Photoinduced charge separation involving an unusual double electron transfer mechanism in a donor-bridge-acceptor molecule

Article abstract of DOI:10.1021/ja001298w

A series of rodlike donor-bridge-acceptor (D-B-A) molecules was synthesized to study the role of bridge energy levels on electron transfer (ET) rates. In these compounds, a 4-aminonaphthalene-1,8-imide (ANI) electron donor is linked to a 1,8:4,5-naphthalenediimide acceptor (NI) via the 1,4 positions on a phenyl bridge. The phenyl bridge is substituted at the 2 and 5 positions with methyl or methoxy groups to yield ANI-diMe-NI and ANI-diMeO-NI. These molecules differ only in the energy levels of the bridge molecular orbitals. Other parameters affecting ET rates such as donor-acceptor distance, orientation, and driving force are constant between the two systems. The rate constants for charge separation (CS) and charge recombination (CR) within ANI-diMeO-NI in toluene are 32 and 1400 times larger, respectively, than the corresponding rate constants for ANI-diMe-NI. Solvents of higher polarity diminish these differences in rate constants, making them comparable to those observed for ANI-diMe-NI. The relative energies of the ion pair states suggest that it is possible for the reaction ¹*D-B-A → D-B⁺-A^- to occur via a double electron-transfer process that is somewhat analogous to Dexter energy transfer. The lowest excited singlet state of the donor, ¹*ANI, possesses about 70% charge-transfer character, so that significant positive charge is localized on its amine nitrogen, whereas significant negative charge is localized on its naphthalene-1,8-imide ring. Electron transfer from the naphthalene-1,8-imide ring of ¹*ANI to NI is concomitant with electron transfer from the p-dimethoxybenzene bridge to the electron-deficient amine nitrogen atom in ¹*ANI. A series of reference molecules in which the p-dimethoxybenzene bridge moiety is attached only to ANI or NI alone is used to establish the structural and electronic requirements for this unusual charge separation mechanism.

Full text of DOI:10.1021/ja001298w

7802
J. Am. Chem. Soc. 2000, 122, 7802-7810
Photoinduced Charge Separation Involving an Unusual Double
Electron Transfer Mechanism in a Donor-Bridge-Acceptor Molecule
Scott E. Miller, Aaron S. Lukas, Emily Marsh, Patrick Bushard, and
Michael R. Wasielewski*
Contribution from the Department of Chemistry, Northwestern UniVersity, EVanston, Illinois 60208-3113
ReceiVed April 13, 2000
Abstract: A series of rodlike donor-bridge-acceptor (D-B-A) molecules was synthesized to study the role
of bridge energy levels on electron transfer (ET) rates. In these compounds, a 4-aminonaphthalene-1,8-imide
(ANI) electron donor is linked to a 1,8:4,5-naphthalenediimide acceptor (NI) via the 1,4 positions on a phenyl
bridge. The phenyl bridge is substituted at the 2 and 5 positions with methyl or methoxy groups to yield
ANI-diMe-NI and ANI-diMeO-NI. These molecules differ only in the energy levels of the bridge molecular
orbitals. Other parameters affecting ET rates such as donor-acceptor distance, orientation, and driving force
are constant between the two systems. The rate constants for charge separation (CS) and charge recombination
(CR) within ANI-diMeO-NI in toluene are 32 and 1400 times larger, respectively, than the corresponding rate
constants for ANI-diMe-NI. Solvents of higher polarity diminish these differences in rate constants, making
them comparable to those observed for ANI-diMe-NI. The relative energies of the ion pair states suggest that
it is possible for the reaction 1*D-B-A f D-B+-A- to occur via a double electron-transfer process that is
somewhat analogous to Dexter energy transfer. The lowest excited singlet state of the donor, 1*ANI, possesses
about 70% charge-transfer character, so that significant positive charge is localized on its amine nitrogen,
whereas significant negative charge is localized on its naphthalene-1,8-imide ring. Electron transfer from the
naphthalene-1,8-imide ring of 1*ANI to NI is concomitant with electron transfer from the p-dimethoxybenzene
1
bridge to the electron-deficient amine nitrogen atom in *ANI. A series of reference molecules in which the
p-dimethoxybenzene bridge moiety is attached only to ANI or NI alone is used to establish the structural and
electronic requirements for this unusual charge separation mechanism.
Introduction
In a typical donor-bridge-acceptor molecule (D-B-A) the
superexchange mechanism for electron transfer involves mixing
of electronic states of the bridge molecule with those of the
donor and acceptor. This mixing depends critically on both the
spatial overlap of the molecular orbitals of B with those of D
and A and the vertical energy gap between 1*D-B-A and the
energetically higher lying D+-B--A state.11 In principle,
several different states of the bridge molecule can contribute to
the overall electronic coupling between the donor and the
acceptor, their relative contributions being determined by the
electronic couplings and the energy gaps between the various
states.12-14 In addition to P+-BChl--H, several other states
have been invoked in an attempt to describe the primary ET
step in the reaction center. Bixon et al. analyzed femtosecond
spectroscopic data on the reaction center assuming that the
sequential and superexchange mechanism operate competi-
tively.5 The relative contributions of the two mechanisms were
The bacterial photosynthetic reaction center has been studied
extensively due to its ability to convert light energy into
chemical energy with near-unity quantum yield.1 The initial
events of charge separation consist of electron transfer (ET) from
the initially excited bacteriochlorophyll dimer (P) to the
bacteriopheophytin (H) over a distance of 17 Å in 3 ps. The
mechanism of this electron transfer is of particular interest, and
has been studied to determine the function of the intermediary
bacteriochlorophyll (BChl) that is positioned between P and
H.2-6 Discussions in the literature center on whether a two-
step, sequential electron transfer occurs, yielding the intermedi-
ate P+-BChl--H, or whether the BChl participates indirectly
via a superexchange mechanism.7-10
* Address correspondence to this author. E-mail: wasielew@
chem.nwu.edu.
(1) Deisenhofer, J.; Epp, O.; Miki, K.; Huber, R.; Michel, H. J. Mol.
Biol. 1984, 180, 385-398.
(2) Kirmaier, C.; Holten, D. Proc. Natl. Acad. Sci. U.S.A. 1990, 87,
3552-3556.
(3) Holzapfel, W.; Finkele, U.; Kaiser, W.; Oesterhelt, D.; Scheer, H.;
Stilz, H. U.; Zinth, W. Chem. Phys. Lett. 1989, 160, 1-7.
(4) Martin, J. L.; Breton, J.; Hoff, A. J.; Migus, A.; Antonetti, A. Proc.
Natl. Acad. Sci. U.S.A. 1986, 83, 957.
(5) Bixon, M.; Jortner, J.; Michel-Beyerle, M. E. Biochim. Biophys. Acta
1991, 1056, 301-315.
(6) Kirmaier, C.; Holten, D. Electron Transfer and Charge Recombina-
tion Reactions in Wild-Type and Mutant Bacterial Reaction Centers;
Deisenhofer, J., Norris, J. R., Eds.; Academic Press: San Diego, 1993; Vol.
2, pp 49-71.
(7) Ogrodnik, A.; Michel-Beyerle, M. E. Z. Naturforsch. 1989, 44a, 763-
764.
1
found to depend on the energy gap between the initial *P-
BChl-H state and the intermediate P+-BChl--H state. The
authors suggested that the presence of the two mechanisms
(8) DiMagno, T. J.; Norris, J. R. Initial Electron-Transfer EVents in
Photosynthetic Bacteria; Deisenhofer, J., Norris, J. R., Eds.; Academic
Press: San Diego, 1993; Vol. 2, pp 105-133.
(9) Sumi, H.; Kakitani, T. Chem. Phys. Lett. 1996, 252, 85-93.
(10) Plato, M.; Mobius, K.; Michel-Beyerle, M. E.; Bixon, M.; Jortner,
J. J. Am. Chem. Soc. 1988, 110, 7279-7285.
(11) McConnell, H. M. J. Chem. Phys. 1961, 35, 508-515.
(12) Kuznetsov, A. M.; Ulstrup, J. J. Chem. Phys. 1981, 75, 2047-2055.
(13) Tang, J.; Wang, Z.; Norris, J. R. J. Chem. Phys. 1993, 99, 979-
984.
(14) Tang, J.; Norris, J. R. J. Chem. Phys. 1994, 101, 5615-5622.
10.1021/ja001298w CCC: $19.00 © 2000 American Chemical Society
Published on Web 07/28/2000

The Role of Bridge Energy LeVels on Electron Transfer Rates
J. Am. Chem. Soc., Vol. 122, No. 32, 2000 7803
provide redundancy, which allows efficient charge separation
under many different conditions. Won and Friesner15 suggested
that the initial excited state mixes strongly with a charge
resonance state of the special pair dimer, increasing the
superexchange coupling. Fischer proposed that the involvement
of a virtual oxidized bridge state, P-BChl+-H-, is consistent
with the energies of the reaction center chromophores,16 and
further proposed a real state called the “trip-trip singlet” in
which triplet states on both P and the intermediate BChl interact
in what is formally a singlet state.17 More recently, Sumi and
Kakatani have suggested that a more useful approach would be
to describe bridge-mediated ET as a single, unique process rather
than attempting to characterize the relative contributions of
sequential ET versus superexchange.9
Several D-B-A systems have been synthesized and used
as experimental tests of superexchange. Sessler and co-workers
synthesized a series of triads consisting of two porphyrins and
a quinone bonded in various orientations.18 This group was able
to adjust the ion pair state energies of the bridging porphyrin
via selective metalation of one of the two macrocycles.
Photoinduced charge separation (CS) occurred in these Zn
porphyrin-quinone dyads on time scales of e1 ps, while thermal
charge recombination (CR) occurred in 3-6 ps. In free-base
porphyrin-zinc porphyrin-quinone (HP-ZnP-Q) triads, ex-
citation of ZnP formed a transient species whose 55-75 ps
lifetime showed little temperature dependence. This species was
identified as the HP+-ZnP-Q- ion pair formed by the
superexchange mechanism.
Warman, Paddon-Row, and Verhoeven examined the distance
dependence of photoinduced ET through norbornyl bridges of
varying length in an extensive series of D-B-A molecules.19,20
In addition, ester groups were added to the bridge molecules in
an attempt to adjust their orbital energies. However, no effect
was observed on the rates for either charge separation or
recombination. Unusually fast charge recombination was seen
in a triad of the form D2-D1-A, which undergoes two-step
CS to yield the giant-dipole state D2+-D1-A-.21 The observed
rapid CR reaction was rationalized in the context of a super-
exchange mechanism involving D-B+-A- virtual states. The
competition between sequential vs superexchange charge re-
combination was also considered in a series of triads studied
by the same group.22
Gust, Moore, and Moore have addressed the role of super-
exchange in a series of carotenoid-porphyrin-quinone triads.23
The ET rate was observed to decrease when rotation about a
single bond linking the porphyrin and quinone was hindered
by the steric influence of adjacent methyl groups. Furthermore,
the direction of the amide linkage between the porphyrin donor
and quinone acceptor demonstrated a larger, almost 30-fold
effect on the ET rates in these systems.
Wasielewski et al.24 have addressed the effects of changing
both structural isomers and the energy levels of the bridging
molecule. Several compounds were examined which employ a
pentiptycene spacer between a ZnP donor and a 1,4-naphtho-
quinone acceptor. Addition of two methoxy groups to the central
benzene ring of the pentiptycene spacer did not change the rate
of CS, but produced a 3- to 4-fold increase in the CR rate. By
comparing the intermediate state energies for the unsubstituted
and methoxy-substituted bridge molecules, the D-B+-A- state
was suggested as the virtual state that mediates the fast CR
reaction.
A series of chlorophyll-porphyrin-quinone triads that use
zinc chlorophylls as the primary electron donor has also been
studied.25,26 The bridge species is either a zinc or free-base
porphyrin and the electron acceptor moiety is a triptycene-1,4-
naphthoquinone. A comparison of the ET rates in compounds
with a ZnP bridge to those with the corresponding HP bridge
shows that the CS and CR rates increase 11- and 3-fold,
respectively, when the ZnP bridge is used. These effects were
observed despite the fact that HP is approximately 0.25 V easier
to reduce than ZnP. These results flatly contradict the expected
behavior if a D+-B--A virtual state mediates the CS reaction.
On the other hand, the data are consistent with the possible
involvement of a virtual D-B+-A- state in this reaction, based
on the corresponding relative ease of oxidation of the bridge
molecule.
Osuka et al. synthesized several multicomponent chlorophyll
and porphyrin systems in an effort to model the photosynthetic
RC.27,28 Chlorophyll-porphyrin-pyromellitimide (ZC-HP-
PI) triads, similar to the structures of Johnson et al.,25 were
studied in which phenyl rings separated each of the components
and pyromellitimide (PI) was the acceptor. In these systems
superexchange was ruled out as the CS mechanism based on
the similarity between the photophysical behavior of the ZC-
HP-PI triad and the ZC-HP dyad. When the bridge molecule
is HP, sequential CS results: 1*ZC-HP-PI f ZC+-HP--PI
f ZC+-HP-PI-. This mechanism was confirmed by direct
observation of the HP anion radical by transient absorption
spectroscopy. In an extension of this work, a 1,2-phenylene-
bridged ZnP dimer (D) was used as the primary chromophore
in an attempt to mimic the dimeric electron donor in the RC. A
superexchange mechanism for CR was implicated in the
D-HP-PI triad as evidenced by the biexponential decay of
D+-HP-PI-, in which one of the decay components displayed
no temperature dependence. Another compound, D-ZnP-PI,
exhibited CS in 500 ps in THF, yielding the distal ion pair D+-
ZnP-PI-. Unlike other compounds in this study, the reduced
state of the bridge was not detected, leading the researchers to
conclude that the electron transfer was superexchange-mediated.
Zimmt et al. have extended studies of superexchange in
structurally well-defined systems to include electron transfer
through trapped solvent in a series of C-clamp molecules.29
These molecules are based on extended polycyclic norbornyl
(15) Won, Y.; Friesner, R. A. Biochim. Biophys. Acta 1988, 935, 9-18.
(16) Fischer, S. F.; Scherer, P. O. J. Chem. Phys. 1987, 115, 151-158.
(17) Fischer, S. F.; Scherer, P. O. J. Eur. Biophys. J. 1997, 26, 477-
483.
(18) Sessler, J. L.; Johnson, M. R.; Lin, T.-Y. Tetrahedron 1989, 45,
4767-4784.
(24) Wasielewski, M. R.; Niemczyk, M. P.; Johnson, D. G.; Svec, W.
A.; Minsek, D. W. Tetrahedron 1989, 45, 4785-806.
(25) Johnson, D. G.; Niemczyk, M. P.; Minsek, D. W.; Wiederrecht, G.
P.; Svec, W. A.; Gaines, G. L., III; Wasielewski, M. R. J. Am. Chem. Soc.
1993, 115, 5692-701.
(19) Verhoeven, J. W. AdV. Chem. Phys. 1999, 106, 603-644.
(20) Warman, J. M.; Smit, K. J.; de Haas, M. P.; Jonker, S. A.; Paddon-
Row, M. N.; Oliver, A. M.; Jan, K.; Oevering, H.; Verhoeven, J. W. J.
Phys. Chem. 1991, 95, 1979-1987.
(26) Wiederrecht, G. P.; Watanabe, S.; Wasielewski, M. R. Chem. Phys.
1993, 176, 601-14.
(21) van Dijk, S. I.; Wiering, P. G.; van Staveren, R.; van Ramesdonk,
H. J.; Brouwer, A. W.; Verhoeven, J. W. Chem. Phys. Lett. 1993, 214,
502-506.
(27) Osuka, A.; Marumo, S.; Mataga, N.; Taniguchi, S.; Okada, T.;
Yamzaki, I.; Nishimura, Y.; Ohno, T.; Nozaki, K. J. Am. Chem. Soc. 1996,
118, 155-167.
(28) Osuka, A.; Mataga, N.; Okada, T. Pure Appl. Chem. 1997, 69, 797-
802.
(29) Kumar, K.; Lin, Z.; Waldeck, D. H.; Zimmt, M. B. J. Am. Chem.
Soc. 1996, 118, 243-244.
(22) Roest, M. R.; Lawson, J. M.; Paddon-Row, M. N.; W., V. J. Chem.
Phys. Lett. 1994, 230, 536-542.
(23) Kuciauskas, D.; Liddell, P. A.; Hung, S.-C.; Lin, S.; Stone, S.; Seely,
G. R.; Moore, A. L.; Moore, T. A.; Gust, D. J. Phys. Chem. B 1997, 101,
429-440.

7804 J. Am. Chem. Soc., Vol. 122, No. 32, 2000
Miller et al.
systems that have anthracene donors and dimethyl maleate
acceptors at opposite ends of the C-shaped molecule. Entrapment
of benzonitrile in the cleft results in a much larger electron-
transfer rate than does the more difficult to reduce acetonitrile.
This suggests that the lower energy D+-B--A virtual state
involving benzonitrile results in a faster rate of electron transfer
via the superexchange mechanism.
Chart 1
In the work presented here new D-B-A molecules based
on a 4-aminonaphthalene-1,8-imide (ANI) donor and a 1,8:4,5-
naphthalenediimide (NI) acceptor have been synthesized to study
the influence of bridging group energetics and orbital symmetry
on the formation of D+-B-A-. These particularly simple
structures are prepared from building blocks that lend themselves
readily to such studies.30-34 Synthetic details are provided in
the Supporting Information. In 1a and 1b the 1,4-phenyl bridges
between ANI and NI are substituted at the 2 and 5 positions
with methyl or methoxy groups (diX), respectively. These
substituents alter the energies of the highest occupied and lowest
unoccupied molecular orbitals (HOMO and LUMO, respec-
tively) of the bridging groups without changing other significant
factors such as the donor-acceptor distance and orientation
which affect electron-transfer rates. In addition, the π system
of the bridge is constrained by steric interactions to an
orientation that is nearly orthogonal to those of ANI and NI,
which places the π systems of the donor and acceptor in the
same plane. The molecular triads 2a and 2b are analogous to
the dyads, except that an additional electron donor moiety,
4-methoxyaniline (MeOAn), is attached to ANI via a piperazine
spacer. These compounds were synthesized to study the effects
of bridge energetics on charge separation when its mechanism
involves the initial migration of positive charge away from the
bridge. Excitation of ANI within 2a and 2b results in rapid
electron transfer from 4-methoxyaniline to 1*ANI to yield
MeOAn+-ANI--diX-NI. Thus, the initial CS step within triads
2a and 2b is the reduction of 1*ANI, whereas that for dyads 1a
1
and 1b is the oxidation of *ANI. Formation of the MeOAn+-
ANI-diX-NI- state within 2a and 2b occurs via a secondary
charge shift reaction.
been characterized previously in detail.35 The ground-state
optical spectrum of ANI in toluene exhibits a broad absorption
centered at 398 nm, which possesses about 70% charge-transfer
character. The ground-state absorption spectra of the bridged
dyads 1a and 1b in toluene show the absorption band due to
ANI as well as bands due to the NI acceptor at 343, 363, and
382 nm.36 The lowest energy NI absorption band partially
overlaps the blue edge of the ANI charge-transfer absorption
band, Figure 1.30,37 The ground state optical spectra of these
compounds are approximately the sum of the spectra of the
isolated chromophores, N-(p-tolyl)-4-(1-piperidinyl)naphthalene-
1,8-dicarboximide (ANI) and N,N′-di-(n-octyl)-1,8:4,5-naph-
thalene-tetracarboxydiimide (NI), respectively, suggesting weak
electronic interaction between the ANI and NI moieties in all
of these molecules. Increasing the solvent polarity results in no
spectral shift of the NI peaks, while the ANI charge-transfer
band shifts to 405 nm in butyronitrile (PrCN), further resolving
it from the NI vibronic bands. The corresponding ground-state
absorption spectra for the triads 2a and 2b exhibit analogous
behavior.
Compounds 3a and 3b were synthesized to determine what
1
role, if any, the bridge itself plays in quenching the *ANI
excited state, and to better understand the behavior of the
D-B-A dyads. They consist of identical donor and bridge
moieties, but replace NI with a terminal naphthalene-1,8-
monoimide (NMI), which cannot be reduced by 1*ANI;
however, its presence does provide a substituent effect similar
to that of NI on the electronic structure of the phenyl bridge. In
an analogous manner, it is possible to study the properties of
the B+-A- state by substituting NMI for ANI. The reference
molecules 4a and 4b, which lack the ANI donor, use NMI again
to maintain the effect of an imide substituent on the bridge
electronic structure.
Results
Steady-State Spectroscopy. The photophysics of the 4-(N-
piperidinyl)naphthalene-1,8-imide (ANI) chromophore have
(30) Korol’kova, N. V.; Val’kova, G. A.; Shigorin, D. N.; Shigalevskii,
V. A.; Vostrova, V. N. Russ. J. Phys. Chem. 1990, 64, 206-209.
(31) Alexiou, M. S.; Tychopoulos, V.; Ghorbanian, S.; Tyman, J. H. P.;
Brown, R. G.; Brittain, P. I. J. Chem. Soc., Perkin Trans. 2 1990, 837-
842.
(32) Dietz, T. M.; Stallman, B. J.; Kwan, W. S. V.; Penneau, J. F.; Miller,
L. L. J. Chem. Soc., Chem. Commun. 1990, 367-9.
(33) Debreczeny, M. P.; Svec, W. A.; Wasielewski, M. R. New J. Chem.
1996, 20, 815-828.
(34) Gosztola, D.; Wang, B.; Wasielewski, M. R. J. Photochem.
Photobiol., A 1996, 102, 71-80.
The charge transfer state of ANI decays radiatively in all
solvents, with an emission maximum at 497 nm in toluene. The
(35) Greenfield, S. R.; Svec, W. A.; Gosztola, D.; Wasielewski, M. R.
J. Am. Chem. Soc. 1996, 118, 6767-6777.
(36) Adachi, M.; Murata, Y.; Nakamura, S. J. Phys. Chem. 1995, 99,
14240-14246.
(37) deSilva, A. P.; Gunaratne, H. Q. N.; Habib-Jiwan, J.-L.; McCoy,
C. P.; Rice, T. E.; Soumillin, J.-P. Angew. Chem., Int. Ed. Engl. 1995, 34,
1728-1731.

The Role of Bridge Energy LeVels on Electron Transfer Rates
J. Am. Chem. Soc., Vol. 122, No. 32, 2000 7805
fluorescence quenching data for 2b, Table 2, suggests that the
corresponding methoxy-substituted derivative may be undergo-
ing the same reaction.35 Moreover, the data suggest that the
electron-transfer pathways that occur in 2b are most likely quite
different from those that occur in the corresponding dyad, 1b
(vide infra).
Compounds 4a and 4b have the structure NMI-B-A where
NMI is substituted for ANI, while NI is the isolated 1,8:4,5-
naphthalenediimide chromophore substituted at the amide
nitrogens with n-octyl groups. Excitation within the central
vibronic band of NI with 360 nm light results in formation of
the 1*NI excited state, whose emission maximum is 406 nm in
toluene. In 4a and 4b this emission is quenched significantly,
which suggests that B+-A- may form in these molecules. The
fluorescence quantum yields for these compounds are listed in
Table 3.
Redox Potentials. AC voltammetry was used to determine
the oxidation potentials of the ANI donor and the p-dimethoxy-
phenyl bridge in the D-B-A molecules, Table 4. The value
of Eox ) 1.41 V vs SCE for p-dimethoxybenzene obtained in
this study agrees reasonably well with the 1.32 V vs SCE
reported in the literature,38 as does Eox ) 1.20 V for ANI.35
Compound 1b undergoes two one-electron oxidations at 1.22
and 1.67 V. Compound 4b, lacking the ANI donor, undergoes
a single one-electron oxidation at 1.63 V. These results show
that imide groups bound to the 2 and 5 positions of p-
dimethoxybenzene increase its oxidation potential by about 0.25
V relative to that of p-dimethoxybenzene.
The similarity between the oxidation potentials of the bridge
molecules in 1b and 4b indicates that the positive change in
the oxidation potential of the p-dimethoxyphenyl bridge relative
to that of p-dimethoxybenzene is not due to the presence of the
ANI electron donor. There are two possible explanations for
this observed change. The imide substituents are somewhat
electron withdrawing, and are known to raise the oxidation
potentials of other systems.39 A second possibility is that steric
hindrance between the methoxy groups and the oxygen atoms
of the imides forces the methoxy groups out of the plane of the
bridge phenyl ring, which reduces the conjugation of the oxygen
lone pair with the π system of the phenyl, presumably making
it more difficult to oxidize. Geometry-optimized AM1 calcula-
tions40,41 predict that in 1b the average torsional angle that the
C-O bonds of the two p-dimethoxyphenyl groups make with
the π system of the phenyl is 24.5°. This compares with a
torsional angle of 8.6° in p-dimethoxybenzene. Thus, it is likely
that the increased oxidation potential of the p-dimethoxybenzene
bridge observed in 1b and 4b is a combination of steric and
electronic effects. The measured redox potentials for the donors
and acceptors, as well as those of selected bridge molecules,
were used to calculate the ion pair state energies of the various
intermediates observed in this study (vide infra).
Figure 1. UV-vis absorption spectra of the indicated molecules in
toluene.
Table 1. Spectroscopic Properties of the ANI Chromophore
solvent
λ
abs (nm)
λ
em (nm)
ΦF
ES (eV)
τf (ns)
toluene
MTHF
PrCN
398
397
405
497
514
530
0.910
0.790
0.230
2.81
2.76
2.70
8.5
7.5
1.5
spectroscopic properties of the isolated ANI chromophore in
the solvents toluene, MTHF, and PrCN are given in Table 1.
As the polarity of the solvent increases, the emission maximum
shifts to red and the fluorescence quantum yield decreases. These
1
effects are consistent with stabilization of the *ANI excited
state in higher polarity media, due to its charge-transfer
character.
The steady-state spectroscopic properties of 3a and 3b are
listed in Table 2. Compound 3a exhibits a ground-state
absorption spectrum in which the ANI CT absorption feature
dominates. A weaker absorption due to NMI appears at 340
nm. The absorption and emission spectra of 3a are virtually
identical to those of ANI and 1a in all solvents, which shows
that NMI has an effect similar to NI on the electronic structure
of the donor and bridge. Interestingly, the absorption maximum
of ANI within 3b exhibits a solvent-dependent red shift of 5 to
10 nm relative to both 3a and ANI itself, which is also reflected
in the emission spectra. The quantum yields of 3b also display
solvent-dependent behavior not observed in the isolated chro-
mophores. In toluene and MTHF the fluorescence quantum
yields of 3b are close to that of ANI, but in PrCN the
fluorescence quantum yield is quenched significantly relative
to that of ANI itself. These observations indicate that an
interaction between the donor and bridge occurs, whose
magnitude increases with solvent polarity and lowering of the
excited-state energy.
Time-Resolved Spectroscopy. Ultrafast transient absorption
spectroscopy allows for direct observation of intermediate states
formed during photoinduced charge separation reactions. Tran-
sient absorption measurements on the reference molecules 3a
and 3b in toluene display two features, illustrated for 3a in
1
The fluorescence emission from *ANI is quenched in the
presence of nearby electron acceptors. In the case of 1a this
quenching has been shown previously to arise from photo-
1
induced electron transfer from *ANI to NI.35 Fluorescence
1
Figure 2a. The positive ∆A at 450 nm is the *ANI excited-
quantum yields and emission maxima for the D-B-A dyads
1a and 1b are given in Table 2. Emission from 1b is quenched
more strongly than that of 1a in all solvents.
The absorption and emission features of the triads 2a and 2b
are very similar to those of the dyads, and are dominated by
the ANI and NI absorptions. Earlier data obtained for 2a show
that rapid electron transfer occurs from p-methoxyaniline to the
adjacent 1*ANI excited state, quenching the fluorescence. The
(38) Meites, L.; Zuman, P.; Scott, W. J.; Campbell, B. H.; Kardis, A.
M. Electrochemical Data; John Wiley and Sons: New York, 1974; Vol. 1.
(39) Wiederrecht, G. P.; Svec, W. A.; Wasielewski, M. R. Springer Ser.
Chem. Phys. 1994, 60, 452-3.
(40) Dewar, M. J. S.; Zoebisch, E. G.; Healy, E. F.; Stewart, J. J. P. J.
Am. Chem. Soc. 1985, 107, 3902-9.
(41) Ion pair distances were estimated from structures calculated using
the MM+ force field and AM1 MO calculations performed within
Hyperchem V5.01a (Hypercube, Waterloo Ontario).

7806 J. Am. Chem. Soc., Vol. 122, No. 32, 2000
Miller et al.
Table 2. Steady State Spectroscopic Properties of Bridged Compounds 1, 2, and 3
toluene
ΦF
MTHF
ΦF
PrCN
ΦF
compd
λabs
λem
τf (ns)
λabs
λem
τf (ns)
λabs
λem
τf (ns)
1a
1b
2a
2b
3a
3b
384
384
384
384
399
404
500
496
499
499
497
501
0.168
0.012
0.006
0.007
0.911
0.928
379
378
378
379
398
403
523
511
500
501
510
515
0.070
0.015
0.004
0.004
0.807
0.684
383
383
383
383
406
416
533
530
497
498
531
536
0.044
0.015
0.002
0.002
0.213
0.019
10.0
6.3
7.1
6.2
1.82
0.17
Table 3. Spectroscopic Properties of NI and X-B-A Reference
Compounds in Toluene
toluene
compd
λabs
λem
Φf
τ
CS (ps)
τ
CR (ps)
NI
4a
4b
382, 362, 342
381, 362, 340
381, 362, 340
406
405
406
0.05
0.0005
0.0006
0.63
0.63
125
114
Table 4. Redox Potentials (vs SCE)
compd
Eox1 (V)
Eox2 (V)
Ered (V)
NI
ANI
-0.53
-1.40
1.20
1.41
1.20
1.63
p-dimethoxybenzene
1b
4b
1.67
Figure 3. Comparison of the transient absorption spectra of 1a and
1b in toluene at the indicated times following a 400 nm, 130 fs laser
flash.
transient absorption spectrum of 1b in PrCN, Figure 2b, exhibits
a peak at 485 nm characteristic of the p-dimethoxybenzene
radical cation.42
The electron-transfer dynamics of the D-B-A compounds
1a and 1b were studied in several solvents to directly determine
their rates of charge separation and recombination. Upon
excitation in toluene their transient spectra display an absorption
band at 450 nm and stimulated emission at 535 nm. The spectra
of 1a and 1b both evolve with time to reveal absorption bands
at 480 and 607 nm characteristic of NI-.35 Figure 3 shows the
signals from 1a and 1b in toluene at the times at which their
maximum ion pair populations occur. The spectrum of 1b shows
1
a residual absorption at 450 nm from the *ANI excited state
and the intense absorption band due to the NI radical anion
(λmax ) 475 nm, ꢀ ) 28000 l‚M-1). The dielectric constant of
the solvent significantly affects the electron-transfer rates.
Monitoring the formation and decay of ∆A for NI- allows for
direct determination of the CS and CR time constants. The CS
and CR rates for the D-B-A dyads listed in Table 5 show
that in toluene these processes occur much faster in 1b than in
1a. However, this difference diminishes in higher polarity media
and converges to a common set of values.
The transient absorption spectra of 2a and 2b in toluene are
shown in Figure 4. Following excitation, a broad absorption
band appears near 500 nm, which is characteristic of the ANI
radical anion and the MeOAn cation radical.43,44 This spectral
feature is indicative of the formation of the MeOAn+-ANI--
diX-NI state, and evolves over several hundred picoseconds to
a spectrum characteristic of MeOAn+-ANI-diX-NI-, with peaks
at 480 and 607 nm. The time constants for each of these
Figure 2. Transient absorption spectra of (A) 3a in toluene and (B)
3b in PrCN at the noted times following a 400 nm, 130 fs laser flash.
state absorption, while the negative ∆A at 535 nm is due to
stimulated emission from this state. Both features appear in 3a
and 3b within the 180 fs instrument response function. As the
solvent polarity increases, the stimulated emission red shifts,
mirroring the steady-state spectra. While no electron transfer
occurs within either 3a or 3b in toluene and MTHF, the
fluorescence for 3b in PrCN is quenched and its excited-state
lifetime is much shorter. For 3b in PrCN the p-dimethoxyphenyl
(42) Shida, T. Electronic absorption spectra of radical ions; Elsevier:
Amsterdam, 1988.
(43) Kimura, K.; Yoshinaga, K.; Tsubomura, H. J. Phys. Chem. 1967,
71, 4485-4491.
(44) Hester, R. E.; Williams, K. P. J. J. Chem. Soc., Perkin Trans. 2
1982, 559-563.
1
bridge is a good enough electron donor to reduce *ANI. The

The Role of Bridge Energy LeVels on Electron Transfer Rates
J. Am. Chem. Soc., Vol. 122, No. 32, 2000 7807
Table 5. Electron Transfer Time Constants for Compounds 1 and 2
toluene
MTHF
PrCN
compd
τCS1 (ps)
τ
CS2 (ps)
τ
CR (ps)
τ
CS1 (ps)
τ
CS2 (ps)
τ
CR (ps)
τ
CS1 (ps)
τ
CS2 (ps)
τ
CR (ps)
1a
1b
2a
2b
1500
47
8
170000
119
230000
25000
516
152
2
2410
594
1350
170
86
85
1.5
1
456
350
4
430
600
370
95
8
2
10
Figure 5. Comparison of the transient absorption spectra for 4a and
4b in toluene at 25 ps following a 360 nm, 130 fs laser flash.
1
the solvent dependence of the excited-state energy of *ANI.
Thus, to a reasonable approximation the calculation of the ion
pair energy depends only on the donor-acceptor distance, and
differences in redox potentials. Using this model, no additional
solvation terms are material to the calculation, wherein the use
of these terms in the dielectric continuum model of Weller45
consistently underestimates the degree to which a polarizable
solvent, such as toluene, stabilizes an ion pair.46 Using the
1
spectroscopic data for *ANI and the redox potentials of the
Figure 4. Transient absorption spectra of (A) 2a and (B) 2b in toluene
donor and acceptor, the ion pair state energy of ANI+-diMe-
NI- in toluene is 2.42 eV.35 The energy of the ANI+-diMeO-
NI- ion pair is also 2.42 eV in toluene because the donor,
acceptor, and ion pair distance are identical. The energies of
the ion pair states involving the bridge molecules, ANI+-B--
NI and ANI-B+-NI-, in toluene were calculated relative to
the energy of ANI+-diMe-NI- using the relationship
at the noted times following a 400 nm, 130 fs, laser flash.
electron-transfer events are listed in Table 5. Interestingly, the
kinetics of the charge shift reaction MeOAn+-ANI--diX-NI f
MeOAn+-ANI-diX-NI- show only a weak dependence on the
substituent X.
The reference compounds NI, 4a, and 4b have no significant
ground-state absorption at 400 nm, so that transient absorption
data were acquired by direct excitation of the NI component at
360 nm. The excited-state spectrum of 1*NI has an asymmetric
absorption centered at 510 nm, Figure 5, which decays with τ
) 3 ns. This lifetime is significantly reduced in 4a and 4b. In
each of these compounds, the transient spectra evolve rapidly
in <1 ps to yield the absorption spectrum characteristic of NI-,
Figure 5. This can be attributed to electron transfer from the
e2
1
1
∆GIP2 ) ∆GIP1 + |E2 - E1| +
-
(1)
(
)
ꢀs rIP1 rIP2
where ∆GIP1 and rIP1 are the ion pair energy and distance,
respectively, in compound 1a, ∆GIP2 and rIP2 are the ion pair
energy and distance,41 respectively, for ANI+-B--NI or ANI-
B+-NI-, E2 is the oxidation or reduction potential of the bridge
molecule, E1 is the oxidation potential of ANI or the reduction
potential of NI, e is the electronic charge, and ꢀs is the static
dielectric constant of the solvent. The ion pair energies in MTHF
and PrCN were calculated directly using the oxidation and
reduction potentials of the donor and acceptor, respectively, and
the Coulomb stabilization of the ion pair
1
adjacent phenyl ring to *NI, forming the ion pair NMI-B+-
A-. The lifetimes of these ion pair states are listed in Table 3.
It should be noted that the ion pairs formed within 4b and 1b
have nearly identical CR rates, which suggests that the same
charge separated state may be formed in both compounds.
Discussion
Energetics. The lowest excited-state energies of ANI in
e2
ꢀs‚rDA
toluene, MTHF, and PrCN are listed in Table 1. Greenfield et
∆GIP ) EOX - ERED
-
(2)
1
al. showed that *ANI possesses about 70% charge-transfer
character.35 Using this fact and the solvent dependence of the
energy of this state, electron transfer from 1*ANI to NI can be
treated as a charge shift reaction. Energy changes due to
solvation of the charge distribution are taken into account by
(45) Weller, A. Z. Z. Phys. Chem. 1982, 133.
(46) Warman, J. M.; Smit, K. J.; Jonker, S. A.; Verhoeven, J. W.;
Oevering, H.; Kroon, J.; Paddon-Row, M. N.; Oliver, A. M. Chem. Phys.
1993, 170, 359-380.

7808 J. Am. Chem. Soc., Vol. 122, No. 32, 2000
Miller et al.
Figure 6. Energy levels for 3b in the indicated solvents.
Figure 8. Energy levels for 1a and 1b in toluene.
diMeO+-NMI ion pair state lies well above the excited state of
the ANI chromophore, while in PrCN the ion pair and excited
state are nearly isoenergetic. The transient spectrum of 3b in
PrCN (Figure 2b) supports this mechanism. A peak at 485 nm,
1
which is resolved from the short-lived absorption of *ANI at
450 nm, appears in neither 3a in PrCN nor 3b in MTHF, where
the D--B+ state is not energetically favored. Thus, the observed
485 nm peak is attributed to the p-dimethoxyphenyl radical
cation,42 whose formation is strictly dependent on solvent
polarity.
The dyads 1a and 1b are analogous to the model com-
pounds 3a and 3b except that the NI electron acceptor in 1a
1
and 1b can be reduced by *ANI. These compounds demon-
strate that methoxy substituents on the phenyl bridge dramati-
cally increase both the CS and CR rates relative to those
observed for the dimethyl-substituted phenyl bridge. Again,
solvent polarity plays a major role in determining the mecha-
nisms of CS and CR. Previous work has demonstrated that
electron transfer in 1a occurs predominantly through the
σ-bonded framework of the bridge in all solvents.35 The data
are consistent with small donor-bridge and bridge-acceptor
electronic couplings due to the orthogonal arrangement of the
bridge relative to both the donor and the acceptor. In low polarity
media this superexchange mechanism results in slow rates for
both CS and CR, while both processes become faster in higher
polarity media.
Figure 7. Energy level diagram featuring the frontier molecular orbitals
of D, B, and A within 1a (- - -) and 1b (s). All orbitals above 2.0 eV
are σ orbitals, while those below 2.0 eV are π orbitals.
where rDA is the distance41 between the separated ions. The
corresponding values of the free energies of CS and CR are
∆GCS ) ∆GIP - ES
where ES is the lowest excited-state energy of the donor, and
∆GCR ) -∆GIP (4)
(3)
In toluene the CS and CR rates for 1b increase by factors of
32 and 1400, respectively, over those for 1a, despite the
seemingly small electronic coupling due to the orthogonal
bridging group. If electron transfer in 1b also occurs predomi-
nantly through the σ-bonded framework of the bridge, then one
might expect an enhancement of the electron-transfer rates due
to the inductive electron withdrawing effect of the methoxy
groups on the bridge. The electronic coupling matrix element
for an electron-transfer reaction mediated by a superexchange
interaction involving the bridge is given by
The calculated CS free energies predict that electron transfer
is thermodynamically possible in all three solvents for all
compounds having both the ANI donor and the NI acceptor. A
selection of these data relevant to the arguments discussed below
are presented in Figures 6 and 8.
Photoinduced Charge Separation and Recombination
Mechanisms. Model compounds 3a and 3b contain the terminal
NMI group that thermodynamically cannot be reduced by
1*ANI. In toluene and MTHF both compounds have fluores-
cence quantum yields comparable to or greater than that of ANI
alone, suggesting that electron transfer does not occur. In the
polar solvent PrCN, however, the fluorescence quantum yield
of 3b is reduced by more than an order of magnitude relative
to those of ANI and 3a. This suggests the possibility of bridge-
to-donor electron transfer, forming ANI--diMeO+-NMI. The
energies of this state in toluene, MTHF, and PrCN were
calculated using eq 2, the reduction potential of ANI, and the
oxidation potential of the p-dimethoxyphenyl bridge, Figure 6.
The calculations indicate that in toluene and MTHF the ANI--
V
DBVBA
VDA
)
(5)
∆E
where VDB and VBA are the electronic coupling matrix elements
for the donor-bridge and bridge-acceptor interactions, and ∆E
is the vertical energy difference between the potential energy
surfaces of the donor and bridge states.15 Given that the electron-
transfer rate constant kDA V2DA, and assuming that VDB and
V
BA are similar for 1a and 1b, the ratio of electron-transfer rate

The Role of Bridge Energy LeVels on Electron Transfer Rates
constants k1b/k1a can be expressed as
J. Am. Chem. Soc., Vol. 122, No. 32, 2000 7809
2
k1b ∆E1a
)
(6)
2
k1a
∆E1b
It is not possible to obtain the relevant energies of the σ
orbitals of B to include in eq 6 using electrochemical data
because the LUMO of B, which would accept the electron, is
a π orbital. Even the energy of the π symmetric LUMO of
benzene itself is difficult to obtain from electrochemical data
because its reversible thermodynamic reduction potential is very
negative, E1/2 ) -3.4 V vs SCE.47 As an alternative, the orbital
energies of the frontier MOs localized on D, B, and A within
1a and 1b were calculated using the AM1 model and are shown
in Figure 7. The calculations show that the lowest energy
unoccupied orbitals in both compounds are π orbitals localized
on A. The lowest unoccupied π orbitals localized on B in 1a
and 1b are 1.4 and 1.2 eV above the HOMO of D, respectively,
while the corresponding σ orbitals of B are 5.0 and 3.9 eV above
the HOMO of D. This is the usual energy ordering for D-B-A
molecules. The orientation of ANI relative to the bridge suggests
that electron transfer from the HOMO of ANI to NI involves
the σ orbitals of B in a superexchange interaction. Using the
energy difference between the HOMO of ANI and the lowest
unoccupied σ orbital of B for 1a and 1b in eq 6 yields k1b/k1a
) 1.6, while the experimentally determined ratio is 32. Thus,
the inductive effect of the methoxy groups and a superexchange
mechanism that uses the σ bonds of the bridge cannot account
for the observed increase in CS and CR rates for 1b relative to
those for 1a.
Figure 9. Proposed orbital interactions for the double electron-transfer
mechanism.
to the donor and the donor to the acceptor. In addition, these
two electron-transfer processes must be coupled to each other,
otherwise electron transfer from 1*D to A that uses the LUMO
of the bridge would be rate limiting. The highly CT nature of
1*ANI may promote such coupling, as visualized in Figure 9.
The positive charge residing in the lone pair orbital of the amine
1
nitrogen within *ANI has a convenient route to the π-system
of the bridge via σ orbitals of the aryl C-H bonds of the
naphthalene-1,8-imide and the lone pair orbitals of the carbonyl
groups. The negative charge of the CT state, which resides
primarily in the π-symmetric orbitals of the imide group,
likewise can undergo electron transfer directly to the π orbital
of NI. Interestingly, this mechanism is observed only in toluene,
while the difference in both the CS and CR rates between dyads
1a and 1b is less pronounced in MTHF, and nearly absent in
PrCN. This is a consequence of the fact that in both MTHF
and PrCN the D-B+-A- state within both 1a and 1b is higher
in energy than the corresponding D+-B-A- state and the
electron transfer proceeds via the conventional superexchange
mechanism.
On the other hand, if the π orbitals of the bridge are involved
in the electron transfer, k1b/k1a should be influenced significantly
by replacing the methyl substituents on the phenyl bridge with
the more π electron-releasing methoxy substituents. Experi-
mentally, p-dimethoxybenzene is easier to oxidize than p-
dimethylbenzene by about 0.4 V,48 which is reflected in the
ordering of the energies of the calculated HOMOs for the bridges
and ANI in 1a and 1b, Figure 7. In 1a, the HOMO is localized
on ANI, while the next highest occupied orbital (HOMO-1) is
localized on the p-dimethylphenyl bridge. The energy gap
between these two orbitals is 0.3 eV. The situation is reversed
in 1b where the HOMO is localized on the p-dimethoxyphenyl
bridge, and the energy difference between it and the highest
occupied orbital localized on ANI is 0.3 eV. Notwithstanding
the fact that these calculations were made on isolated molecules
in the gas phase, these results predict that bridge-to-donor
This mechanistic picture is further supported by examination
of the CS dynamics in the triads 2a and 2b. As previously
mentioned, the fundamental difference between the dyads and
triads centers on the initial electron-transfer reaction; in the
dyads 1*ANI acts as an electron donor, whereas in the triads, it
functions as an electron acceptor. In the triads, ANI is reduced
in the initial photoinduced electron transfer to yield MeOAn+-
ANI--diX-NI. In this intermediate a full negative charge resides
on the naphthalene-1,8-imide ring of ANI, while the positive
charge is moved farther away and resides on the oxidized
p-methoxyaniline donor. Thus, only the charge shift from ANI-
to NI occurs. The coplanar orientation of the π systems of ANI
and NI within 2a and 2b suggests that only direct interactions
of these orbitals and/or participation of the σ orbitals of the
phenyl bridge should be important in transferring the electron
from ANI- to NI. If this model is correct, changing the
substituents on the phenyl bridge should have a minor impact
on the charge shift rate constant. The spectral (Figure 4) and
kinetic (Table 5) data both show that there is no significant
difference between the dynamics observed in 2a and 2b in any
solvent. These data lend further support to the arguments
presented earlier that the inductive effect of the methoxy groups
on the bridge molecule have only a small impact on electron-
transfer rates that employ the σ-bonded pathway through the
bridge. Thus, the unusual double electron-transfer mechanism
proposed for 1b depends primarily on the energy and symmetry-
dependent electronic interaction between the HOMO of 1*ANI
and that of the p-dimethoxyphenyl bridge.
1
electron transfer is possible in 1b, but not in 1a. Thus, *D-
B-A may undergo a double electron transfer in which electron
transfer from 1*D to A occurs concomitant with electron transfer
from B to 1*D, as shown in Figure 7. The energy of the
hypothetical intermediate D-B+-A- state in the CS reactions
of 1a and 1b can be calculated using eq 1. The results of this
calculation for 1a and 1b in toluene are shown in Figure 8.
The D-B+-A- state in 1a is higher in energy than D+-B-
A-, while the energy of the D-B+-A- state in 1b lies below
that of D+-B-A- by 0.06 eV as well as significantly below
1
1
that of *D-B-A. These data show that the reaction *D-
B-A f D-B+-A- is thermodynamically possible in 1b.
While thermodynamic arguments favor the formation of a
D-B+-A- state within 1b, the rapid formation of this state
requires an electronic interaction that couples both the bridge
Additional support for the double electron-transfer mechanism
in 1b is presented by transient absorption data on the reference
compounds 4a and 4b. These systems lack the ANI electron
donor and thus present the possibility of forming the B+-A-
(47) Mortensen, J.; Heinze, J. Angew. Chem. 1984, 96, 64-65.
(48) Howell, J. O.; Goncalves, J. M.; Amatore, C.; Klasinc, L.;
Wightman, R. M.; Kochi, J. K. J. Am. Chem. Soc. 1984, 106, 3968-3976.
1
state independent of the production of *ANI. If this state is
formed in the dyad 1b, then identical CR rates should be

7810 J. Am. Chem. Soc., Vol. 122, No. 32, 2000
Miller et al.
The femtosecond transient absorption apparatus for time-resolved
measurements has been described in detail elsewhere.49 Transient
lifetimes of longer than 5 ns were determined using a 10-Hz Nd:YAG
laser system. The frequency-tripled output of the Nd:YAG laser (355
nm, ∼7 ns) was focused in a pressure cell with 100 PSI H2, and the
Raman line at 416 nm was used to excite the sample and could be
blocked by a computer-controlled shutter. The probe light was generated
by a xenon flashlamp (EG&G) driven by a homemade circuit, and
detected with a Hamamatsu R928 photomultiplier tube with only four
dynodes connected. Typically, 10 laser shots each with the pump beam
on and pump beam off were averaged by a LeCroy 9384 digital
oscilloscope and sent to a computer. Lifetimes of up to 50 µs could be
determined with this apparatus. Samples had an optical density of 0.2-
0.4 in a 1-cm cuvette. All samples were deoxygenated by bubbling
with N2 for 10 min prior to the experiment.
observed for 1b and 4b in toluene. As noted previously, when
the NI chromophore is photoexcited with 150 fs, 360 nm laser
flashes, the transient spectrum evolves rapidly with time to show
the excited state with an asymmetric band centered at 520 nm,
Figure 5. However, in 4a and 4b identical excitation shows the
presence of NMI-diX+-NI-. The most likely mechanism by
which this ion pair forms is electron donation from the adjacent
phenyl to 1*NI. This reaction occurs in <1 ps in both in 4a and
4b, Table 3. The NMI-diMe+-NI- state undergoes CR to the
ground state with τ ) 50 ps, while the NMI-diMeO+-NI- ion
pair decays with τ ) 114 ps. This latter time constant agrees
very well with the τ ) 119 ps decay observed for the ANI-
diMeO+-NI- ion pair proposed to form in 1b.
Conclusions
Absorption measurements were made on a Shimadzu spectrometer
(UV-1601), and fluorescence measurements were made on a single-
photon-counting fluorimeter (PTI), exciting at 400 or 340 nm. Samples
for quantum yield measurements had absorptions of 0.1 ( 0.05 at the
excitation wavelength. Quantum yields for 400 nm excitation were
determined by comparison to ANI (æF ) 0.91), while samples excited
at 340 nm were determined by comparison to pyrene (æF ) 0.7).
Electrochemistry was performed using a model 173 potentiostat, a
model 179 digital coulometer, and a model 175 universal programmer,
all from Princeton Applied Research. A small AC signal was applied
on top of the voltage sweep using an Ithaco Dynatrac 391 lock-in
amplifier, and the response was plotted. All AC voltammetry was
performed in 0.1 M tetra-n-butylammonium perchlorate in twice-
distilled PrCN using a platinum working electrode, a platinum mesh
counter electrode, and a saturated calomel reference electrode (SCE).
Addition of π electron-donating methoxy groups to the bridge
of the ANI-diX-NI molecule increases the rates of charge
separation and recombination relative to those of the corre-
sponding dimethyl-substituted bridge. The effect is particularly
strong in toluene and decreases as the solvent polarity increases.
A mechanism consistent with the observed spectroscopic and
kinetic behavior is the formation of a charge-separated state
ANI-diMeO+-NI- via a double electron transfer mechanism that
is somewhat analogous to Dexter energy transfer. This mech-
anism depends critically on the orbital energies and symmetries
of the donor relative to those of the bridge. In polar media the
energies change sufficiently to return the system to a normal
ET mechanism in which electron transfer occurs by means of
bridge-mediated superexchange involving the virtual D+-B--
A state. These observations suggest that the scope of mechanistic
possibilities for bridge-mediated electron transfer within organic
molecules is very broad, and that given a set of appropriate
conditions, unusual mechanisms can strongly dominate the
overall electron-transfer reaction.
Acknowledgment. This work was supported by the National
Science Foundation (CHE-9732840). The authors also thank Dr.
Martin Debreczeny for carrying out preliminary experiments
in this study.
Supporting Information Available: The synthesis and
characterization of all new molecules (PDF). This material is
available free of charge via the Internet at http://pubs.acs.org.
Experimental Section
All solvents used were either HPLC or spectrophotometric grade
except for THF and MTHF. The latter two were distilled over lithium
aluminum hydride prior to use. The synthesis and characterization of
the new molecules studied in this paper are given in the Supporting
Information.
JA001298W
(49) Lukas, A. S.; Miller, S. E.; Wasielewski, M. R. J. Phys. Chem. B
2000, 104, 931-940.