Synthesis and photophysics of porphyrin-fullerene donor-acceptor dyads with conformationally flexible linkers

Article abstract of DOI:10.1016/j.tet.2005.07.127

The synthesis and photophysics of a series of porphyrin-fullerene (P-C ₆₀) dyads in which the two chromophores are linked by conformationally flexible polyether chains is reported. Molecular modeling indicates the two moieties adopt a stacked c

Full text of DOI:10.1016/j.tet.2005.07.127

Tetrahedron 62 (2006) 1928–1936
Synthesis and photophysics of porphyrin–fullerene
donor–acceptor dyads with conformationally flexible linkers
a,
a
Shaun MacMahon, Dirk M. Guldi, Luis Echegoyen
b
c
David I. Schuster,
*
d
and Silvia E. Braslavsky
a
Department of Chemistry, New York University, New York, NY 10003, USA
Institute for Physical and Theoretical Chemistry, University of Erlangen, 91058 Erlangen, Germany
b
c
Department of Chemistry, Clemson University, Clemson, SC 29634, USA
d
Max Planck Institute for Bioinorganic Chemistry (formerly Strahlenchemie), D 45413 M u¨ lheim, Germany
Received 7 June 2005; revised 26 July 2005; accepted 30 July 2005
Available online 15 December 2005
Abstract—The synthesis and photophysics of a series of porphyrin–fullerene (P–C ) dyads in which the two chromophores are linked by
6
0
conformationally flexible polyether chains is reported. Molecular modeling indicates the two moieties adopt a stacked conformation in which
the two chromophores are in close proximity. Photoexcitation of the free base dyads in polar solvents such as tetrahydrofuran and
benzonitrile, causes electron transfer (ET) to generate charge-separated radical pair (CSRP) states, which were directly detected using
transient absorption (TA) techniques. In nonpolar solvents such as toluene, where CSRP states were not directly detected, fullerene triplet
state states were formed, according to TA studies as well as singlet oxygen sensitization measurements. The low value of the quantum
1
efficiency for sensitized formation of singlet molecular oxygen [O ( D )] in toluene and chloroform indicates that singlet energy transduction
g
2
1
to give H P– C *, followed by intersystem crossing to H P– C * and energy transfer to O , is not the operative mechanism. Rather, a
3
3
2
60
2
60
2
mechanism is proposed involving ET to give CSRP states followed by exergonic charge recombination to eventually generate fullerene
triplets. Such a mechanism has been demonstrated experimentally for structurally related P–C60 dyads. For the corresponding ZnP–C60 dyads
with flexible linkers, only photoinduced ET to generate long-lived CSRP states is observed. Photoinduced charge separation in these dyad
systems is extremely rapid, consistent with a through space rather than through-bond mechanism. Charge recombination is up to three orders
of magnitude slower, indicating this process occurs in the inverted region of the Marcus curve that relates ET rates to the thermodynamic
driving force. These observations once again demonstrate the advantages of incorporating fullerenes as electron acceptor components in
photosynthetic model systems.
q 2005 Elsevier Ltd. All rights reserved.
1
. Introduction
reorganization energy (l) accompanying electron transfer
in fullerenes. Since the negative charge on the fullerene
produced after photoexcitation of the donor moiety is
delocalized over the entire p-system, the charge density is
much lower than it would be on acceptors such as quinones,
where the charge is concentrated mainly on the oxygens.
The structural inflexibility of fullerene-based systems
Fullerenes, C60 in particular, have been found to make ideal
acceptors in model photosynthetic systems, due to their
unique photophysical, electrochemical, and chemical pro-
1
perties. C has been shown to reversibly accept up to six
0
6
electrons in solution due to its three low lying degenerate
lowest unoccupied molecular orbitals (LUMOs), with a first
electrode potential resembling that of the quinone-derived
results in small internal reorganization energies (l
the charge dispersion minimizes the solvent reorganization
), while
v
2
3
electron acceptor in the photosynthetic reaction center.
energy (l ).
s
Fullerenes accelerate charge separation and slow down
charge recombination versus donor–acceptor (DA) dyads
containing traditional two-dimensional electron acceptors
such as quinones and imides. This phenomenon, which is
general, has been rationalized by the small Marcus
For electron DA systems, Marcus theory predicts that as the
0
free energy change (DG ) for electron transfer (ET) becomes
more negative, that is, increasingly exergonic, the ET rates
increase, reaching a maximum when KDGETZl, the
0
reorganization energy. As KDG becomes more negative
than the absolute value of l, the ET rate decreases. For C60
dyads, the maximum in the Marcus curve is reached at smaller
ET
Keywords: Fullerenes; Porphyrins; Dyads.
*
Corresponding author. Tel.: C1 212 998 8447; fax: C1 212 260 7905;
e-mail: david.schuster@nyu.edu
0
values of KDG compared to two-dimensional acceptors.
ET
0040–4020/$ - see front matter q 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2005.07.127

D. I. Schuster et al. / Tetrahedron 62 (2006) 1928–1936
1929
Thus, charge separation (CS) rates for most fullerene-based
DA dyads are located along the upward slope of the Marcus
curve, near the maximum, leading to enhanced CS rates,
whereas the large thermodynamic driving force for charge
recombination (CR) places this process on the downward
side of the Marcus parabola in the inverted region. This
phenomenon causes a noticeable enhancement in the lifetime
of charge-separated radical pair (CSRP) states in C -based
solubility properties, as well as higher yields in the coupling
reactions. Improved solubility would in turn allow more
facile structural characterization. In the first series of
reactions, summarized in Scheme 1, fullerene acid 1 was
condensed with glycols 2a and 2b to give 3a and 3b,
respectively, which were then condensed with 5,10,15,20-
tetraphenylporphyrin carboxylic acid 4 to give dyads 5a and
5b, respectively. Details of these reactions are provided in
an earlier paper dealing with the synthesis of a variety of
6
0
4
DA systems versus analogous two-dimensional DA systems.
1
0
P–C60 dyads.
Porphyrins are frequently used as donors in artificial
photosynthetic systems. Porphyrins are synthetically
accessible in a laboratory, (although their synthesis can be
O
O
5
a rather arduous task), and they absorb light much more
O
(
O
)
O
1
effectively than fullerenes in the visible region between 400
and 600 nm. Favorable van der Waals interactions between
the curved p surface of the fullerene and the planar p
surface of the porphyrin assist in the formation of
supramolecular structures, overcoming the necessity to
HO
H
n
OH
O
2
a) n = 3
b) n = 4
DCC,
DMAP,
BtOH
6
match a concave-shaped host to a convex-shaped guest.
This unique intermolecular electronic interaction is
observed in condensed media as well as in the solid state.
O
OH
O
O
H
(
)O
N
O
O
O
n
Attempts to suppress CR have led to the development of
multi-chromophoric triads, tetrads and pentads, where ET
can occur through a multi-step process, yielding CSRP with
N
HH N
O
N
3
a) n = 3
n = 4
1
c,d,7
4
microsecond to second lifetimes.
An inherent problem
EDC,
DMAP
is the synthetic challenge involved in the synthesis of such
complicated assemblies. These multicomponent materials
are far less efficient in the storage of energy than dyad
systems, since sequential ET dissipates more and more of
the initial excitation energy, lowering the capacity for
energy storage.
O
O
(
O)
O
O
O
O
N
n
O
Herein, we report on simple porphyrin–C60 (P–C ) dyads
60
5a n=3
5b n=4
N
H
with flexible linkers, which are easily synthesized and
which have many of the desirable properties of DA systems
with respect to energy storage and CSRP state lifetimes.
N
HN
2
. Free base porphyrin–fullerene dyads (H P–C )
60
Scheme 1.
2
2
.1. Synthesis and structural characterization
1
H NMR spectra of 5a and 5b showed characteristic
porphyrin bands between 7.5 and 9 ppm, glycol bands at
Our initial foray into this area centered on the synthesis of
porphyrin–fullerene (H P–C ) dyads with flexible crown
ether-like glycol linkers by a convergent route. Steady-
state fluorescence spectra demonstrated rapid and efficient
quenching of the porphyrin S state by the fullerene, but the
CS and CR kinetics were not measured on these phase I
dyads. Among the synthetic problems encountered were the
extremely low yields (i.e., w30%) when using simple
methanofullerene carboxylic acid, due to the acid poor
solubility properties. Coupling of tetraphenylporphyrin
carboxylic acid to the polyether linker also proved difficult,
making it extremely challenging to generate dyads on a
milligram scale.
3.5–4 ppm, and distinctive peaks from the C60 synthon 1.
Computer modeling using Insight II provided a prediction of
the H P–C geometry in the ground state (see Fig. 1).
2
Because of favorable porphyrin–fullerene p–p interactions,
the glycol linker curves around, placing the porphyrin and
fullerene within van der Waals contact range.
2
60
8
1
1
60
1
UV–vis spectra of dyads 5a and 5b in chloroform
demonstrated strong ground state interactions. The por-
phyrin Soret bands at 422 nm were red shifted by
approximately 5–10 nm relative to methyl ester 8, indicat-
ing that the fullerene was close enough to the porphyrin in
the ground state to perturb its absorption spectrum. Similar
spectral shifts were also visible in the porphyrin Q-band
region in both dyads. The broadness of the absorption in
both the Soret and Q-band regions was surprising, and could
indicate the presence of more than one conformation of
these dyads in solution.
In order to circumvent these problems, fullerene carboxylic
acid 1, originally reported by Diederich and Nierengarten,
was used as the fullerene synthon. The strategy was to
replace methanofullerene carboxylic acid by 1, which would
lead to flexibly linked H P–C dyads with much better
9
2
60

1930
D. I. Schuster et al. / Tetrahedron 62 (2006) 1928–1936
4
-carboxymethylbenzaldehyde to give porphyrin methyl
ester 8. Overall yields of purified porphyrin were 17%.
Hydrolysis gave carboxylic acid 6 in 94% yield.
Scheme 3 depicts the synthesis of the solubilized flexibly
linked H P–C dyads. The initial coupling reaction to
2
60
fullerene acid 1 was carried out as before. Coupling of 3a
and 3b to the porphyrin acid 6 using 1-(3-dimethylamino-
propyl)-3-ethylcarbodiimide/dimethyl-aminopyridine
(EDC/DMAP) gave the desired dyads 9a and 9b in 50%
yield.
Intramolecular electronic interactions were again monitored
using UV–vis spectroscopy. Bathochromic shifts of the
porphyrin Soret band were pronounced, ranging from
6
–10 nm depending on the solvent. The UV–vis spectra
for dyads 9a and 9b were collected in solvents of greatly
varying relative permittivity: toluene (2.0), THF (7.6),
CH Cl (9.1), and benzonitrile (PhCN) (26.0). An increase
2
2
in the porphyrin–fullerene interaction was seen with
increasing solvent polarity, particularly in 9b. In relatively
nonpolar solvents such as toluene and THF, the interaction
between the chromophores in 9a was stronger than in 9b,
according to shifts in the Soret band. As the solvent became
more polar, the interaction in 9b became comparable to that
in 9a. This may be because the chromophores are not as
effectively solvated in polar media. It is remarkable that the
interaction persists even in toluene, possibly because both
the porphyrin and fullerene moieties are effectively solvated
in aromatic solvents, and therefore must overcome a large
entropic barrier to interact with each other. Such effects
should be more pronounced with 9a and 9b versus 5a and 5b
due to the presence of three di-tert-butylphenyl groups.
Figure 1. Insight II molecular mechanics calculations for dyads 5a and 5b.
Whereas substantial improvements in yield were observed
in the coupling yields using fullerene carboxylic acid 1, the
porphyrin coupling step still proved to be extremely
difficult. Moreover, the solubility of dyads 5a and 5b,
though somewhat improved over the first generation dyads,
still severely restricted the choice of solvents for photo-
physical studies. In order to improve the solubility
characteristics of the porphyrin component, tert-butyl
substituents were incorporated on the benzene rings, as in
a large body of work in this field. The stepwise synthetic
route to the solubilized porphyrin carboxylic acid 6 shown
in Scheme 2 was developed. The first step involved
preparation of a dipyrromethane derivative 7 from
pyrrole and 3,5-di-t-butyl benzaldehyde, followed by
condensation with 1 equiv each of this aldehyde and
2
.2. Electrochemistry
For 9a, the first reductive redox couple of the C60 center
appears at K0.55 V versus Ag/AgCl, which is in good
agreement with values typically observed for C deriva-
6
0
1
2
tives. The first one-electron oxidation of the porphyrin
CHO
CHO
CHO
pyrrole
,
COOMe
TFA, CH Cl
2
2
NHHN
TFA, CH Cl
2
2
7
N
O
NaOH, iPr-OH
HN
O
N
∆
HN
NH N
OMe
NH N
OH
8
6
Scheme 2.

D. I. Schuster et al. / Tetrahedron 62 (2006) 1928–1936
O
1931
excitation in the Soret band to the porphyrin S2 state.
The rate constant for fluorescence deactivation was higher
in chloroform, which we ascribe to the increased rate of
electron transfer quenching expected in the more polar
solvent (vide infra). An increase in fluorescence quenching
as a function of increasing solvent relative permittivity was
observed. The quenching efficiency for dyads 5a and 5b
both increase upon going from o-dichlorobenzene (ODCB)
to PhCN to N,N-dimethylformamide (DMF).
OH
O
O
N
H
(
)O
n
N HH
N
N
O
O
O
O
3
3
a n = 3
b n = 4
6
EDC,
DMAP
O
O
(
O)
n
O
O
O
O
O
N
H
N
N
HN
9
9
a n=3
b n=4
Scheme 3.
center occurs at C1.10 V, which is a bit more positive than
that for tetraphenylporphyrins (typically C1.0 V). The
small shifts are attributable to the intramolecular inter-
actions in the dyad. Dyad 9b showed a weak wave for
the oxidation of the porphyrin center at 1.08 V, but no
discernible peak for the reduction of the C60 moiety was
observed, probably due to the small amount of sample
available. Similar measurements were not performed on 5a
and 5b.
Figure 3. Fluorescence emission for dyads 5a and 5b in CHCl
The porphyrin standard was methyl ester 8.
3
(10 mM).
The fluorescence properties of 9a and 9b were measured
against that of the methyl ester of 6 as the porphyrin
standard, with a fluorescence lifetime of 9 ns. Again, rapid
and efficient quenching of porphyrin fluorescence in 9a and
9b was observed. The rate and efficiency of fluorescence
quenching increased with increasing solvent polarity, as was
seen with 5a and 5b.
2
.3. Steady-state fluorescence
The steady-state fluorescence properties of the flexibly
linked H P–C dyads were measured against that of a
2.4. Quantum yields for singlet oxygen formation
2
60
porphyrin standard, namely the methyl ester 8, which has a
fluorescence lifetime of 9 ns. Strong quenching in dyads 5a
and 5b was seen relative to the standard in both benzene
1
Quantum yields of sensitized singlet oxygen [O ( D )]
formation (F ) by dyads 5a and 5b upon excitation of
D
2
g
air-saturated solutions at 532 nm are shown in Table 1. Full
1
(
Fig. 2) and chloroform (Fig. 3). To avoid signal saturation
3
details of the procedure are given elsewhere. Quantum
yields are reported relative to tetraphenylporphine (TPP) for
due to excessive absorption of the incident excitation light,
the excitation was carried out at 550 nm, a region well
separated from the Soret peak where the light is still
absorbed exclusively by the porphyrin moiety. The other
option available was to use extremely dilute solutions with
1
4
which F Z0.67G0.14. The F value is an indirect way
D
D
3
3
to measure C formation, since C transfer energy to O
3
60
60
2
1
to generate O ( D ) occurs with unit efficiency, and for
2
g
1
5
simple fullerene derivatives F is generally O0.9. The FD
D
values for 5a and 5b are lower in chloroform than in toluene,
where the values are the same for both dyads. The low FD
value in toluene indicates that singlet energy transduction
1
to give H P– C *, followed by intersystem crossing to
2
60
3
H P– C * and energy transfer to O , is not operative.
3
2
60
2
1
2 g
Table 1. Quantum yields for O ( D ) formation
Compound
F (relative) in
PhMe
F (relative) in
CHCl
3
Porphyrin standard (TPP)
Dyad 5a
Dyad 5b
0.89
0.34
0.34
0.94
0.18
0.09
Figure 2. Fluorescence emission for dyads 5a and 5b in benzene (10 mM).
The porphyrin standard was methyl ester 8.
Excitation of air-saturated solutions at 532 nm. For procedural details, see
Ref. 13.

1932
D. I. Schuster et al. / Tetrahedron 62 (2006) 1928–1936
If such a mechanism were operative, the values of F
metastable singlet excited states in H P, leading to the
2
D
1
5
should have been much higher, in the range 0.9–1.0. Such
a sequence has indeed been proposed for some H P–C
corresponding triplet manifold, which we see with our
slower experiments (i.e., 10 ns laser pulses). Characteristic
absorption for C triplets was seen at 720 nm (data not
2
60
1
6,17
dyads in nonpolar solvents.
However, another possible
6
0
1
9
mechanism involves electron transfer in nonpolar solvents
followed by back electron transfer to generate lower lying
porphyrin and fullerene triplet states. Indeed, recent time-
resolved electron paramagnetic resonance (TREPR) studies
shown). For experimental details, see previous work from
1
3
this collaborative team. Time-resolved transient absorp-
tion measurements with the free base dyads 9a and 9b
completely corroborated the steady-state fluorescence
1
8
have established that the following stepwise mechanism is
operative for a parachute-shaped ZnP–C60 dyad system in
toluene: (1) electron transfer to give a charge-separated
experiments. The rate of decay of the S singlet excited
1
state of H P in THF and PhCN is much faster than the
2
typical intersystem crossing rate of H P, using ps excitation
2
3
state, (2) back electron transfer to give ZnP*–C , and
0
and detection in the ps and ns range. Moreover, the
differential absorption changes at the end of the fast decay
bear no resemblance with any of the typical singlet/triplet
6
finally (3) energy transfer to give the lower lying
ZnP– C *. Although we have no TREPR data as yet on
3
6
0
1
9
dyads 5a and 5b, we propose that a similar mechanism is
operative in the present case. It is interesting to note that the
triethylene glycol-linked dyad 5a generates twice as much
excited state characteristics for H P or C . Instead, broad
2 60
2
H P p-radical cation absorption in the visible and C
0
2
60
2
1
p-radical anion absorption in the near-infrared are
observed (see Fig. 4), which were then employed as
convenient markers to quantify the CS efficiency and CR
dynamics.
1
O ( D ) as the tetraethylene glycol-linked compound 5b in
2
g
chloroform, which would be difficult to explain if the energy
transduction pathway was operative, but makes sense in
terms of a stepwise mechanism for O ( D ) formation where
1
2
g
three distinct steps of variable efficiency are involved.
Unfortunately, because of lack of material, similar studies
were not done with dyads 9a and 9b.
In this context, it is worth noting that generation of fullerene
triplets in high yield by an electron transfer/charge
recombination pathway has also been reported by Armaroli
et al. for a quite different type of dyad, namely one in which
C60 is covalently linked to a bipyridine (bipy) moiety
1
9
complexed with Ru(II). Excitation of the metal complex
moiety gives the lowest triplet metal-to-ligand-charge-
transfer (MLCT) excited state, which is quenched by
electron transfer (ET) to give the C60 radical anion in
quantitative yield. This is followed by charge recombination
3
to generate C *, again in quantitative yield, as determined
6
0
Figure 4. Differential absorption spectrum (visible and near-infrared)
by luminescence, as above. It was noted that triplet–triplet
energy transfer (EnT) from MLCT* to C * does not
K5
obtained upon nanosecond flash photolysis (532 nm) of w1.0!10
solutions of 9b in nitrogen-saturated THF with a time delay of 50 ns
M
3
3
6
0
compete with ET, because the EnT process is highly
exergonic and is therefore located in the Marcus inverted
region.
at room temperature. The spectrum corresponds to the radical pair
%K
%C
P
H
2
–C60 .
1
9
The CS and CR dynamics in dyads 5a and 5b, along with the
quantum yields for formation of the charge-separated states
2
.5. Transient absorption measurements
%
C
%K
H2P –C60 in five different solvents (F ), are summar-
CS
ized in Table 2. The lifetimes associated with CS decrease
as the solvent polarity increases. Thus, the CS rates are
highest in the most polar solvent, DMF, which is consistent
with normal Marcus region behavior. Given that the CS rate
constants vary only slightly with changes in solvent
polarity, that is, they differ by less than a factor of three
for both dyads, it is likely that CS in 5a and 5b is occurring
Characteristic transient singlet and triplet spectra were
recorded in the pico-, nano-, and microsecond time regime
following 532 nm laser excitation. Typically, in our ultrafast
experiments (i.e., 20 ps laser pulses) it is possible to see
singlet excited states formed instantaneously, namely, with
1
0 K1
kinetics faster than 5!10
s
. A fast intersystem crossing
8
K1
process (i.e., k w1.0!10 s ) governs the fate of the
isc
Table 2. Charge separation and recombination data for dyads 5a and 5b
Solvent
Dyad 5a
Dyad 5b
a
a
b
a
a
b
tCS (ns)
tCR (ns)
F
CS
tCS (ns)
t
CR (ns) ‘
F
CS
Toluene
THF
ODCB
PhCN
DMF
0.188
0.145
373
288
246
115
83
0.19
0.24
0.50
0.37
0.17
0.222
0.166
396
298
270
156
99
0.33
0.34
0.67
0.43
0.19
0.09
0.083
0.099
0.095
a
%K
From formation and decay of C60 absorption at 980–1020 nm.
From absorption at 1000 nm measured 20 ns after the flash, relative to reference dyad.
b

D. I. Schuster et al. / Tetrahedron 62 (2006) 1928–1936
Table 3. Electron transfer dynamics for dyads 9a and 9b
1933
a
b
Compound
t
F
(ns)
t
F
(ns)
tCS (ns)
THF
PhCN
PhMe
THF
PhCN
THF
PhCN
8
9.6
9.5
9.8
9
9
a
b
0.78
0.94
0.74
0.83
1.06
1.93
0.95
1.81
0.83
1.28
490
450
725
a
Measured directly by disappearance of porphyrin S
1
state.
From steady-state fluorescence quenching data and fluorescence lifetime of porphyrin 6 methyl ester.
b
near the top of the Marcus curve, where rates are relatively
insensitive to the thermodynamic driving force. However,
CR is clearly occurring in the inverted region of the Marcus
curve, as indicated by the fact that kCR is slowest in the least
polar solvent, toluene, and is highest in the most polar
solvent, DMF. This is exactly the opposite of what would be
expected if CR were occurring in the normal region of
the Marcus curve. The value of FCS is highest in
o-dichlorobenzene.
the longer linker, 9bZn. Eventually, in the most polar
solvents studied, the UV–vis spectra of ZnP–C60 display
nearly identical interactions to those in the H P–C
2
60
analogs. It is known that fullerene–porphyrin interactions
are not as strong in metallated as in free base porphyrins,
consistent with a larger distance between the chromophores in
9
fullerene–metalloporphyrin co-crystals. Interchromophoric
interactions prevail also in chelating solvents such as
benzonitrile and THF, as clear shifts and bleaching in all
absorption bands are observed upon comparison to the
Zn-porphyrin standard.
The ET dynamics for dyads 9a and 9b are summarized in
Table 3, along with fluorescence data for tetraphenyl-
porphyrin methyl ester 8. H P singlet excited state lifetimes,
2
3
.2. Electrochemistry
the inverse of the rates of charge separation, were measured
in two different ways. The first column shows transient
absorption data, following the rate of disappearance of
For 9aZn, the first oxidation (0.95 V) of the porphyrin
center is shifted negatively by 150 mV and the reduction
singlet–singlet absorption of H P in the visible. The second
2
(
K0.65 V) of C60 is shifted negatively by 100 mV relative
set of lifetimes derives from measurements of the
fluorescence lifetime of the H P standard, and determination
to appropriate standards. For 9bZn, the first one-electron
oxidation of the porphyrin center occurs at 0.92 V and the
first one-electron reduction of the C60 center appears at
K0.59 V. There is a negative shift of 160 mV for the
porphyrin center from 9b to 9bZn, consistent with the well
known greater ease of oxidation of zinc versus free base
porphyrins.
2
of the extent of quenching in the dyads in three different
solvents. The lifetime of the H P singlet excited state is then
2
calculated by assuming that I/I is equal to t/t . Although
0
0
the lifetimes determined by the two methods differ slightly,
the trend is the same, namely kCS increases with increasing
solvent polarity, consistent with normal Marcus region
behavior. The CR process, on the other hand, is clearly
occurring in the Marcus inverted region, as the lifetime in
THF far exceeds that in benzonitrile, a more polar solvent
in which the CSRP state is lower in energy relative to the
ground state than it is in THF. The very long lifetime of
the CSRP state of 9b in THF, namely 725 ns, is one of
the longest ever measured for a simple P–C dyad in
3.3. Steady-state fluorescence
Experiments for 9aZn and 9bZn were conducted using an
excitation wavelength of 520 nm. The relative quenching
efficiencies, relative to that of the ZnP standard, are similar
to those observed in the corresponding H P–C . However,
2
60
6
0
1
6,22
since the fluorescence lifetime of ZnP is typically on
solution.
3
. Zinc porphyrin–fullerene dyads (ZnP–C )
60
3
.1. Synthesis and structural characterization
Complexation of dyads 9a and 9b with zinc proceeded
quantitatively using Zn(OAc)2 in methanol, generating
ZnP–C60 9aZn and 9bZn. Since zinc porphyrins are in
general better electron donors than their free base
counterparts, 9aZn and 9bZn were expected to generate
CSRP states more rapidly and with a higher quantum yield
than their free base analogs. Electronic ZnP–C60 inter-
actions in 9aZn and 9bZn, as monitored by UV–vis
spectroscopy, are similar to those in 9a and 9b. Whereas
red shifts in the absorption spectra were observed, the trend
as a function of solvent polarity was not quite as clear as it
Figure 5. Differential absorption spectrum (visible and near-infrared)
K5
obtained upon nanosecond flash photolysis (532 nm) of w1.0!10
solutions of 9bZn in nitrogen–saturated THF with a time delay of 50 ns
M
was with the H P–C . An increase in solvent polarity
2
60
at room temperature. The spectrum corresponds to the radical pair
%
C
%K
ZnP –C60 .
increased the interaction, particularly in the dyad with

1934
D. I. Schuster et al. / Tetrahedron 62 (2006) 1928–1936
Table 4. Charge transfer dynamics for dyads 9aZn and 9bZn
a
Compound
t
F
(ns)
tCS (ns)
PhMe
THF
PhCN
THF
PhCN
8
9
9
Zn
aZn
bZn
2.1
0.24
0.56
2.1
0.21
0.50
2.1
0.11
0.49
490
440
180
150
a
Measured by comparing steady-state fluorescence results to lifetime of Zn-porphyrin standard.
the order of w2 ns, compared to 9 ns for the corresponding
In these systems, the rapidity of the photoinduced ET
process is best rationalized in terms of the molecular
conformations of these dyads shown in Figure 1. It was
H P, the intramolecular CS rates in the ZnP–C are clearly
faster. As with the H P–C analogues, increased quenching
2
60
2
60
2
4
with increasing solvent polarity is observed for the
ZnP–C .
reported by Armaroli et al. that a doubly-linked face-to-
face porphyrin–fullerene system and a related singly linked
system show remarkably similar photophysical behavior,
indicating that the donor and acceptor moieties in both these
systems are in close proximity, and that through space rather
than through-bond ET mechanisms dominate. On the other
hand, there are some significant differences between the
results of the present study and those reported for analogous
6
0
3
.4. Transient absorption measurements
The fate of the ZnP singlet excited state in 9aZn and 9bZn
was also examined by pico-, nano-, and microsecond
transient absorption spectroscopies. At early times (i.e.,
1
3
doubly-linked dyads with parachute topology. The linkers
in both cases are polyethers. The rates of charge separation
in the present system are slightly slower than those in the
parachute system by up to an order of magnitude, while the
rates of charge recombination in the flexibly linked dyads
are considerably slower, by closer to three orders of
magnitude. For example, in THF the fluorescence lifetime
of 5a is 0.15 ns compared to 0.02–0.08 ns for the parachute
dyad in the same solvent, while the CSRP lifetime of 5a is
5
0–100 ps), these are practically identical to those of the
ZnTPP, disclosing strong bleaching at 550 nm and attesting
to the formation of the ZnP singlet excited state. At a delay
time of ca. 200 ps, a new transition around 640 nm starts to
grow in, accompanied by another absorption in the near-
infrared. Based on a spectral comparison, we ascribe the
former to the ZnP p-radical cation, while the latter band
belongs to the C60 p-radical anion. The CSRP absorption is
persistent on the picosecond time scale and decays in the
nano- to microsecond time regime (see Fig. 5).
3
00 ns, compared to 0.3 ns for the parachute compound.
This difference can be rationalized in terms of significant
differences in topology of the systems, but also suggests that
through-bond ET processes may compete with through
space mechanisms in the flexibly linked dyads, particularly
with respect to charge recombination.
The ET dynamics in several solvents are tabulated in
Table 4. The rates of charge separation in 9aZn and 9bZn
are consistently faster than those in 9a and 9b, by up to as
much as an order of magnitude. This is hardly surprising
since CS in the Zn-dyads is thermodynamically more
favorable, which would lead to faster rates in an ET process
occurring in the normal region of the Marcus curve. The
charge recombination data again display inverted Marcus
region behavior, as longer lifetimes are observed in THF
than in benzonitrile.
We conclude that the forward ET process in these flexibly
linked donor–acceptor systems occurs near the peak of the
Marcus parabola relating ET rates and thermodynamic
driving force, while BET occurs deep in the inverted region,
resulting in charge-separated radical pair (CSRP) states with
lifetimes approaching the microsecond time regime. This
has obvious implications with respect to use of such systems
2
2
in energy storage devices.
4. Conclusions
The synthesis of a series of porphyrin–C60 dyads with
flexible polyether linkers of varying length has been
accomplished using a standard set of reactions. Introduction
of alkyl substituents on the porphyrin greatly increased the
solubility and ease of handling of these compounds. Using
steady-state and time resolved techniques, it was found that
the rate constants for forward electron transfer in both free
5. Experimental
All starting materials and solvents were purchased from
Sigma-Aldrich and used without purification unless other-
wise noted. H NMR spectra were collected using a Gemini-
1
200 MHz spectrometer. UV–vis spectra were collected
using a Perkin-Elmer Lambda 35 UV–vis spectrometer.
Fluorescence spectra were acquired using a Shimadzu RF-
5301 PC fluorescence spectrophotometer. A gas chroma-
tography mass spectrometer with an HP 5890 mass selective
detector was used to acquire GC–MS data. MALDI-MS data
were collected using an Omniflex MALDI-TOF mass
spectrometer. FAB-MS data were collected in collaboration
with the Michigan State Mass Spectrometry Facility. HPLC
was performed using a Milton-Roy 4000 obtained courtesy
base (H P) and ZnP systems, are on average approximately
2
three orders of magnitude greater than the rate constants
associated with back electron transfer. The formation of
C60 triplet states in toluene for the free base dyads is
attributed to ET followed by charge recombination, rather
than a direct energy transfer mechanism, analogous to the
situation in some other types of dyads. Fullerene triplets are
not detected for the ZnP dyads in either polar or nonpolar
solvents.

D. I. Schuster et al. / Tetrahedron 62 (2006) 1928–1936
1935
of Prof. Ira Krull at Northeastern University. The HPLC
column was a Buckyclutcher 1, using 15 mL injections.
Fluorescence lifetime data was collected using a Pico-Quant
Fluo-Time 100 time-resolved spectrophotometer. Pico-
second laser flash photolysis experiments were carried out
with 532 nm laser pulses from a mode-locked, Q-switched
Quantel YG-501 DP Nd:YAG laser system (18 ps pulse
width, 2–3 mJ/pulse). Nanosecond Laser Flash Photolysis
experiments were performed with laser pulses from a
Quanta-Ray CDR Nd:YAG system (532 nm, 6 ns pulse
width) in a front face excitation geometry.
was added dropwise, and the solution was heated at
reflux for 6 h. The solvent was removed under reduced
pressure, the residue was dissolved in CH Cl and then
2
2
extracted sequentially with 1 M HCl and water. After drying
over Na SO , the solvent was removed under reduced
pressure, and the final porphyrin carboxylic acid 6 was
2
4
purified via flash chromatography (yield 94%) (SiO , 1:9
2
MeOH/CH Cl , R Z0.5). The material was characterized
2
2
f
1
by H NMR and MALDI-MS analysis.
23
5.1.3. Glycol–C60 adducts 3a and 3b. General procedure:
bromobenzene was added to an Ar-purged flask
containing fullerene carboxylic acid 1 (1 equiv), DMAP
5
.1. Modified synthesis of 3,5-di-tert-butylbenzaldehyde
(0.2 equiv), 1-hydroxybenzotriazole hydrate (1.1 equiv) and
A solution of 3,5-di-tert-butyltoluene (21.84 g, 107 mmol)
and N-bromosuccinimide (26.66 g, 150 mmol) in benzene
dicyclohexylcarbodiimide (DCC) (10 equiv). The appropriate
diol (10 equiv)was added in DMSO, and the reaction was held
at 50 8C with stirring for 18 h. Bromobenzene was removed
under reduced pressure and the crude product was dissolved in
CH Cl and extracted sequentially with water, 10% aqueous
(
55 mL) was heated at reflux under visible light irradiation.
The reaction was then cooled, filtered, the benzene was
removed under reduced pressure and the residue was added
to a solution of hexamethylenetetramine (45.00 g,
2
2
CaSO and brine. The final solution was dried (Na SO ) and
4
2
4
3
21 mmol) in a 1:1 H O/EtOH mixture (64 mL). The
2
the solvent was removed under reduced pressure. The residue
was purified by flash chromatography (SiO , 10% EtOAc/
PhMe) to give 3a (61%) and 3b (64%), respectively.
solution was heated at reflux for 4 h, concd HCl was added
21 mL) and heating at reflux was continued for 30 min. The
2
(
ethanol was removed under reduced pressure, and the
remaining aqueous layer was extracted with ether. The ether
layer was dried (MgSO ) and the solvent was removed
under reduced pressure. Recrystallization from EtOH
yielded the desired product as white crystals (66%).
5
.1.4. Synthesis of dyads 9a and 9b. General procedure for
EDC/DMAP coupling: to a solution of porphyrin acid 6
1.01 equiv), EDC (1.01 equiv)and DMAP (1.01 equiv) in the
minimum amount of dry CH Cl , a solution of C carboxylic
4
(
2
2
60
acid derivative 3a or 3b (1 equiv) in CH Cl was added at
2
2
5
.1.1. 3,5-Di-tert-butylphenyl dipyrromethane 7.
room temperature. The mixture was stirred for 18 h, diluted
with CH Cl (20 mL), washed with a solution of saturated
aqueous NH Cl (3!20 mL) and brine (2!20 mL), dried
A solution of the above aldehyde (3.581 g, 16.4 mmol)
and freshly distilled pyrrole (28.4 mL, 409 mmol) was
stirred under Ar at room temperature. Trifluoroacetic acid
2
2
4
(Na SO ) and finally evaporated to give the crude dyad, which
was purified via flash chromatography to give the pure
2 4
(
0.13 mL, 16.4 mmol) was added dropwise, and the solution
was stirred an additional 30 min. An excess of triethylamine
0.3 mL) was added to quench the reaction, and the pyrrole
material. Dyad 9a: 51% (10% EtOAc/PhMe, R Z0.25); Dyad
f
(
9
b: 47% (10% EtOAc/PhMe, R Z0.2). The final products
f
1 23
was removed under reduced pressure. The crude product
was dissolved in CH Cl and extracted with 0.1 M aqueous
NaOH and water, dried (Na SO ) and the solvent was
were characterized by H NMR and MALDI-MS analysis.
2
2
2
4
5.2. Singlet molecular oxygen emission measurements
removed under reduced pressure. Flash chromatography
yielded the dipyrromethane derivative as a sticky light
brown solid (51%) (SiO , 85:14:1 Cyclohexane/EtOAc/
These experiments were done following the procedure
described in a previous paper in this series.
1
3
2
Et N, R Z0.5). The material was characterized by its
3
f
1 13
00 MHz H NMR spectrum, 50 MHz C NMR spectrum
2
and GC/MS analysis.
2
3
Acknowledgements
0
5
.1.2. 5-(4 -Carboxymethylphenyl)-10,15,20-tris-(3,5-di-
D.I.S. and S.M. are grateful to the US National Science
Foundation for support of their work at NYU, under grant
CHE-009789. L.E. also gratefully acknowledges NSF
support for the work at Clemson.
tert-butyl-phenyl)porphyrin 6. A solution of dipyrro-
methane 6 (3.00 g, 9.0 mmol), 3,5-di-tert-butylbenzalde-
hyde (0.979 g, 4.5 mmol), and 4-carboxymethyl
benzaldehyde (0.736 g, 4.5 mmol) in dry CH Cl (900 mL)
2
2
was stirred at room temperature under Ar. Trifluoroacetic
acid (0.69 mL, 16 mmol) was added dropwise and the
mixture was stirred an additional 1 h. Excess 2,3-dichloro-
5
,6-dicyano-1,4-benzoquinone (DDQ) (3 g) was added and
References and notes
the solution was stirred open to the air for an additional 2 h.
The solvent was removed under reduced pressure and the
product was purified via flash chromatography (17%) (6:4
1. (a) Imahori, H.; Sakata, Y. Adv. Mater. 1997, 9, 537. (b) Prato,
M. J. Mater. Chem. 1997, 7, 1097. (c) Martin, N.; S a´ nchez, L.;
Illescas, B.; P e´ rez, I. Chem. Rev. 1998, 98, 2527. (d) Imahori,
H.; Sakata, Y. Eur. J. Org. Chem. 1999, 2445. (e) Diederich,
F.; G o´ mez-L o´ pez, M. Chem. Soc. Rev. 1999, 28, 263. (f)
Guldi, D. M. Chem. Commun. 2000, 321. (g) Reed, C. A.;
Bolskar, R. D. Chem. Rev. 2000, 100, 1075. (h) Gust, D.;
Hexanes/EtOAc, R Z0.5). The porphyrin ester 8 was
f
1
characterized by H NMR and MALDI-MS analysis.
23
A solution of 8 (250 mg) in isopropanol was heated at
reflux. A solution of KOH (1.4 g) in water (12.5 mL)

1936
D. I. Schuster et al. / Tetrahedron 62 (2006) 1928–1936
Moore, T. A.; Moore, A. L. J. Photochem. Photobiol., B 2000,
8, 63. (i) Gust, D.; Moore, T. A.; Moore, A. L. Acc. Chem.
Res. 2001, 34, 40. (j) Guldi, D. M.; Martin, N. J. Mater. Chem.
11. Schuster, D. I.; Jarowski, P. D.; Kirschner, A. N.; Wilson, S. R.
J. Mater. Chem. 2002, 12, 2041.
5
12. Bourgeois, J.-P.; Diederich, F.; Echegoyen, L.; Nierengarten,
J.-F. Helv. Chim. Acta 1998, 81, 1835.
2002, 12, 1978–1992. (k) Guldi, D. M. Chem. Soc. Rev. 2002,
3
1, 22. (l) Imahori, H.; Mori, Y.; Matano, Y. J. Photochem.
13. Schuster, D. I.; Cheng, P.; Jarowski, P. D.; Guldi, D. M.;
Luo, C.; Echegoyen, L.; Pyo, S.; Holzwarth, A. R.; Braslavsky,
S. E.; Williams, R. M.; Klihm, G. J. Am. Chem. Soc. 2004, 126,
Photobiol., C 2003, 4, 51. (m) Nierengarten, J. F. Top. Curr.
Chem. 2003, 228, 87. (n) Guldi, D. M. Pure Appl. Chem. 2003,
7
257.
75, 1069. (o) El-Khouly, M. E.; Ito, O.; D’Souza, F.
1
1
4. Ogilby, P. R.; Foote, C. S. J. Am. Chem. Soc. 1982, 104, 2069.
5. Arbogast, J. W.; Darmanyan, A. P.; Foote, C. S.; Rubin, Y.;
Diederich, F. N.; Alvarez, M. M.; Anz, S. J.; Whetten, R. L.
J. Phys. Chem. 1991, 95, 11. Bensasson, R. V.; Bienvenue, E.;
Dellinger, M.; Leach, S.; Seta, P. J. Phys. Chem. 1994, 98,
J. Photochem. Photobiol., C 2004, 5, 79. (p) Guldi, D. M.;
Prato, M. Chem. Commun. 2004, 2517.
. Xie, Q.; Perez-Cordero, E.; Echegoyen, L. J. Am. Chem. Soc.
2
1992, 114, 3978.
3
. Imahori, H.; Hagiwara, K.; Akiyama, T.; Aoki, M.; Taniguchi,
3
492.
S.; Okada, T.; Shirakawa, M.; Sakata, Y. Chem. Phys. Lett.
1
1
6. For a review, see: Bracher, P. J.; Schuster, D. I. In Fullerenes:
From Synthesis to Optoelectronic Properties; Guldi, D. M.,
Martin, N., Eds.; Kluwer Academic: Dordrecht, 2002;
pp 163–212.
1
996, 263, 545.
4
. (a) Guldi, D. M.; Luo, C.; Prato, M.; Dietel, E.; Hirsch, A.
Chem. Commun. 2000, 373. (b) Imahori, H.; Tamaki, K.;
Guldi, D. M.; Luo, C.; Fujitsuka, M.; Fukuzumi, S. J. Am.
Chem. Soc. 2001, 123, 2607. (c) Imahori, H.; Yamada, H.;
Guldi, D. M.; Endo, Y.; Shimomura, A.; Kundu, S.; Yamada,
K.; Okada, T.; Sakata, Y.; Fukuzumi, S. Angew.Chem., Int. Ed.
7. Kuciauskas, D.; Lin, S.; Seely, G. R.; Moore, A. L.; Moore,
T. A.; Gust, D.; Drovetskaya, T.; Reed, C. A.; Boyd, P. D. W.
J. Phys. Chem. 1996, 100, 15926–15932.
1
1
8. Galili, T.; Regev, A.; Levanon, H.; Schuster, D. I.; Guldi,
D. M. J. Phys. Chem. A 2004, 108, 10632–10639.
2002, 41, 2344. (d) Ohkubo, K.; Imahori, H.; Shao, J. G.; Ou,
Z. P.; Kadish, K. M.; Chen, Y. H.; Zheng, G.; Pandey, R. K.;
9. Armaroli, N.; Accorsi, G.; Felder, D.; Nierengarten, J.-F.
Chem. Eur. J. 2002, 8, 2314–2323.
Fujitsuka, M.; Ito, O.; Fukuzumi, S. J. Phys. Chem. A 2002,
106, 10991. (e) Fukuzumi, S.; Ohkubo, K.; Imahori, H.; Guldi,
2
0. Guldi, D. M.; Prato, M. Acc. Chem. Res. 2000, 33, 695.
1. (a) Guldi, D. M.; Luo, C.; Da Ros, T.; Prato, M.; Dietel, E.;
Hirsch, A. Chem. Commun. 2000, 375. (b) Luo, C.; Guldi,
D. M.; Imahori, H.; Tamaki, K.; Sakata, Y. J. Am. Chem. Soc.
D. M. Chem. Eur. J. 2003, 9, 1585.
2
5
6
. (a) Lindsey, J. S.; Schreiman, I. C.; Hsu, H. C.; Kearney, P. C.;
Marguerettaz, A. M. J. Org. Chem. 1987, 52, 827. (b) Lee,
C. H.; Lindsey, J. S. Tetrahedron 1994, 50, 11427. (c) Littler,
B. J.; Ciringh, Y.; Lindsey, J. S. J. Org. Chem. 1999, 64, 2864.
2
000, 122, 6535. (c) Guldi, D. M.; Luo, C.; Prato, M.; Dietel,
E.; Hirsch, A. Chem. Commun. 2000, 373. (d) Imahori, H.;
Tamaki, K.; Guldi, D. M.; Luo, C.; Fujitsuka, M.; Ito, O.;
Sakata, Y.; Fukuzumi, S. J. Am. Chem. Soc. 2001, 123, 2607.
(
d) Smith, K. M. J. Porphyrins Phthalocyanines 2000, 4, 319.
. Boyd, P. D. W.; Hodgson, M. C.; Rickard, C. E. F.; Oliver,
A. G.; Brothers, P. J.; Chaker, L.; Bolskar, R. D.; Tham, F. S.;
Reed, C. A. J. Am. Chem. Soc. 1999, 121, 10487.
(
e) Imahori, H.; Guldi, D. M.; Tamaki, K.; Yoshida, Y.; Luo, C.;
Sakata, Y.; Fukuzumi, S. J. Am. Chem. Soc. 2001, 123, 6617. (f)
Imahori, H.; Yamada, H.; Guldi, D. M.; Endo, Y.; Shimomura,
A.; Kundu, S.; Yamada, K.; Okada, T.; Sakata, Y.; Fukuzumi, S.
Angew. Chem., Int. Ed. 2002, 41, 2344. (g) Guldi, D. M.;
Scheloske, M.; Dietel, E.; Hirsch, A.; Troisi, A.; Zerbetto, F.;
Prato, M. Chem. Eur. J. 2003, 9, 4968. (h) Balbinot, D.; Atalick,
S.; Guldi, D. M.; Hatzimarinaki, M.; Hirsch, A.; Jux, N. J. Phys.
Chem. 2003, 107, 13273.
7
. Imahori, H.; Guldi, D. M.; Tamaki, K.; Yoshida, Y.; Luo, C.;
Sakata, Y.; Fukuzumi, S. J. Am. Chem. Soc. 2001, 123, 6617.
. Baran, P. S.; Monaco, R. R.; Khan, A. U.; Schuster, D. I.;
Wilson, S. R. J. Am. Chem. Soc. 1997, 119, 8363. Schuster,
D. I. Carbon 2000, 38, 1607.
8
9
. Nierengarten, J.-F.; Herrmann, A.; Tykwinski, R. R.;
Ruttimann, M.; Diederich, F.; Boudon, C.; Gisselbrecht,
J. P.; Gross, M. Helv. Chim. Acta 1997, 80, 293.
22. Guldi, D. M. Chem. Soc. Rev. 2002, 31, 22.
23. For details, see: MacMahon, S. Ph.D. Dissertation, New York
University, 2004.
1
0. (a) MacMahon, S.; Wilson, S. R.; Schuster, D. I. Proc.
Electrochem. Soc. 2000, 8, 155. (b) MacMahon, S.; Fong, R.;
Baran, P. S.; Safonov, I.; Wilson, S. R.; Schuster, D. I. J. Org.
Chem. 2001, 66, 5449.
24. Armaroli, N.; Marconk, G.; Echegoyen, L.; Bourgeois, J.-P.;
Diederich, F. Chem. Eur. J. 2000, 6, 1629–1645.