Using meta conjugation to enhance charge separation versus charge recombination in phenylacetylene donor-bridge-acceptor complexes

Article abstract of DOI:10.1021/ja054543q

A pair of donor-bridge-acceptor electron-transfer complexes, with a carbazole donor and a naphthalimide acceptor connected by either a para- or meta-conjugated phenylacetylene bridge, are synthesized and studied using time-resolved and steady-state spectroscopy. These experiments show that the charge separation times, which depend on the coupling of the donor and acceptor through the excited bridge moiety, are similar for the two molecules (Meta and Para). The charge recombination time, however, is a factor of 10 slower for Meta than for Para. These results are related to changes in the electronic coupling of the bridge depending on its electronic state, and show that meta-conjugated bridges provide a possible motif for the design of asymmetric molecular wires. Copyright

Full text of DOI:10.1021/ja054543q

Published on Web 11/04/2005
Using Meta Conjugation To Enhance Charge Separation versus Charge
Recombination in Phenylacetylene Donor-Bridge-Acceptor Complexes
Alexis L. Thompson,† Tai-Sang Ahn,§ K. R. Justin Thomas,‡ S. Thayumanavan,*,‡
Todd J. Mart´ınez,*,† and Christopher J. Bardeen*,§
Departments of Chemistry, UniVersity of Illinois, Urbana, Illinois 61801, UniVersity of Massachusetts,
Amherst, Massachusetts 01003, and UniVersity of California, RiVerside, California 92521
Received July 8, 2005; E-mail: thai@chem.umass.edu (S.T.); tjm@spawn.scs.uiuc.edu (T.J.M.); christob@ucr.edu (C.J.B.)
Table 1. Excited-State Properties at Absorption and Emission
Geometries in Polar (DMF) and Nonpolar (Toluene/Cyclohexane)
Solventsa
A promising approach to high-efficiency solar energy conversion
uses molecules consisting of an electron donor-acceptor pair
separated by a chemical bridge as components of solar electro-
chemical cells.1 Conjugated organic bridges in such donor-bridge-
acceptor (DBA) compounds have been shown2 to act as molecular
“wires”. However, the ideal bridge would be an asymmetric wire
or “diode”spromoting electron transfer (ET) from donor to acceptor
while resisting ET in the reverse direction and thus suppressing
charge recombination. Our recent studies of excitation energy trans-
fer (EET) in meta-linked phenylacetylene (PA) dendrons suggest
a route to this goal. It is well-known that meta linkages lead to
electronic decoupling of the substituents in the ground electronic
state. We showed that this same meta linkage undergoes enhanced
electronic coupling in the excited state.3,4 The similarity of the
electronic coupling matrix elements in EET and ET further suggests
that DBA complexes with a meta-substituted bridge could act as
gated wires. The bridge would be a wire or resistor when in its
excited or ground electronic state, respectively. If de-excitation of
the bridge coincides with the forward ET event, the effective result
is a “one-electron wire”. Here, we demonstrate that such an asym-
metric wire can indeed be created. Both para- and meta-linked PA
bridges act as good conduits for forward excited-state ET, but the
meta bridge effectively prevents recombination, leading to the
asymmetric behavior. While ET through para-meta linkages in
model DBA compounds has been studied previously in terms of
both through-bond5 and through-space mechanisms,6 these studies
have focused on the ground state of the bridge.
Meta
Para
polar
nonpolar
polar
nonpolar
Eabs (eV)
3.77 (3.17)
2.5/10
2.37 (n/m)
0.5/63
3.79 (3.15)
2.5/9
2.53 (2.65)
0.3/60
3.51 (3.05)
3.9/16
2.07 (1.86)
3.4/53
3.56 (3.03)
3.8/13
2.76 (2.56)
4.8/26
TDM/DM
E
em (eV)
TDM/DM
a Experimental values in parentheses. Transition dipole moment (TDM)
and dipole moment (DM) lengths are in debye.
Figure 1. Absorption and emission spectra of Meta (top) and Para (bottom)
molecules. Emission in neat toluene (black), 10% CH2Cl2 (red), 25% CH2-
Cl2 (green), 50% CH2Cl2 (blue), and 100% CH2Cl2 (violet).
The insets in Figure 1 show the two DBA molecules (denoted
Meta and Para) that are the subject of this Communication.
Carbazole (donor) and naphthalimide (acceptor) units are linked
by a para- or meta-conjugated PA bridge. Steady-state spectral
properties of both are shown in Figure 1, where the solvent polarity
is tuned using mixtures of toluene and CH2Cl2. The PA bridge is
partially conjugated to the donor-acceptor moieties, as can be seen
from the fact that both absorptions are considerably red-shifted
compared to those of naphthalimide (340 nm) and carbazole (365
nm). While absorption does not change with solvent composition,
the emission red-shifts, broadens, and decreases in intensity with
increasing solvent polarity. The different dependences of absorption
and emission on solvent polarity indicate that they originate from
two different electronic states. Multireference semiempirical quan-
tum mechanical/molecular mechanical calculations (Table 1) show
that absorption corresponds to a low-lying neutral excited state
(DBA)* for both Para and Meta in low- and high-polarity solvents,
cyclohexane and dimethylformamide (DMF). A higher excited state
with charge-transfer (CT) character, D+BA-, becomes the lowest-
energy emitting state after excited-state relaxation. Both the steady-
Figure 2. Three-state model for the DBA compounds used in this study.
The variables are as defined in eqs 1 and 2 in the text.
state spectra and theoretical calculations suggest a three-state model
where the initial absorption event between neutral states, DBA to
(DBA)*, which are largely insensitive to solvent polarity, is
followed by electron transfer to D+BA-, whose energy depends
strongly on polarity. This is shown schematically in Figure 2.7
We use time-resolved spectroscopy to determine the forward and
backward ET rates for Para and Meta in toluene. Using femto-
second transient absorption measurements, with 150 fs time
resolution and excitation at 400 nm, we observe a broad excited-
state absorption ranging across the visible spectrum in both
molecules. This spectrum, shown in Figure 3a for Meta, rapidly
converts into a double-peaked spectrum. A global fit of the data at
probe wavelengths from 475 to 775 nm reveals two principal
components, shown in Figure 3b, which interconvert with a single-
exponential time: 4.5 ps in Meta and 5.4 ps in Para. We take this
to be the formation time of the D+BA- state after excitation to the
(DBA)* state. After this initial rapid decay, we observe only very
† University of Illinois.
‡ University of Massachusetts.
§ University of California, Riverside.
9
16348
J. AM. CHEM. SOC. 2005, 127, 16348-16349
10.1021/ja054543q CCC: $30.25 © 2005 American Chemical Society

C O M M U N I C A T I O N S
kr
n3
) 3.137 × 10-7(V∆µ)2νf
(1)
while in the second case, V* . V and kr is given by
kr 3.137 × 10-7(V*µ*)2νf3
Figure 3. (a) Pump-probe transient signals at 200 fs (---) and 200 ps (s)
for Meta. (b) Pump-probe transient signals monitoring the decay of the
neutral excited-state absorption for Meta (555 nm, black) and Para (614
nm, red). The decay times from a global fit are 4.5 ps for Meta and 5.4 ps
for Para. (c) Fluorescence decays for Meta (black, lifetime ) 16.8 ns)
and Para (red, lifetime ) 1.8 ns).
)
(2)
n3
(νa - νf)3 + V*2
where n is the refractive index, νa and νf are the peaks of the
absorption and fluorescence spectra, respectively, ∆µ is the change
in permanent dipole between the DBA and D+BA- states, and µ*
is the transition dipole between DBA and (DBA)*. Using either
measured or calculated values for νa, νf, µ*, and ∆µ in eqs 1 and
2, only modest changes in kr less than a factor of 2 can be obtained.
Therefore, the 30-fold enhancement in kr for Para cannot be solely
due to variations in νa, νf, µ*, or ∆µ, but must result from either V
or V*, or both, being larger for Para than for Meta. In other words,
the enhanced radiative rate that leads to fast charge recombination
in Para is a direct result of its larger ET matrix elements between
the CT state and at least one neutral state.
In conclusion, we have shown that the two isolectronic DBA
compounds have similar forward but very different backward ET
rates. The origin of this difference lies in the excited-state enhance-
ment of electronic coupling through the meta bridge: it acts like a
wire in the excited state, allowing facile forward ET, while in its
ground state the meta linkage acts like an insulator, preventing elec-
tronic communication. To exploit the asymmetry of the meta bridge,
the initially absorbed photon must both provide energy for the CT
event and place the bridge in its excited state. In our molecules,
this is accomplished by having the bridge partially conjugated to
the donor-acceptor groups. It is an interesting question whether
more weakly coupled bridge structures can exhibit similar behavior.
Acknowledgment. This work was supported by DOE grants
DEFG-01ER15270 and DEFG-05ER15747. C.J.B. is a Sloan
Research Fellow, A.L.T. is a NSF predoctoral fellow, and T.J.M.
is a Packard Fellow and Dreyfus Teacher-Scholar.
minor changes in the spectral shape over the course of hundreds
of picoseconds. Preliminary pump-probe experiments in more polar
solvents such as CH2Cl2 and DMF show the same initial decay,
accelerated by at least a factor of 2, providing further evidence
that it reflects the forward electron-transfer event.
Figure 3c shows the measured fluorescence decays of Meta and
Para in toluene. In both molecules, fluorescence measurements in
toluene reveal emission from the initially excited neutral state which
disappears within the 15 ps instrument response time, replaced by
a much longer-lived, red-shifted CT emission. Unlike the forward
charge separation times, the charge recombination times extracted
from these data are quite different: 16.8 ns for Meta and 1.8 ns
for Para. Thus, the ratio of forward/backward ET rates is ∼300 in
Para and ∼3500 in Meta, making Meta better suited for photo-
voltaic applications.
The greater rate asymmetry in Meta results from two factors:
the similar forward ET rates and the more rapid recombination in
Para. The fact that the rates of CT state formation differ by only
about 20% is surprising considering the conventional picture that
meta conjugation prevents electronic communication. But when the
donor and/or acceptor is electronically coupled to the bridge moiety,
as is the case here, the excited-state properties of the bridge must
be taken into account. Both theory and experiment have demon-
strated that phenyl compounds in their lowest excited state exhibit
much stronger coupling between meta substituents than in the
ground state.8 Recent ab initio calculations showed how excitonic
coupling between meta-linked PA segments is enhanced by a factor
of 7.3 in the excited state, versus a factor of 1.3 for para-linked
PAs.4 This provides a qualitative explanation of why Meta’s
forward ET rate is comparable to that of Para.
Supporting Information Available: Synthetic details, transient
spectra, calculation details, and optimized geometries (PDF, PDB). This
material is available free of charge via the Internet at http://pubs.acs.org.
References
The second factor is the slower recombination rate in Meta. After
forward ET, the bridge is presumably no longer in its excited state,
and the electronic coupling decreases. Neglecting vibrational
Franck-Condon effects, this decreased coupling results in a smaller
recombination rate. To confirm this decreased electronic coupling
in Meta, we need to determine the electron-transfer matrix elements
between the D+BA- state and the ground state and the neutral
excited state (denoted V and V*, as shown in Figure 2). By meas-
uring the fluorescence quantum yields, we determine the radiative
and nonradiative components of the total fluorescence decay rate,
(1) Gust, D.; Moore, T. A.; Moore, A. L. Acc. Chem. Res. 2001, 34, 40-48.
Wasielewski, M. R. Chem. ReV. 1992, 92, 435-461.
(2) Davis, W. B.; Svec, W. A.; Ratner, M. A.; Wasielewski, M. R. Science
1998, 396, 60-63.
(3) Gaab, K. M.; Thompson, A. L.; Xu, J.; Martinez, T. J.; Bardeen, C. J. J.
Am. Chem. Soc. 2003, 125, 9288-9289.
(4) Thompson, A. L.; Gaab, K. M.; Xu, J.; Bardeen, C. J.; Martinez, T. J. J.
Phys. Chem. A 2004, 108, 671-682.
(5) Patoux, C.; Coudret, C.; Launay, J. P.; Joachim, C.; Gourdon, A. Inorg.
Chem. 1997, 36, 5037-5049. Rovira, C.; Ruiz-Molina, D.; Elsner, O.;
Vidal-Gancedo, J.; Bonvoisin, J.; Launay, J. P.; Veciana, J. Chem. Eur.
J. 2001, 7, 240-250. Shaakov, S.; Galili, T.; Stavitski, E.; Levanon, H.;
Lukas, A.; Wasielewski, M. R. J. Am. Chem. Soc. 2003, 125, 6563-
6572.
(6) Stockmann, A.; Kurzawa, J.; Fritz, N.; Acar, N.; Schneider, S.; Daub, J.;
Engl, R.; Clark, T. J. Phys. Chem. A 2002, 106, 7958-7970. Tkachenko,
N. V.; Lemmetyinen, H.; Sonoda, J.; Ohkubo, K.; Sato, T.; Imahori, H.;
Fukuzumi, S. J. Phys. Chem. A 2003, 107, 8834-8844.
(7) Yonemoto, E. H.; Saupe, G. B.; Schmehl, R. H.; Hubig, S. M.; Riley, R.
L.; Iverson, B. L.; Mallouk, T. E. J. Am. Chem. Soc. 1994, 116, 4786-
4795.
(8) Lewis, F. D.; Yang, J.-S. J. Am. Chem. Soc. 1997, 119, 3834-3835.
Zimmerman, H. E. J. Am. Chem. Soc. 1995, 117, 8988-8991.
(9) Gould, I. R.; Young, R. H.; Mueller, L. J.; Albrecht, A. C.; Farid, S. J.
Am. Chem. Soc. 1994, 116, 8188-8199. Bixon, M.; Jortner, J.; Verhoeven,
J. W. J. Am. Chem. Soc. 1994, 116, 7349-7355.
(10) Creutz, C.; Newton, M. D.; Sutin, N. J. Photochem. Photobiol. A 1994,
82, 47-59.
kr and knr. For Para, kr ) 4.2 × 108 s-1 and knr ) 1.1 × 108 s-1
,
and for Meta, kr ) 1.4 × 107 s-1 and knr ) 4.7 × 107 s-1. kr for
Para is 30 times larger than that of Meta, while the knr values
differ only by a factor of 2.2, and this larger kr leads to rapid
radiative recombination. The radiative rate of a CT state is deter-
mined by its coupling to both the ground state and nearby neutral
excited states, and the most general expression is quite complicated.9
We examine two limiting cases where the solutions are well-known
and show that in either one, the factor of 30 increase in kr can be
explained only by a larger electronic matrix element V or V*. In
the first case, V . V* and kr is given by10
JA054543Q
9
J. AM. CHEM. SOC. VOL. 127, NO. 47, 2005 16349