Intramolecular electron transfer through the 20-position of a chlorophyll a derivative: An unexpectedly efficient conduit for charqe transport

Article abstract of DOI:10.1021/ja058233j

Suzuki cross-coupling reactions have afforded 20-phenyl-substituted Chlorophyll a derivatives (ZCPh) in good yields and significant quantities from readily available Chl a. A series of donor-acceptor dyads was synthesized in which naphthalene-1,8:4,5-bis(

Full text of DOI:10.1021/ja058233j

Published on Web 03/15/2006
Intramolecular Electron Transfer through the 20-Position of a
Chlorophyll a Derivative: An Unexpectedly Efficient Conduit
for Charge Transport
Richard F. Kelley, Michael J. Tauber, and Michael R. Wasielewski*
Contribution from the Department of Chemistry and International Institute of Nanotechnology,
Northwestern UniVersity, EVanston, Illinois 60208-3113
Received December 4, 2005; E-mail: m-wasielewski@northwestern.edu
Abstract: Suzuki cross-coupling reactions have afforded 20-phenyl-substituted Chlorophyll a derivatives
(ZCPh) in good yields and significant quantities from readily available Chl a. A series of donor-acceptor
dyads was synthesized in which naphthalene-1,8:4,5-bis(dicarboximide) or either of two perylene-3,4:9,-
10-bis(dicarboximide) electron acceptors is attached to the para position of the 20-phenyl group.
Comparisons with the analogous dyads based on a zinc 5,10,15-tri(n-pentyl)-20-phenylporphyrin donor
show that, for a given acceptor and solvent, the rates of photoinduced charge separation and recombination
as well as the calculated electronic coupling matrix elements, V, for these reactions differ by less than a
factor of 2. However, EPR and ENDOR spectroscopy corroborated by DFT calculations show that the
highest occupied MO of ZCPh+• has little spin (charge) density at the 20-carbon atom, whereas Z3PnPh+•
has significant spin (charge) density there, implying that V, and therefore the electron-transfer rates, should
differ significantly for these two macrocyclic donors. DFT calculations on ZCPh+• and Z3PnPh+•, with two
-0.5 charges located where the nearest carbonyl oxygen atoms of the acceptor would reside in the donor-
acceptor dyads, show that the presence of the negative charges significantly shifts the charge density of
both ZCPh+• and Z3PnPh+• from the macrocycle onto the phenyl rings. Thus, the presence of adjacent
covalently linked radical anions at a fixed location relative to each of these radical cations results in nearly
identical electronic coupling matrix elements for electron transfer and therefore very similar rates.
Introduction
Chlorophyll a (Chl a) and its derivatives self-assemble in
vitro into large aggregates using two principal structural
motifs.7,8 First, in the absence of other nucleophiles, the 131-
ketone of one Chl a serves as an axial ligand for the Mg of
another Chl a. Second, the presence of a hydrogen-bonding
nucleophile, such as water or an alcohol, ROH, results in ligation
of the Mg of one Chl a by the oxygen atom of ROH, while the
H atom of ROH hydrogen-bonds to the 131-ketone of another
Chl a. Both motifs result in J-type aggregates having red-shifted
optical transitions.9 In most biomimetic Chl-based systems
synthesized previously, electron acceptors have been attached
to positions on the Chl macrocycle that limit the ability of the
system to self-assemble into supramolecular structures by either
changing the steric requirements for assembly or introducing
additional intermolecular interactions. For example, the 3b-vinyl
position of Chl a has been functionalized with bulky triptycene-
quinone,10 arylene bis(dicarboximide),11,12 and C6013-15 electron
acceptors. The shapes and sizes of these acceptors prevent the
often desirable self-association of the Chl a derivatives along
Photosynthetic antenna and reaction center proteins serve as
important models for developing synthetic light-harvesting and
photochemical charge separation systems.1,2 Chlorophylls have
strong electronic absorptions in both the blue and red regions
of the visible spectrum, thus providing excellent solar spectral
coverage. The intense lowest energy Qy electronic transition in
chlorophylls is strongly polarized along the ring A-ring C axis,
and the coupling of this transition dipole to those of other
chlorophylls is largely responsible for many of the spectral forms
that are observed for chlorophylls in proteins. The spatial
relationships between chlorophylls bound to proteins are
maintained using weak interactions, such as ligation of their
central magnesium atoms to histidines and hydrogen bonding
of their oxygen-containing functional groups to nearby amino
acids. Notable exceptions to this binding motif are the chloro-
phylls present in the peripheral antenna system of green sulfur
bacteria, which engage in direct chlorophyll-chlorophyll
association to produce large cylindrical arrays of pig-
ments.3-6
(5) Blankenship, R. E.; Olson, J. M.; Miller, M. In Anoxygenic Photosynthetic
Bacteria; Blankenship, R. E., Madigan, M. T., Bauer, C. E., Eds.; Kluwer:
Dordrecht, 1995; pp 399-435.
(1) Wasielewski, M. R. Chem. ReV. 1992, 92, 435-461.
(2) Gust, D.; Moore, T. A.; Moore, A. L. Acc. Chem. Res. 1993, 26, 198-
205.
(3) Staehelin, L. A.; Golecki, J. R.; Drews, G. Biochim. Biophys. Acta 1980,
589, 30-45.
(4) Worcester, D. L.; Michalski, T. J.; Tyson, R. L.; Bowman, M. K.; Katz, J.
J. Physica B (Amsterdam, Neth.) 1989, 156-157, 502-504.
(6) van Rossum, B. J.; Steensgaard, D. B.; Mulder, F. M.; Boender, G. J.;
Schaffner, K.; Holzwarth, A. R.; de Groot, H. J. M. Biochemistry 2001,
40, 1587-1595.
(7) Katz, J. J. Inorganic Biochemistry; Elsevier: Amsterdam, 1973; Vol. 2.
(8) Katz, J. J.; Shipman, L. L.; Cotton, T. M.; Janson, T. R. Porphyrins 1978,
5, 401-458.
(9) Kasha, M.; Rawls, H. R. Pure Appl. Chem. 1965, 11, 371-392.
9
10.1021/ja058233j CCC: $33.50 © 2006 American Chemical Society
J. AM. CHEM. SOC. 2006, 128, 4779-4791
4779

A R T I C L E S
Kelley et al.
the ring A-ring C axis. Holzwarth et al.13 compared the self-
assembly characteristics of pyro-Chl a having a C60 rigidly
linked to its 3-position to those of the corresponding molecule
having a C60 flexibly attached at the less sterically hindered
173-position. They concluded that the self-assembly character-
istics of the pyro-Chl a should only be preserved by attaching
It is well known that the rates of both Dexter-type energy
transfer and electron transfer to or from metalloporphyrins
depend on the position of the macrocycle to which the donor
or acceptor is attached.24-28 The rates of these transfers, kDA
,
are given by kDA VDA2(FCWD), where VDA is the electronic
coupling matrix element and FCWD is the Franck-Condon
weighted density of states (see below).29-31 Since the matrix
element VDA depends on orbital overlap between the frontier
molecular orbitals of the metalloporphyrin and those of the donor
or acceptor,32 rates of Dexter-type energy transfer and electron
transfer depend on the electron density distributions within these
orbitals.24,33 The two highest occupied molecular orbitals
(HOMO and HOMO-1) within metalloporphyrins are nearly
degenerate and have electron density distributions that differ
greatly.34-36 The a1u orbital has significant electron density at
the â carbons of the porphyrin with little density at the meso
positions, while the opposite is true for the a2u orbital. The
energetic ordering of these two orbitals in the metalloporphyrin
depends on the nature and the position of the substituents
attached to the porphyrin.36 Substituting electron-donating
groups, e.g., phenyl rings or alkyl chains, at the meso positions
(5,10,15,20) of the porphyrin raises the energy of the a2u orbital
above that of the a1u orbital, making the a2u orbital the HOMO.
On the other hand, substitution of the â positions with similar
electron-releasing groups results in the a1u orbital being the
HOMO. This is consistent with EPR data obtained for zinc
meso-tetraphenylporphyrin radical cation, which shows that its
meso positions have significantly greater spin (and charge)
density compared to those at the â positions.36 It is also
consistent with EPR measurements on related zinc porphyrins
having a mix of both alkyl and aryl substituents at their meso
positions,37 which all show that the a2u orbital is the zinc
porphyrin HOMO. Generally, increased electron density at
particular peripheral carbon atoms of the porphyrin results in
stronger electronic coupling to molecules attached to those
positions.32 If the attached molecule can accept energy and/or
electrons from the porphyrin, the increased electronic coupling
leads to increased rates of energy and/or electron transfer.24,33
For example, Hayes et al.26 showed that a zinc porphyrin having
a pyromellitimide electron acceptor attached to a meso phenyl
group has a charge recombination rate that is 3 times faster than
that of a porphyrin in which the acceptor is attached to a
â-phenyl group.
C
60 to the 173-position. The ability to attach electron acceptors
to the Chl macrocycle in a rigid fashion without interfering with
functional groups that are essential to self-assembly is important
for developing photofunctional structures that carry out rapid,
directional energy and charge transport.1,2,16-20
The 20-position of Chl a offers a potential site for rigid
attachment of either redox or energy-transfer components that
should not interfere with self-assembly. Using the 20-position
of Chl a as a conduit for charge transport is unprecedented,
although several related synthetic chlorin-acceptor molecules
have been previously studied. Lindsey has prepared a series of
synthetic oxochlorin dyads in which zinc and magnesium
oxochlorins are linked to the corresponding free bases using
phenyl and phenylethynyl linkers.21-23 Laser flash photolysis
studies of these molecules show that energy transfer from the
metallo-oxochlorins to the free base is slower than that observed
for the corresponding porphyrin systems. This difference is
attributed to through-bond Dexter-type energy transfer, which
depends critically on the electron density distribution in the
frontier molecular orbitals of the donor and acceptor,21 as well
as the properties of the linker group.22 Kirmaier et al.23 have
studied electron transfer from similar metallo-oxochlorins to a
perylene-3,4:9,10-bis(dicarboximide) (PDI) molecule linked to
the 10-position of the oxochlorin using a diphenylacetylene
spacer. Photoexcitation of PDI in these dyads results in
competitive singlet-singlet energy transfer to the oxochlorin
1
and electron transfer from the oxochlorin to *PDI. Electron
transfer from the photoexcited oxochlorin to PDI is relatively
slow, even in polar solvents, due primarily to free energy
considerations. Direct comparisons of the electron-transfer
characteristics of the oxochlorin dyads with the corresponding
porphyrin dyads were not reported.
(10) Johnson, D. G.; Niemczyk, M. P.; Minsek, D. W.; Wiederrecht, G. P.;
Svec, W.; Gaines, G. L., III; Wasielewski, M. R. J. Am. Chem. Soc. 1993,
115, 5692-5701.
(11) Wiederrecht, G. P.; Niemczyk, M. P.; Svec, W.; Wasielewski, M. R. J.
Am. Chem. Soc. 1996, 118, 81-88.
(12) Wasielewski, M. R.; Johnson, D. G.; Svec, W. A. NATO ASI Ser., Ser. C
1987, 214, 255-266.
(13) Holtzwarth, A. R.; Katterle, M.; Muller, M. G.; Ma, Y.-Z.; Prokhorenko,
V. Pure Appl. Chem. 2001, 73, 469-474.
(14) Fukuzumi, S.; Ohkubo, K.; Imahori, H.; Shao, J.; Ou, Z.; Zheng, G.; Chen,
Y.; Pandey, R. K.; Fujitsuka, M.; Ito, O.; Kadish, K. M. J. Am. Chem.
Soc. 2001, 123, 10676-10683.
(24) Holten, D.; Bocian, D. F.; Lindsey, J. S. Acc. Chem. Res. 2002, 35, 57-
69.
(15) Gust, D.; Moore, T. A.; Moore, A. L. Acc. Chem. Res. 2001, 34, 40-48.
(16) Jordens, S.; De Belder, G.; Lor, M.; Schweitzer, G.; Van der Auweraer,
M.; Herrmann, A.; Wiesler, U. M.; Mu¨llen, K.; De Schryver, F. C.
Photochem. Photobiol. Sci. 2003, 2003, 1118-1124.
(25) Redmore, N. P.; Rubtsov, I. V.; Therien, M. J. J. Am. Chem. Soc. 2003,
125, 8769-8778.
(26) Hayes, R. T.; Walsh, C. J.; Wasielewski, M. R. J. Phys. Chem. A 2004,
108, 2375-2381.
(17) Furuta, P.; Brooks, J.; Thompson, M. E.; Fre´chet, J. M. J. Am. Chem. Soc.
2003, 123, 13165-13172.
(27) Strachan, J.-P.; Gentemann, S.; Seth, J.; Kalsbeck, W. A.; Lindsey, J. S.;
Holten, D.; Bocian, D. F. J. Am. Chem. Soc. 1997, 119, 11191-11201.
(28) Osuka, A.; Tanabe, N.; Kawabata, S.; Yamazaki, I.; Nishimura, Y. J. Org.
Chem. 1995, 60, 7177-7185.
(18) Aratani, N.; Osuka, A.; Cho, H. S.; Kim, D. J. J. Photochem. Photobiol.,
C 2002, 3, 25-52.
(19) Wu¨rthner, F.; Thalacker, C.; Diele, S.; Tschierske, C. Chem. Eur. J. 2001,
7, 2245-2253.
(29) Hopfield, J. J. Proc. Natl. Acad. Sci. U.S.A. 1974, 71, 3640-3644.
(30) Jortner, J. J. Chem. Phys. 1976, 64, 4860-4867.
(31) Marcus, R. J. Chem. Phys. 1984, 81, 4494-4500.
(32) Plato, M.; Moebius, K.; Michel-Beyerle, M. E.; Bixon, M.; Jortner, J. J.
Am. Chem. Soc. 1988, 110, 7279-7285.
(20) Wagner, R. W.; Lindsey, J. S.; Seth, J.; Palaniappan, V.; Bocian, D. F. J.
Am. Chem. Soc. 1996, 118, 3996-3997.
(21) Taniguchi, M.; Kim, H.-J.; Ra, D.; Schwartz, J. K.; Kirmaier, C.; Hindin,
E.; Diers, J. R.; Prathapan, S.; Bocian, D. F.; Holten, D.; Lindsey, J. S. J.
Org. Chem. 2002, 67, 7329-7342.
(33) Tsai, H.; Simpson, M. C. Chem. Phys. Lett. 2002, 353, 111-118.
(34) Gouterman, M. Porphyrins 1978, 3, 1-165.
(35) Weiss, C.; Kobayashi, H.; Gouterman, M. J. Mol. Spectrosc. 1965, 16,
415-450.
(22) Hindin, E.; Kirmaier, C.; Diers, J. R.; Tomizaki, K.-y.; Taniguchi, M.;
Lindsey, J. S.; Bocian, D. F.; Holten, D. J. Phys. Chem. B 2004, 108, 8190-
8200.
(36) Fajer, J.; Borg, D. C.; Forman, A.; Dolphin, D.; Felton, R. H. J. Am. Chem.
Soc. 1970, 92, 3451-3459.
(23) Kirmaier, C.; Hindin, E.; Schwartz, J. K.; Sazanovich, I. V.; Diers, J. R.;
Muthukumaran, K.; Taniguchi, M.; Bocian, D. F.; Lindsey, J. S.; Holten,
D. J. Phys. Chem. B 2003, 107, 3443-3454.
(37) Atamian, M.; Wagner, R. W.; Lindsey, J. S.; Bocian, D. F. Inorg. Chem.
1988, 27, 1510-1512.
9
4780 J. AM. CHEM. SOC. VOL. 128, NO. 14, 2006

Intramolecular ET in a Chlorophyll a Derivative
A R T I C L E S
Electronic structure calculations and electron-nuclear double-
resonance (ENDOR) spectroscopy have shown that the electron
density distributions within chlorins differ significantly from
those of metalloporphyrins. Given the observed dependence of
both energy- and electron-transfer rates on the detailed electron
density distributions in porphyrins, it is important to carefully
examine similar considerations for chlorins. Semiempirical
molecular orbital calculations on the corresponding chlorin
radical cations have predicted a modest 0.25 eV energy gap
between the HOMO and HOMO-1 orbitals,38 while ab initio
calculations on the ethyl chlorophyllide a radical cation have
estimated a 0.65 eV gap,39 which is considerably larger than
that observed in metalloporphyrins. However, more recently,
O’Malley et al.40 reported density functional theory (DFT)
calculations on chlorophyll cations with water-ligated metal
centers that give a gap of only ∼0.03 eV between these orbitals.
Direct measurements of the spin density distributions in the
radical cations of Chl a and several of its derivatives using
ENDOR spectroscopy show high spin densities at the â carbon
atoms of the pyrroles, whereas the spin densities at the meso
positions, in particular the 20-position, are relatively low.38,41-43
This electron density distribution is very similar to that of
metalloporphyrins having â alkyl substituents in which the a1u
orbital is the HOMO. Mo¨ssler et al.44 have reported ENDOR
spectra of a series zinc meso-tetraphenylchlorins in which one
phenyl was either converted into a benzoquinone electron
acceptor or replaced by a 1,4-cyclohexyl group bearing the
benzoquinone. In particular, they studied the system in which
the quinone or cyclohexylquinone occupies the 20-position of
the macrocycle. They found that the attachment of these groups
to the macrocycle did not significantly perturb the spin (and
charge) distribution in the chemically generated radical cation.
The only observed effects were small changes in the dihedral
angles of the â-protons on the reduced ring, which resulted in
changes in the hyperfine coupling constants (hfcc’s) of these
protons.
axis but also enhance the self-association of the dyads by adding
an additional π-π interaction. Arylene bis(dicarboximide)s are
easy to reduce, yielding stable radical anions with intense and
distinct electronic transitions that make them easy to identify
in transient optical experiments.45-50 PDIs have proven to be
excellent complementary chromophores for porphyrin-based
systems by filling the spectral gap between the Soret and Q
bands, thus serving as a light-harvesting component of the
system.50 The PDI chromophore should prove even more useful
in a chlorophyll-based system, since the intense Chl a Qy band
is shifted ∼100 nm to the red of the lowest electronic transition
in metalloporphyrins.
For comparison, an analogous group of three porphyrin-
containing donor-acceptor dyads was synthesized and charac-
terized. The zinc 5,10,15-tri(n-pentyl)-20-phenylporphyrin was
chosen to match both the single phenyl linkage to the acceptor
and the free energy for charge recombination relative to that of
the chlorophyll series. The photophysical studies presented here
are conducted on the monomeric, disaggregated state of the
dyads and will be compared to the aggregated systems in a
subsequent publication. We also characterize the spin density
distribution in the radical cation of this new arylated chlorin
using ENDOR spectroscopy in fluid solution to probe the
influence of the 20-phenyl group on the structure of the radical
cation and its electronic coupling to the reduced acceptor. To
explain the results of the photophysical and ENDOR studies, a
series of DFT calculations were performed on both chlorin and
porphyrin model systems.
Experimental Section
Synthesis. The syntheses of zinc methyl 3-ethyl-20-phenylpyro-
chlorophyllide a (4, ZCPh) and the donor-acceptor dyads having
arylenebis(dicarboximide)s attached to the corresponding 20-(p-
aminophenyl) derivative (9, ZCPh-NI; 10, ZCPh-PDIa; and 11, ZCPh-
PDIb), as well as the intermediates leading to them, are detailed in the
Supporting Information. The syntheses of zinc 5-phenyl-10,15,20-tri-
(n-pentyl)porphyrin (13, Z3PnPh) and its corresponding 20-aminophenyl
derivatives (19, Z3PnPh-NI; 20, Z3PnPh-PDIa; and 21, Z3PnPh-PDIb)
used for comparisons to the chlorophyll derivatives are also detailed
in the Supporting Information. Each chlorin was synthesized from
chlorophyll a, which was extracted from the alga Spirulina maxima
and subsequently converted to methyl 3-ethylpyropheophorbide a using
literature procedures.51,52 The porphyrins were produced from pyrrole
and the appropriate aldehydes using the Lindsey method.53 Character-
ization was performed with Varian 400 MHz NMR and PE Voyager
DE-Pro MALDI-TOF mass spectrometers. High-resolution electrospray
and fast atom bombardment mass spectra were obtained with the 70-
SE-4F and Q-Tof Ultima mass spectrometers at the University of Illinois
at Champaign-Urbana.
In this study we report the synthesis and photophysics of three
donor-acceptor dyads in which arylene bis(dicarboximide)
acceptors are attached to the 20-position of chlorophyll donors
via phenyl groups (Figure 1). The benefits of using arylene bis-
(dicarboximide)s as electron acceptors in self-assembling sys-
tems have been discussed recently.45-48 The attachment of an
arylene bis(dicarboximide) acceptor directly to the 20-position
of the chlorin macrocycle would not only preserve the ability
of the chlorin to direct aggregation along its ring A-ring C
(38) Davis, M. S.; Forman, A.; Fajer, J. Proc. Natl. Acad. Sci. U.S.A. 1979, 76,
4170-4174.
Electrochemistry. Electrochemical measurements were performed
using a CH Instruments model 622 electrochemical workstation.
Reversible, half-wave potentials were determined in butyronitrile
(PrCN) containing 0.1 M n-Bu4N+BF4- electrolyte. A 1.0-mm-diameter
platinum disk electrode, platinum wire counter electrode, and Ag/AgxO
reference electrode were employed. Ferrocene/ferrocenium (Fc/Fc+,
0.52 V vs SCE) was used as an internal reference for all measurements.
(39) Petke, J. D.; Maggiora, G. M.; Shipman, L. L.; Christoffersen, R. E.
Photochem. Photobiol. 1980, 31, 243-257.
(40) O’Malley, P. J.; Collins, S. J. J. Am. Chem. Soc. 2001, 123, 11042-11046.
(41) Norris, J. R.; Scheer, H.; Druyan, M. E.; Katz, J. J. Proc. Natl. Acad. Sci.
U.S.A. 1974, 71, 4897-4900.
(42) Scheer, H.; Katz, J. J.; Norris, J. R. J. Am. Chem. Soc. 1977, 99, 1372-
1381.
(43) Ka¨ss, H.; Lubitz, W.; Hartwig, G.; Scheer, H.; Noy, D.; Scherz, A.
Spectrochim. Acta, Part A 1998, 54A, 1141-1156.
(44) Mo¨ssler, H.; Wittenberg, M.; Niethammer, D.; Mudrassagam, R. K.;
Kurreck, H.; Huber, M. Magn. Reson. Chem. 2000, 38, 67-84.
(45) Rybtchinski, B.; Sinks, L. E.; Wasielewski, M. R. J. Phys. Chem. A 2004,
108, 7497-7505.
(49) Gosztola, D.; Niemczyk, M. P.; Svec, W.; Lukas, A. S.; Wasielewski, M.
R. J. Phys. Chem. A 2000, 104, 6546-6551.
(50) Prathapan, S.; Yang, S. I.; Seth, J.; Miller, M. A.; Bocian, D. F.; Holten,
D.; Lindsey, J. S. J. Phys. Chem. B 2001, 105, 8237-8248.
(51) Smith, K. M.; Goff, D. A.; Simpson, D. J. J. Am. Chem. Soc. 1985, 107,
4946-4954.
(46) Rybtchinski, B.; Sinks, L. E.; Wasielewski, M. R. J. Am. Chem. Soc. 2004,
126, 12268-12269.
(47) Ahrens, M. J.; Sinks, L. E.; Rybtchinski, B.; Liu, W.; Jones, B. A.; Giaimo,
J. M.; Gusev, A. V.; Goshe, A. J.; Tiede, D. M.; Wasielewski, M. R. J.
Am. Chem. Soc. 2004, 126, 8284-8294.
(52) Kenner, G. W.; McCombie, S. W.; Smith, K. M. J. Chem. Soc., Perkin
Trans. 1 1973, 2517-2523.
(48) van der Boom, T.; Hayes, R. T.; Zhao, Y.; Bushard, P. J.; Weiss, E. A.;
Wasielewski, M. R. J. Am. Chem. Soc. 2002, 124, 9582-9590.
(53) Lindsey, J. S.; Schreiman, I. C.; Hsu, H. C.; Kearney, P. C.; Marguerettaz,
A. M. J. Org. Chem. 1987, 52, 827-836.
9
J. AM. CHEM. SOC. VOL. 128, NO. 14, 2006 4781

A R T I C L E S
Kelley et al.
Figure 1. Structures of molecules studied here.
Optical Spectroscopy. UV-vis absorption measurements were made
on a Shimadzu spectrometer (UV1601). Steady-state fluorescence
measurements were made on a single-photon-counting fluorimeter
(PTI). All measurements were performed at room temperature. Optical
densities of the solutions were kept below 0.1 at 655 nm in a 1-cm
cuvette, and excitation/emission geometry was at right angles. Toluene
was purified by passing it through a series of CuO and alumina columns,
while tetrahydrofuran (THF) was purified by passing it through two
alumina columns (GlassContour) immediately prior to use.
Femtosecond transient absorption measurements were made with a
Ti:sapphire laser system as detailed in the Supporting Information. The
instrument response function (IRF) for the pump-probe experiments
was 150 fs. Typically 5 s of averaging was used to obtain the transient
spectrum at a given delay time. Cuvettes with a 2 mm path length
were used, and the samples were irradiated with 0.5 µJ/pulse focused
to a 200-µm spot. The optical density at λex was kept between 0.2 and
0.4 to prevent aggregation. Analysis of the kinetic data was performed
at multiple wavelengths using a Levenberg-Marquardt nonlinear least-
squares fit to a general sum-of-exponentials function with an added
Gaussian to account for the finite instrument response.
Fluorescence lifetimes were determined using a frequency-doubled,
cavity-dumped Ti:sapphire laser as the excitation source and a
Hamamatsu C4780 picosecond fluorescence lifetime measurement
system as described in the Supporting Information. The energy of the
400 nm, 25 fs pulses was attenuated to approximately 1.0 nJ/pulse for
all fluorescence lifetime experiments. The total IRF of the streak camera
system was 25 ps. The samples were prepared in glass cuvettes, and
the optical density at the excitation wavelength was typically 0.020-
0.035. At λmax of the samples, the optical density was typically 0.080-
0.105. All fluorescence data were acquired in single-photon-counting
mode using the Hamamatsu HPD-TA software. The data were fit using
the Hamamatsu fitting module and deconvoluted using the laser pulse
profile. Measurements were generally repeated three times and the
resulting time constants were averaged. Deviation from the mean was
generally no worse than (5%.
EPR and ENDOR Spectroscopy. All samples and solvents were
handled in an inert atmosphere glovebox (MBraun Unilab). Stock
solutions of ZC, ZCPh, and the I2 oxidant were prepared in dry CH2-
Cl2/THF (9:1 v/v). Solutions of the ZC and ZCPh radical cations were
formed by oxidation with a 10- or 20-fold excess (mol/mol) of I2. The
9
4782 J. AM. CHEM. SOC. VOL. 128, NO. 14, 2006

Intramolecular ET in a Chlorophyll a Derivative
A R T I C L E S
Figure 2. Synthetic scheme.
resulting 0.5 mM samples of oxidized chlorins were then loaded into
1.4- and 2.4-mm-i.d. quartz tubes and frozen in liquid nitrogen pending
analysis. EPR and ENDOR spectra were acquired with a Bruker
ELEXSYS E580 spectrometer, fitted with the DICE ENDOR accessory,
EN801 resonator, and radio frequency (RF) power amplifier (ENI
A-500). Applied RF powers ranged from 180 to 700 W across the 20
MHz scanned range, and microwave power was typically 31 mW. The
sample temperature was controlled by a liquid nitrogen flow system,
and spectra were obtained at 170, 195, and 220 K.
Computational Methods. Geometry optimizations of compounds
4 and 13 in both their neutral and cationic states were performed using
Gaussian 0354 using the Perdew-Wang 1991 nonlocal functional55 with
the 6-31G* basis set. No symmetry constraints were enforced, and all
open-shell calculations were performed using unrestricted methods. The
spin contamination in the radical species was found to deviate from
the expected value by less than 5% (calculated S2 e 0.78; expected,
0.75). Single-point calculations used to determine internal reorganization
energies and the effects of bis(dicarboximide) radical anions on the
electronic structures of ZCPh+• and Z3PnPh+• were performed using
the above geometry-optimized structures. The locations of the partial
negative point charges relative to the cationic macrocycles were
determined from geometry-optimized neutral structures of models
containing the bis(dicarboximide) acceptors (not shown).
formation of the 20-phenyl derivative 3 in 40% yield, while
the analogous reaction of 2 with the pinacol ester of p-
aminophenylboronic acid gave the 20-(p-aminophenyl) deriva-
tive 5 in 76% yield. The arylene monoimide monoanhydride
precursors to the NI, PDIa, and PDIb electron acceptors were
then condensed with the amine in 5 to yield the corresponding
arylene bis(dicarboximide) dyads 6-8.11,47,57 Finally, the free-
base pheophorbides 3 and 6-8 were metalated using zinc acetate
in CH2Cl2/CH3OH to give the corresponding zinc chlorophyl-
lides, 4 and 9-11. For spectroscopic comparisons to the
chlorophyll derivatives, an analogous set of porphyrins was
prepared. The syntheses of the porphyrin derivatives are outlined
in Figure S1 in the Supporting Information. 5-Phenyl-10,15,-
20-tri(n-pentyl)porphyrin (12) and the corresponding 20-(p-
nitrophenyl) derivative 14 were prepared by a mixed conden-
sation of n-hexanal with either benzaldehyde or p-nitrobenz-
aldehyde, respectively, using Lindsey’s method.53 Reduction of
the nitro group in 14 gave the 20-(p-aminophenyl) derivative
15 in 73% yield. Condensation of 15 with the arylene mono-
imide monoanhydride precursors to the NI, PDIa, and PDIb
electron acceptors gave free-base porphyrin acceptors 16-18.
The free-base porphyrins 12 and 16-18 were metalated using
zinc acetate in CH2Cl2/CH3OH to give the corresponding zinc
porphyrins, 13 and 19-21.
Results
Synthesis. The syntheses of 4 and 9-11 are outlined in Figure
2. Chlorophyll a was extracted from the alga Spirulina maxima
and subsequently converted to methyl 3-ethylpyropheophorbide
a (1) according to known procedures.51 Selective bromination
of the 20-position of 1 results in the formation of the bromo
compound 2.52 Suzuki cross-coupling of 2 with phenylboronic
acid using the mild conditions outlined by Shi and Boyle56 for
use with free-base porphyrins resulted in the
Steady-State Spectroscopy. The UV-vis spectra of the
ZCPh series of donor-acceptor dyads are compared to the
spectrum of ZCPh in toluene in Figure 3, while the correspond-
ing spectra of the Z3PnPh series of dyads are compared to the
spectrum of Z3PnPh in Figure 4. The corresponding spectra
obtained in THF (Figures S2 and S3) are nearly identical. In
the ZCPh spectra, the chlorin Soret bands occur at 428 nm and
(54) Frisch, M. J.; et al. Gaussian; Gaussian, Inc.: Wallingford, CT, 2004.
(55) Perdew, J. P.; Chevary, J. A.; Vosko, S. H.; Jackson, K. A.; Pederson, M.
R.; Singh, D. J.; Fiolhais, C. Phys. ReV. B: Condens. Matter Mater. Phys.
1992, 46, 6671-6687.
(56) Shi, B.; Boyle, R. W. J. Chem. Soc., Perkin Trans. 1 2002, 11, 1397-
1400.
(57) Holman, M. W.; Liu, R.; Adams, D. M. J. Am. Chem. Soc. 2003, 125,
12649-12654.
9
J. AM. CHEM. SOC. VOL. 128, NO. 14, 2006 4783

A R T I C L E S
Kelley et al.
(Figures 3 and 4 insets and Figure S4), only weak ground-state
electronic interactions occur between all redox pairs.58
The fluorescence maxima and quantum yields for both the
ZCPh and Z3PnPh series of donor-acceptor dyads as well as
those of reference compounds ZCPh, zinc methyl 3-ethylpyro-
chlorophyllide a (Zn EtPChlide a), and Z3PnPh in toluene and
THF are presented in Table 1. The values obtained in THF are
nearly identical to those given for toluene. For all molecules,
the emission features are approximately mirror images of the
corresponding Qy(0,0) and Qy(0,1) absorption features. The
fluorescence quantum yield of ZCPh is only slightly lower than
that of Zn EtPChlide a.59 The fluorescence of all molecules with
the arylene bis(dicarboximide) electron acceptors attached is
significantly quenched, suggesting that electron transfer occurs
in these molecules (see below).
Energetics. The one-electron oxidation and reduction poten-
tials for both the ZCPh and Z3PnPh series of donor-acceptor
dyads as well as those of several reference compounds are
summarized in Table 2. Neither the oxidation potentials of the
donors nor the reduction potentials of the acceptors used to
construct the dyads are significantly affected by covalent
linkages between the two redox-active centers. The redox
potentials of the acceptors change only slightly when they are
linked to the chlorin or porphyrin donor, in accord with small
substituent effects observed earlier.60 The electrochemical data
reinforces the conclusion suggested by the UV-vis spectra that
the ground-state electronic interactions between the donor and
acceptors are weak.
Figure 3. Ground-state electronic absorption spectra of the indicated
compounds in toluene. Inset: The spectrum of ZCPh-PDIa together with
the spectra of its component chromophores and the simple sum of these
spectra.
The energies of the radical pair states for the donor-acceptor
dyads were calculated using the experimentally determined
redox potentials along with computed ionic radii and donor-
acceptor distances. The ion-pair distances and ionic radii were
estimated using MM+ geometry-optimized models (Table S1,
Supporting Information).61 Equation 1 yields estimates of the
energies of the solvated ion pairs based on the Born dielectric
continuum model of solvent,62
e2
DAꢀS
1
1
1
1
Figure 4. Ground-state electronic absorption spectra of the indicated
compounds in toluene. Inset: The spectrum of Z3PnPh-PDIa together with
the spectra of its component chromophores and the simple sum of these
spectra.
∆GIP ) EOX - ERED
-
+ e2
+
-
(
)(
)
r
2rD 2rA ꢀS ꢀSP
(1)
where EOX and ERED are the redox potentials of the donor and
acceptor, respectively, determined in a polar solvent with
dielectric constant ꢀSP, ꢀS is the dielectric constant of the solvent
in which the electron transfer takes place, rD and rA are the
ionic radii, and rDA is the donor-acceptor distance. The free
energies for charge separation and charge recombination, ∆GCS
and ∆GCR, respectively (see Table 3), were then determined
using eqs 2 and 3.
the Qx(0,1), Qx(0,0), Qy(0,1), and Qy(0,0) bands at 521, 564,
609, and 655 nm, respectively. In the porphyrin spectra, the
porphyrin Soret bands appear at 424 nm and the Qy(0,1) and
Qy(0,0) bands at 556 and 595 nm, respectively. The ground-
state absorption spectra of both ZCPh and Z3PnPh are es-
sentially unperturbed by the attachment of either of the three
electron acceptors at the para position of the 20-phenyl group.
The spectra of ZCPh-NI and Z3PnPh-NI contain additional
peaks at 361 and 381 nm due to the NI acceptors, while in
ZCPh-PDIa, Z3PnPh-PDIa, ZCPh-PDIb, and Z3PnPh-PDIb, the
visible spectra between 450 and 600 nm exhibit absorption
features of the PDI derivatives. Local maxima for PDIa appear
at 459, 491, and 528 nm and for PDIb at 479, 510, and 546
nm.47,49 Deviations in the absorption features exhibited by both
donors in the 350-400 nm region are due to lower intensity
transitions in the PDIa and PDIb chromophores.47,49 Since the
spectra of the covalently linked systems can be closely ap-
proximated using a weighted sum of the component absorptions
∆GCS ) ∆GIP - ES1
∆GCR ) -∆GIP
(2)
(3)
The lowest excited-state energies (ES1) for the dyads were
(58) Greenfield, S. R.; Svec, W.; Gosztola, D.; Wasielewski, M. R. J. Am. Chem.
Soc. 1996, 118, 6767-6777.
(59) Wasielewski, M. R.; Johnson, D. G.; Niemczyk, M. P.; Gaines, G. L., III;
O’Neil, M. P.; Svec, W. J. Am. Chem. Soc. 1990, 112, 6482-6488.
(60) Viehbeck, A.; Goldberg, M. J.; Kovac, C. A. J. Electrochem. Soc. 1990,
137, 1460-1466.
(61) Hyperchem, MM+, and PM3 calculations were performed using Hyperchem
software from Hypercube, Inc., 115 NW 1114th St., Gainesville, FL 32601.
(62) Weller, A. Z. Phys. Chem. 1982, 133, 93-98.
9
4784 J. AM. CHEM. SOC. VOL. 128, NO. 14, 2006

Intramolecular ET in a Chlorophyll a Derivative
A R T I C L E S
Table 1. UV-Vis Absorption and Emission
toluene
THF
a
b
c
d
a
b
c
d
compd
λabs
λem
ES
ΦF
λabs
λem
ES
ΦF
Zn EtPChlide a
ZCPh
ZCPh-NI
ZCPh-PDIa
ZCPh-PDIb
657
656
655
655
655
662
658
658
658
658
1.88
1.89
1.89
1.89
1.89
0.15
0.13
<0.0005
<0.0005
<0.0005
657
653
654
654
654
662
657
657
657
657
1.88
1.89
1.89
1.89
1.89
0.15
0.15
<0.0005
<0.0005
<0.0005
Z3PnPh
595
595
595
595
605
605
605
605
2.07
2.07
2.07
2.07
0.03
603
603
603
603
610
610
610
610
2.05
2.05
2.05
2.05
0.03
Z3PnPh-NI
Z3PnPh-PDIa
Z3PnPh-PDIb
<0.0005
<0.0005
<0.0005
<0.0005
<0.0005
<0.0005
a Ground-state absorption maxima corresponding to the Qy(0,0) band. b Steady-state fluorescence maxima following 638 (ZCPh) and 550 nm (Z3PnPh)
excitation. c First singlet excited-state energies calculated using the average energies of the absorption and emission maxima. d Fluorescence quantum yields
of ZCPh derivatives use Zn EtPChlide a as a reference, while those of Z3PnPh derivatives use ZnTPP as reference.
Table 2. Redox Potentials (V vs SCE)a
comps
EOX
ERED
ZCPh-NI
ZCPh-PDIa
ZCPh-PDIb
0.66
0.66
0.66
-0.51
-0.45
-0.55
Z3PnPh-NI
Z3PnPh-PDIa
Z3PnPh-PDIb
0.65
0.65
0.65
-0.51
-0.45
-0.55
a All values obtained in PrCN/0.1 M TBABF4.
derived from the average energy of the Qy(0,0) absorption and
fluorescence maxima of the respective porphyrin or chlorophyll
donor (Table 1).58 Equations 1 and 2 predict that photoinduced
charge separation should occur in all six dyads in both toluene
and THF.
Transient Absorption Spectroscopy. Excitation of ZCPh
in toluene with 655 nm, 120 fs pulses yields the transient
absorption spectrum of *ZCPh within the 150 fs instrument
Figure 5. Transient absorption spectra of ZCPh in toluene following
excitation with 655 nm, 120 fs laser pulses. Inset: Transient absorption
kinetics for ZCPh (blue) at 631 nm and (red) at 659 nm. Nonlinear least-
squares fits to the data are also shown.
1
response function of the apparatus (Figure 5). The bleach of
1
the Qy(0,0) band occurs at 655 nm, while *ZCPh has a broad
absorption between 440 and 640 nm with local maxima
occurring at 455, 587, and 629 nm. Stimulated emission from
1*ZCPh can be detected as a negative feature centered around
723 nm. Transient absorption kinetics monitored at 631 and 659
nm (inset to Figure 5) show complex decays over the 5 ns time
using 414 nm, 120 fs pulses, followed by rapid, instrument-
limited internal conversion, produces transient absorption spectra
identical to those attained using 655 nm excitation (not shown).
Attachment of arylene bis(dicarboximide)s to ZCPh provides
an efficient, nonradiative deactivation pathway for the 1*ZCPh
excited state. Thus, the transient spectra in Figure 6 confirm
the initial formation of 1*ZCPh (features at 440-640 nm), while
the growth of the sharp NI-• absorption band at 475 nm49
indicates that the radical ion pair ZCPh+•-NI-• is formed.
Kinetic traces obtained at 478 and 657 nm (inset to Figure 6)
reveal the time constants for charge separation (τCS) leading to
1
window as *ZCPh decays competitively to ground state and
the long-lived 3*ZCPh. The corresponding spectra and kinetics
for ZCPh in THF are given in Figure S5. The lifetimes of
1*ZCPh in toluene (τ ) 3.2 ns) and THF (τ ) 4.0 ns) were
determined independently from its fluorescence decay (not
shown). Excitation of ZCPh into its second excited state (S2)
Table 3. Charge Separation and Charge Recombination Data
1
1
compd
τCS (ps)
∆G
CS (eV)
V
CS (cm-
)
τCR (ps)
∆G
CR (eV)
V
CR (cm-
)
Toluene
31
17
22
47
ZCPh-NI
21 ( 2
12 ( 1
25 ( 1
16 ( 1
14 ( 2
19 ( 1
-0.31
-0.43
-0.33
-0.50
-0.62
-0.52
1150 ( 50
2000 ( 200
3300 ( 200
1100 ( 100
1850 ( 50
3050 ( 50
-1.58
-1.46
-1.56
-1.57
-1.45
-1.55
88
36
25
116
65
32
ZCPh-PDIa
ZCPh-PDIb
Z3PnPh-NI
Z3PnPh-PDIa
Z3PnPh-PDIb
22
31
THF
23
21
20
42
ZCPh-NI
13 ( 1
14 ( 1
19 ( 1
5 ( 1
-0.66
-0.73
-0.63
-0.82
-0.90
-0.80
25 ( 2
31 ( 2
43 ( 2
18 ( 1
24 ( 2
37 ( 2
-1.23
-1.16
-1.26
-1.25
-1.17
-1.27
54
37
46
99
60
76
ZCPh-PDIa
ZCPh-PDIb
Z3PnPh-NI
Z3PnPh-PDIa
Z3PnPh-PDIb
7 ( 2
42
31
9 ( 1
9
J. AM. CHEM. SOC. VOL. 128, NO. 14, 2006 4785

A R T I C L E S
Kelley et al.
Figure 6. Transient absorption spectra of ZCPh-NI in toluene following
excitation with 655 nm, 120 fs laser pulses. Inset: Transient absorption
kinetics for ZCPh-NI (blue) at 478 nm and (red) at 657 nm. Nonlinear
least-squares fits to the data are also shown.
Figure 8. Transient absorption spectra of ZCPh-PDIb in toluene following
excitation with 655 nm, 120 fs laser pulses. Inset: Transient absorption
kinetics for ZCPh-PDIb (blue) at 657 nm and (red) at 727 nm. Nonlinear
least-squares fits to the data are also shown.
ground-state bleaches occur at 510 and 546 nm, and the PDI-•
absorption feature occurs at 725 nm. Kinetic data monitored at
657 and 727 nm (inset to Figure 8 and Table 3) give the rates
of charge separation and recombination. An analogous set of
transient absorption data obtained for ZCPh-NI, ZCPh-PDIa,
and ZCPh-PDIb in THF (Figures S6-S8) shows spectra similar
to those obtained in toluene. Charge separation and recombina-
tion time constants in THF are presented in Table 3. As
expected, in the more polar solvent the time constants for both
charge separation and recombination are faster than those
observed in toluene.
Excitation of Z3PnPh with 414 nm, 120 fs pulses yields
1*Z3PnPh, following rapid, instrument-limited relaxation from
the second singlet excited state (S2). The major features include
broad absorptions at 450-550 and 575-720 nm as well as
ground-state bleaching of the Qy(0,1) and Qy(0,0) bands at 564
and 595 nm, respectively (Figures S9 (toluene) and S10 (THF)).
Once again, due to the complexity of the transient absorption
Figure 7. Transient absorption spectra of ZCPh-PDIa in toluene following
excitation with 655 nm, 120 fs laser pulses. Inset: Transient absorption
kinetics for ZCPh-PDIa (blue) at 657 nm and (red) at 704 nm. Nonlinear
least-squares fits to the data are also shown.
1
decay of *Z3PnPh over the limited 5 ns time window, the
1
lifetimes of *Z3PnPh in toluene (τ ) 2.3 ns) and THF (τ )
ZCPh+•-NI-• and its subsequent recombination (τCR) (Table
3). The ZCPh-PDIa and ZCPh-PDIb dyads have donors (ZC)
and acceptors (PDI) that both absorb in the visible, resulting in
complex transient absorption spectra. Selective excitation of the
chlorin within ZCPh-PDIa in toluene using 655 nm, 120 fs
pulses results in the transient absorption spectra shown in Figure
1.9 ns) were determined from its fluorescence decay (not
shown). 1*Z3PnPh can also be directly accessed using 595 nm
pulses as evidenced by acquisition of transient absorption spectra
identical to those attained using 414 nm excitation (not shown).
Attachment of arylene bis(dicarboximide)s to Z3PnPh also
provides efficient, nonradiative deactivation pathways for the
1*Z3PnPh excited state. Selective excitation of the metallo-
porphyrin within Z3PnPh-NI, Z3PnPh-PDIa, and Z3PnPh-PDIb
in toluene using 414 nm, 120 fs pulses results in the transient
absorption spectra shown in Figures S11-S13. The only
significant difference between these transient absorption spectra
and those of the ZCPh dyads is the position and intensity of
the ground-state bleaches due to Z3PnPh (see above). The
kinetic data obtained in toluene at wavelengths corresponding
to the radical anion absorbance maxima and ground-state bleach
minima for the three donor-acceptor dyads are shown in the
insets of Figures S11-S13, and the values are listed in Table
3. An analogous set of transient absorption data obtained for
Z3PnPh-NI, Z3PnPh-PDIa, and Z3PnPh-PDIb in THF shows
spectra similar to those obtained in toluene (Figures S14-S16).
Charge separation and recombination time constants in THF
1
7. The initial spectrum due to *ZCPh evolves in time to that
of ZCPh+•-PDIa-•, which is comprised of ground-state bleaches
of PDI at 491 and 527 nm and ZC at 657 nm as well as the
appearance of an absorption at 701 nm due to the formation of
PDIa-•.49 The kinetic data obtained in toluene near these
wavelengths are shown in the inset to Figure 7 and are given
in Table 3. The electron-accepting PDIb chromophore is
functionalized with 3,5-di-tert-butylphenoxy groups, which
enhance the solubility of the PDI core and result in a one-
electron reduction potential that is 0.1 V more negative than
that of PDIa. In addition, the electronic absorption spectra of
both the ground state and radical anion of PDIb are red-shifted
and broadened relative to those of PDIa.47 The radical pair state,
ZCPh+•-PDIb-•, is obtained following selective excitation of
ZC using 655 nm, 120 fs laser pulses (Figure 8). The PDIb
9
4786 J. AM. CHEM. SOC. VOL. 128, NO. 14, 2006

Intramolecular ET in a Chlorophyll a Derivative
A R T I C L E S
are presented in Table 3. Once again, in the more polar solvent
the time constants for both charge separation and recombination
are faster than those observed in toluene.
The PDI electron acceptors serve as excellent antenna
chromophores for both ZC and Z3PnP by absorbing light in
the spectral window between the Soret and Q bands of these
macrocycles and efficiently funneling excitation energy to them.
Ultrafast energy transfer from the PDI derivatives in ZCPh-
PDIa, ZCPh-PDIb, Z3PnPh-PDIa, and Z3PnPh-PDIb to the
electron donors, ZC and Z3PnPh, is demonstrated by selective
excitation of the PDI chromophores using 510 nm, 120 fs laser
pulses. In ZCPh-PDIa and Z3PnPh-PDIa, excitation of PDIa in
toluene results in the instrument-limited appearance of 1*PDIa,
as evidenced by concurrent bleaching of the 491 and 527 nm
ground-state absorption bands and the growth of the excited-
state feature centered at 690 nm. Decays of the transient spectral
features of PDIa, monitored at 526 and 706 nm, respectively,
reveal that energy transfer occurs with τEnT ) 400 fs in ZCPh-
PDIa (Figure S17) and τEnT ) 550 fs in Z3PnPh-PDIa (Figure
S18). Energy transfer in ZCPh-PDIb and Z3PnPh-PDIb occurs
with similar time constants, τEnT ) 350 (Figure S19) and 500
fs (Figure S20), respectively, as determined by monitoring their
transient absorption features at 507 and 729 nm. The lifetime
of the lowest excited singlet state of both isolated PDI
chromophores is τ ) 4.5 ns with a fluorescence quantum yield
of unity.48 Given that the lifetime of 1*PDI is approximately 3
orders of magnitude longer than the experimentally determined
energy-transfer rates, energy transfer from PDIa or PDIb to ZC
and Z3PnP occurs with a quantum yield of unity. Following
energy transfer, ZCPh-PDIa, ZCPh-PDIb, Z3PnPh-PDIa, and
Z3PnPh-PDIb each undergo electron transfer with the same rates
obtained by selective photoexcitation of either ZCPh or Z3PnPh
with 150 fs laser pulses tuned to their respective absorbance
maxima.
Figure 9. 1H ENDOR spectra of ZC+• (black) and ZCPh+• (red), 0.5 mM
in CH2Cl2/THF (9:1 v:v) at 195 K. Experimental conditions: frequency
modulation, 25 kHz; modulation depth, 100 kHz; time constant, 5.12 ms;
scan time, 244
180-700 W.
×
21 s; microwave power, 31 mW; RF power,
Table 4. Proton hfcc’s for Chlorin Cation Radicals in DCM/THF
(9:1)a
proton assignment
compdb,c
var.
â,
γ
20
2a, 7a, 8a, 5, 10
12a
18
17
ZC+•
0.6
0.5
1.02
1.33
3.38
3.34
3.29, 3.87
2.98
7.11 10.61 11.62
ZCPh+•
7.17
7.3
6.46
10.8
9.93
7.84
10.8
9.93
Zn-Chl a+•
pyro-Chl a+•
1.02
0.80
6.83
a Values in MHz. b Data for ZC+• and ZCPh+• obtained at 195 K. c Data
for Zn-Chl a+• and pyro-Chl a+• taken from Table 2 of ref 38, obtained at
T ) 140 K.
220 K to 1.26 MHz at 170 K, and it is possible that the hfcc
measured for ZC+• would be in closer agreement at the 140 K
temperature of the previous studies. A second difference between
the spectra of ZC+• and ZCPh+• is the reduction of the â-proton
hfcc’s of carbons 17 and 18, due to substitution of the phenyl
group at the 20-position. The nature of this change is discussed
below.
ENDOR Spectroscopy. The CW-EPR spectra of ZC+• and
ZCPh+• are inhomogeneously broadened due to the large number
of hfcc’s, resulting in a single unresolved line for each spectrum
(not shown). Isotropic hfcc’s were obtained from ENDOR
spectroscopy in liquid CH2Cl2/THF solution using the ENDOR
resonance condition ν(ENDOR ) |νn ( a/2|, where ν(
are
ENDOR
the ENDOR transition frequencies, νn is the proton Larmor
frequency, and a is the isotropic hfcc.63 The ENDOR spectra
of ZC+• and ZCPh+• obtained at 195 K are presented in Figure
9. The ZC+• spectrum shows six well-defined peak pairs, and
the ZCPh+• spectrum shows five. The assignments listed in
Table 4 are based on previous studies of a series of substituted
chlorophyll radical cations, including pyro-Chl a+•, whose
substituents most closely resemble those of the present work.43
One significant difference between the spectra of ZC+• and
ZCPh+• is the loss of the peak hfcc at 1.33 MHz, which strongly
suggests that this resonance is due to the proton attached to the
20-position of ZC. This hfcc is somewhat higher than the value
of ∼0.80 MHz previously measured for pyro-Chl a+•;43
however, part of this difference is attributed to the replacement
of Zn for Mg, which is known to cause ∼0.20 MHz increases
for the R-protons at carbons 10 and 20 for Zn-Chl a+• and Zn-
BChl a+•, relative to their Mg counterparts. Additionally, the
frequency of this ZC+• resonance decreases from 1.48 MHz at
Discussion
The data presented in Table 3 show that the rates of electron
transfer though a phenyl group attached to the 20-position of a
chlorophyll are very rapid, even at relatively modest values of
∆GCS, and result in quantitative charge separation. A comparison
of the electron-transfer rates for the ZCPh dyads with those of
the corresponding Z3PnPh dyads shows that the rates of both
charge separation and recombination differ by less than a factor
of 2. Electron-transfer rate constants depend on several param-
eters involving both the donor-acceptor molecule and its
interaction with the surrounding medium. Semiclassical electron-
transfer theory shows that the electron-transfer rate constant, k,
depends on the electronic coupling matrix element V for the
reaction and the free energy of reaction ∆G:29-31
k )
-(∆G + λS + npω)2
4λSkBT
Sn
exp(-S) exp
n!
∞
2π
1
2
|V|
∑
n)0
B
[
]
p
4πλ k T
x
S
(63) Kurreck, H.; Kirste, B.; Lubitz, W. Electron Nuclear Double Resonance
Spectroscopy of Radicals in Solution; VCH: Weinheim, 1988.
(4)
9
J. AM. CHEM. SOC. VOL. 128, NO. 14, 2006 4787

A R T I C L E S
Kelley et al.
where hω is the vibrational quantum, assumed to be ∼1500
cm-1, S ) λI/pω, λI is the internal reorganization energy of the
donor-acceptor molecule, λS is the solvent reorganization
energy, and the summation n is done over the quantum number
of the high-frequency vibrational mode. For a given acceptor,
∆GCS for the Z3PnPh dyad is ∼0.2 eV more negative than that
for the corresponding ZCPh dyad (Table 3); therefore, eq 4
predicts that the charge separation rate constant, kCS, should be
somewhat larger for the porphyrin dyad than for the chlorin
dyad, provided that λS, λI, and VCS for charge separation are
comparable.
0.10 for Z3PnPh+•/Z3PnPh (see Supporting Information). The
similarity of the internal reorganization energies is reasonable,
given that the positive charge within each radical cation is
delocalized over a relatively large π system. These calculations
show that λI does not change significantly between the chlorin
and the porphyrin.
Given that λS and λI are essentially the same for both charge
separation and recombination for the ZCPh and Z3PnPh dyads
with a given acceptor in a particular solvent, the differences
observed for the charge separation and recombination rates must
be due to differences in ∆G and/or V. Unfortunately, there is
no independent way to estimate the differences in V between
the chlorin and porphyrin series. Nevertheless, using the
measured values of k and ∆G along with the values of λS and
λI given above, eq 4 can be used to estimate V. The data
An assessment of the solvent reorganization energy can be
made using the Marcus expression for λS based on treating the
solvent as a dielectric continuum:64
1
1
1
1
1
λS ) e2
+
-
-
(5)
presented in Table 3 show that all the values of VCS and VCR
,
(
)(
)
2rD 2rA rDA ꢀ0 ꢀS
comparing the ZCPh and Z3PnPh dyads with a given acceptor
in a particular solvent, are within a factor of 2 of each other,
suggesting that electronic coupling for photoinduced charge
separation and recombination through the 20-position of the
chlorin is similar to that of the porphyrin.
where ꢀS and ꢀ0 are the static and high-frequency dielectric
constants of solvent, respectively, rD and rA are the ionic radii,
and rDA is the donor-acceptor distance. Due to the structural
similarities between pairs of ZCPh and Z3PnPh dyads having
a given acceptor, the ionic radii and the donor-acceptor
distances are essentially identical (see Supporting Information).
For toluene, ꢀS ) 2.38 and ꢀ0 ) 2.25, so that eq 5 yields λS )
0.03 eV, while for THF, ꢀS ) 7.58 and ꢀ0 ) 1.98, so that λS )
0.43 eV.
The Z3PnPh porphyrin was chosen specifically as a reference
molecule because its one-electron oxidation potential differs by
only 0.01 V from that of ZCPh, so that the ∆GCR values for the
charge recombination reactions for ZCPh and Z3PnPh dyads
having a given acceptor in a particular solvent are very similar.
This similarity assumes that the entire radical-ion-pair population
recombines directly to the singlet ground state and is not
complicated by radical pair intersystem crossing,65-68 in which
a substantial fraction of the population could recombine to give
Evaluating λI for photoinduced charge separation reactions
is difficult because there is no independent way to assess the
degree of electronic interaction between 1*ZCPh and 1*Z3PnPh
with a given acceptor. For example, accurate excited-state
electronic structure calculations on molecules of this size are
not practical. However, photoexcitation of molecules with large
π systems frequently results in only small structural changes
on going from the ground state to the relaxed excited state. The
total reorganization energy, λ, where λ ) λS + λI, for formation
3
3
3
3*Z3PnPh or *ZC. Since the energies of *Z3PnPh and *ZC
are ∼1.7 and ∼1.3 eV, respectively, charge recombination to
these triplet states would have very different values of ∆GCR
.
This fact, combined with the presence of two parallel charge
recombinations pathways, would make further comparisons of
VCR between the chlorin and porphyrin series for these reactions
1
1
of the relaxed *ZCPh and *Z3PnPh states from their corre-
sponding ground states can be estimated from the Stokes shifts
between their optical absorption and emission spectra, where λ
) (Eabs - Eem)/2. Using the data in Table 1, λ e 0.02 eV for
the formation of both 1*ZCPh and 1*Z3PnPh, so that the
structural changes that occur upon formation of these excited
states are neglible. This means that λI for both charge separation
and recombination can be calculated by considering only
geometry changes between the ground state and the radical-
ion-pair state. For a given acceptor, its contribution to λI for
both charge separation and recombination is the same within
each pair of chlorin and porphyrin dyads, so that the entire
difference in λI is due solely to the differences between λI for
the chlorin and porphyrin radical cations. For each radical cation,
λI was calculated using eq 6:
difficult. The data in Table 3 show that the radical-ion-pair
energies of the ZCPh dyads in toluene are all slightly above
that of 3*ZC, while those of the Z3PnPh dyad are slightly below.
In THF, all of the radical-ion-pair energies are below those of
the triplet states of the macrocyclic donors. This suggests that
charge recombination rates from a triplet radical ion pair to the
neutral triplet states should be quite slow and noncompetitive
with charge recombination to the singlet ground state. The
transient absorption data in Figures 6-8 and Figures S6-S8,
S11-S16 confirm that charge recombination in all the dyads
presented here does not produce any significant yield of long-
lived triplet products.
Another potential source of variability in the calculated values
of V due to small differences in ∆G may arise specifically in
low-polarity solvents, such as toluene. Whenever λS is very
small, as is the case for toluene, rate constants calculated using
eq 4 show oscillations with a period of pω near the top of the
rate vs free energy profile as well as in the inverted region. For
small variations in the magnitude of ∆G, these oscillations can
λI ) E0+ - E0
(6)
where E0+ is the energy of the ground state at the radical cation
geometry and E0 is the energy of the ground state in its energy-
minimized conformation. These energies were calculated by
unrestricted DFT using the PW91 functional and the 6-31G*
basis set,55 to give λI ) 0.14 eV for ZCPh+•/ZCPh and λI )
(65) Weller, A.; Staerk, H.; Treichel, R. Faraday Discuss., Chem. Soc. 1984,
78, 271-278.
(66) Hoff, A. J.; Gast, P.; van der Vos, R.; Franken, E. M.; Lous, E. J. Z. Phys.
Chem. 1993, 180, 175-192.
(67) Steiner, U. E.; Ulrich, T. Chem. ReV. 1989, 89, 51-147.
(68) Till, U.; Hore, P. J. Mol. Phys. 1997, 90, 289-296.
(64) Marcus, R. A. J. Chem. Phys. 1965, 43, 679-701.
9
4788 J. AM. CHEM. SOC. VOL. 128, NO. 14, 2006

Intramolecular ET in a Chlorophyll a Derivative
A R T I C L E S
result in significant differences in the calculated value of V.
When λS is large, as is the case for THF, there are no oscillations
in the rate vs free energy profile, so that small variations in
∆G will not affect V. To determine whether this issue affects
the results for the molecules examined here, we obtained a
complete set of charge separation and recombination time
constants in THF (Table 3). Comparing the data for the ZCPh
and Z3PnPh dyads with a given acceptor in THF, both VCS and
tively. The reduction in the hfcc’s is consistent with a small
distortion of the geometry in ring D around the 17- and 18-
positions due to steric hindrance of the 20-phenyl group.
As noted previously,43 the effect of substitution does not cause
the long-range changes that would be expected if the molecular
orbital occupation were different for ZC+• and ZCPh+•. The
ENDOR data allow us to conclude that the addition of the 20-
phenyl does not change the electronic structure of the zinc
chlorin radical cation significantly, which is completely con-
sistent with the HOMO having approximate a2 symmetry. The
hfcc’s observed using ENDOR are also consistent with the spin
(and charge densities) calculated using DFT (Table S3), which
show that FC ) 0.03 for the 20-positions of both ZC+• and
ZCPh+•. These results leave us with a significant discrepancy.
The charge distributions within ZCPh+• and Z3PnPh+• are both
consistent with established patterns present in related macro-
cycles, which leads us to expect that VCR and kCR for the Z3PnPh
dyads should be significantly larger than that for the corre-
sponding ZCPh dyads. This, however, is not observed. To
examine this point further, we used DFT calculations to compare
the electronic structures of ZCPh+• and Z3PnPh+•.
VCR do not vary by more than a factor of 2, similar to the results
in toluene, so that the observed differences of VCS and VCR
between the chlorins and porphyrins in toluene are most likely
not due to oscillatory behavior in the rate vs free energy profile.
The value of VCR can be related to several parameters
associated with the molecular orbitals on the donor and acceptor
by the approximate expression32
DA
VCR ) K
c D cjA Sij
(7)
∑∑ i
i
j
where ciD and cjA are the atomic orbital coefficients of the
DA
relevant MOs on the donor atoms i and acceptor atoms j, Sij
is the overlap integral of atomic orbitals on D and A, and K is
an empirical constant. As discussed above, the Z3PnPh HOMO
has a2u symmetry,37 while the ZCPh HOMO should have a2
symmetry, which has a charge distribution similar to that of
the metalloporphyrin a1u orbital. This should place much more
charge density at the 20-position of ZnPnPh+• than on that of
ZCPh+•, resulting in a much larger value of VCR (eq 7) and
consequently a much faster rate of charge recombination for
the porphyrin-containing radical ion pairs. Since this is contrary
to what is observed, we used ENDOR spectroscopy to establish
the electron density distribution within ZCPh+•.
ENDOR spectroscopy of Chl a+• in solution shows that the
hfcc’s of the peripheral alkyl groups on Chl a+• are large, while
the splittings of the protons at the meso positions (5, 10, and
20) are very small.43 This is consistent with the HOMO of Chl
a+• having a2 symmetry. Very similar electron density distribu-
tions are observed for both ZC+• and ZCPh+•. The only
significant difference between the hfcc’s observed for ZC+• and
ZCPh+• involves the protons at the 17 and 18 carbon atoms.
The hfcc’s for the remaining protons do not change significantly.
The â-proton hfcc’s (in MHz) are related to the dihedral angle
between the π-orbital bearing the spin density and C-H bond
by eq 8:38,69,70
DFT calculations (PW-91, 6-31G*) were used to obtain the
charge (and spin) density distributions of the HOMOs of ZCPh+•
and Z3PnPh+• (Figure 10A,B). The Z3PnPh+• HOMO has an
electron density distribution typical of a porphyrin a2u orbital,
which is consistent with previous DFT calculations and with
well-established experimental evidence in the porphyrin field.71,72
In contrast, the electron density distribution within the HOMO
of ZCPh+• indicates that the HOMO has a2 symmetry, which
appears as a mixture of the distributions characteristic of
metalloporphyrin a1u and a2u orbitals. This phenomenon has been
observed previously for water-ligated chlorophyll molecules and
is attributed to a Jahn-Teller distortion of the macrocycle
resulting from ligation of the metal center.40,73,74
A contributing factor to the similarity of the values of VCR
for the ZCPh and Z3PnPh dyads might be a discrepancy between
the dihedral angles that the π-system of the 20-phenyl group
makes with the π-systems of ZCPh+• and Z3PnPh+•. If this
dihedral angle is defined as 0° when the two π-systems are
coplanar, the DFT calculations show that the dihedral angles
for ZCPh+• and Z3PnPh+• are 75° and 60°, respectively. This
suggests that π-π overlap of the phenyl group with the
macrocycle is slightly larger for Z3PnPh+•, so for a given
acceptor, the calculations predict that VCR for Z3PnPh+• should
be larger than that for ZCPh+•. Thus, the calculated differences
in the dihedral angle of the 20-phenyl group do not account for
the similar values of VCR observed for the ZCPh+• and Z3PnPh+•
dyads.
aH ) Fc(-9.2 + 96.7 cos2 θ)
(8)
where Fc is the spin density on the R-carbon of the π-system
and θ is the dihedral angle between the axis of the pz orbital of
the R-carbon in the π-system and the Câ-Hâ bond. Unrestricted
DFT spin density calculations (Table S3) show that FC for the
16- and 19-positions of ZC+•, which are the relevant R-carbons,
are both reduced upon incorporation of the 20-phenyl group in
ZCPh+•. Using eq 8 and the data in Tables 4 and S3, and
assuming that the calculated values of FC are dominated by the
π spin density at the R-carbon, the calculated dihedral angles,
θ, for the 17 and 18 â-protons in ZCPh+• are 28° and 30°,
respectively, while those for ZCPh+• are 31° and 42°, respec-
In essentially all previous considerations of the influence of
electronic coupling on rates of both energy and electron transfer
to or from porphyrins and chlorophylls, the analysis of rate
changes as a function of electronic structure focuses on the
electron density distribution in the relevant orbitals of the
macrocycles. For electron-transfer reactions, the presence of the
(71) Spellane, P. J.; Gouterman, M.; Antipas, A.; Kin, S.; Liu, Y. C. Inorg.
Chem. 1980, 19, 386-391.
(72) Vangberg, T.; Lie, R.; Ghosh, A. J. Am. Chem. Soc. 2002, 124, 8122-
8130.
(69) Heller, C.; McConnell, H. M. J. Chem. Phys. 1960, 32, 1535-1539.
(70) Stock, L. M.; Wasielewski, M. R. J. Am. Chem. Soc. 1973, 95, 2743-
2744.
(73) Song, H.; Reed, C. A.; Scheidt, W. R. J. Am. Chem. Soc. 1989, 111, 6868-
6870.
(74) Prendergast, K.; Spiro, T. G. J. Phys. Chem. 1991, 95, 9728-9736.
9
J. AM. CHEM. SOC. VOL. 128, NO. 14, 2006 4789

A R T I C L E S
Kelley et al.
Figure 10. HOMOs of model compounds for Z3PnPh+• (A) and ZCPh+• (B). The perturbation caused by the presence of negative partial point charges are
shown in (C, D) (see text).
partner donor or acceptor is often neglected. For cases involving
energy transfer, where the donor and acceptor are both neutral
molecules, this assumption generally causes few difficulties.
However, photoinduced electron transfer within covalently
linked donor-acceptor molecules often places the photo-
oxidized donor at a highly restricted distance and orientation
relative to the reduced acceptor. Under these circumstances, the
charge distributions of each partner ion can have a significant
influence on the overall charge distribution within the radical
ion pair. At the simplest level, this distortion is due to Coulombic
attraction of the positive and negative charges. Measurements
of charge distributions within the chemically oxidized chlorin
or porphyrin donors using EPR or ENDOR measurements of
their spin distributions do not capture these distortions. The
counterions that are present following chemical or electrochemi-
cal generation of the chlorin and porphyrin radical cations do
not occupy a single unique position relative to the macrocycle,
so their electrostatic influence on the charge distributions of
the radical cations is averaged out in solution.
Previous DFT calculations performed on NI-• and PDI-•
reveal that the majority of the negative charge density resides
on the carboximide oxygen atoms.75 While it is not yet possible
to model the overall radical-ion-pair charge distribution using
current computational methods, we have modeled the influence
of the negative charge density due to the radical anion on the
electronic structure of both ZCPh+• and Z3PnPh+•. DFT
calculations were performed on these radical cations with two
-0.5 charges located where the nearest carbonyl oxygen atoms
of the acceptor would reside in the complete donor-acceptor
dyad. Although the aforementioned DFT calculations on PDI-•
alone place only ∼25% of the negative charge density on each
oxygen atom, we assume that Coulombic interactions between
the radical pairs result in a larger portion of the negative charge
density residing on the dicarboximide closest to the electron
donor. The calculations with the two fixed -0.5 charges reveal
that the presence of the negative charges significantly shifts the
charge density of both radical cations from the macrocyclic core
onto their bridging phenyl rings (Figure 10C,D). As a result,
the atomic orbital coefficients on the donor atoms i, ciD,76,77
relevant to the electronic coupling matrix element, VCR (eq 7),
are actually quite similar for the different donor chromophores
(see Table S4). This effect should be more pronounced in low-
polarity solvents, e.g. toluene used in this study, than it would
be in highly polar solvents in which the charges are more
effectively screened from one another by the high dielectric
constant of the medium. Therefore, the presence of an adjacent
covalently linked radical anion at a fixed location relative to
the radical cation significantly alters the electron densities on
both ZCPh and Z3PnPh, which results in nearly identical values
of VCR, and consequently kCR for charge recombination. We are
currently performing a more extensive DFT computational study
on macrocyclic radical cations in the presence of charge
distributions characteristic of adjacent acceptors at fixed dis-
tances and orientations relative to the macrocycle to better
simulate the charge distributions of the donor-acceptor pairs
and VCR
.
Conclusions
Suzuki cross-coupling reactions have afforded 20-phenyl-
substituted Chl a derivatives in good yields and significant
quantities from readily available Chl a. The versatility of this
(76) Gorelsky, S. I. AOmix: Program for Molecular Orbital Analysis; York
University: Toronto, Canada (http://www.sg-chem.net), 1997.
(77) Gorelsky, S. I.; Lever, A. B. P. J. Organomet. Chem. 2001, 635, 187-
196.
(75) Weiss, E. A.; Wasielewski, M. R.; Ratner, M. A. J. Chem. Phys. 2005,
123, 064504/064501-064504/064508.
9
4790 J. AM. CHEM. SOC. VOL. 128, NO. 14, 2006

Intramolecular ET in a Chlorophyll a Derivative
A R T I C L E S
new functionality has resulted in the synthesis and characteriza-
tion of a series of ZCPh-arylene bis(dicarboximide) donor-
acceptor dyads. In addition to synthetic flexibility, arylation of
the 20-position preserves the 132 keto and 173 ester groups that
direct its self-association, while keeping the electron acceptor
at a position that should not interfere with known chlorophyll
aggregation motifs. Comparing these molecules to an analogous
zinc 5,10,15-tri(n-pentyl)-20-phenylporphyrin-acceptor (Z3PnPh-
acceptor) series shows that the rates of photoinduced charge
separation and recombination differ by less than a factor of 2.
However, EPR and ENDOR spectroscopy corroborated by DFT
calculations show that the highest occupied MO of ZCPh+• has
little spin (charge) density at the 20-carbon atom, whereas
Z3PnPh+• has significant spin (charge) density. DFT calculations
on ZCPh+• and Z3PnPh+• with two -0.5 charges located where
the nearest carbonyl oxygen atoms would reside in the complete
donor-acceptor dyad show that the presence of the negative
charges significantly shifts the charge density of both ZCPh+•
and Z3PnPh+• from the macrocycle onto the phenyl rings. These
results suggest that the use of measured and/or calculated charge
distributions in free radical ions to estimate electronic coupling
matrix elements for electron transfer most likely gives erroneous
results when covalently linked radical ions are positioned across
a short spacer at a fixed distance from one another.
Acknowledgment. The authors thank Mr. J. A. Vura-Weis
and Dr. B. Rybtchinski for many valuable discussions, along
with Drs. M. J. Ahrens and S. E. Miller for synthetic and
spectroscopic assistance. M.J.T. acknowledges the donors of
the American Chemical Society Petroleum Research Fund for
partial support of this research. This research was supported by
the Division of Materials Sciences, Office of Basic Energy
Sciences, U.S. Department of Energy, under Contract No. W-31-
109-ENG-38. The Bruker E580 spectrometer was purchased
with partial support from NSF Grant No. CHE-0131048. The
authors also thank BASF AG for a generous gift of bis-
(pinacolato)diboron.
Supporting Information Available: Details regarding the
synthesis and characterization of the molecules used in this
study; the ∆G/λi calculations; additional transient absorption
data, including Figures S1-S20 and Tables S1-S4; and the
complete list of authors for ref 54. This material is available
free of charge via the Internet at http://pubs.acs.org.
JA058233J
9
J. AM. CHEM. SOC. VOL. 128, NO. 14, 2006 4791