(bpyph
4
DQ)
4+
are reported as an example in Figure S2
of the Supporting Information.
Emission Measurements. At room temperature, for all the
dyads studied, the long-lived
3
MLCT emission typical of the ruthe-
nium(II) polypyridine complexes is very efficiently (>99%)
quenched relative to that observed for the corresponding non-
quaternarized Ru(bpy)
2
(bpyph
n
bpy)
2+
models (practically
independent of n,Φ= 0.01, τ= 220 ns, aerated CH
3
CN solu-
tion). The emission lifetimes were measured in CH
3
CN and
CH
2
Cl
2
solutions by single-photon counting technique. For the
dyads with n=35inCH
3
CN and n=25inCH
2
Cl
2
, the
decays observed are always monoexponential
21
with lifetimes
(Table 2) remarkably shorter with respect to those of the non-
quaternarized models. The decay for Ru(bpy)
2
(bpyph
3
DQ)
4+
in CH
3
CN is presented in Figure S3 as an example. In the case of
the dyads with n= 1 and n= 2 in CH
3
CN and with n=1in
CH
2
Cl
2
, the lifetime is shorter than the time resolution of the
instrument used (ca. 250 ps).
Transient Absorption Measurements. All the dyads were
investigated by ultrafast time-resolved absorption spectroscopy
(λ
exc
= 400 nm) in CH
3
CN and in CH
2
Cl
2
. For comparison,
experiments were also performed for the nonquaternarized
models under the same experimental conditions.
Ru(bpy)
2
(bpyphDQ)
4+
.The transient behavior observed
upon 400 nm excitation of CH
3
CN solutions of the Ru(bpy)
2
-
(bpyphbpy)
2+
model is summarized in Figure 4. The transient
spectrum is characterized by a ground-state bleach around 450 nm
and a broad positive absorption at λ> 500 nm. While the instan-
taneous (<1 ps) bleach of the ground-state absorption at 450 nm
remains constant over the whole kinetic range of the experiment (1
1000 ps), the positive absorption, after the initial instantaneous rise
(<1 ps), undergoes a further increase with a time constant of ca. 11 ps
(Figure 4). On a longer time scale, this absorption remains constant
over the time window of the instrument.
The ultrafast results obtained for quaternarized dyad Ru-
(bpy)
2
(bpyphDQ)
4+
in CH
3
CN are depicted in Figure 5.
The behavior observed is signicantly dierent from that of the
Ru(bpy)
2
(bpyphbpy)
2+
model in the same solvent. The ini-
tial spectrum, taken immediately after the excitation pulse (t=1ps),
exhibits, besides a bleach of the ground-state absorption at
450 nm and a broad positive absorption at λ> 550 nm, a sharp
absorption maximum at 515 nm (Figure 5a). The transient spec-
tral changes are biphasic, with a rst fast process with time
constant of 3.7 ps (Figure 5b), in which the 515 nm maximum
disappears and the 450 nm bleach undergoes a partial (ca. one-
third) recovery, while the broad absorption at λ> 550 nm
remains constant. The subsequent spectral changes (7120 ps)
exhibit a clean decay to the baseline (Figure 5c), with a time
constant of 18 ps (Figure 5d).
Ultrafast results obtained in CH
2
Cl
2
(Figure 6) clearly indicate
that the polarity of solvent has a strong eect on the kinetics of
the transient spectral changes. The features of the transient
spectrum (a bleach at 450 nm, a broad positive absorption at
λ> 550 nm, and a sharp absorption maximum at 520 nm) are
qualitatively very similar to those observed in CH
3
CN, but in this
solvent the formation of the positive absorption maximum at
520 nm is time-resolved (Figure 6a). This formation is biphasic
Table 2. Photophysical Data for Dyads Ru(bpy)
2
(bpyph
n
DQ)
4+
(n=15)
τ
em
,ns
a
τ
abs
,ns
b
k
el
,s
1e
dyad CH
3
CN CH
2
Cl
2
CH
3
CN CH
2
Cl
2
CH
3
CN CH
2
Cl
2
Ru(bpy)
2
(bpyphDQ)
4+
<0. 25 <0.25 <0.001
c
0.002
d
>1.0 10
12
5.0 10
11
Ru(bpy)
2
(bpyph
2
DQ)
4+
<0.25 0.76 0.080 0.680 1.2 10
10
1.4 10
9
Ru(bpy)
2
(bpyph
3
DQ)
4+
1.00 3.20 1.0 >2.0 1.0 10
9
3.1 10
8
Ru(bpy)
2
(bpyph
4
DQ)
4+
1.24 1.36 1.5 >2.0 8.1 10
8
7.3 10
8
Ru(bpy)
2
(bpyph
5
DQ)
4+
2.70 2.60 2.0 >2.0 3.7 10
8
3.8 10
8
a
Measured by single-photon counting technique.
b
Measured by ultrafast transient absorption spectroscopy.
c
Estimated from the prompt formation of
the 520 nm absorption; see Discussion.
d
obtained from the rise of the 520 nm absorption; see Discussion.
e
Calculated as the reciprocal of the lifetimes
obtained in transient absorption (n= 1, 2) or emission (n=35) measurements.
Figure 4. Ultrafast spectroscopy of the Ru(bpy)
2
(bpyphbpy)
2+
model in acetonitrile.
124 dx.doi.org/10.1021/jp209858p |J. Phys. Chem. A 2012, 116, 119–131
The Journal of Physical Chemistry A ARTICLE
(Figure 6b) with an ultrafast component (ca. 30%, time constant
of ca. 2 ps) and a slower component (ca. 70%, time constant of
15 ps). In this early time scale, the bleach and the broad
absorption remain almost constant. In the longer time scale
(23800 ps, Figure 6c), the transient spectrum decays cleanly to
the baseline. A value of 190 ps for the time constant is obtained by
kinetic analysis of the bleach at 455 nm as well as of the
absorption at 510 nm (Figure 6d).
Ru(bpy)
2
(bpyph
n
DQ)
4+
(n = 25). In ultrafast spectros-
copy, the qualitative behavior of the Ru(bpy)
2
(bpyph
n
DQ)
4+
dyads with n=25 is rather homogeneous within the series.
The decay for Ru(bpy)
2
(bpyph
3
DQ)
4+
(in CH
3
CN, 400 nm
excitation wavelength) is reported as an example in Figure 7. The
transient spectrum observed after a few picoseconds
22
is the same
observed for the respective nonquaternarized model compounds
(see, e.g., Figure S4 of the Supporting Information), i.e., the
typical spectrum of the MLCT excited state with a bleach at 450 nm
and a broad absorption at λ> 550 nm. This transient spectrum
decays cleanly to the ground state (isosbestic points at ΔOD = 0)
with a time constant dependent on the number of phenyl units of
the bridge. The lifetimes obtained by kinetic analysis are sensitive
to solvent and span a wide range of values (Table 2). The values
for n= 3, 4 in CH
3
CN and for n=3,4,5inCH
2
Cl
2
, too long for
an accurate measurement in the time window of the experiment,
are only approximate estimates. No transient absorption signals
can be observed in nanosecond flash photolysis experiments on
the same systems (355 nm excitation, time resolution g8 ns).
Computational Results. Gas-phase ground-state optimized
geometries have been calculated for the Ru(bpy)
2
(bpyph
n
DQ)
4+
dyads with n=14. In all cases, donor/bridge dihedral
angles have been calculated at ca. 30, while phenylene/phenyl-
ene dihedral angles lie in the range between 28 and 35and
bridge/acceptor dihedral angles are ca. 17. These values are
relevant to the electronic couplings along the modular bridge and
between components. Molecular orbitals and their relative energies
have been calculated in gas phase and CH
3
CN for n=14
(Figures S6S10 of the Supporting Information) and also in
CH
2
Cl
2
for n= 4. Introducing solvent via the polarizable contin-
uum model increases the HOMOLUMO gap (CH
3
CN >
CH
2
Cl
2
> gas phase) but does not alter the orbital relative
positioning on an energy scale. Gas-phase geometry optimization
of the lowest triplet state for n= 4 gives a structure where donor/
bridge dihedral angle is 32, phenylene/phenylene dihedral
angles lie in the range between 20 and 28, and the bridge/
acceptor dihedral angle is 27. Singly occupied orbitals calculated
for n= 4 in gas phase, CH
2
Cl
2
, and CH
3
CN, are found to be
bridge-localized and DQ-localized. The resulting spin density is
confined to the bridgeDQ region (Figure S11 of the Support-
ing Information). Some relevant computational results are re-
called below in the Discussion Section.
DISCUSSION
Absorption Spectra. The absorption spectra of related model
systems, such as the nonquaternarized analogues Ru(bpy)
2
-
(bpyph
n
bpy)
2+
(Figure 2b) or the homobinuclear complexes
Ru(bpy)
2
(bpyph
n
bpy)Ru(bpy)
24+
,
23
are a simple superposi-
tion of the ligand-centered (LC) bands (at ca. 290 nm), the
metal-to-ligand charge-transfer (MLCT) bands of the ruthenium
polypyridine units (at ca. 460 nm), and the ππ* bands of the
oligophenylene bridge (at 320350 nm, depending on n). The
spectra of the Ru(bpy)
2
(bpyph
n
DQ)
4+
dyads, however, in
addition to the above-mentioned types of bands, display a set of
new bands in the 350450 nm range (Figure 2a) that cannot be
attributed to any of the molecular components.
24
These bands
are solvatochromic (Figure 3) and disappear upon electrochem-
ical reduction of the DQ unit (Figure S2 of the Supporting
Information). Thus, it is reasonable to assign them to some kind
of charge-transfer transitions involving the DQ unit as acceptor.
Deeper insight into the nature of such transitions can be obtained
Figure 5. Ultrafast spectroscopy of the Ru(bpy)
2
(bpyphDQ)
4+
dyad in CH
3
CN.
125 dx.doi.org/10.1021/jp209858p |J. Phys. Chem. A 2012, 116, 119–131
The Journal of Physical Chemistry A ARTICLE
from the DFT calculations performed on this series of dyads.
Some relevant orbitals of Ru(bpy)
2
(bpyph
2
DQ)
4+
and Ru-
(bpy)
2
(bpyph
4
DQ)
4+
, labeled according to their predomi-
nant localization, are shown as an example in Figure 8.
25
The
figure also shows how the energy of these orbitals changes along
the n=24 series. It is seen that, while the filled metal d-orbital
and the vacant orbitals localized on the bipyridine fragment of the
bridging ligand or on the DQ acceptor unit remain appreciably
constant along the series, the highest filled orbital localized on
the bridge undergoes a steady increase in energy with increasing
nvalue. In fact, the filled bridge orbital crosses the metal d-orbital
and becomes the HOMO for ng3. Although one-electron orbital
energy differences cannot simply be translated into spectroscopic
excitation energies, it is evident that charge-transfer transitions
having the DQ unit as the acceptor and the oligophenylene bridge
as the donor are expected to occur at low energies in these
systems. Taking place between MOs localized on adjacent fragments
with good overlap (which is not the case, e.g., when the orbitalsare
localized on Ru and DQ), such charge-transfer transitions are also
expected to have high oscillator strength. Thus, we assign the new
bands in the 350450 nm range of the absorption spectra of the
Ru(bpy)
2
(bpyph
n
DQ)
4+
dyads (Figure 2a) as π(ph
n
)fπ*-
(DQ), bridge-to-acceptor CT transitions.
Photophysical Behavior. A typical energy level scheme for
the Ru(bpy)
2
(bpyph
n
DQ)
4+
dyads, as obtained from spectro-
scopic
26,23
and electrochemical data (Table 1)
27,28
in acetonitrile, is
shown in Figure 9. This diagram holds with minor variations for
all the complexes in the series, showing that electron transfer from
the MLCT excited states of the Ru complex donor unit to the
appended DQ
2+
acceptor is thermodynamically favored by 0.33 (
0.02 eV. Therefore, a conventional electron-transfer quenching
scheme, with moderately exergonic charge separation followed
by strongly exergonic charge recombination,
29
is expected to be
effective in these systems. The following discussion will show that
Figure 6. Ultrafast spectroscopy of the Ru(bpy)
2
(bpyphDQ)
4+
dyad in CH
2
Cl
2
.
Figure 7. Ultrafast spectroscopy of the Ru(bpy)
2
(bpyph
3
DQ)
4+
dyad in acetonitrile.
126 dx.doi.org/10.1021/jp209858p |J. Phys. Chem. A 2012, 116, 119–131
The Journal of Physical Chemistry A ARTICLE
the situation can be more complex than suggested by the simple
scheme of Figure 9.
Since the qualitative behavior of the Ru(bpy)
2
(bpyph
n
DQ)
4+
dyad with n= 1 is clearly dierent from that of the other
members of the series (n=25), the photophysics of these two
classes of systems will be discussed separately.
Ru(bpy)
2
(bpyphDQ)
4+
.In order to discuss the photophys-
ics of this dyad, a comparison with the behavior of the nonqua-
ternarized model system Ru(bpy)
2
(bpyphbpy)
2+
is valuable.
For the Ru(bpy)
2
(bpyphbpy)
2+
model, the emission
energy and lifetime (626 nm and 220 ns, in aerated acetonitrile)
are quite typical of Ru(II) polypyridine complexes. The slight
(ca. 10 nm) red shift relative to the standard Ru(bpy)
32+
complex
indicates that, as in related systems,
16
the emitting MLCT excited
state is on the phenyl-substituted bipyridine ligand, lying at
slightly lower energy than those localized on the unsubstituted
bipyridines. The behavior in ultrafast time-resolved spectrosco-
py, however, is signicantly dierent from that of Ru(bpy)
32+
.In
particular (Figure 4), while the instantaneous bleach of the ground-
state MLCT band at 450 nm remains constant over the 11000
ps time range, the positive ligand radical anion absorption in the
λ> 500 nm range, after the initial instantaneous rise, undergoes a
further increase with time constant of 11 ps. As shown by pre-
vious nanosecond work,
23
complexes of phenyl-substituted bi-
pyridine exhibit larger transient absorption in this spectral range
than the simple Ru(bpy)
32+
complex. Therefore, the biphasic in-
crease observed here can be attributed to MLCT excitation
followed by the relaxation of states involving the bpy ligands to
the lowest state involving the phenyl-substituted bipyridine and
provides a direct spectroscopic measurement of the rate of inter-
ligand electron transfer (ILET) in the complex. A time constant of
47 ps was estimated for ILET in Ru(bpy)
32+
in acetonitrile by
Malone and Kelley
30
using time-resolved absorption polarization
spectroscopy. Subsequently, with a better instrumental resolution,
Papanikolas and co-workers
31
measured a lifetime of 8.7 ps for
ILET in Os(bpy)
32+
. The time constant, 11 ps, obtained here for
Ru(bpy)
2
(bpyphbpy)
2+
is close to this last gure.
32
In Ru(bpy)
2
(bpyphDQ)
4+
, the MLCT emission is very
eciently (>99%) quenched relative to the model compound,
with an emission lifetime shorter than the time resolution of the
single-photon counting instrument used (Table 2). The behavior
observed in ultrafast spectroscopy (Figures 5 and 6) is complex
and solvent-dependent and deserves a careful examination.
In CH
3
CN, the initial spectrum (1.2 ps in Figure 5) is clearly
dierent from that of the Ru(bpy)
2
(bpyphbpy)
2+
model
(Figure 4), featuring, besides the usual bleach of the ground-state
absorption at 450 nm and the broad positive absorption at λ>
550 nm, also a sharp maximum at 515 nm. This maximum is
characteristic of the reduced form of the DQ acceptor (λ=
514 nm, ε= 4.400 M
1
cm
1
),
33
suggesting that intramolecular
electron transfer with formation of Ru(III) and reduced DQ
2+
takes place promptly, i.e., in the subpicosecond time domain. As
this time scale is much shorter than that for interligand electron
hopping (11 ps in the Ru(bpy)
2
(bpyphbpy)
2+
model,
see above), the prompt electron-transfer quenching is likely to
involve only about one-third of the MLCT states, i. e., those
localized on the quencher-functionalized bipyridine. Thus, in the
initial transient spectrum, the 450 nm bleach can be attributed by
ca. one-third to the charge-separated state and by ca. two-thirds
to the MLCT states localized on the outer bipyridine ligands.
The subsequent transient spectral changes are biphasic, with a
rst fast process, with time constant of 3.7 ps, in which the
515 nm maximum disappears, the 450 nm bleach undergoes a
partial (ca. one-third) recovery, and the broad absorption at λ>
550 nm remains constant. This behavior is exactly as expected for
charge recombination of the electron-transfer state, considering
that the various MLCT states do not interconvert in this time
scale. The nal spectrum of this step (6.9 ps in Figure 5) is thus a
Figure 8. Relevant molecular orbitals of the Ru(bpy)
2
(bpyph
n
DQ)
4+
dyads (n= 2, 3, 4) (for a more complete series, see the Supporting
Information). The central diagram shows the energy changes undergone by orbitals with the same predominant localization along the n=24 series.
Figure 9. General energy level diagram for the Ru(bpy)
2
(bpyph
n
DQ)
4+
dyads, where D = Ru(bpy)
2
bpy
2+
,B=ph
n
,andA=DQ
2+
.
127 dx.doi.org/10.1021/jp209858p |J. Phys. Chem. A 2012, 116, 119–131
The Journal of Physical Chemistry A ARTICLE
typical MLCT excited-state spectrum, representing the MLCT
states localized on the outer bipyridine ligands. The subsequent
decay of the spectral changes leads cleanly to the baseline, with
time constant of 18 ps. Since this gure is close to that for inter-
ligand hopping,
34
it seems likely to associate this process to a
rate-limiting conversion of the MLCT states localized on
the outer bipyridine ligands to that localized on the quencher-
functionalized one, followed by fast electron transfer and charge
recombination to the ground state. The photophysical mechan-
ism is sketched in Figure 10.
When the solvent is changed from CH
3
CN to CH
2
Cl
2
, the
energy of the charge-separated state is expected to increase (by
ca. 0.1 eV, from Weller-type
28
calculation) and the reorganiza-
tional energy to decrease. This has important consequences in
the time-resolved spectroscopic behavior (Figure 6). The de-
crease in driving force, only partially compensated (in the Marcus
normal region) by the lower reorganizational energy, makes now
the charge separation step somewhat slower. The consequence is
that the formation of the 520 nm maximum characteristic of the
reduced DQ unit is now time-resolved, with a fast component
(1.8 ps) assigned to direct charge separation (electron transfer
within the quencher-functionalized bipyridine) and a slower
component (ca. 15 ps) assigned to ILET from the outer bipyrid-
ines. In CH
2
Cl
2
, the charge recombination process, which lies in
the Marcus inverted region, undergoes a strong decrease in rate
(190 ps), under the combined eects of the increase in driving
force and decrease in reorganizational energy. As this process is
now much slower than both direct charge separation and ILET,
in time-resolved spectroscopy the charge-separated state rst
accumulates and then undergoes a clean decay to the baseline.
The kinetic dierences between acetonitrile and dichloromethane
are summarized in Figure 11.
The very fast rate observed for charge separation (<1 ps) in
Ru(bpy)
2
(bpyphDQ)
4+
suggests that the donoracceptor
electronic coupling is relatively strong through this short bridge.
This is in line with the results of DFT calculations, which show
that the phenyl spacer is only slightly twisted relative to the adjacent
pyridyl ring of DQ, so that (Figure 12) the π* DQ acceptor
orbitals (LUMO and LUMO+1) have large amplitudes at the
phenyl spacer and undergo signicant direct overlap with the π*
bpy excited-state donor orbital (LUMO+2). The fact that charge
separation is always faster than charge recombination likely
reects a combination of nuclear (normal vs inverted Marcus
region) and electronic factors (less ecient coupling for charge
recombination, involving π* DQ LUMO and d Ru HOMO).
Ru(bpy)
2
(bpyph
n
DQ)
4+
(n = 25). The behavior of the
Ru(bpy)
2
(bpyph
n
DQ)
4+
dyads with n=25 is qualitatively
rather homogeneous. The emission intensity and lifetime are
always strongly (>99%) quenched relative to those of the
respective nonquaternarized model compounds. The transient
spectral changes obtained by ultrafast spectroscopy (after a short,
ca. 10 ps, delay)
22,35
apparently exhibit a clean decay of the
typical triplet MLCT excited-state spectrum down to the ground
state. Where both measurements are feasible (i.e., for n=3in
CH
3
CN and n= 2 in CH
2
Cl
2
), the lifetimes obtained from
emission decay and transient spectroscopy are comparable
(Table 2). Thus, the overall picture is that of a series of dyads
in which a rate-determining quenching step is followed by faster
processes leading to the ground state, without accumulation of
intermediate transient products. Within this general qualitative
behavior, the n=25 systems differ in kinetics, spanning a wide
range of lifetimes.
In an electron-transfer quenching mechanism (holding true
for this series of dyads with the possible exception, see below, of
n= 4, 5), the observed behavior implies a change in kinetic regime
from n= 1 (charge separation faster than charge recombination)
to ng2 (charge recombination faster than charge separation).
The reasons can be discussed in terms of electron-transfer theory.
As the energetics (MLCT energy and redox potentials) are
virtually the same for the two systems, the nuclear factors
(FCWD in eq 1) for charge separation and recombination are
also expected to be practically the same for n= 1 and n=2.
36
Therefore, the reasons for the change in kinetic regime must be
sought in the electronic factors. For n=1,wehavealreadyremarked
Figure 10. Schematic representation of the electron-transfer processes involved in the photophysical mechanism for Ru(bpy)
2
(bpyphDQ)
4+
in
CH
3
CN.
128 dx.doi.org/10.1021/jp209858p |J. Phys. Chem. A 2012, 116, 119–131
The Journal of Physical Chemistry A ARTICLE
that the very fast charge separation can be related to the ecient
overlap between the donor and acceptor orbitals, which provides
strong direct donoracceptor electronic coupling. In this case,
second-order interaction via virtual states can be disregarded.
37
For ng2, on the other hand, direct overlap is much less ecient
(Figure 8 and Figures S7S10 of the Supporting Information),
and a superexchange mechanism (eq 2) is appropriate to describe
the donoracceptor electronic couplings. In general terms (Figure 1),
both electron- and hole-transfer virtual states involving the bridge
can be considered as superexchange mediators. When elec-
tron transfer is photochemically induced, however, dierent super-
exchange mechanisms apply to charge separation and recombina-
tion. When excitation takes place at the donor (Figure 13), both
electron- and hole-transfer virtual states can in principle mediate
charge recombination, but only the electron-transfer virtual state
can aect charge separation (the D*BA excited state and the
DB
+
A
hole-transfer state dier by a two-electron shift and
cannot mix eectively). Thus, for charge separation only the rst
term of eq 2 should be retained, while for charge recombination
both terms could in principle operate. In practice, for oligo-
phenylene bridges, the electron-transfer virtual states lie at much
higher energy than the hole-transfer ones (Table 3), so that only
the hole-transfer superexchange pathway is expected to dominate
charge recombination. The energy gaps with the appropriate virtual
states estimated
38
for the two pathways (Table 3) clearly show that,
in terms of the denominators in eq 2, charge recombination is highly
favored over charge separation. Of course, electronic coupling is not
the only factor determining the rates (eq 1). The experimental
evidence that no accumulation of the charge-separated state takes
place, however, seems to indicate that in these systems the eect of
the nuclear factors (which, per se, would favor charge separation) is
overcome by that of the superexchange electronic factors.
As discussed above, the rate constants for ng2reect the
primary quenching step that leads to disappearance of the MLCT
excited state. When these rate constants are plotted against n
(Figure 14), however, the behavior is quite dierent from the
regular exponential decay expected on the basis of simple theory
(eq 4) and found in several studies of distance dependence of
electron transfer.
3
We suggest that the main reason for the
nonideal behavior of this series of dyads lies in the breakdown
of the main assumption embedded in eq 2, i.e., that the energy of
the virtual charge-transfer states involving the bridge is essentially
constant. In fact, as inferred from Table 3, these states are
expected to undergo quite substantial energy changes along the
series. Besides being responsible, as discussed above, for the fast
charge recombination occurring in these systems, the presence of
low-lying bridge-localized charge-transfer states is likely to have a
deep eect also on the trend observed for n=25. The main
point is that, as the bridge is elongated, charge-transfer states
involving the bridge may become suciently low in energy as to
have an active, direct role in the quenching of the excited state.
A rough idea of the relative energy orderings of the relevant
electronic states in the n=24 series can be obtained from the
Figure 11. Schematic summary of the kinetic behavior of Ru(bpy)
2
(bpyphDQ)
4+
in acetonitrile and dichloromethane.
Figure 12. Frontier molecular orbitals of Ru(bpy)
2
(bpyphDQ)
4+
.
Figure 13. Schematic representation of the superexchange pathways
available to photoinduced charge separation (cs, blue) and charge
recombination (cr, red).
129 dx.doi.org/10.1021/jp209858p |J. Phys. Chem. A 2012, 116, 119–131
The Journal of Physical Chemistry A ARTICLE
orbital energies in Figure 15. As already noticed, the lled bridge
orbital undergoes a steady increase in energy along the series and,
at ca. n= 3, raises above the metal-centered orbitals, becoming
the HOMO. In an admittedly very crude approximation, where
relative excited-state energies are estimated from MO energy
dierences (Figure 8), it is predicted that at ca. n= 3 two relevant
state crossings may occur: (i) charge-transfer states of bridge-
tobpy type (D
B
+
A) may become comparable or lower in
energy than the MLCT excited state, and (ii) the lowest excited
state of the dyads switches from charge-separated state (D
+
BA
) to bridge-to-DQ charge-transfer state (DB
+
A
).
39
From this viewpoint, it is no wonder that the decrease in
quenching rate (elongation in lifetimes) of the MLCT state does
not follow the standard exponential behavior. In particular from
point (i), the abrupt stop taking place around n= 3 can be most
likely attributed to a change in quenching mechanism: from
π*(bpy) fπ*(DQ) electron transfer (*DBAfD
+
B
A
)toπ(bridge) fd(Ru) electron transfer (*DBAf
D
B
+
A). As to the presence of low-lying states of the
DB
+
A
type, they are probably of little relevance to the decay
of the MLCT excited state, since a *DBAfDB
+
A
process, though thermodynamically allowed (Figure 15), would
involve two simultaneous electron-transfer steps: π*(bpy) f
π*(DQ), π(bridge) fd(Ru.). Such type of states can, on the
other hand, be very ecient mediators of the D
+
BA
f
DBA charge recombination, either as superexchange hole-
transfer virtual states (Figure 13) or, perhaps, for the longer dyads
(point ii), even as real intermediate states.
CONCLUSIONS
The main conclusions of this work can be summarized as
follows:
(i) In the Ru(bpy)
2
(bpyph
n
DQ)
4+
dyads, new spectro-
scopic transitions are observed, in addition to those
characteristic of the donor, bridge, and acceptor frag-
ments. DFT calculations point out the presence of lled
bridge-localized orbitals at high energy and suggest the
assignment of such bands as bridge-to-acceptor (πph
n
)f
(π* DQ) charge-transfer transitions.
(ii) The Ru(bpy)
2
(bpyphDQ)
4+
dyad (n=1)under-
goes very fast (few picosecond) photoinduced electron
transfer followed by slower charge recombination. The
ultrafast charge separation is a consequence of good over-
lap between the π* bpy and π* DQ orbitals, which provides
good, direct donoracceptor electronic coupling. The
ultrafast electron-transfer step takes place from the MLCT
state localized on the functionalized ligand and is time-
resolved with respect to the slower interligand hopping
processes. Charge separation is kinetically favored over
charge recombination by both nuclear (normal vs inverted
region) and electronic (better coupling) factors.
(iii) In the Ru(bpy)
2
(bpyph
n
DQ)
4+
dyads with n=25,
MLCT excited-state quenching takes place without ac-
cumulation of charge-separated product, suggesting that
charge recombination is faster than charge separation. This
behavior can be rationalized in termsof the superexchange
Table 3. Redox Properties of Oligo-p-phenylenes and Estimated Energies of the Electron- and Hole-Transfer States in the
Ru(bpy)
2
(bpyph
n
DQ)
4+
Dyads
Ru(bpy)
2
(bpyph
n
DQ)
4+
=DBA
ph
n
[Ru(bpy)
2
bpy=D,ph
n
=B,DQ = A]
nE
ox a,b
(V) E
red a,c
(V) E(DB
+
A
) (eV)
d
E(D
+
B
A) (eV)
d
ΔE(h
+
)cs
d,e
ΔE(e
)cs
d,f
2 1.81 2.72 2.33 4.01 0.53 1.93
3 1.56 2.44 2.08 3.73 0.28 1.65
4 1.43 2.32 1.95 3.61 0.15 1.53
5 1.36
g
2.24
g
1.88 3.53 0.08 1.45
a
Versus SCE, adapted from Meerholz, K; Heinze, J. Electrochimica Acta 1996,41, 1839.
b
Dichloromethane.
c
Dmethylamine.
d
Neglecting electrostatic
work term corrections.
e
ΔE(e
)cs = E(DB
+
A
)E(*DBA).
f
ΔE(h
+
)cs = E(D
+
B
A) E(D
+
BA
).
g
Estimated from data for
(ph)
1
(ph)
4
and (ph)
6
.
Figure 14. Logarithmic plot of the rate constants for excited-state decay
as a function of the number of spacers in the bridge.
Figure 15. Approximate energy ordering of the various types of excited
states in Ru(bpy)
2
(bpyph
n
DQ)
4+
dyads (n=25), as obtained
from MO energy dierences.
130 dx.doi.org/10.1021/jp209858p |J. Phys. Chem. A 2012, 116, 119–131
The Journal of Physical Chemistry A ARTICLE
couplings for the two processes: with this type of bridges,
the electron-transfer virtual states required for charge
separation lie at very high energy, whereas charge recom-
bination is eciently mediated by low-energy virtual states
of the hole-transfer type.
(iv) The kinetics of MLCT quenching in the Ru(bpy)
2
-
(bpyph
n
DQ)
4+
dyads with n=15 does not follow
the usual exponential fallowith number of spacers
(or distance): after a regular decrease from n= 1 to 3,
the rate constants become almost insensitive to bridge
length for n=35. Now, n= 3 is precisely the dyad
length at which a lled bridge-localized orbital, increasing
steadily in energy along the series, replaces the Ru
d-orbitals as the HOMO of the system. Thus, the
uncommon kinetic behavior is rationalized by a change
in quenching mechanism with increasing bridge length
from oxidative quenching by the DQ acceptor to reduc-
tive quenching by the bridge.
(v) Generally speaking, oligo p-phenylene bridges have sub-
stantial drawbacks from the viewpoint of achieving e-
cient, long-lived charge separation upon donor excitation
in donorbridge-acceptor dyads. These bridges are char-
acterized by relatively high-energy lled orbitals, and, in
this type of dyads, this (a) will generally favor charge
recombination via hole-transfer superexchange pathways
and (b) may, for long bridges, switch on new quenching
mechanisms other than electron transfer to the acceptor.
The same type of bridges are expected to perform much
better, on the other hand, when electron transfer in the
dyad is promoted by excitation of the acceptor.
6,8,12,17
In
this case, the situation is reversed, with hole-transfer
superexchange pathways favoring charge separation.
ASSOCIATED CONTENT
b
SSupporting Information.
1
H NMR spectrum of Ru(bpy)
2
-
(bpyph
2
DQ)
4+
; spectroelectrochemistry of Ru(bpy)
2
(bpy
phDQ)
4+
and of Ru(bpy)
2
(bpyph
4
DQ)
4+
; emission decay
for Ru(bpy)
2
(bpyph
3
DQ)
4+
; ultrafast spectroscopy of Ru(bpy)
2
-
(bpyph
3
bpy)
2+
model; ultrafast spectroscopy of Ru(bpy)
2
-
(bpyph
2
DQ)
4+
as a function of excitation wavelength, relevant
MOs of Ru(bpy)
2
(bpyph
n
DQ)
4+
(n=14); spin density plot
for the lowest triplet state of Ru(bpy)
2
(bpyph
4
DQ)
4+
.This
material is available free of charge via the Internet at http://pubs.
acs.org.
AUTHOR INFORMATION
Corresponding Author
*E-mail: snf@unife.it (F.S.); decola@uni-muenster.de (L.D.C.).
Present Addresses
)
Chemical Technologies International s.r.l, 44100 Ferrara, Italy.
3
Universite Joseph Fourier Grenoble 1/CNRS, Departement de
Chimie Moleculaire, UMR 5250, Laboratoire de Chimie Inorga-
nique Redox, Institut de Chimie Moleculaire de Grenoble FR-
CNRS-2607, BP 53, 38041 Grenoble Cedex 9, France.
4
European Patent Oce, Patentlaan 2, 2288EE Rijswijk,
Netherlands.
1
Molecular Stamping s.r.l., Fondazione Bruno Kessler, 38123
Trento, Italy.
ACKNOWLEDGMENT
Financial support by the EC (Grant G5RD-CT-2002-00776,
MWFM) and by the Italian MIUR (PRIN 20085ZXFEE) is
gratefully acknowledged.
REFERENCES
(1) (a) Lewis, N. S.; Nocera, D. G. Proc. Natl. Acad. Sci. U.S.A. 2006,
103, 1572915735. (b) Wasielewski, M. R. J. Org. Chem. 2006,
71, 50515066. (c) Hambourger, M.; Moore, G. F.; Kramer, D. M.;
Gust, D.; Moore, A. L.; Moore, T. A. Chem. Soc. Rev. 2009,38,2535.
(2) (a) Molecular Electronics; Jortner, J., Ratner, M., Eds.; Blackwell
Science: London, U.K., 1997. (b) Aviram, A; Ratner, M. A. Ann. N.Y.
Acad, Sci. 1998,852 (Special Volume on Molecular Electronics).
(c) Joachim, C.; Gimzewski, J. K.; Aviram, A. Electronics using
hybrid-molecular and mono-molecular devices. Nature 2000,408, 541.
(d) Tour, J. M. Acc. Chem. Res. 2000,33, 791804. (e) Pease, A. R.;
Jeppesen, J. O.; Stoddart, J. F.; Luo, Y.; Collier, C. P.; Heath, J. R. Acc.
Chem. Res. 2001,34, 433.
(3) (a) Paddon-Row, M. N. In Electron Transfer in Chemistry;
Balzani, V., Ed.; Wiley-VCH: Weinheim, Germany, 2001; Vol. III,
Chapter 2.1, p 179. (b) Scandola, F.; Chiorboli, C.; Indelli, M. T.;
Rampi, M. A. In Electron Transfer in Chemistry; Balzani, V., Ed.; Wiley-
VCH: Weinheim, Germany, 2001; Vol. III, Chapter 2.3, p 337.
(c) Petersson, K.; Wiberg, J.; Ljungdahl, T.; Martensson, J.; Albinsson, B
J. Phys. Chem. A 2006,110, 319. (d) Albinsson, B.; M
artensson, J.
J. Photochem. Photobiol., C 2008,9, 138155.
(4) (a) Marcus, R. A. Annu. Rev. Phys. Chem. 1964,15, 155. (b) Sutin,
N. Prog. Inorg. Chem. 1983,30, 441. (c) Miller, J. R.; Beitz, J. V.;
Huddleston, R. K. J. Am. Chem. Soc. 1984,106, 5057. (d) Marcus, R. A.;
Sutin, N. Biochim. Biophys. Acta 1985,811, 265. (e) Jortner, J. J. Chem.
Phys. 1976,64, 4860. (f) Bixon, M.; Jortner, J. Adv. Chem. Phys. 1999,
106, 35. (g) Newton, M. D. Chem. Rev. 1991,91, 767.
(5) (a) Halpern, J.; Orgel, L. E. Discuss. Faraday Soc. 1960,29, 32.
(b) McConnell, H. M. J. Chem. Phys. 1961,35, 508. (c) Mayoh, B.; Day,
P. J. Chem. Soc., Dalton Trans. 1974, 846. (d) Miller, J. R.; Beitz, J. V.
J. Chem. Phys. 1981,74, 6746. (e) Richardson, D. E.; Taube, H. J. Am.
Chem. Soc. 1983,105, 40.
(6) For a clear demonstration of such a dependence, see, e.g.:
(a) Hanss, D.; Wenger, O. S. Inorg. Chem. 2008,47, 90819084.
(b) Hanss, D.; Wenger, O. S. Inorg. Chem. 2009,48, 671680.
(7) Davis, W. B.; Svec, W. A.; Ratner, M. A.; Wasielewski, M. R.
Nature 1998,396, 60.
(8) Weiss, E. A.; Ahrens, M. J.; Sinks, L. E.; Ratner, M. A.; Wasielewski,
M. R. J. Am. Chem. Soc. 2004,126, 5577.
(9) Zojer, E.; Cornil, J.; Leising, G.; Bredas, J. L. Phys. Rev. B 1999,
59, 7957.
(10) Onuchic, J. N.; Beratan, D. N. J. Am. Chem. Soc. 1987,109,
6771.
(11) (a) Helms, A.; Heiler, D.; McLendon, G. J. Am. Chem. Soc.
1991,113, 4325. (b) Helms, A.; Heiler, D.; McLendon, G. J. Am. Chem.
Soc. 1992,114, 6227.
(12) (a) Weiss, E. A.; Ahrens, M. J.; Sinks, L. E.; Ratner, M. A.;
Wasielewski, M. A. J. Am. Chem. Soc. 2004,126, 9510. (b) Weiss, E. A.;
Tauber, M. J.; Kelley, R. F.; Ahrens, M. J.; Ratner, M. A.; Wasielewski,
M. R. J. Am. Chem. Soc. 2005,127, 11842.
(13) (a) Barigelletti, F.; Flamigni, L.; Guardigli, M.; Juris, A.; Beley,
M.; Chodorowski-Kimmes, S.; Collin, J.-P.; Sauvage, J.-P. Inorg. Chem.
1996,35, 136. (b) Barigelletti, F.; Flamigni, L.; Collin, J.-P.; Sauvage,
J.-P. Chem. Commun. 1997, 333. (c) Collin, J.-P.; Gavina, P.; Heitz, V.;
Sauvage, J.-P. Eur. J. Inorg. Chem. 1998,1.
(14) (a) Schlicke, B.; Belser, P.; De Cola, L.; Sabbioni, E.; Balzani, B.
J. Am. Chem. Soc. 1999,121, 4207.(b) De Cola, L.; Belser, P. In Electron
Transfer in Chemistry; Balzani, V., Ed.; Wiley-VCH, Verlag GmbH:
Weinheim, Germany, 2001; Vol. 5, p 97. (c) Welter, S.; Salluce, N.;
Belser, P.; Groeneveld, M.; De Cola, L. Coord. Chem. Rev. 2005,249,
1360.
131 dx.doi.org/10.1021/jp209858p |J. Phys. Chem. A 2012, 116, 119–131
The Journal of Physical Chemistry A ARTICLE
(15) Welter, S.; Lafolet, F.; Cecchetto, E.; Vergeer, F.; De Cola, L.
Chem. Phys. Chem. 2005,6, 2417.
(16) Indelli, M. T.; Chiorboli, C.; Flamigni, L.; De Cola, L.;
Scandola, F. Inorg. Chem. 2007,46, 56305641.
(17) Wenger, O. Acc. Chem. Res. 2011,44,2535.
(18) Chiorboli, C.; Rodgers, M. A. J.; Scandola, F. J. Am. Chem. Soc.
2003,125, 483.
(19) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.;
Robb, M. A.; Cheeseman, J. R.; Montgomery, J. A., Jr.; Vreven, T.;
Kudin, K. N.; Burant, J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi, J.;
Barone, V.; Mennucci, B.; Cossi, M.; Scalmani, G.; Rega, N.; Petersson,
G. A.; Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.;
Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.;
Klene, M.; Li, X.; Knox, J. E.; Hratchian, H. P.; Cross, J. B.; Bakken, V.;
Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.;
Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Ayala, P. Y.;
Morokuma, K.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Zakrzewski,
V. G.; Dapprich, S.; Daniels, A. D.; Strain, M. C.; Farkas, O.; Malick,
D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.; Ortiz, J. V.; Cui,
Q.;Baboul,A.G.;Cliord, S.; Cioslowski, J.; Stefanov, B. B.; Liu, G.;
Liashenko, A.; Piskorz, P.; Komaromi, I.; Martin, R. L.; Fox, D. J.; Keith, T.;
Al-Laham, M. A.; Peng, C. Y.; Nanayakkara, A.; Challacombe, M.; Gill,
P. M. W.; Johnson, B.; Chen, W.; Wong, M. W.; Gonzalez, C.; Pople, J. A.
Gaussian 03, revision C.02; Gaussian, Inc.: Wallingford CT, 2004.
(20) Elliott, C. M.; Freitag, R. A.; Blaney, D. D. J. Am. Chem. Soc. 1985,
107, 46474655.
(21) A minor (<5%) long-lived component, with lifetime identical to
that of the corresponding model compound, is attributed to impurities
of nonquaternarized species.
(22) The appearance of the MLCT transient is preceded by some fast
spectral changes in the <10 ps time scale (Figure S5a in the Supporting
Information). Such initial processes are likely related to the fact that the
400 nm pump light does not give direct MLCT excitation (as was the case
for n= 1) but rather excitation in the new bands (Figure 2) of a dierent
orbital origin (see Discussion). As expected in this hypothesis, the initial
fast processes disappear (the formation of the MLCT transient becomes
instantaneous) when the excitation is performed at 500 nm (Figure S5b in
the Supporting Information).
(23) Welter, S.; Salluce, N.; Benetti, A.; Rot, N.; Belser, P.; Sonar, P.;
Grimsdale, A. C.; Mu1llen, K.; Lutz, M.; Spek, A. L.; De Cola, L. Inorg.
Chem. 2005,44, 47064718.
(24) Quaternarized bipyridines typically absorb in the UV, with
absorption maxima at wavelengths <290 nm.
(25) The orbitals depicted in Figure 8 for Ru(bpy)
2
(bpyph
2
DQ)
4+
are HOMO3, HOMO, LUMO, LUMO+2 (the predominant
localization of the other orbitals in this energy range, see Figure S10 of
the Supporting Information, is as follows: HOMO2 and HOMO1, d
Ru; LUMO+1, π* DQ). The orbitals depicted in Figure 8 for Ru(bpy)
2
-
(bpyph
4
DQ)
4+
are HOMO1, HOMO, LUMO, LUMO+2 (the
predominant localization of the other orbitals in this energy range, see
Figure S10 of Supporting Information, is as follows: HOMO3 and
HOMO2, d Ru; LUMO+1, π* DQ).
(26) The energy of the triplet MLCT state is obtained from the 77 K
emission of the symmetric Ru(II) dyads.
23
(27) The energy of the electron-transfer state, containing an oxi-
dized ruthenium center and a reduced DQ
2+
unit, is obtained from the
measured redox potentials, considering the appropriate correction for
the electrostatic work terms.
28
(28) (a) Rehm, D.; Weller, A. Ber. Bunsen-Ges. Phys. Chem. 1969,
73, 834839. (b) Weller, A. Z. Phys. Chem. 1982,133,9398.
(29) The terms charge separationand charge recombinationare
strictly appropriate only for neutral dyads. We use them conventionally,
however, to denote the forward and back electron-transfer processes.
(30) Malone, R. A.; Kelley, D. F. J. Chem. Phys. 1991,95, 89708976.
(31) Shaw, G. B.; Brown, C. L.; Papanikolas, J. N. J. Phys. Chem. A
2002,106, 14831495.
(32) In complexes of lower symmetry, ILET is expected to be faster
because of the nite driving force of the process. For instance, going
from Os(bpy)
32+
to a heteroleptic analogue with a driving force of ca.
0.1 eV, a reduction in lifetime from 8.7 to 1.5 ps was observed.
31
In our
case, the driving force for ligand-to-ligand electron transfer is expected to
be very small (ca. 0.03 eV, as estimated from electrochemical
measurements
23
on related systems).
(33) (a) Elliott, C. M.; Freitag, R. A. J. Chem. Soc., Chem. Commun.
1985, 156.
(34) The detailed energy levels of the various MLCT states could be
somewhat dierent in Ru(bpy)
2
(bpyphDQ)
4+
and in Ru(bpy)
2
-
(bpyphbpy)
2+
.
(35) The spectral changes in this initial time interval (bleaching
of the MLCT absorption, decay of positive absorption at 520 nm,
Figure S5a of the Supporting Information) very likely represent the
relaxation from the bridge-toDQ charge-transfer states populated by
400 nm light absorption to the MLCT triplet manifold, besides any
equilibration between the MLCT states.
(36) In principle, a decrease in FCWD is expected to occur from n=
1ton= 2, because of the distance dependence of the solvent reo-
rganizational energy. The eect, however, is expected to be small. In fact,
the eect depends on the relative weight of solvent and inner modes in
the overall of reorganization. An upper limiting value of 20% can be
calculated upon complete neglect of inner mode displacements.
(37) With a single p-phenylene spacer, the energy of the electron-
transfer virtual state is expected to be so high (>4 eV) to make any
superexchange contribution to charge separation negligible.
(38) Strictly speaking, the energy dierences in eq 2 should be taken
at the geometry where reactant and product states are isoenergetic
(intersection geometry). Given the substantially exergonic nature of
both processes, the energy of the reactant state can be used as a decent
approximation.
(39) This prediction is conrmed by DFT calculations of the lowest
triplet state for n= 4, where the singly occupied orbitals are bridge-
localized and DQ-localized MOs, and the spin density is conned to the
bridgeDQ region (Figure S11 of the Supporting Information)
  • Article
    Polyoxometalate (POM)-associated charge-separated states, formed by photoinduced oxidation of a covalently attached photosensitizer and reduction of the POM, have attracted much attention due to the remarkable catalytic properties of reduced POMs. However, short lifetimes of the POM-associated charge-separated state, which in some cases lead to backward electron transfer being more rapid than the formation of the charge-separated state itself, are generally observed. Recently, we reported on the first example of a relative long-lived (τ = 470 ns) charge-separated state in a Ru(II) bis(terpyridine)-POM molecular dyad. In this manuscript, further studies on extended molecular structures – two molecular triads – which contain an additional electron donor, phenothiazine (PTZ) or π-extended tetrathiafulvalene (exTTF), are discussed. We show that excitation of the photosensitizer leads to the population of two distinct MLCT states, which differ in the distribution of excess electron density on the two distinct tpy ligands. These two MLCT states decay separately and, thus, constitute the starting points for distinct intramolecular electron-transfer pathways leading to the simultaneous population of two partially charge-separated states, i.e. PTZ·+-Ru(tpy)2·--POM and PTZ-RuIII(tpy)2-POM·-. These independent decay pathways are unaffected by the choice of the electron donor. Thus, the initial charge distribution within the coordination environment of the photocenter determines the nature of the subsequent (partially) charge separated state that is formed in the triads. These results might open new avenues to design molecular interfaces, in which the directionality of electron transfer can be tuned by the choice of initial excitation.
  • Article
    Subtractively normalized interfacial Fourier transform infrared spectroscopy (SNIFTIRS) and electrochemical monitoring demonstrated how insertion of Os or Ru bipiridyl substituents into the backbone of regio-regular polythiophenes provided low energy access to the polymer backbone. The good coordination between the polymeric backbone and the metal complex effectively enables delocalization of charge across the polymeric structure whilst the metallic centers act as "electronic gates" (during MLCT) or "hole injectors" (during LMCT) thus enabling the polymers to be redox switched (and hence rendered conductive) at lower potentials, and at faster rate, than their unsubstituted counterparts.
  • Article
    Ligand field (LF) states have been present in discussions on the photophysics and photochemistry of ruthenium-iminic chromophores for decades, although there is very little documented direct evidence of them. We studied the picosecond transient absorption (TA) spectroscopy of four {RuII(imine)} complexes that respond to the formula trans-[Ru(L)4(X)2], where L is either pyridine (py) or 4-methoxypyridine (MeOpy) and X is either cyanide or thiocyanate. Dicyano compounds behave as most ruthenium polypyridines and their LF states remain silent. In contrast, in the dithiocyanate complexes we found clear spectroscopic evidence of the participation of LF states in the MLCT decay pathway. These states are of donor and acceptor character simultaneously and this is manifested in the presence of MLCT and LMCT transient absorption bands of similar energy. Spectroelectrochemical techniques supported the interpretation of the absorption features of MLCT states, and DFT methods helped to assign their spectroscopic signatures and provided strong evidence on the nature of LF states.
  • Article
    Photoinduced electron transfer continues to be a key process for the design of artificial systems capable to perform an efficient solar energy conversion. In particular, linearly-arranged donor–bridge–acceptor dyads have greatly contributed to shine light on the various factors that must be taken into account when designing systems for obtaining long-lived charge separation, a useful property on the route to artificial photosynthesis. Here we summarized the results we recently obtained on the photoinduced electron transfer processes occurring in Os(II)-bis(terpyridine)-(bi)pyridinium dyads. In particular, we will focus on the role of the bridge in forward and backward electron transfer processes, and on the possibility of obtaining efficient photoinduced charge separation even when the driving force for the electron transfer process approaches zero. This latter point can be of considerable interest when several electron transfer steps are considered to ultimately yield long-range charge-separated state, with minimal energy losses from the initial, light-prepared localized excited state.
  • Article
    New dyad and triad systems based on a zinc porphyrin (ZnP), a naphthalenediimide (NDI), and a ferrocene (Fc) as molecular components, linked by 1,2,3-triazole bridges, ZnP-NDI (3) and Fc-ZnP-NDI (4), have been synthesized. Their photophysical behavior has been investigated by both visible excitation of the ZnP chromophore and UV excitation of the NDI unit. Dyad 3 exhibits relatively inefficient quenching of the ZnP singlet excited state, slow charge separation, and fast charge recombination processes. Excitation of the NDI chromophore, on the other hand, leads to charge separation by both singlet and triplet quenching pathways, with the singlet charge-separated (CS) state recombining in a subnanosecond time scale and the triplet CS state decaying in ca. 90 ns. In the triad system 4, primary formation of the Fc-ZnP+-NDI– charge-separated state is followed by a secondary hole shift process from ZnP to Fc. The product of the stepwise charge separation, Fc+-ZnP-NDI–, undergoes recombination to the ground state in 1.9 μs. The charge-separated states are always formed more efficiently upon NDI excitation than upon ZnP excitation. DFT calculations on a bridge–acceptor fragment show that the bridge is expected to mediate a fast donor-to-bridge-to-acceptor electron cascade following excitation of the acceptor. More generally, triazole bridges may behave asymmetrically with respect to photoinduced electron transfer in dyads, kinetically favoring hole-transfer pathways triggered by excitation of the acceptor over electron-transfer pathways promoted by excitation of the donor.
  • Article
    Three tetrapodal ligands 1,2,4,5‐tetrakis[4‐(4,5‐diazafluoren‐9‐ylimino)phenoxymethyl]benzene (L1), 1,2,4,5‐tetrakis[2‐(4,5‐diazafluoren‐9‐ylimino)phenoxymethyl]benzene (L2), and 1,2,4,5‐tetrakis[(4,5‐diazafluoren‐9‐ylimino)methyl]benzene (L3), and their corresponding RuII polypyridyl complexes [{Ru(bpy)2}4(μ4‐L1–3)](PF6)8 (bpy = 2,2′‐bipyridine) were synthesized and characterized. The spectroscopic behavior of the three complexes was investigated by UV/Vis absorption and emission spectroscopy. They display metal‐to‐ligand charge transfer (MLCT) absorptions at around 443 nm in CH3CN solution at room temperature, and emission at around 575 nm in EtOH/MeOH (4:1, v/v) glassy matrix at 77 K. Electrochemical studies of the three complexes show one RuII‐based oxidation at around 1.35 V and three ligand‐based reductions.
  • Article
    Photoinduced electron transfer plays key roles in many areas of chemistry. Superexchange is an effective model to rationalize photoinduced electron transfer, particularly when molecular bridges between donor and acceptor subunits are present. In this tutorial review we discuss, within a superexchange framework, the complex role played by the bridge, with an emphasis on differences between thermal and photoinduced electron transfer, oxidative and reductive photoinduced processes, charge separation and charge recombination. Modular bridges are also considered, with specific attention to the distance dependence of donor-acceptor electronic coupling and electron transfer rate constants. The possibility of transition, depending on the bridge energetics, from coherent donor-acceptor electron transfer to incoherent charge injection and hopping through the bridge is also discussed. Finally, conceptual analogies between bridge effects in photoinduced electron transfer and optical intervalence transfer are outlined. Selected experimental examples, instrumental to illustration of the principles, are discussed.
  • Article
    Photoinduced electron transfer is a topical issue in chemistry. In multicomponent donor-bridge-acceptor systems, electron transfer is usually discussed within the frame of superexchange theory, which takes into account electronic coupling mediated by virtual states involving bridge orbitals. However, the schematization used for superexchange in thermal electron transfer processes is not suitable to immediately understand some intriguing aspects of photoinduced charge separation and recombination processes, which are only uncovered by analyzing the virtual states involved in forward and backward excited-state electron transfer. In particular, for oxidative photoinduced electron transfer, a low-energy virtual state which cannot mediate the forward charge separation can efficiently mediate charge recombination via the hole-transfer superexchange route, whereas for reductive photoinduced electron transfer, a low-energy virtual state which cannot mediate the forward process can efficiently mediate charge recombination via electron-transfer superexchange. As a consequence, to obtain long-lived charge-separated states upon oxidative photoinduced electron transfer in donor-bridge-acceptor systems it is preferable to avoid easy-to-oxidize bridges, whereas easy-to-reduce bridges should better be avoided in reductive photoinduced charge separation. These considerations, exemplified by the analysis of some literature cases, can be useful hints for the design of long-lived charge-separated states.
  • Article
    This review summarises the literature reported in 2011 on the photophysical properties of metal complexes and their polynuclear supramolecular assemblies.
  • Article
    Two iridium(III) complexes displaying for one a high HOMO-LUMO gap and for the other a weaker gap were linked in a controlled and logical manner to closo-p-carborane spacers. The bridging ligand is composed of 5-ethynyl-2,2'-bipyridine units, and the peripherical Ir-ligands are orthometalated 2',4'-difluoro-2-phenylpyridine (dfppy) (λabs at 400 nm for the "Ir(dfppy)2(bpy')") for the energy donor fragment and dibenzo[a,c]phenazine (dbpz) (λabs at 525 nm for "Ir(dbpz)2(bpy')") for the energy acceptor fragment.Redox, spectroscopic, and photophysical properties for models and the donor-carborane-acceptor complex were determined. Efficient energy transfer from the "Ir(dfppy)2(bpy')" moiety to the "Ir(dbpz)2(bpy')" fragment is occurring with a rate constant of 3.3 × 108 s-1 despite weak electronic coupling through the inert p-carborane spacer. From flash photolysis experiments it is shown that, by excitation of the donor, a low lying triplet state localized on the acceptor bridging ligand side is formed which decays by conversion to the 3MLCT of the acceptor fragment which phosphoresces at 644 nm.
  • Article
    The solvent dependent free enthalpies of exciplex and radical ion pair formation in solution have been calculated on the basis of the experimentally obtained thermodynamic and spectroscopic values and with the aid of theoretical considerations concerning solvent polarity effects in general.
  • Article
    The role of bridging groups in promoting electron transfer between metal ions in solution is discussed. Interactions between orbitals of the metal ions through those of the bridging groups give rise to two possible exchange mechanisms-double exchange and superexchange. Expressions for the exchange frequencies associated with these are derived and their dependence on various factors such as orbital symmetry, overlap and redox potentials is discussed. Particular consideration is given to electron transfer through bridging groups containing conjugated π-electron systems.
  • Article
    Full-text available
    This paper considers electron transfer between biological molecules in terms of a nonadiabatic multiphonon nonradiative decay process in a dense medium. This theoretical approach is analogous to an extended quantum mechanical theory of outer sphere electron transfer processes, incorporating the effects of both low-frequency medium phonon modes and the high-frequency molecular modes. An explicit, compact and useful expression for the electron transfer probability is derived, which is valid throughout the entire temperature range, exhibiting a continuous transition from temperature independent tunneling between nuclear potential surfaces at low temperatures to an activated rate expression at high temperatures. This result drastically differs at low temperatures from the common, semiclassical, Gaussian approximation for the transition probability. The experimental data of De Vault and Chance [Biophys. J. 6, 825 (1966)] on the temperature dependence of the rate of electron transfer from cytochrome to the chlorophyll reaction center in the photosynthetic bacterium Chromatium are properly accounted for in terms of the present theory.
  • Article
    Pulse radiolysis has been used to observe and measure the kinetics for intermolecular positive charge (hole) transfer from biphenyl+ or pyrene+ ions to TMPD molecules in rigid 2-chlorobutane glass at 77 K. These hole transfers occur over distances of about 17 A˚ at 10−6 s, increasing to about 34 A˚ at 102 s. The kinetic data are interpreted in terms of current theories which treat electron transfer processes as radiationless transitions. Estimates of the required electron exchange interactions based on the usual electron tunneling models can not explain the fast reactions observed, even when Coulombic effects on the ’’barrier’’ are considered. A superexchange model is proposed which involves interactions propagated by both negative and positive ion states of the solvent. This model adequately interprets the data in terms of a dominant role of the solvent positive ion states, and is also applicable to negative charge transfer in the condensed phase. In samples containing only one solute (biphenyl or pyrene), ion recombination with Cl− removed about one third of the positive ions of the solute between 10−6 and 102 s. The data on intermolecular hole transfer between two solutes must be corrected for the effects of ion recombination. The correction is simple and quantitative only when it is possible to observe decay of the reactants without substantial spectral overlap from the products. Because the effects of ion recombination on product growths are complex, it is not presently possible to use the growths to measure reliably the kinetics intermolecular positive charge transfer.
  • Article
    A non-luminescent ruthenium tris(bipyridyl) complex containing a covalently attached diquat electron has been prepared, and its electrochemical, spectroelectrochemical, and photochemical properties have been investigated.
  • Electron-transfer reactions between ions and molecules in solution have been the subject of considerable experimental study during the past three decades. Experimental results have also been obtained on related phenomena, such as reactions between ions or molecules and electrodes, charge-transfer spectra, photoelectric emission spectra of ionic solutions, chemiluminescent electron transfers, electron transfer through frozen media, and electron transfer through thin hydrocarbon-like films on electrodes.
  • Article
    Dinuclear complexes containing ruthenium and osmium-based bis-terpyridyl chromophoric termini are prepared and their luminescence properties investigated. The two chromophoric units are connected by 1,4-phenylenes only, or phenylene and bicyclooctane spacers. In this way complete control of the geometry of the dinuclear complexes is achieved and these rigid species resemble molecular rods or girders featuring intermetal distances in the range 11-24 Angstrom. The Ru --> Os transfer of electronic excitation is energetically allowed and we have studied the effect on this process both of the intermetal separation and the electronic properties of the spacers, The main conclusions are that the phenylene spacers are very efficient in transmitting the intermetal electronic communication but an important role is also played by the spatial localization of the metal-to-ligand charge-transfer excited states involved in the excitation-transfer process.
  • Article
    A method is described for calculating coupling integrals between donor and acceptor ions in class II mixed valence compounds. Metal–ligand–metal interactions are simulated by a perturbation model, which is then used to calculate valence delocalization coefficients from empirical and theoretical information On metal → ligand and ligand → metal charge transfer in the component single-valence complexes. The model is tested by applying it to FeII,III cyanides (Prussian Blue and its discrete molecular analogues) and silicates (e.g. biotite micas). Good agreement is found between observed and calculated intensities of the mixed valence charge transfer transitions.
  • Article
    We have synthesized nine rodlike compounds of nanometric dimension with general formula [M(bpy)3-(ph)n-M‘(bpy)3]4+ (M = M‘ = Ru(II); M = M‘ = Os(II); M = Ru(II), M‘ = Os(II); bpy = 2,2‘-bipyridine; ph = 1,4-phenylene; n = 3, 5, 7; the central phenylene unit bears two alkyl chains for solubility reasons; the metal-to metal distance is 4.2 nm for the longest spacer). The absorption spectra and the luminescence properties (emission spectra, quantum yields, and excited-state lifetimes) of the nine dinuclear complexes have been investigated in acetonitrile solution at 293 K and in butyronitrile rigid matrix at 77 K. The results obtained have been compared with those found for the separated chromophoric units ([Ru(bpy)3]2+, [Os(bpy)3]2+, and oligophenylene derivatives). The absorption spectrum of each dinuclear complex is essentially equal to the sum of the spectra of the component species, showing that intercomponent electronic interactions are weak. In the homodinuclear compounds, the strong fluorescence of the oligophenylene spacers is completely quenched by energy transfer to the metal-based units, which exhibit their characteristic metal-to-ligand charge-transfer (MLCT) phosphorescence. In the heterodinuclear compounds, besides complete quenching of the fluorescence of the oligophenylene spacers, a quenching of the phosphorescence of the [Ru(bpy)3]2+ chromophoric unit and a parallel sensitization of the phosphorescence of the [Os(bpy)3]2+ chromophoric unit are observed, indicating the occurrence of electronic energy transfer. The rate of the energy-transfer process from the [Ru(bpy)3]2+ to the [Os(bpy)3]2+ unit is practically temperature independent and decreases with increasing length of the oligophenylene spacer (in acetonitrile solution at 293 K, ken = 6.7 × 108 s-1 for n = 3; ken = 1.0 × 107 s-1 for n = 5; ken = 1.3 × 106 s-1 for n = 7). It is shown that such an energy-transfer process takes place via a Dexter-type mechanism (superexchange interaction) with an attenuation coefficient of 0.32 per Å and 1.5 per interposed phenylene unit.
  • Article
    We have investigated the geometry and the nature of optically allowed transitions in neutral and charged phenylene-based oligomers by means of Hartree-Fock calculations. Geometry optimizations are performed using the semiempirical Austin Model 1 (AM1) method for oligomers containing from two (2P) to twelve (12P) benzene rings. The transition energies and related intensities of the optical-absorption spectra are calculated by means of the intermediate neglect of differential overlap Hamiltonian that is combined with a single configuration interaction technique in order to include electron correlation effects in the description of the excited states. The calculations show that two subgap absorption features appear in short oligomers carrying a single charge (polaron), whereas a single intense peak is observed in the presence of two charges (bipolaron). These results are consistent with a wide range of experimental and theoretical data obtained for various conjugated oligomers. Interestingly, the appearance of a second subgap feature is predicted in the spectra of long doubly oxidized chains as well as for chains supporting interacting bipolarons.