Hole-transfer dyads and triads based on perylene monoimide, quaterthiophene, and extended tetrathiafulvalene

Article abstract of DOI:10.1002/chem.201000640

Two families of dyad and triad systems based on perylene monoimide (PMI), quaterthiophene (QT), and 9,10-bis(1,3-dithiol-2-ylidene)-9,10-dihydroanthracene (extended tetrathiafulvalene, exTTF) molecular components have been designed and synthesized. The dyads (D1 and D2) are of the PMI-QT type and the triads (T1 and T2) of the PMI-QT-exTTF type. The two families differ in the saturated or unsaturated nature of the linker groups (ethynylene in D1 and T1, ethylene in D2 and T2) that bridge the molecular components. The dyads and triads have been characterized by electrochemical, photophysical, and computational methods. Both the experimental and the computational (DFT) results indicate that in the unsaturated systems strong intercomponent interactions lead to substantial perturbation of the properties of the subunits. In particular, in T1, delocalization is particularly effective between the QT and exTTF units, which would be better viewed combined as a single electronic subsystem. For the dyad systems, the photophysics observed following excitation of the PMI unit is solvent-dependent. In moderately polar solvents (dichloromethane, diethyl ether) fast charge separation is followed by recombination to the ground state. In toluene, slow conversion to the chargeseparated state is followed by intersystem crossing and recombination to yield the triplet state of the PMI unit. The behavior of the triads, on the other hand, is remarkably similar to that of the corresponding dyads, which indicates that, after primary charge separation, hole shift from the oxidized QT component to exTTF is quite inefficient. This unexpected result has been rationalized on the basis of the anomalous (simultaneous two-electron oxidation) electrochemistry of exTTF and with the help of DFT calculations. In fact, although exTTF is electrochemically easier to oxidize than QT by around 0.6 V, the one-electron redox orbitals (HOMOs) of the two units in triad T2 are almost degenerate.

Full text of DOI:10.1002/chem.201000640

DOI: 10.1002/chem.201000640
Hole-Transfer Dyads and Triads Based on Perylene Monoimide,
Quaterthiophene, and Extended Tetrathiafulvalene
Julien Boixel,[a] Errol Blart,[a] Yann Pellegrin,[a] Fabrice Odobel,*[a] Nicola Perin,[b]
Claudio Chiorboli,[c] Sandro Fracasso,[b] Marcella Ravaglia,[b] and Franco Scandola*[b, d]
Abstract: Two families of dyad and
triad systems based on perylene monoi-
mide (PMI), quaterthiophene (QT),
and 9,10-bis(1,3-dithiol-2-ylidene)-9,10-
dihydroanthracene (extended tetrathia-
fulvalene, exTTF) molecular compo-
nents have been designed and synthe-
sized. The dyads (D1 and D2) are of
the PMI–QT type and the triads (T1
and T2) of the PMI–QT–exTTF type.
The two families differ in the saturated
or unsaturated nature of the linker
groups (ethynylene in D1 and T1, eth-
ylene in D2 and T2) that bridge the
molecular components. The dyads and
triads have been characterized by elec-
trochemical, photophysical, and com-
putational methods. Both the experi-
mental and the computational (DFT)
results indicate that in the unsaturated
systems strong intercomponent interac-
tions lead to substantial perturbation
of the properties of the subunits. In
particular, in T1, delocalization is par-
ticularly effective between the QT and
exTTF units, which would be better
viewed combined as a single electronic
subsystem. For the dyad systems, the
photophysics observed following exci-
tation of the PMI unit is solvent-depen-
dent. In moderately polar solvents (di-
chloromethane, diethyl ether) fast
charge separation is followed by re-
combination to the ground state. In tol-
uene, slow conversion to the charge-
separated state is followed by intersys-
tem crossing and recombination to
yield the triplet state of the PMI unit.
The behavior of the triads, on the other
hand, is remarkably similar to that of
the corresponding dyads, which indi-
cates that, after primary charge separa-
tion, hole shift from the oxidized QT
component to exTTF is quite ineffi-
cient. This unexpected result has been
rationalized on the basis of the anoma-
lous (simultaneous two-electron oxida-
tion) electrochemistry of exTTF and
with the help of DFT calculations. In
fact, although exTTF is electrochemi-
cally easier to oxidize than QT by
around 0.6 V, the one-electron redox
orbitals (HOMOs) of the two units in
triad T2 are almost degenerate.
Keywords: charge separation
donor–acceptor systems
·
·
electron transfer · photophysics ·
time-resolved spectroscopy
Introduction
[a] Dr. J. Boixel, Dr. E. Blart, Dr. Y. Pellegrin, Dr. F. Odobel
Chimie et Interdisciplinaritꢀ: Synthꢁse, Analyse, Modꢀlisation
(CEISAM)
The rational molecular design of photomolecular assemblies
for efficient light-induced charge separation is a topic of
considerable importance owing to its fundamental and prac-
tical implications for solar energy conversion (artificial pho-
tosynthesis[1–5] and photovoltaics[6–8]). Towards this goal, the
general approach, which is essentially inspired by the strat-
egy adopted by Nature in photosynthesis, involves creating a
multistep electron-transfer chain along a molecular assembly
that exhibits a defined redox gradient that allows the vecto-
rial spatial separation of the photogenerated charges. In this
context, many covalently linked donor–bridge–acceptor (D–
B–A) systems have been prepared.[5,9–12] To produce a long-
lived charge-separated state with high quantum yield, it is
essential to maximize the rate of forward charge separation
while reducing the rate of charge recombination by separat-
ing the two photogenerated charges over a large distance.
So far the vast majority of triads designed for multistep elec-
Universitꢀ de Nantes, CNRS UMR CNRS n8 6230
2, rue de la Houssiniꢁre
BP 92208–44322 Nantes Cedex 3 (France)
Fax : (+33)251125402
E-mail: Fabrice.Odobel@univ-nantes.fr
[b] N. Perin, S. Fracasso, Dr. M. Ravaglia, Prof. F. Scandola
Dipartimento di Chimica, Universitaꢂ di Ferrara
via L. Borsari 46, 44100 Ferrara (Italy)
Fax : (+39)0532240709
E-mail: snf@unife.it
[c] C. Chiorboli
ISOF-CNR, Sezione di Ferrara, via L. Borsari 46
44100 Ferrara (Italy)
[d] Prof. F. Scandola
Centro di Ricerca Interuniversitario per la Conversione Chimica
dell’Energia Solare
Sezione di Ferrara, 44100 Ferrara (Italy)
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201000640.
9140
ꢃ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 9140 – 9153

FULL PAPER
tron transfer are of the type
shown in Figure 1 a,b, that is,
based on a light-absorbing unit
(photosensitizer, P) sandwiched
between a donor (D) and an ac-
ceptor (A) to form a triad of
the type D–P–A or a photosen-
sitizer covalently linked to sev-
eral electron acceptors to give a
triad of type P–A1–A2.[9–12] In
the first case, the charge-sepa-
ration process takes place by
sequential electron- and hole-
transfer processes,[5,11–16] where-
as in the second, the charge
separation takes place through
plore the possibility of extending the lifetime of the charge-
separated state by a hopping mechanism. In this study the
first electron donor is quaterthiophene (QT) because it is a
rigid p-conjugated molecule, the electronic properties of
which have been proved to be particularly suitable for
acting as an effective bridging group or electron donor in
molecular systems for photoinduced charge separation. For
example, Otsubo and co-workers reported several interest-
ing fullerene- or porphyrin-based photochemical systems
connected to oligothiophenes, mostly oriented towards the
development of plastic solar cells.[6,26–29] Perylene monoimide
(PMI) is an attractive light collector to use for our purpose
because it displays quite high molar absorptivity and strong
electron-acceptor characteristics suitable in its photoexcited
state for pulling away an electron from the quaterthiophene
unit.[30] Furthermore, the radical anion of PMI exhibits a
clear spectroscopic signature, which will ease the photophys-
ical characterization.[31] In addition, PMI was previously con-
nected to an oligothiophene chain and photoinduced charge
separation was evidenced.[19,32–34] Finally, the 9,10-bis(1,3-di-
thiol-2-ylidene)-9,10-dihydroanthracene, also known as ex-
tended tetrathiafulvalene (exTTF), is an excellent electron
donor, which has previously been used as a component in
photomolecular arrays.[8,35–40] Another attractive feature of
exTTF is its bis-electron-donating capability, which makes it
a potentially useful molecule to serve as a hole reservoir for
photoaccumulative electron transfer.[41–46]
We describe herein the synthesis and extensive characteri-
zation of original photomolecular triads and demonstrate
that photoinduced hole transfer between PMI and quater-
thiophene is a very fast process, irrespective of the solvent
and the linkage between these units. Surprisingly, in spite of
the presumably large driving force, the secondary hole-shift
reaction (ꢀP–D1+–D2!ꢀP–D1–D2+) was not observed in
these triads. Quantum chemical calculations have been used
to rationalize this unexpected behavior, namely by showing
that the first oxidation potential of exTTF is underestimated
when determined by cyclic voltammetry.
Figure 1. Schematic represen-
tation of the photoinduced
charge separation in different
types of triads.
two sequential electron-transfer steps.[17,18]
The reverse design, that is, the development of triads con-
sisting of an electron-acceptor photosensitizer linked to dif-
ferent electron donors (such as P–D1–D2, Figure 1c) in
which the photosensitizer decays by reductive quenching fol-
lowed by a cascade hole transfer chain, are rare.[19–22] We
have previously designed and reported bis-porphyrin-based
photomolecular systems that undergo fast single-step photo-
induced electron transfer over distances of 20–45 ꢄ by a su-
perexchange mechanism.[23–25] However, the very fast and ef-
ficient forward photoinduced charge separation was ob-
tained at the expense of an also fast charge recombination,
which precludes the formation of a long-lived charge-sepa-
rated state. Herein, we describe two new molecular triads of
the type P–D1–D2 (T1 and T2; Scheme 1), designed to ex-
Results and Discussion
Synthetic procedures: The synthesis of T1 and T2 is shown
in Scheme 2 and relies on iterative Sonogashira cross-cou-
pling reactions. The bromo-perylene imide 1, which is one
of the key synthons, was prepared by O-arylation of the cor-
responding trisbromo-perylene imide by copper-catalyzed
Ullmann reaction, as described previously.[47] Imide 1 was
subjected to the first Sonogashira cross-coupling reaction
with trimethylsilylacetylene to afford 2 in a yield of 87%
using classic conditions for this transformation. The trime-
thylsilyl protecting group was cleaved by using potassium
carbonate and the resulting perylene imide 3 was treated
with the known iodo-quaterthiophene-triisopropylsilylacety-
lene 4[24] in a second Sonogashira cross-coupling reaction
using palladium tetrakis(triphenylphosphine) and copper
iodide as the catalytic system to give 5 in a yield of 49%.
Scheme 1. Structures of the dyads and triads studied in this work.
Chem. Eur. J. 2010, 16, 9140 – 9153
ꢃ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
9141

F. Odobel, F. Scandola et al.
lene in a Sonogashira cross-cou-
pling reaction to give 10. Subse-
quently the triisopropylsilyl
protecting group in 10 was
cleaved with tetrabutylammoni-
um fluoride and the resulting
acetylene 11 was connected to
bromo-perylene imide 1 in a
second Sonogashira cross-cou-
pling reaction to give D1. Dyad
D1 was finally reduced with hy-
drogen to afford the expected
dyad D2. The new compounds
were satisfactorily characterized
by 1H NMR spectroscopy and
high-resolution mass spectrom-
etry
Electronic absorption spectra:
Figure 2 shows a superimposi-
tion of the spectra of triad T1
and dyad D1 along with those
of their constituents and the
spectroscopic data of the com-
pounds are presented in
Table 1.
The spectra of the two com-
pounds T1 and D1 are dominat-
ed by two broad absorption
bands located at around 430
and 560 nm, but each band is
composed of several transitions
that occur in the different chro-
Scheme 2. Preparation of the triads T1 and T2. a) CuI, [Pd
508C (87%); b) K2CO3, CH2Cl2, MeOH, RT (100%); c) CuI, [PdAHCTUNGTRENNUNG
ACHTUNGTRENNUNG
THF, RT (100%); e) BuLi, THF, ꢀ788C (49%); f) CuI, [Pd
ACHTUNGTRENNUNG
(100%).
mophores. The perylene imide
dye displays an intense absorp-
Removal of the triisopropylsilyl group with tetrabutylammo-
nium fluoride afforded 6 in an almost quantitative yield. 2-
Iodo-9,10-bis(4,5-dimethyl-1,3-dithiol-2-ylidene)-9,10-dihy-
droanthracene (9) was prepared by the Wittig–Horner reac-
tion by condensation of the 2-iodoanthraquinone (7) with
the carbanion generated from dimethyl (4,5-dimethyl-1,3-di-
thiol-2-yl)phosphonate (8) and butyllithium.[48–50] The result-
ing iodo-substituted extended TTF 9 was finally grafted
onto 6 in the last Sonogashira cross-coupling reaction to
afford the first triad T1 in a yield of 63% after purification.
The second triad T2 was obtained by reduction of the
triple bonds in triad T1 under a hydrogen atmosphere using
palladium on charcoal as the catalyst (Scheme 2). Note that
triad T2 was systematically contaminated with a small
amount of not fully reduced triad (<15%), as judged by the
presence of a small signal for [Mꢀ2H]+ in the mass spec-
trum of T2. The separation of these two compounds was not
possible by column chromatography and therefore all the
analyses were conducted on the mixture.
tion band at around 530 nm, which corresponds to a p!p*
transition, the vibronic bands of which are still visible in the
spectrum of trimethylsilylacetylene-perylene imide 2.[51,52]
Quaterthiophene exhibits
a
broad p!p* transition at
The synthesis of dyads D1 and D2 is outlined in
Scheme 3. It starts with the iodo-quaterthiophene-triisopro-
pylsilylacetylene 4, which was first capped with phenylacety-
Scheme 3. Preparation of the dyads D1 and D2. a) CuI, [Pd
THF, Et3N, 80 8C (69%); b) Bu4NF, CH2Cl2, MeOH, RT (100%); c) CuI,
[Pd(PPh3)4], THF, Et3N, 80 8C (7%); d) H2, Pd/C, 408C (100%).
ACHTUNGTRENNUNG(PPh3)4],
AHCTUNGTRENNUNG
9142
www.chemeurj.org
ꢃ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 9140 – 9153

Hole-Transfer Dyads and Triads
FULL PAPER
lapping contribution of the two transitions of exTTF
(Figure 2).
Further clear evidence for ground-state electronic cou-
pling between the units can be found by comparison of the
absorption spectra of the two triads T1 and T2, which differ
significantly in the larger absorbance in the PMI region for
T1 (Figure 3). In the triad T1, there is significant electronic
communication between the linked chromophores, which is
Figure 2. Absorption spectra of dyad D1, triad T1, and of their molecular
components in dichloromethane (absorbance, arbitrary units; molar ab-
sorptivity values are given in Table 1). 10’=5-(triisopropylsilylethyny-
lene)-5’’’-(trimethylsilylethynylene)-3,3’’’-dioctyl[2,2’:5’,2’’:5’’,2’’’]quater-
thiophene.
Table 1. Characteristics absorption data for the dyads, triads, and molec-
ular components recorded in dichloromethane.
labs [nm] (e [mꢀ1 cmꢀ1])
D1
D2
T1
T2
2
276 (45000), 383 (29000), 418 (39000), 528 (44000), 556 (49000)
274 (42000), 393 (24000), 488 (29000), 519 (33000)
273 (37000), 427 (36000), 464 (32000), 554 (36000)
268 (35000), 399 (20000), 489 (13000), 525 (19000)
275 (29000), 494 (22000), 522 (33000)
Figure 3. Absorption spectra of dyad D2, triad T2, and of their molecular
components in dichloromethane (absorbance, arbitrary units; molar ab-
sorption values are given in Table 1). 10’’.
9
288 (30000), 383 (17000), 430 (19000), 453 (24000)
256 (8600), 289 (5800), 404 (21000)
10’[a]
10’’[b]
promoted by the ethynylene linkages that create a p-conju-
gated pathway throughout the molecule. Conversely, in the
triad T2, the presence of saturated ethylene linkages be-
tween each unit interrupts the conjugation and disrupts the
communication between the linked elements. As a result,
the three units behave independently and the spectrum of
T2 can be considered as a reasonable sum of the absorban-
ces of each individual chromophore (Figure 3). In the spec-
trum of triad T2, the tail on the red edge of the perylene
monoimide absorption band is due to a small amount (from
spectral fitting, ca. 15%) of an unsaturated compound that
results from the incomplete reduction of T1 in the synthetic
procedure. After repeated attempts at purification by vari-
ous means (see Synthetic procedures), this impurity proved
to be unavoidable.
398 (28000)
[a] 10’=5-(Triisopropylsilylethynylene)-5’’’-(trimethylsilylethynylene)-
3,3’’’-dioctyl[2,2’:5’,2’’:5’’,2’’’]quaterthiophene. [b] 10’’=3,3’’’-Dioctyl[2,2’:
5’,2’’:5’’,2’’’]quaterthiophene.
around 400 nm located in the same region as those of the
extended TTF. The absorption bands of exTTF are also at-
tributed to p!p* transitions characterized by significant
charge transfer from the dithiolium units to the central aro-
matic core.[36] Clearly the spectra of dyad D1 and triad T1
are not a mere linear combination of the spectra of each
chromophore. This indicates significant ground-state elec-
tronic interactions between each unit in D1 and T1. For ex-
ample, the absorption band of the parent perylene imide 2,
initially located at 522 nm, is red-shifted, respectively, to 556
and 554 nm in dyad D1 and triad T1.
This shift can be the consequence of a decrease in the
HOMO–LUMO gap within the PMI unit as a consequence
of increased intercomponent interactions and/or the appear-
ance of a new band arising from charge transfer from the
quaterthiophene unit to PMI. Both hypotheses are consis-
tent with DFT calculations (see below). Tails attributed to
charge-transfer transitions have previously been observed
by Bꢅuerle and co-workers for perylene imide linked to oli-
gothiophene units.[32–34] The spectrum of triad T1 differs to
that of dyad D1 first by the enhanced intensity of the band
at around 420 nm and secondly by the presence of a should-
er at 464 nm. These changes are brought about by the over-
Electrochemical study: The elementary electron-transfer
processes that can take place following excitation of the
PMI chromophore in the dyad and triad systems are defined
in Scheme 4. To determine the thermodynamic requirements
of the various electron-transfer processes, the redox poten-
tials of the units comprising the dyads and triads were deter-
mined by cyclic voltammetry and square-wave stripping in
dichloromethane. The half-wave potentials are collected in
Table 2. The assignments of the redox processes in the
dyads and triads were made by comparison with the redox
potentials of individual units such as 2, 9, and 10’, which
serve as references for PMI, exTTF, and QT, respectively.
The cathodic region shows two reversible reduction pro-
cesses attributed to the stepwise introduction of an electron
Chem. Eur. J. 2010, 16, 9140 – 9153
ꢃ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
9143

F. Odobel, F. Scandola et al.
this wave is attributed to a two-electron process involving
the direct transformation of the neutral exTTF into the di-
cation.[36,40,53] The reason for the two-electron process (and
for the lack of an intermediate one-electron step) lies in the
large stabilization accompanying the transformation of the
saddle-shaped neutral molecule, in which the central ring is
not aromatic, into a planar anthracenyl system with the 1,3-
dithiolium cations lying almost orthogonal to the newly
formed aromatic core. At higher potentials, we observe the
two stepwise oxidations of quaterthiophene and finally that
of PMI.
The potentials measured in the conjugated systems (D1
and T1) differ significantly to those with the saturated link-
ages (D2 and T2). This confirms the conclusion drawn from
the UV/Vis absorption study that there are significant elec-
tronic interactions between each unit when they are linked
through an ethynylene spacer.
Scheme 4. Elementary electron-transfer processes following excitation of
the dyads and triads.
Table 2. Redox potentials of the dyads and triads measured in dichloro-
methane with 0.15m of Bu4NPF6 as the supporting salt.[a]
The above redox potentials, together with the absorption
and fluorescence spectra, permit the free-energy changes ex-
pected for the electron-transfer processes in Scheme 4 to be
estimated. The free energy of the various charge-transfer
states can be estimated by using the classic Weller formalism
given by Equation (1), whereas the energy of the excited
singlet state of the PMI unit (E00) can be obtained as an
average of the absorption and emission maxima of the
dyads and triads, or of model systems thereof. In Equa-
tion (1), e is the charge of the electron, e0 is the permittivity
of a vacuum, RDA is the donor–acceptor center-to-center dis-
tance, r+ and rꢀ are the radii of the positive and negative
ions, es is the dielectric constant of the solvent used (diethyl
ether: es =4.33; toluene: es =2.38; dichloromethane: es =
8.93), and eref is the dielectric constant of the solvent in
which the electrochemical data are obtained (in this case, di-
chloromethane). The RDA distances and the effective radii
r+ and rꢀ were estimated from DFT-optimized structures
(see below and the Supporting Information). The free-
energy changes calculated for the electron-transfer processes
in Scheme 4 are collected in Table 3.
ERed2
(PMI1ꢀ/2ꢀ
[V]
ERed1
(PMI0/1ꢀ
[V]
EOx1
EOx2
(QT1+/0
[V]
EOx3
(QT2+/1+
[V]
EOx4
(PMI1+/0
)
[V]
G
)
G
)
(exTTF2+/0
[V]
)
E
)
N
)
G
2
9
10’
ꢀ1.36
ꢀ0.95
1.25
0.38
0.94
0.84
0.82
0.83
0.80
1.20
1.12
1.05
1.15
1.00
D1 ꢀ1.29
D2 ꢀ1.30
T1 ꢀ1.26
T2 ꢀ1.36
ꢀ0.93
ꢀ1.04
ꢀ0.94
ꢀ1.01
1.37
1.22
1.37
1.20
0.42
0.34
[a] All the potentials are referenced versus the saturated calomel elec-
trode (SCE).
into the perylene monoimide to form PMIꢀ and PMI2ꢀ, re-
spectively, in which the spin density is mostly delocalized on
the aromatic system of the perylene core (see the molecular
orbital representations below). In the unsaturated systems
(D1 and T1), these reduction processes are shifted anodical-
ly owing to the longer p conjugation through the ethynylene
linkages, which induce a stabilization of the LUMO orbitals
(see the Computational study section below). Conversely,
the inductive effect of the ethylene linkers in D2 and T2
makes the PMI more difficult to reduce than the parent
compound 2.
The anodic region exhibits multiple waves because all
three units are redox active in the electrochemical window
explored under our conditions. In the dyads, the two first
one-electron reversible waves at around 0.82 and 1.1 V cor-
respond, respectively, to the formation of the mono- and di-
cation of the quaterthiophene. They are followed by a rever-
sible oxidation reaction located on the PMI at around 1.3 V.
The presence of the ethynylene linkage in D1 raises the
energy of the HOMO orbital and also increases the coulom-
bic electrostatic repulsion when several positive charges ac-
cumulate in the same molecule. As a consequence, the half-
wave potentials are shifted towards more positive values in
D1 with respect to D2.
e2
DGðAꢀDþÞ ¼ e½EoxðDÞ ꢀ EredðAÞꢁ ꢀ
4pe0esRDA
ð1Þ
ꢀ
ꢁꢀ
ꢁ
e2
1
1
1
1
þ
þ
ꢀ
es eref
8pe0 rþ rꢀ
Photophysical study: The photophysical behavior of the
dyads and triads was studied by emission and time-resolved
absorption techniques in solvents of different polarity. The
results will be presented and discussed in some detail for the
saturated systems D2 and T2, which are better defined from
a supramolecular viewpoint. The unsaturated analogues D1
and T1 will be then be discussed on a comparative basis.
Detailed experimental results are given in the Supporting
Information.
In the triads the additional exTTF unit gives its electro-
chemical signature at around 0.4 V. Referring to published
works reporting the electrochemical properties of exTTF,
Model PMI: The absorption and emission spectra of the
model compound 2 in dichloromethane are shown in Fig-
9144
www.chemeurj.org
ꢃ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 9140 – 9153

Hole-Transfer Dyads and Triads
FULL PAPER
Table 3. Excited-state energies and free-energy changes[a] associated with
the electron-transfer processes presented in Scheme 4.
thane, is shown in Figure 4b. The bleaching observed in the
470–620 nm range with a maximum at around 530 nm is due
to a combination of ground-state absorption bleaching and
stimulated emission, whereas the positive feature with a
maximum at around 730 nm represents an excited-state ab-
sorption. As expected on the basis of the emission lifetime,
this difference spectrum is manifestly constant in the 1 ns
time frame of the ultrafast experiment. No transient was ob-
served in the nanosecond laser photolysis of 2 (timescale
>10 ns), which is consistent with high fluorescence efficien-
cy and the negligible intersystem crossing yield of this mole-
cule.
Dichloromethane
Diethyl ether
Toluene
E00
2.31[b]
ꢀ0.54
ꢀ1.77
2.18[b]
ꢀ0.50
ꢀ1.68
2.31[b]
ꢀ0.59
ꢀ1.72
ꢀ0.42
ꢀ1.30
2.18[b]
ꢀ0.50
ꢀ1.68
ꢀ0.37
ꢀ1.31
2.31[b]
ꢀ0.35
ꢀ1.960
2.18[b]
ꢀ0.31
ꢀ1.87
2.31[b]
ꢀ0.40
ꢀ1.91
ꢀ0.37
ꢀ1.52
2.18[b]
ꢀ0.31
ꢀ1.87
ꢀ0.33
ꢀ1.54
2.33
ꢀ0.05
ꢀ2.28
2.12
D2
D1
DGCS
DGCR
E00
DGCS
DGCR
E00
DGCS
DGCR1
DGHS
DGCR2
E00
0.07
ꢀ2.19
2.30
ꢀ0.08
ꢀ2.22
ꢀ0.34
ꢀ1.88
2.12
T2
T1
[c]
DGCS
DGCR1
DGHS
DGCR2
0.07
Dyad and triad with saturated linkers
[c]
ꢀ2.19
ꢀ0.29
ꢀ1.90
[c]
[c]
Dyad D2: The energy-level diagram for dyad D2 in di-
chloromethane, obtained from spectroscopic and electro-
chemical data (Table 2), is shown in Figure 5. The energies
of the triplet states of the two molecular components have
been taken from literature data.[54,55]
[a] Calculated from Equation (1). [b] Given the complete quenching of
the fluorescence of the PMI unit, the excited-state energy was estimated
from the absorption of the dyad/triad and emission from the model
system 2. [c] The physical meaning of these formal labels is limited by
the strong delocalization found in this system in the QT–exTTF moiety
(see below).
ure 4a. The intense emission with lmax =564 nm has a quan-
tum yield close to unity and a lifetime of 5.0 ns. The transi-
ent difference spectrum of the singlet excited state of PMI,
as obtained by ultrafast spectroscopy of 2 in dichlorome-
Figure 5. Energy-level diagram for dyad D2 in dichloromethane.
In dyad D2 in dichloromethane (dielectric constant, e=
8.93), the emission of the PMI unit is completely
quenched.[56] On the basis of the energy-level diagram in
Figure 5, the emission quenching is very likely due to photo-
induced electron transfer from the QT unit to the excited
PMI. Ultrafast spectroscopy (Figure 6) permits direct obser-
vation of the process. The first spectrum, taken immediately
after the pump pulse, is identical to that of the model
system 2 (Figure 4b) and is thus assigned to the S1 state of
the PMI component. It consists of bleaching at l<600 nm
(bleaching of the ground-state absorption plus some stimu-
lated emission) and of a positive absorption peaking at
660 nm. The transient spectral changes are clearly biphasic.
In the first process (Figure 6a), the bleaching corresponding
to the PMI ground-state absorption at around 530 nm is es-
sentially maintained, whereas that corresponding to stimu-
lated emission in the 550–650 nm range disappears and the
Figure 4. Photophysical properties of PMI model system 2 in dichlorome-
thane. a) Absorption and emission spectra. b) Transient spectrum ob-
tained in ultrafast spectroscopy (excitation at 490 nm); constant in the
time window 0–1 ns.
Chem. Eur. J. 2010, 16, 9140 – 9153
ꢃ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
9145

F. Odobel, F. Scandola et al.
combination step (DG=ꢀ1.77 eV, likely to lie in the Marcus
inverted region).[61] Consistent with the finding that charge
recombination leads efficiently to the ground state, no long-
lived transient was observed in laser flash photolysis experi-
ments (timescale>10 ns) in dichloromethane. The photo-
physical behavior of D2 in dichloromethane is summarized
in Figure 5.
In diethyl ether (e=4.33), the PMI fluorescence is again
completely quenched in the D2 dyad.[56] The transient spec-
tral changes (Figure SI2) are qualitatively the same as in di-
chloromethane, with the formation and subsequent decay of
a two-peak structure corresponding to the radical ions. The
kinetics for the formation (45 ps) and decay (680 ps) of the
transient are slower than in dichloromethane. This can be
easily explained on the basis of the expected rise in energy
of the charge-separated state in the less polar solvent and
the consequent changes in the driving force for the charge
separation and recombination processes.[62] The former, in
the normal free-energy region, is expected to slow down as
the driving force decreases, whereas the latter is pushed fur-
ther into the Marcus inverted region by the increase in the
driving force.[63] Again, no long-lived transient was observed
in laser flash photolysis experiments (timescale>10 ns) in
diethyl ether.
In toluene (e=2.38), in contrast to the other solvents, the
PMI emission of dyad D2 is only partially quenched: The
lifetime is 3.0 ns and the quantum yield is around 60% of
that of the PMI model 2. Also, the transient behavior
(Figure 7) is very different to that observed in the other sol-
vents. The initial spectrum is again that of the singlet excited
state of PMI (see Figure SI3), but the spectral changes are
qualitatively similar to, but much slower (ca. 2 ns) than
those observed initially in the other solvents and no transi-
ent decay is observed in the time window (0–1 ns) of the ex-
periment. On a longer timescale, flash photolysis experi-
ments revealed the presence of a long-lived transient
(Figure 8) with an oxygen-dependent lifetime (deaerated,
2.6 ms; aerated, 350 ns), which, by comparison with literature
data,[55] can be unequivocally assigned to the triplet state of
the PMI unit.
The sharply different behavior observed in toluene is very
likely determined by the low polarity of this solvent, which
raises the charge-separated state to energies comparable to
that of the singlet state of the PMI unit (Table 3,
Figure 9).[64] In this case, the small fluorescence quenching
and the slow transient spectral changes are consistent with
the PMI excited singlet state undergoing conversion to (or
equilibration with) a charge-separated state, a process with
little, if any, driving force. As to the efficient formation of
the PMI triplet state observed on the longer timescale,
direct S1 T1 intersystem crossing within the PMI unit can be
ruled out because no triplet state can be observed following
excitation of the highly fluorescent model 2. Therefore a
likely mechanism is triplet formation upon charge recombi-
nation. For this mechanism to be efficient, the charge-sepa-
rated state (initially formed by photoinduced electron trans-
fer in a singlet spin state) must be able to undergo spin in-
Figure 6. Ultrafast spectroscopy of dyad D2 in dichloromethane (excita-
tion at 490 nm).
maximum at 660 nm evolves into a broader two-featured
profile (maximum at 690 nm, shoulder at 610 nm). These
spectral changes are exactly as expected for photoinduced
electron transfer based on spectroelectrochemistry observa-
tions in which the 610 nm absorption of the PMI radical
anion and the 690 nm absorption of the QT radical cation
are observed upon reduction and oxidation of the D2 dyad,
respectively (Figure SI1 of the Supporting information). In
the second process (Figure 6b), the spectral features of the
charge-separated state decay monotonically with isosbestic
points close to the baseline, consistent with a charge recom-
bination process leading back to the ground state (the resid-
ual absorption, with the same profile as the initial spectrum,
is assigned to the same impurity of free PMI observed in
emission[56]). The kinetics observed (Figure 6c) agree with
standard electron-transfer theory,[57–60] which predicts a very
high rate for the moderately exergonic charge-separation
step (DG=ꢀ0.54 eV, likely to be close to the activationless
regime) and a lower rate for the highly exergonic charge-re-
9146
www.chemeurj.org
ꢃ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 9140 – 9153

Hole-Transfer Dyads and Triads
FULL PAPER
Figure 9. Energy-level diagram for dyad D2 in toluene.
Triad T2: The energy-level diagram for triad T2 in dichloro-
methane, obtained from the spectroscopic and electrochemi-
cal data (Table 3), is shown in Figure 10. The only triplet
state included in this energy-level diagram is the lowest one,
localized on the PMI unit. As for the PMIꢀ–QT–exTTF+
charge-separated state, a lower limiting energy value (see
discussion in the Computational study section) is indicated
by the full line, based on the experimental two-electron oxi-
dation potential of the exTTF unit. For the triad T2 in di-
chloromethane, the PMI emission is totally quenched rela-
tive to model 2. Similarly to the corresponding dyad D2, the
quenching is very likely due to photoinduced electron trans-
fer from the directly linked QT unit. The ultrafast spectros-
copy experiments (Figure 11) confirm this assumption. The
very fast (1.2 ps) process with bleach recovery in the 550–
630 nm range (stimulated emission) and the positive change
with the formation of a broad absorption maximum at
longer wavelengths (Figure 11a) is attributed to the primary
charge-separation step 1*PMI–QT–exTTF!PMIꢀ–QT+–
exTTF. The subsequent spectral evolution shows complete
Figure 7. Ultrafast spectroscopy of dyad D2 in toluene (excitation at
520 nm).
Figure 8. Nanosecond flash photolysis of dyad D2 in toluene (excitation
at 532 nm).
version (intersystem crossing) before recombination. This
process can be efficient in toluene because of the intrinsic
slowness of the competing spin-allowed recombination to
the ground state brought about by its highly exergonic in-
verted nature. This recombination mechanism in solvents of
low polarity is common to other dyad systems involving aro-
matic imides.[65,66] Whether the PMI triplet is formed directly
by charge recombination or by the intermediate formation
of the QT triplet followed by fast triplet energy transfer is
not possible to determine on the basis of the experimental
results. The photophysical behavior of D2 in toluene is sum-
marized in Figure 9.
Figure 10. Energy-level diagram for triad T2 in dichloromethane. For the
PMIꢀ–QT–exTTF+ charge-separated state, the lower limiting energy
value calculated using the experimental 2eꢀ, oxidation potential of
exTTF is shown (full line; see text).
Chem. Eur. J. 2010, 16, 9140 – 9153
ꢃ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
9147

F. Odobel, F. Scandola et al.
nent (17%) assuming similar molar absorption coefficients
of the primary and long-range charge-separated states at the
examined wavelength. The strong similarity between the be-
havior of triad T2 and that observed for dyad D2 is main-
tained in solvents of lower polarity. In diethyl ether, the re-
combination of PMI occurs in the short ps timescale and
leads again to the triplet state of the PMI unit (Figure SI5).
Dyad and triad with unsaturated linkers
Dyad D1: The photophysical behavior of dyad D1 is qualita-
tively similar to that described above for D2. In particular,
complete quenching of the PMI emission is observed in di-
chloromethane and diethyl ether, whereas only partial
quenching is observed in toluene. The transient spectral
changes of D1 (Figure 12) are again biphasic and indicate
the occurrence of charge separation and recombination with
the eventual formation of ground-state products in dichloro-
methane and diethyl ether and of the PMI triplet state in
toluene (Figure SI6). Differences with respect to D2 (e.g., in
the spectral shape of emission and transient absorption) can
be traced back to the previously mentioned large intercom-
ponent interaction taking place in the dyad across the unsa-
turated ethynylene linkage. In particular, the emission ob-
served in toluene, which is significantly red-shifted and
much broader than that of the PMI model 2 and dyad D2
(Figure SI7), can be attributed to a strongly perturbed PMI
fluorescence or, alternatively, to a new emission with
charge-transfer characteristics. The appreciable delocaliza-
tion of the MOs of PMI and QT on the neighboring unit in
D1, as indicated by the DFT calculations (see below), indi-
cates that, indeed, emission from the PMIꢀ–QT+ charge-
separated state may have appreciable oscillator strength.
Charge-transfer emission has previously been observed by
De Schryver and co-workers in a directly linked PMI-bis-
thiophene dyad.[34]
Figure 11. Ultrafast spectroscopy of triad T2 in dichloromethane (excita-
tion at 530 nm).
decay of the spectral changes back to the baseline (Fig-
ure 11b). The process is not a simple one, as indicated by
the lack of a clear isosbestic point and by the biexponential
kinetics (Figure 11c) but, overall, it is remarkably similar to
that observed for dyad D2. This suggests that the decay of
the primary charge-separated state involves as the main pro-
cess fast (23 ps) charge recombination to the ground state
(PMIꢀ–QT+–exTTF!PMI–QT–exTTF). Note, however,
that clear spectroscopic discrimination between primary
(PMIꢀ–QT+–exTTF) and long-range (PMIꢀ–QT–exTTF+)
charge-separated states is difficult because of the largely
similar signatures expected for the radical cations of QT
(lmax =690 nm, Figure SI1) and exTTF (lmax =680 nm[41]).
Therefore the occurrence of some charge shift (PMIꢀ–QT+–
exTTF!PMIꢀ–QT–exTTF+) in competition with the main
primary charge recombination process cannot be ruled out.
In this hypothesis, the minor, longer-lived (150 ps) compo-
nent of the overall decay can be tentatively associated with
a long-range recombination process. The yield of this pro-
cess can be estimated from the weight of the slow compo-
Triad T1: In line with the observations made for the dyad
systems, the behavior of the unsaturated triad T1 is rather
similar to that of the saturated analogue T2. The transient
spectral changes observed in dichloromethane are shown in
Figure 13. Again, after a very fast (Figure 13a, 1.4 ps, inset
Figure 13c) process similar to that exhibited by the corre-
sponding dyad and attributed to charge separation, a two-
component decay (major time constant 26 ps, minor time
constant 160 ps, Figure 13b,c) restores the original baseline.
As for the saturated analogue, it is likely that the main re-
combination path is similar to that of the dyad, that is, fast
recombination to the ground state with any other processes
constituting, at best, a minor pathway. The solvent effects,
including the efficient formation of triplet PMI in toluene
(Figure SI8), also follow closely those observed for T2.
Energy levels: Computational study: Overall, the results ob-
tained with the triads are somewhat deceiving as the pri-
mary charge-separated state seems to recombine largely by
primary charge recombination with no substantial gain in
9148
www.chemeurj.org
ꢃ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 9140 – 9153

Hole-Transfer Dyads and Triads
FULL PAPER
Figure 13. Ultrafast spectroscopy of triad T1 in dichloromethane (excita-
tion at 560 nm).
Figure 12. Ultrafast spectroscopy of dyad D1 in dichloromethane (excita-
tion at 530 nm).
charge-separation lifetime with respect to the dyads. This
may be surprising considering the almost ideal driving force
(ca. 0.4 eV) estimated for charge shift in both T2 and T1
(Figure 10 and Table 3). Given the similar driving force and
the identical spacers involved, charge shift is expected to be
as fast as primary charge separation (few ps), that is, should
be able to compete effectively with the slower primary
charge recombination process. This not being the case seems
to point to a fundamental problem in the evaluation of the
energetics of the triads. In fact, it should be kept in mind
that the energy of the long-range charge-separated state in
Table 3 and Figure 10 was calculated, as usual,[67] by using,
faute de mieux, the experimental two-electron oxidation po-
tential of the exTTF unit. The fact that only a two-electron
oxidation is experimentally observed for this unit is attribut-
ed to the strong stabilization (gain in aromaticity) experi-
enced by the dication upon planarization of the central poly-
acenic moiety.[68] Thus, the one-electron oxidation potential
of exTTF is not known, except for the fact that it must be
equal or higher than the experimental two-electron value.
The observed disproportionation of the radical monocation,
generated radiolytically[69] or by flash photolysis,[41] strongly
suggests that the one-electron oxidation potential must be
higher than the experimental two-electron value. This is an
example of “potential inversion”, a phenomenon extensively
studied by Evans[70] in organic molecules undergoing struc-
tural change coupled with electron transfer. In appropriate
cases, the extent of potential inversion (the difference be-
tween the first and second redox potentials) can be estimat-
ed by voltammetry or by EPR measurement of the compro-
portionation equilibrium. For exTTF, a lower limit of 0.28 V,
determined by voltammetry, was confirmed by quantitative
EPR measurements.[71] Interestingly, in cyclophane-type de-
rivatives of exTTF, in which planarization of the dication ox-
idation is prevented by steric constraints, two stepwise one-
electron oxidation processes are observed and the radical
cation can be electrochemically generated and spectroscopi-
cally characterized.[72] Thus, because of the above-mentioned
Chem. Eur. J. 2010, 16, 9140 – 9153
ꢃ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
9149

F. Odobel, F. Scandola et al.
problems, the energetics of the triads estimated from elec-
trochemical data are intrinsically uncertain. In an attempt to
circumvent the experimental problem, and possibly to un-
derstand the reasons for inefficient charge shift, we per-
formed the DFT computational work described below.
Calculations on the electronic ground state of the dyad
(D1 and D2) and triad (T1 and T2) model systems were car-
ried out by using the B3LYP functional within the density
functional theory method.[73–75] The “pruned” molecular
models were tailored for the computational study (aryl and
aliphatic substituents replaced with phenyl and methyl resi-
dues, respectively) to reduce complexity and cost while still
providing a reliable description of the molecular systems in-
vestigated experimentally. The geometry (see Figure SI9;
the coordinates are presented in the Supporting Informa-
tion) was fully optimized, without symmetry constraints, by
describing the systems with a Dunning–Huzinaga valence
“double-z” quality basis set.[76] Single-point energy calcula-
tions with the standard basis set 6-31G**[77–83] were accom-
plished to account for the polarization effects. Additional
calculations on T2 were performed to include the solvent
effect using the polarizable continuum model[84] (PCM) to
simulate dichloromethane (e =8.93), that is, the solvent
used to collect the spectra of the present molecular systems.
All calculations were performed by using the Gaussian 03
software package.[85]
units, the unsaturated system exhibits a larger degree of de-
localization through the ethynylene spacers. Thus, the fron-
tier orbitals in D1, although still being predominantly local-
ized on either the PMI or QT unit, also have significant am-
plitudes on the neighboring unit. The calculated relative en-
ergies of the MOs are consistent with the spectroscopic data
for the dyads (Figures 2 and 3), in which the PMI absorption
(HOMO-1!LUMO transition) is red-shifted in D1 relative
to D2, and the electrochemical data for the dyads (Table 2),
with D1 easier to reduce but more difficult to oxidize than
D2. In very much the same way, Figure 15 shows the MO
energy diagrams for triads T1 and T2. As observed for the
dyads, the unsaturated linkers between fragments allow effi-
cient p conjugation and therefore delocalization of the MOs
in the case of T1. As for the dyads, this is reflected by sta-
tionary absorption spectroscopy (Figures 2 and 3) in which
the PMI region of the spectra is strongly influenced by the
degree of ground-state electronic coupling between the
units.
The calculated energy-level diagrams are also valuable in
interpreting the photophysical results. For the dyads,
Figure 14 clearly shows that the lowest excited state (corre-
sponding to the HOMO!LUMO transition) is a PMIꢀ–
QT+ charge-separated state and that the driving force for
populating such a state from the PMI local excited state is
large (to a first approximation, the HOMO–HOMO-1
energy difference). These findings are fully consistent with
the very fast charge separation observed for both dyads in
dichloromethane.
The MO energy diagrams attained at the DFT level for
dyads D1 and D2 are shown in Figure 14, with each MO la-
beled by the prominent localization on the system. Whereas
for the saturated dyad D2 the MOs, including the frontier
orbitals, are strongly localized on either the PMI or QT
The most interesting result relates to the triads, however.
In the energy-level diagram of T2 (Figure 15) it is clearly
Figure 14. MO energy diagrams of the dyads D1 and D2 calculated in the gas phase at the B3LYP/6-31G** level of theory, along with HOMO and
LUMO contour plots.
9150
www.chemeurj.org
ꢃ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 9140 – 9153

Hole-Transfer Dyads and Triads
FULL PAPER
Figure 15. MO energy diagram of the triads T1 (gas phase) and T2 (dichloromethane) calculated at the B3LYP/6-31G** level of theory, along with
LUMO, HOMO, and HOMO-1 contour plots.
seen that the highest-occupied MOs are a pair of very close-
ly spaced orbitals, one localized on QT and the other on the
exTTF unit. The two orbitals are quasi-degenerate, with pQT
slightly higher than pexTTF in dichloromethane (quasi-degen-
erate also in the gas phase, but in the reverse order). The
similar energies of the two orbitals imply that the unknown
one-electron potential for the oxidation of the exTTF unit
must be similar to that of the QT unit, that is, substantially
higher (by ca. 0.4 V) than the experimental two-electron po-
tential (Table 2) so that the driving force for charge shift in
triad T2 in dichloromethane must be, if any, very small. The
energy state schematized by the dashed level in Figure 10
accounts for the experimental finding that the primary
charge-separated state deactivates largely by charge recom-
bination rather than by charge shift and the fact that the
minor decay component, attributed to long-range charge re-
combination, is only moderately slower than the primary
process (a two-step process through the almost isoenergetic
QT unit is expected to be much faster than a direct transfor-
mation).
The results for T1 show again the presence of two closely
spaced highest-occupied MOs. The main difference with re-
spect to the saturated triad lies in the strong delocalization
of both orbitals on the QT and exTTF units. Note that the
extent of delocalization between neighboring units through
the ethynylene linkages is much greater between QT and
exTTF (Figure 15) than between PMI and QT (see, for ex-
ample, the corresponding dyad in Figure 14). In terms of
through-bond perturbation, this can be explained by the
smaller energy difference between the interacting orbitals of
the QT–exTTF fragment relative to PMI–QT. From the
photophysical point of view, the strong delocalization in the
QT–exTTF fragment suggests that for T1 some of the terms
used in the above discussion should be reconsidered. In par-
ticular, the picture of two distinct charge-separated states
(primary and long-range) is inappropriate and should be re-
placed by that of two charge-separated states containing a
reduced PMI unit and a collectively oxidized QT–exTTF
fragment.
Conclusion
Two dyad and triad systems of the PMI–QT and PMI–QT–
exTTF type, differing in the saturated (ethylene) or unsatu-
rated (ethynylene) nature of the linker groups, have been
designed, synthesized, and characterized electrochemically,
photophysically, and computationally. All the experimental
results indicate that, although the saturated systems have
weakly interacting, essentially unperturbed molecular subu-
nits (i.e., are strictly supramolecular in nature), in the unsa-
turated systems strong intercomponent interactions lead to
substantial perturbation of the properties of the subunits.
The DFT study on D2 and T2 indicates MOs largely local-
ized within the molecular components, whereas for D1 and
T1 more or less extensive delocalization of the MOs over
neighboring units takes place across the ethynylene linkers.
In T1, the delocalization is particularly effective between
the QT and exTTF units, which should be more appropriate-
ly viewed combined as a single electronic subsystem. This
implies that for T1 the definition of “triad”, as well as no-
tions such as charge shift and primary versus long-range
charge separation and recombination, lose much of their
meaning.
Chem. Eur. J. 2010, 16, 9140 – 9153
ꢃ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
9151

F. Odobel, F. Scandola et al.
de Silva, I. Gould, Electron Transfer in Chemistry, Wiley-VCH,
Weinheim, 2001.
[10] V. Balzani, F. Scandola, Supramolecular Photochemistry, Ellis Hor-
wood, Chischester, 1991.
[11] D. Gust, T. A. Moore, A. L. Moore, Acc. Chem. Res. 1993, 26, 198–
205.
[12] M. R. Wasielewski, Chem. Rev. 1992, 92, 435–461.
[13] E. Baranoff, J. P. Collin, L. Flamigni, J. P. Sauvage, Chem. Soc. Rev.
2004, 33, 147–155.
[14] D. M. Guldi, Pure Appl. Chem. 2003, 75, 1069–1075.
[15] D. M. Guldi, H. Imahori, K. Tamaki, Y. Kashiwagi, H. Yamada, Y.
Sakata, S. Fukuzumi, J. Phys. Chem. A 2004, 108, 541–548.
[16] A. Harriman, J. P. Sauvage, Chem. Soc. Rev. 1996, 25, 41.
[17] M. R. Wasielewski, J. Org. Chem. 2006, 71, 5051–5066.
[18] A. Osuka, S. Marumo, N. Mataga, S. Taniguchi, T. Okada, I. Yama-
zaki, Y. Nishimura, T. Ohno, K. Nozaki, J. Am. Chem. Soc. 1996,
118, 155–168.
For the dyad systems, the photophysics observed following
excitation of the PMI unit is solvent-dependent. In moder-
ately polar solvents, such as dichloromethane and diethyl
ether, fast charge separation is followed by recombination
to the ground state. In the less polar toluene, slow conver-
sion to (or equilibration with) the charge-separated state is
followed by intersystem crossing and recombination to yield
the triplet state of the PMI unit. This difference in behavior
is related to the solvent effect on the energy of the charge-
separated states.
The behavior of the triads is remarkably similar to that of
the corresponding dyads, with only minor differences. This
indicates that, after primary charge separation, the charge
shift to yield a long-range charge-separated state is quite in-
efficient. At first sight, this result was surprising given the
large driving force calculated for charge shift from the elec-
trochemical data. In fact, the problem lies in the electro-
chemistry of exTTF for which, because of the strong gain in
aromaticity achieved in forming the dicationic species, only
a single two-electron oxidation process is observed. The
matter is clarified by DFT calculations on triad T2, which
show that, although from electrochemistry exTTF looks
much easier to oxidize than QT, the highest-occupied MOs
(one-electron redox orbitals) of the two units are almost de-
generate. Such a result quantifies the potential inversion[70]
of exTTF at around +0.4 V. This value may be of interest
for the design of charge-separation systems that include this
type of unit, particularly when small driving forces are rele-
vant.[86]
[19] A. Petrella, J. Cremer, C. L. De, P. Bauerle, R. M. Williams, J. Phys.
Chem. A 2005, 109, 11687–11695.
[20] A. Marcos Ramos, E. H. A. Beckers, T. Offermans, S. C. J. Meskers,
R. A. J. Janssen, J. Phys. Chem. A 2004, 108, 8201–8211.
[21] M. Holzapfel, C. Lambert, J. Phys. Chem. C 2008, 112, 1227–1243.
[22] C. Lambert, J. Schelter, T. Fiebig, D. Mank, A. Trifonov, J. Am.
Chem. Soc. 2005, 127, 10600–10610.
[23] J. Fortage, J. Boixel, E. Blart, L. Hammarstrom, H. C. Becker, F.
Odobel, Chem. Eur. J. 2008, 14, 3467–3480.
[24] J. Fortage, J. Boixel, E. Blart, H. C. Becker, F. Odobel, Inorg. Chem.
2009, 48, 518–526.
[25] J. Fortage, E. Goransson, E. Blart, H. C. Becker, L. Hammarstrom,
F. Odobel, Chem. Commun. 2007, 4629–4631.
[26] T. Nakamura, J. Y. Ikemoto, M. Fujitsuka, Y. Araki, O. Ito, K. Taki-
miya, Y. Aso, T. Otsubo, J. Phys. Chem. B 2005, 109, 14365–14374.
[27] T. Nakamura, H. Kanato, Y. Araki, O. Ito, K. Takimiya, T. Otsubo,
Y. Aso, J. Phys. Chem. A 2006, 110, 3471–3479.
[28] H. Kanato, K. Takimiya, T. Otsubo, Y. Aso, T. Nakamura, Y. Araki,
O. Ito, J. Org. Chem. 2004, 69, 7183–7189.
[29] T. Nakamura, M. Fujitsuka, Y. Araki, O. Ito, J. Ikemoto, K. Taki-
miya, Y. Aso, T. Otsubo, J. Phys. Chem. B 2004, 108, 10700–10710.
[30] F. Wꢇrthner, Chem. Commun. 2004, 1564–1579.
[31] A. Morandeira, J. Fortage, T. Edvinsson, L. Le Pleux, E. Blart, G.
Boschloo, A. Hagfeldt, L. Hanmiarstrom, F. Odobel, J. Phys. Chem.
C 2008, 112, 1721–1728.
[32] J. Cremer, P. Bꢅuerle, Eur. J. Org. Chem. 2005, 3715–3723.
[33] J. Cremer, E. Mena-Osteritz, N. G. Pschierer, K. Mullen, P. Bꢅuerle,
Org. Biomol. Chem. 2005, 3, 985–995.
Acknowledgements
The authors would like to thank Maria Teresa Indelli for the critical read-
ing of the manuscript and Roberto Argazzi for precious help with the
nanosecond laser photolysis experiments. The Agence Nationale de la
Recherche (ANR) is gratefully acknowledged for financial support (pro-
gram “PhotoCumElec”) and the European Community through COST
D35 program. Financial support from the Italian MIUR (Grants FIRB
RBNE019H9K and PRIN 2006030320) is also gratefully acknowledged.
[34] E. Fron, M. Lor, R. Pilot, G. Schweitzer, H. Dincalp, F. S. De, J.
Cremer, P. Bꢅuerle, K. Mullen, A. M. Van der, F. C. De Schryver,
Photochem. Photobiol. Sci. 2005, 4, 61–68.
[35] C. Atienza-Castellanos, M. Wielopolski, D. M. Guldi, P. C. van der
Pol, M. R. Bryce, S. Filippone, N. Martin, Chem. Commun. 2007,
5164–5166.
[36] M. C. Dꢈaz, B. M. Illescas, C. Seoane, N. Martin, J. Org. Chem. 2004,
69, 4492–4499.
[37] S. S. Gayathri, M. Wielopolski, E. M. Perez, G. Fernandez, L. San-
chez, R. Viruela, E. Orti, D. M. Guldi, N. Martin, Angew. Chem.
2009, 121, 829–834; Angew. Chem. Int. Ed. 2009, 48, 815–819.
[38] F. Giacalone, J. L. Segura, N. Martin, D. M. Guldi, J. Am. Chem.
Soc. 2004, 126, 5340–5341.
[39] F. Giacalone, J. L. Segura, N. Martin, J. Ramey, D. M. Guldi, Chem.
Eur. J. 2005, 11, 4819–4834.
[40] S. Gonzꢆlez, N. Martin, D. M. Guldi, J. Org. Chem. 2003, 68, 779–
791.
[41] D. M. Guldi, L. Sanchez, N. Martin, J. Phys. Chem. B 2001, 105,
7139–7144.
[42] S. Romain, J. C. Lepretre, J. Chauvin, A. Deronzier, M. N. Collomb,
Inorg. Chem. 2007, 46, 2735–2743.
[1] V. Balzani, A. Credi, M. Venturi, ChemSusChem 2008, 1, 26–58.
[2] Y. Kobori, S. Yamauchi, K. Akiyama, S. Tero-Kubota, H. Imahori, S.
Fukuzumi, J. R. Norris Jr. , Proc. Natl. Acad. Sci. USA 2005, 102,
10017–10022.
[3] D. A. Lavan, J. N. Cha, Proc. Natl. Acad. Sci. USA 2006, 103, 5251–
5255.
[4] I. M. Bennett, H. M. Farfano, F. Bogani, A. Primak, P. A. Liddell, L.
Otero, L. Sereno, J. J. Silber, A. L. Moore, T. A. Moore, D. Gust,
Nature 2002, 420, 398–401.
[5] M. Hambourger, G. F. Moore, D. M. Kramer, D. Gust, A. L. Moore,
T. A. Moore, Chem. Soc. Rev. 2009, 38, 25–35.
[6] T. Otsubo, Y. Aso, K. Takimiya, J. Mater. Chem. 2002, 12, 2565–
2575.
[7] L. Sꢆnchez, M. A. Herranz, N. Martin, J. Mater. Chem. 2005, 15,
1409–1421.
[8] N. Martin, L. Sanchez, M. A. Herranz, B. Illescas, D. M. Guldi, Acc.
Chem. Res. 2007, 40, 1015–1024.
[9] V. Balzani, P. Piotrowiak, M. A. J. Rodgers, J. Mattay, D. Astruc,
H. B. Gray, J. Winkler, S. Fukuzumi, T. E. Mallouk, Y. Haas, A. P.
[43] R. Konduri, N. R. de Tacconi, K. Rajeshwar, F. M. MacDonnell, J.
Am. Chem. Soc. 2004, 126, 11621–11629.
[44] L. Hammarstrom, Curr. Opin. Chem. Biol. 2003, 7, 666–673.
9152
www.chemeurj.org
ꢃ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 9140 – 9153

Hole-Transfer Dyads and Triads
FULL PAPER
[45] F. Odobel, M. Severac, Y. Pellegrin, E. Blart, C. Fosse, C. Cannizzo,
C. R. Mayer, K. J. Elliott, A. Harriman, Chem. Eur. J. 2009, 15,
3130–3138.
[46] A. Harriman, K. J. Elliott, M. A. H. Alamiry, L. Le Pleux, M. Sever-
ac, Y. Pellegrin, E. Blart, C. Fosse, C. Cannizzo, C. R. Mayer, F.
Odobel, J. Phys. Chem. C 2009, 113, 5834–5842.
[68] N. Martꢈn, L. Sanchez, C. Seoane, E. Orti, P. M. Viruela, R. Viruela,
J. Org. Chem. 1998, 63, 1268–1279.
[69] A. E. Jones, C. A. Christensen, D. F. Perepichka, A. S. Batsanov, A.
Beeby, P. J. Low, M. R. Bryce, A. W. Parker, Chem. Eur. J. 2001, 7,
973–978.
[70] D. H. Evans, Chem. Rev. 2008, 108, 2113–2144.
[47] J. Fortage, M. Severac, C. Houarner-Rassin, Y. Pellegrin, E. Blart, F.
Odobel, J. Photochem. Photobiol. A 2008, 197, 156–169.
[48] A. J. Moore, M. R. Bryce, J. Chem. Soc. Perkin Trans. 1 1991, 157–
168.
[71] N. E. Gruhn, N. A. Macꢈas-Ruvalcaba, D. H. Evans, Langmuir 2006,
22, 10683.
[72] C. A. Christensen, A. S. Batsanov, M. R. Bryce, J. Am. Chem. Soc.
2006, 128, 10484–10490.
[49] G. J. Marshallsay, M. R. Bryce, J. Org. Chem. 1994, 59, 6847–6849.
[50] A. J. Moore, M. R. Bryce, A. S. Batsanov, J. C. Cole, J. A. K.
Howard, Synthesis 1995, 675–682.
[51] M. Adachi, Y. Nagao, Chem. Mater. 2001, 13, 662–669.
[52] M. Adachi, Y. Nagao, Chem. Mater. 1999, 11, 2107–2114.
[53] A. E. Jones, C. A. Christensen, D. F. Perepichka, A. S. Batsanov, A.
Beeby, P. J. Low, M. R. Bryce, A. W. Parker, Chem. Eur. J. 2001, 7,
973–978.
[73] A. D. Becke, J. Chem. Phys. 1993, 98, 5648–5652.
[74] P. J. Hay, J. Phys. Chem. A 2002, 106, 1634–1641.
[75] C. T. Lee, W. T. Yang, R. G. Parr, Phys. Rev. B 1988, 37, 785–789.
[76] “Gaussian Basis Sets for Molecular Calculations”: T. H. J. Dunning,
P. J. Hay in Methods of Electronic Structure Theory (Ed.: H. F. I.
Schaefer), Plenum Press, New York, 1977.
[77] R. Ditchfield, W. J. Hehre, J. A. Pople, J. Chem. Phys. 1971, 54, 724–
728.
[54] K. Matsumoto, M. Fujitsuka, T. Sato, S. Onodera, O. Ito, J. Phys.
Chem. B 2000, 104, 11632–11638.
[55] R. T. Hayes, C. J. Walsh, M. R. Wasielewski, J. Phys. Chem. A 2004,
108, 3253–3260.
[56] In fact a small residual emission (intensity, ca. 10% of that of the
model) is observed. On the basis of its unquenched lifetime (5 ns),
this emission is attributed to an impurity of free PMI.
[57] J. Jortner, J. Chem. Phys. 1976, 64, 4860–4867.
[58] J. Ulstrup, Charge Transfer Processes in Condensed Media, Springer,
Berlin, 1979.
[59] R. A. Marcus, N. Sutin, Biochim. Biophys. Acta Rev. Bioenerg. 1985,
811, 265–322.
[60] J. R. Miller, J. V. Beitz, R. K. Huddleston, J. Am. Chem. Soc. 1984,
106, 5057–5068.
[61] With the appropriate parameters (see the Supporting Information),
the solvent contribution to the reorganizational energy is estimated
to be about 0.6 eV in dichloromethane.
[62] The energy of the charge-separated state in diethyl ether can be es-
timated by using Equation (1) (using appropriate effective ionic
radii, see the Supporting Information) to be ca. 1.96 eV.
[63] Note that a decrease in solvent polarity does not only increase the
energy of the charge-separated state, but it also decreases the reor-
ganizational energy of the processes. In the inverted region, the two
factors act synergistically to reduce the rates, whereas in the normal
regime partial compensation between the two effects should be ex-
pected.
[78] M. S. Gordon, Chem. Phys. Lett. 1980, 70, 343–349.
[79] P. C. Hariharan, J. A. Pople, Theor. Chim. Acta. 1973, 28, 213–222.
[80] P. C. Hariharan, J. A. Pople, Mol. Phys. 1974, 27, 209–214.
[81] W. J. Hehre, R. F. Stewart, J. A. Pople, J. Chem. Phys. 1969, 51,
2657–2664.
[82] W. J. Hehre, R. Ditchfield, J. A. Pople, J. Chem. Phys. 1972, 56,
2257–2261.
[83] G. A. Petersson, A. Bennett, T. G. Tensfeldt, M. A. Allaham, W. A.
Shirley, J. Mantzaris, J. Chem. Phys. 1988, 89, 2193–2218.
[84] R. Cammi, B. Mennucci, J. Tomasi, J. Phys. Chem. A 2000, 104,
5631–5637.
[85] Gaussian 03, Revision C.02, M. J. Frisch, G. W. Trucks, H. B. Schle-
gel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, J. A. Montgom-
ery, Jr., T. Vreven, K. N. Kudin, J. C. Burant, J. M. Millam, S. S.
Iyengar, J. Tomasi, V. Barone, B. Mennucci, M. Cossi, G. Scalmani,
N. Rega, G. A. Petersson, H. Nakatsuji, M. Hada, M. Ehara, K.
Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda,
O. Kitao, H. Nakai, M. Klene, X. Li, J. E. Knox, H. P. Hratchian,
J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E.
Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W.
Ochterski, P. Y. Ayala, K. Morokuma, G. A. Voth, P. Salvador, J. J.
Dannenberg, V. G. Zakrzewski, S. Dapprich, A. D. Daniels, M. C.
Strain, O. Farkas, D. K. Malick, A. D. Rabuck, K. Raghavachari,
J. B. Foresman, J. V. Ortiz, Q. Cui, A. G. Baboul, S. Clifford, J. Cio-
slowski, B. B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I. Komaro-
mi, R. L. Martin, D. J. Fox, T. Keith, M. A. Al-Laham, C. Y. Peng,
A. Nanayakkara, M. Challacombe, P. M. W. Gill, B. Johnson, W.
Chen, M. W. Wong, C. Gonzalez, J. A. Pople, Gaussian, Inc., Wall-
ingford CT, 2004.
[64] The energy of the charge-separated state in toluene can be estimat-
ed by using Equation (1) (using appropriate effective ionic radii, see
the Supporting Information) to be ca. 2.29 eV.
[65] M. Ghirotti, C. Chiorboli, C. C. You, F. Wurthner, F. Scandola, J.
Phys. Chem. A 2008, 112, 3376–3385.
[66] R. T. Hayes, M. R. Wasielewski, D. Gosztola, J. Am. Chem. Soc.
2000, 122, 5563–5567.
[67] a) M. C. Dꢈaz, M. A. Herranz, B. M. Illescas, N. Martin, N. Godbert,
M. R. Bryce, C. P. Luo, A. Swartz, G. Anderson, D. M. Guldi, J.
Org. Chem. 2003, 68, 7711–7721; b) C. Atienza, N. Martin, M. Wie-
lopolski, N. Haworth, T. Clark, D. M. Guldi, Chem. Commun. 2006,
3202–3204.
[86] Extensive work has been performed by Guldi and Martin[8,35,38–40,67]
on dyads in which fullerene is a photoexcited electron acceptor and
exTTF is the electron donor. For these systems, with an intrinsically
larger driving force (0.7 eV, using the experimental two-electron
value), the problem of the effective one-electron oxidation potential
of exTTF is evidently much less critical than here.
Received: March 12, 2010
Published online: June 30, 2010
Chem. Eur. J. 2010, 16, 9140 – 9153
ꢃ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
9153