Dendritic porphyrin-fullerene conjugates: Efficient light-harvesting and charge-transfer events

Article abstract of DOI:10.1002/chem.200902161

A novel dendritic C₆₀-H₂P(ZnP)₃ (P = porphyrin) conjugate gives rise to the successful mimicry of the primary events in photosynthesis, that is, light harvesting, unidirectional energy transfer, charge transfer, and charge

Full text of DOI:10.1002/chem.200902161

FULL PAPER
DOI: 10.1002/chem.200902161
Dendritic Porphyrin–Fullerene Conjugates: Efficient Light-Harvesting and
Charge-Transfer Events
Sebastian Schlundt,[a] Gregory Kuzmanich,[b] Fabian Spꢀnig,[c]
Gustavo de Miguel Rojas,[c] Christian Kovacs,[a] Miguel A. Garcia-Garibay,*[b]
Dirk. M. Guldi,*[c] and Andreas Hirsch*[a]
Abstract: A novel dendritic C60-H2P-
(ZnP)3 (P=porphyrin) conjugate gives
rise to the successful mimicry of the
primary events in photosynthesis, that
is, light harvesting, unidirectional
energy transfer, charge transfer, and
charge-shift reactions. Owing, however,
to the flexibility of the linkers that con-
nect the C60, H2P, and ZnP units, the
outcome depends strongly on the ri-
gidity/viscosity of the environment. In
an agar matrix or Triton X-100, time-
resolved transient absorption spectro-
scopic analysis and fluorescence-life-
time measurements confirm the follow-
ing sequence. Initially, light harvesting
is seen by the peripheral C60-H2P-
*
G
ZnP moieties, and for which lifetimes
of up to 460 ns are found. On the other
hand, investigations in organic media
(i.e., toluene, THF, and benzonitrile)
reveal a short cut, that is, the peripher-
al ZnP unit reacts directly with C60 to
a unidirectional energy transfer funnels
the singlet excited-state energy to H2P
to form C60-*ACTHNUGTRNE(UNG H2P)-(ZnP)3, which
powers an intramolecular charge trans-
fer that oxidizes the photoexcited H2P
and reduces the adjacent C60 species. In
Cꢀ
form (C60) -H2P-(ZnP)3 +. Substantial
C
the correspondingly formed (C60) -
configurational rearrangements—
(H2P) +-(ZnP)3 conjugate, an intramo-
placing ZnP and C60 in proximity to
each other—are, however, necessary to
ensure the required through space in-
teractions (i.e., close approach). Conse-
lecular charge-shift reaction generates
(C60) -H2P-(ZnP)3 +, in which the radi-
cal cation resides on one of the three
Cꢀ
quently, the lifetime of (C60) -H2P-
+
C
(ZnP)3 is as short as 100 ps in benzo-
nitrile.
Introduction
[a] S. Schlundt, C. Kovacs, Prof. Dr. A. Hirsch
Department of Chemistry and Pharmacy and
Interdisciplinary Center of Molecular Materials (ICMM)
Friedrich-Alexander-Universitꢀt Erlangen-Nꢁrnberg
Henkestrasse 42, 91054-Erlangen (Germany)
Fax : (+49)9131-852-6864
Light conversion in plants and photosynthetic bacteria is
one of the fundamental processes that governs life on earth.
In fact, energy in the form of sunlight is harvested to power
the processes of life and to build up organic material. The
bacterial photosynthetic reaction center provides meaningful
incentives for the optimization of charge-separation process-
es in artificial model systems.[1,2] Common to all these sys-
tems is a relay of short-range energy/electron-transfer reac-
tions that evolve among chlorophyll and quinone moieties.
Importantly, the primary electron-transfer processes of pho-
tosynthesis are characterized by an extremely small reorgan-
ization energy (0.2 eV) attained by the transmembrane pro-
tein environment. Owing to the importance and complexity
of natural photosynthesis, the study thereof necessitates suit-
able simpler models. The ultimate goal is to design and as-
semble synthetic systems that can efficiently convert solar
energy into useful chemical energy.[3–6] Porphyrinoids, espe-
cially metalloporphyrinoid systems with their rich and exten-
E-mail: andreas.hirsch@chemie.uni-erlangen.de
[b] G. Kuzmanich, Prof. Dr. M. A. Garcia-Garibay
Department of Chemistry and Biochemistry
University of California
Los Angeles, CA 90024-1569 (USA)
Fax : (+1)310-825-0767
E-mail: mgg@chem.ucla.edu
[c] F. Spꢀnig, Dr. G. de Miguel Rojas, Prof. Dr. D. M. Guldi
Department of Chemistry and Pharmacy and
Interdisciplinary Center of Molecular Materials (ICMM)
Friedrich-Alexander-Universitꢀt Erlangen-Nꢁrnberg
Egerlandstrasse 3, 91058-Erlangen (Germany)
Fax : (+49)9131-852-7340
E-mail: dirk.guldi@chemie.uni-erlangen.de
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.200902161.
Chem. Eur. J. 2009, 15, 12223 – 12233
ꢂ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
12223

sive absorptions throughout the visible region of the solar
spectrum, hold particularly great promise as integrative
building blocks with increased absorptive cross sections.
This finding led to the establishment of synthetic methodol-
ogies that allowed porphyrins and metalloporphyrins—as
excited-state electron donors—to be linked to electron ac-
ceptors, such as fullerenes.[7–16] Since the initial discovery of
fullerenes, chemists and physicists worldwide have studied
their solid-state properties, ranging from superconductivity
and nanostructured devices to endohedral fullerene chemis-
try. The three-dimensional, spherical structure of fullerenes,
which are made of alternating hexagons (i.e., electron rich)
and pentagons (i.e., electron deficient) with diameters start-
ing at 7.8 ꢃ for C60, evoked a lively interest in relating their
properties to conventional two-dimensional p systems. Their
extraordinary electron-acceptor properties—predicted theo-
retically and confirmed experimentally—have resulted in
noteworthy advances in the areas of light-induced electron-
transfer chemistry and solar-energy conversion.[17,18] It is
mainly the small reorganization energy that fullerenes ex-
hibit in electron-transfer reactions which is accountable for
this breakthrough. In particular, ultrafast charge-separation
and very slow charge-recombination features lead to unpre-
cedented long-lived radical ion pair states formed with high
quantum efficiencies.[19] Only, through the design and devel-
opment of light harvesting, photoconversion, and catalytic
modules capable of self-ordering and self-assembling into an
integrated functional unit will it be possible to realize effi-
cient artificial photosynthetic systems. Should charges, how-
ever, be transported over distances larger than 20 ꢃ alterna-
tive concepts have to be applied. This task becomes particu-
larly relevant when achieving ultralong radical-ion-pair life-
times. A viable option to achieve long-distance, charge-sepa-
rated states makes use of a relay of several short-range
electron-transfer steps along well-designed redox gradients
rather than forcing the electron transfer through a single,
concerted long-range step. Such a “relay concept” has been
successfully realized by combining several redox active
building blocks. Herein, we focused on ZnP, H2P, and C60
species (P=porphyrin). This approach is expected to lead to
a scenario in which the final electron donors (i.e., ZnP) and
tures of the resulting core-shell ensembles. Ultimately, this
approach may lead to high solar-energy conversion efficien-
cies. In fact, we demonstrate that C60-H2P-(ZnP)3 shows the
two basic functions of photosynthesis, namely, light harvest-
ing and charge transfer.
Results and Discussion
Synthesis: Porphyrin 5[21] (Scheme 1) was synthesized by a
statistical condensation of aldehyde 3[22] (3 equiv), aldehyde
4[23] (1 equiv), and pyrrole (4 equiv) under conditions devel-
oped by Lindsey and co-workers[24] in CH2Cl2. Boron tri-
fluoride etherate was used as a Lewis acid catalyst, ethanol
and PPh4Cl as co-catalysts, and 2,3-dichloro-5,6-dicyano-1,4-
benzoquinone (DDQ) as an oxidizing agent. The product
was purified by column chromatography. Cyclopropanation
of C60 was achieved with a modification of the conditions
developed by Bingel.[25] Porphyrin 5,[21] C60, and iodine were
dissolved in toluene, and diluted 1,8-diazabicyclo-
AHCTUNGTREG[NNNU 5.4.0]undec-7-ene (DBU) was added slowly. After two days
of stirring, compound 6[21] was purified by column chroma-
tography (silica, toluene/EtOAc 9:1) and precipitated in
the primary electron acceptors (i.e., C60) form a spatially
+
Cꢀ
C
separated radical ion pair (C60) -H2P-(ZnP)3
.
Light harvesting is another topic that needs to be ad-
dressed in designing artificial photosynthetic systems that
will ultimately power the practical production of solar
fuels.[20] Three light-harvesting complexes LH1, LH2, and
LH3 are found in bacteria. These complexes consist of por-
phyrin arrays that are embedded in a protein matrix that
can harvest light over a broad range of the solar spectrum
and funnel its excited-state energy to the reaction center.
Artificial light harvesting can be achieved by means of den-
dritic architectures attached to C60, which renders these sys-
tems particularly appealing for photoconversion processes.
Placing four chromophores, that is, three ZnP and one H2P
units, in C60-H2P-(ZnP)3 at the fullerene periphery enhances,
for example, the light-collection and light-transduction fea-
Scheme 1. Synthesis of porphyrin 5[21] and dyad 7. i) NaOH, H2O, reflux
(68%); ii) NaOH, H2O, CH2Cl2, tetrabutylammonium bromide; iii) pyri-
dine, CH2Cl2, 08C; iv) pyrrole, boron trifluoride etherate, PPh4Cl, etha-
nol, DDQ, CH2Cl2; v) iodine, DBU, C60, toluene; vi) formic acid. DDQ=
2,3-dichloro-5,6-dicyano-1,4-benzoquinone,
AHCTUNGTREG[NNUN 5.4.0]undec-7-ene.
DBU=1,8-diazabicyclo-
12224
www.chemeurj.org
ꢂ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 12223 – 12233

Charge Transfer in Porphyrin–Fullerene Dendrons
FULL PAPER
pentane. Subsequently, the tert-butyl protecting groups
could be quantitatively removed in formic acid to afford 7,
carbonyl groups are identified by only one set of signals
with two peaks at d=168.82 and 168.78 ppm. Similar
changes were observed for all the other carbon atoms of 11.
For comparison, the dendritic structure H2P-(ZnP)3 12
was synthesized analogously. Therefore, porphyrin 13 was
1
which was characterized by H NMR, 13C NMR, and UV/Vis
spectroscopy and MALDI-TOF mass-spectrometric analysis.
Compound 7 is one of the major building blocks for the syn-
thesis of dendritic porphyrin–fullerene hybrid 11.
Porphyrin 9[26] was also synthesized statistically (by using
pyrrole, boron trifluoride etherate, PPh4Cl, ethanol, DDQ,
and CH2Cl2) under the conditions developed by Lindsey
and co-workers.[24] Subsequently, 10[26] was obtained from its
free-base precursor 9[26] with fivefold excess of Zn-
ACHTUNGTRENNUNG(OAc)2·2H2O heated to reflux in THF for 3 hours
(Scheme 2).
obtained from 5[21] by treatment
with formic acid (Scheme 4).
An additional reference fuller-
ene derivative 14[27] was synthe-
sized according to known pro-
tocols for the following photo-
physical investigations.
Scheme 2. Synthesis of porphyrin 10.[26] i) Pyrrole, boron trifluoride ether-
ate, PPh4Cl, ethanol, DDQ, CH2Cl2; ii) ZnACTHNUTRGNE(UNG OAc)2·2H2O, THF, reflux.
The two building blocks 7 and 10[26] were coupled in THF
using N,N’-dicyclohexylcarbodiimide (DCC) as the coupling
agent (Scheme 3). DCC and porphyrin 10[26] were added
subsequently to a solution of 7,4-dimethylaminopyridine
(DMAP) and 1-hydroxybenzotriazole hydrate (HOBT) dis-
solved in THF at 08C. After stirring at room temperature
for six days in the dark, additional DCC was added at 08C.
Precipitated dicyclohexylurea (DCU) was removed after an-
other day of stirring at room temperature. The product was
purified by column chromatography (silica gel, CH2Cl2/
EtOAc 20:0.75, 1% NEt3) and recrystallization (CH2Cl2/
pentane). The porphyrin–fullerene hybrid 11 was character-
Steady-state absorption spectroscopy: To shed light on the
electronic interactions between the photo- and redox-active
components (i.e., ZnP, H2P, and C60), the ground-state ab-
sorption spectra of the donor–acceptor ensembles were
compared in THF with those of the individual components.
At first glance, the absorption spectra of the ensembles
might be best described as simple superimpositions of the
component spectra that lack notable perturbations or addi-
tional charge-transfer transitions. In particular, the dominat-
ing Soret and Q bands of the porphyrins are seen in the re-
gions l=400–450 and 550–650 nm, respectively. The stron-
gest bands of the C60 units, on the other hand, are at l=220,
265, and 330 nm and extend all the way to the energetically
lowest lying transition at l=690 nm. Nevertheless, a closer
look reveals subtle changes; for example, the porphyrin
Soret and Q bands in 6 are bathochromically shifted by
about Dl=6 nm relative to 5. Moreover, 11 gives rise to
only one intense Soret band at l=425 nm.
1
ized by H NMR, 13C NMR, UV/Vis, and IR spectroscopy,
and MALDI-TOF and ESI mass spectrometry. Because the
NMR spectra were recorded in CDCl3 at room temperature,
signals were seen that were split into multiplets. For exam-
ple, the three carbonyl groups of the linking esters gave rise
to a series of nine peaks in the 13C NMR spectrum between
d=168.63 and 168.00 ppm. A feasible rationale implies the
presence of different conformers based on p–p stacking of
C60/ZnP. Independent confirmation of this assumption came
from 1H and 13C NMR spectra at 1008C in
[D2]tetrachloroethane. At this temperature, the structures
formed through p–p interactions are likely to be broken up.
In line with such structural arrangements, the three linker
Electrochemistry: The electrochemical properties of all the
investigated compounds were studied by cyclic voltammetry
and differential pulse voltammetry (DPV) in dichlorome-
Chem. Eur. J. 2009, 15, 12223 – 12233
ꢂ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
12225

D. M. Guldi, A. Hirsch, M. A. Garcia-Garibay et al.
Scheme 4. Porphyrin building block 13. i) Formic acid, overnight.
thane (Table 1). For 5, the first and second reversible reduc-
tions are seen at ꢀ1.66 and ꢀ1.96 V, respectively. Reversible
oxidation, on the other hand, occurs at +0.58 and +0.91 V.
Both the reduction and oxidation are affected in 10 with
values of ꢀ1.91 and +0.35/+0.75 V, respectively.
In 12, the reversible reduction of H2P is shifted to
ꢀ1.74 V, while that of ZnP is virtually identical to the refer-
ence. Overall, the 3:1 ratio is discernable when comparing
the areas of the integrated peaks. On the oxidative side, pro-
cesses at +0.35, +0.60, +0.68, and +1.00 V correspond to
the first oxidation of ZnP, the first oxidation of H2P, the
second oxidation of ZnP, and the second oxidation of H2P,
respectively.
Turning to the oxidation of 6, only H2P-centered processes
emerge at +0.55 and +1.05 V. This behavior infers only
slight changes relative to the H2P reference. In stark con-
trast, the reduction of 6 involves C60-centered processes at
ꢀ1.12, ꢀ1.47, and ꢀ2.04 V as well as H2P-centered processes
at ꢀ1.67 and ꢀ2.04 V.
With this information in hand, the redox features of 11
evolve as the sum of the C60, H2P, and ZnP redox properties.
Again, oxidation is limited to H2P and ZnP with values at
+0.53/+1.02 and +0.30/+0.65 V, respectively. For the re-
duction, the following values were derived: ꢀ1.17, ꢀ1.50,
and ꢀ2.06 V (C60); ꢀ1.70 V (H2P); ꢀ1.91 V (ZnP).
Scheme 3. Synthesis of dendritic porphyrin–fullerene hybrid 11. i) DMAP,
HOBT, DCC, THF. DCC=N,N’-dicyclohexylcarbodiimide, DMAP=7,4-
dimethylaminopyridine, HOBT=1-hydroxybenzotriazole hydrate.
Steady-state fluorescence spectroscopy: The fluorescence
spectra of all the reference compounds and all the donor–
acceptor conjugates were recorded in solvents of different
Table 1. Electrochemical features of the different donor–acceptor conjugates and the corresponding reference systems recorded in dichloromethane
versus ferrocene/ferrocenium.
2ꢀ
H2P +/H2P2+
H2P/H2P
0.58
ZnP +/ZnP2+
ZnP/ZnP
C60/C60
C60 /C60
C602ꢀ/C60
H2P/H2P
ZnP/ZnP
+
+
C
C
C
C
Cꢀ
Cꢀ
C3ꢀ
Cꢀ
Cꢀ
5
10
12
14
6
0.91
ꢀ1.66
0.75
0.68
0.35
ꢀ1.91
ꢀ1.91
1.00
0.60
0.35[a]
ꢀ1.74
ꢀ1.13
ꢀ1.12
ꢀ1.17
ꢀ1.48
ꢀ1.47
ꢀ1.50
ꢀ2.03
ꢀ2.04
ꢀ2.06
1.05
1.02
0.55
0.53
ꢀ1.67
ꢀ1.70
11
0.65[a]
0.30
ꢀ1.91[a]
[a] Taken from DPV.
12226
www.chemeurj.org
ꢂ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 12223 – 12233

Charge Transfer in Porphyrin–Fullerene Dendrons
FULL PAPER
polarity—toluene, THF, and benzonitrile—and in the ab-
sence of molecular oxygen (Table 2). All porphyrins con-
taining conjugates were excited at l=420 nm. Moreover, se-
lective excitation of ZnP was achieved at l=560 nm, where-
as H2P was excited at l=647 nm.
below our detection limit, were estimated by using Equa-
tion (1).
FReference ꢀ FCompound
k ¼
ð1Þ
tReferenceFCompound
Femtosecond flash photolysis:
Ultrafast techniques were em-
ployed to characterize the tran-
sient intermediates involved in
the excitation of 5, 10, 14, 12, 6,
and 11. In particular, transient
absorption spectroscopic meas-
urements were carried out fol-
lowing femtosecond laser exci-
tation at either l=387 or
420 nm in solvents of different
polarity. The wavelength l=
387 nm was used to photoexcite
C60 nearly exclusively, whereas
excitation at l=420 nm guaran-
tees the excitation of ZnP or
Table 2. Fluorescence features of the different donor–acceptor conjugates and the corresponding reference
systems recorded in solvents of different polarity.
Toluene
Fem [ꢄ10ꢀ2
THF
Fem [ꢄ10ꢀ2
Benzonitrile
Fem [ꢄ10ꢀ2
tem [ns]
E0ꢀ0 [eV]
lem [nm]
]
tem [ns]
]
tem [ns]
]
5
10
1.89
2.04
2.04
1.89
1.79
1.89
2.04
1.89
649, 715
600, 652
600, 652
649, 715
720
649, 715
600, 652
649, 715
11.0
4.0
9.8
2.4
0.21
5.40
11.0
4.0
9.8
2.4
0.22
5.76
11.0
4.0
9.8
2.4
0.22
5.97
0.30[a]
0.21[a]
0.29[a]
12
5.95[a]
0.06
6.11[a]
0.06
6.63[a]
0.06
14
6
1.8
1.8
1.8
0.22
0.200[b]
0.048[b]
0.054[b]
0.20
0.181[b]
0.036[b]
0.054[b]
0.17
0.154[b]
0.030[b]
0.054[b]
0.08[a]
0.09[a]
0.06[a]
0.09[a]
0.05[a]
0.09[a]
11
[a] Determined based on the emission bands at l=600 and 715 nm for ZnP and H2P, respectively. [b] Estimat-
ed from the fluorescence quenching.
Porphyrin derivative 5 exhibits the well-known emission
features of analogous compounds of its class, with fluores-
cence maxima at l=659 and 715 nm, a fluorescence quan-
tum yield F of 0.11, and a fluorescence lifetime t of
9.8 ns.[28–30] Similarly, the fluorescence of 10—lmax =600 and
652 nm, F=0.04, and t=2.4 ns—resembles previously re-
ported values.[23,31]
Upon the excitation of 12, four emission bands at l=649
and 715 nm for H2P and l=600 and 652 nm for ZnP are dis-
cernable in the fluorescence spectra. Nevertheless, the coa-
lescence of the peaks at l=649 and 652 nm (i.e., H2P and
ZnP, respectively) is notable. The ZnP fluorescence is
quenched by a factor of approximately 20, which corre-
sponds to a lifetime of about 220 ps. Less pronounced is the
impact in 12 on the H2P fluorescence in which only a
quenching of 50% is seen with a lifetime of about 5.76 ns in
THF. Changing the solvent polarity from THF to either tol-
uene or benzonitrile exerted a negligible impact on the fluo-
rescence quantum yields and/or the fluorescence lifetimes
(Table 2).
In the C60-containing donor–acceptor conjugates 6 and 11,
the H2P fluorescence is quantitatively quenched (i.e., >55-
fold) without, however, affecting the overall shape of the
fluorescence spectrum. In particular, the H2P fluorescence
quantum yields were in the range 2.0ꢄ10ꢀ3–0.8ꢄ10ꢀ3. Simi-
larly, the ZnP unit in 11 has a negligible emission. For 6 and
11, the fluorescence quantum yield and the fluorescence life-
time are subject to only a slight solvent dependence. No C60
fluorescence, which tends to be intrinsically weak (i.e., 6.0ꢄ
10ꢀ4) even in the C60 reference, was detectable at l=
720 nm. A likely rationale implies that much stronger H2P
and ZnP fluorescence dominates this spectral region.
Table 2 summarizes the excited-state features of the
donor–acceptor ensembles and the corresponding reference
systems in solvents of varying polarity. Lifetimes, close to or
H2P in the region of their Soret bands.
The photophysics of 14 is summarized as follows: the sin-
glet excited state gives rise to a singlet–singlet absorption
that maximizes around l=920 nm with a lowest vibronic
state of 1.79 eV. Once generated, this state is subject to a
rapid and quantitative intersystem crossing process to yield
the energetically lower lying triplet excited state
(1.50 eV).[32] The intersystem crossing takes place with a rate
constant of k=7.4ꢄ108 sꢀ1 and is governed by large spin–
orbit coupling.[33] However, not only the rate is remarkable,
the overall efficiency of the triplet formation is close to
unity. The corresponding triplet excited-state features
maxima in the visible region at l=360 and 720 nm (see Fig-
ure S1 in the Supporting Information).
Quite different are the differential absorption changes as-
sociated with the excitation of 5. The visible region is domi-
nated by intense H2P-centered transitions that embrace
minima due to ground-state bleaching at l=516, 549, 590,
and 648 nm and maxima at l=450, 532, 567, 620, 674, and
1050 nm. Upon formation (i.e., >2 ps), intersystem crossing
governs the photophysics of the singlet excited state
(1.90 eV) of 5 over the course of approximately 10 ns. Simi-
lar to the H2P singlet excited state, the ZnP singlet excited
state forms within 1 ps. The singlet excited state displays a
minimum—due to Q-band bleaching—at l=550 nm and
maxima between l=570 and 740 nm and 900 and 1000 nm.
Within 2.4 ns, the singlet excited state (2.04 eV) converts
into the lower-lying triplet excited state (1.53 eV) through
an efficient intersystem crossing in all measured sol-
vents.[23,31,34]
The differential absorption changes that were recorded
upon excitation of 12 (Figure 1) are dominated by the ZnP
singlet excited-state features, whereas any appreciable con-
tributions from the H2P singlet excited state are insignifi-
cant. Within the next 140 ps, the ZnP singlet excited-state
Chem. Eur. J. 2009, 15, 12223 – 12233
ꢂ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
12227

D. M. Guldi, A. Hirsch, M. A. Garcia-Garibay et al.
Figure 2. Upper: differential absorption spectrum (visible and near-IR
region) obtained upon femtosecond flash photolysis (l=387 nm) of 6
(10ꢀ4 m) in argon-saturated THF with a time delay of 50 ps at room tem-
perature. Lower: time-absorption profile of the spectra shown above l=
1040 nm, in which the charge separation and charge recombination were
monitored.
Figure 1. Upper: differential absorption spectra (visible and near-IR
region) obtained upon femtosecond flash photolysis (l=387 nm) of 12
(10ꢀ4 m) in argon-saturated THF with time delays of 2 ps (filled circles)
and 100 ps (crosses) at room temperature. Lower: time-absorption profile
of the spectra shown above l=553 nm in which the energy transfer and
the singlet-to-triplet intersystem crossing were monitored.
(2.04 eV) features decay and simultaneously we see the H2P
singlet excited-state (1.9 eV) characteristics being formed.
Implicit is a transduction of singlet excited-state energy
driven along the energy gradient, the product of which starts
to decay slowly on the timescale of the instrument, that is,
3 ns. It is notable that there is good agreement between the
kinetic data derived from the transient absorption and the
fluorescence measurements. Intersystem crossing (>3.0 ns),
by which the H2P singlet excited state is transformed into
the corresponding triplet manifold, commences after the
energy transfer.
The excited-state features of 6 (Figure 2) are dominated
in the visible region by H2P-centered transitions, whereas
the excited-state properties of C60 are observed in the near-
IR region. Instead of the slow intersystem crossing processes
that both C60 and H2P give rise to in the reference com-
pounds, the C60–H2P excited state decays rapidly. In particu-
lar, after about 10 ps the growth of a characteristic absorp-
tion band in the near-IR region is discernable. A maximum
at l=1035 nm is in perfect agreement with the fingerprint
absorption of the C60 radical anion.[35] Simultaneously with
the growth at l=1035 nm, a growth of a broad transient in
the visible region (i.e., between l=650 and 700 nm) occurs.
In comparison with the electrochemically and radiolytically
generated features of the H2P radical cation, we reach a
conclusion with the successfully generated radical-ion-pair
state (C60) -(H2P) +. With this information in hand, we de-
rived the charge-separation (i.e., 60 ps, 1.6ꢄ1010 sꢀ1) and
charge-recombination (i.e., 640 ps, 1.5ꢄ109 sꢀ1) kinetics in
THF. By increasing the solvent polarity, the driving forces
for charge separation and recombination increase and de-
crease, thus obtaining lifetimes of 38 and 334 ps (2.6ꢄ1010
and 2.9ꢄ109 sꢀ1, respectively) in benzonitrile.
Cꢀ
C
In contrast to 6, the photophysics of 11 follows a different
deactivation pathway (Figure 3). Immediately upon photo-
excitation in THF, the ZnP singlet excited-state features are
discernable. These features convert rapidly (k=2.7ꢄ1010 sꢀ1)
into a new transient. Interestingly, the new transient is char-
acteristic of the C60 radical anion that evolves in the near-IR
region at l=1035 nm. Concurrently, the corresponding ZnP
radical cation is registered in the visible range, namely, be-
tween l=580 and 800 nm. A spectral comparison with pre-
vious pulse radiolytic investigations on the one-electron oxi-
dation of ZnP confirm the assignment of visible maxima to
+
+
the ZnP ion.[23,31] The (C60) -H2P-(ZnP)3 charge-separat-
ed state recombines in THF unexpectedly quickly (k=2.7ꢄ
109 sꢀ1) and reinstates the singlet ground state (Table 3). The
charge-separation and recombination processes reveal a
C
Cꢀ
C
12228
www.chemeurj.org
ꢂ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 12223 – 12233

Charge Transfer in Porphyrin–Fullerene Dendrons
FULL PAPER
first scenario is based on thermodynamics, which suggest a
Cꢀ
larger driving force for the formation of (C60) -H2P-
(ZnP)3 than for the formation of (C60) -(H2P) +-(ZnP)3.
The second is based on the nature of the linkers, due to
their overall flexibility, which may facilitate a very close ap-
proach between the triad termini so that a direct charge sep-
aration between the C60 and ZnP units is feasible in a non-
linear assembly. In an effort to disrupt such potentially
folded conformations of the triad, the addition of pyridine
(i.e., 1% (v/v) in toluene) to coordinate the ZnP was inves-
tigated. The resulting spectrum was characterized by a red-
+
C
Cꢀ
C
Cꢀ
shift of 10 nm in absorption and a longer lived (C60) -H2P-
+
C
(ZnP)3
species (i.e., 790 vs. 489 ps (1.3ꢄ109 vs. 2.0ꢄ
109 sꢀ1)) in pure toluene. The extended lifetime suggests that
a strong perturbation of the distance and interactions be-
tween the C60 and ZnP units upon coordination of pyridine
to the latter. However, the possible involvement of the
+
C
H2P ion could not be confirmed.
Transient absorption measurements carried out after em-
bedding 11 in an agar matrix led to drastic changes
(Figure 4).[36] Under such experimental conditions, the ZnP
Figure 3. Upper: differential absorption spectrum (visible and near-IR
region) obtained upon femtosecond flash photolysis (l=387 nm) of 11
(10ꢀ4 m) in argon-saturated THF with a time delay of 50 ps at room tem-
perature. Lower: time-absorption profile of the spectra shown above l=
1040 nm, in which the charge separation and charge recombination were
monitored.
Table 3. Charge-separation and charge-recombination kinetics for 11 and
the corresponding driving forces[a] in different environments.
CS I[b]
[sꢀ1
1.7ꢄ1010
2.7ꢄ1010
8.3ꢄ1010
9.8ꢄ109
CS II[c]
[sꢀ1
CR
[sꢀ1
2.0ꢄ109
2.7ꢄ109
1.0ꢄ1010
1.3ꢄ109
ꢀDGoCS
ꢀDGoCR
Figure 4. Differential absorption spectra (visible and near-IR region) ob-
tained upon femtosecond flash photolysis (l=387 nm) of 11 (10ꢀ4 m) em-
bedded in agar with a time delay of 100 ps at room temperature.
G
]
E
]
T
]
[eV]
[eV]
toluene
THF
benzonitrile
toluene/pyridine
ꢀ0.11
0.57
0.79
2.15
1.47
1.25
ACHTUNGTRENNUNG(99:1 v/v)
singlet excited-state features decay through energy transfer
to afford the corresponding H2P singlet excited state. The
latter is then susceptible to charge transfer: the stable
agar[d]
1.5ꢄ109
9.3ꢄ109
1.0ꢄ107
2.1ꢄ106
Triton X-100[e]
[a] The driving forces were estimated by using the dielectric continuum
Cꢀ
C
(C60) -(H2P) +-(ZnP)3 radical-ion-pair state is seen within
657 ps (1.5ꢄ109 sꢀ1) on the femtosecond scale. Complemen-
tary nanosecond experiments disclose a radical ion pair
state with a lifetime of 100 ns (1.0ꢄ107 sꢀ1; Figure 5). How-
Cꢀ
model (see the Supporting Information for details). [b] (C60) -H2P-
+
(ZnP)3 . [c] (C60) -(H2P) +-(ZnP)3. [d] Agar (0.75 vol%) was added to
an aqueous buffer solution (pH 7) and heated until the agar dissolved
completely; after cooling to room temperature, a transparent jelly was
formed; the compound, dissolved in Triton X-100, was added into the so-
lution at 608C. [e] Degassed (Ar) Triton X-100.
C
Cꢀ
C
Cꢀ
ever, close examination indicates that it is not the (C60) -
(H2P) +-(ZnP)3 radical-ion-pair state that is observed in this
C
Cꢀ
case, but rather one corresponding to the (C60) -H2P-
weak solvent dependence with values of 59/489 ps (1.7ꢄ1010/
2.0ꢄ109 sꢀ1) and 12/100 ps (8.3ꢄ1010/1.0ꢄ1010 sꢀ1) in toluene
and benzonitrile, respectively. However, no spectroscopic/ki-
(ZnP)3 +. This result indicates a cascade of charge-transfer
C
+
Cꢀ
C
reactions that convert the initially formed (C60) -(H2P)
-
Cꢀ
C
(ZnP)3 into (C60) -H2P-(ZnP)3 +. The underlying charge-
shift reaction must have taken place on a timescale between
3 and 10 ns (i.e., the time resolution of the femtosecond and
nanosecond setups, respectively).
Cꢀ
netic evidence was found in favor of forming the (C60) -
(H2P) +-(ZnP)3 charge-separated state. Among several sce-
C
narios to explain this observation, two appear credible. The
Chem. Eur. J. 2009, 15, 12223 – 12233
ꢂ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
12229

D. M. Guldi, A. Hirsch, M. A. Garcia-Garibay et al.
Figure 5. Upper: differential absorption spectra (visible and near-IR
region) obtained upon nanosecond flash photolysis (l=355 nm) of 11
(10ꢀ4 m) embedded in agar with a time delay of 200 ns at room tempera-
ture. Lower: time-absorption profile of the spectra shown above l=
1040 nm, in which the charge recombination was monitored.
Figure 6. Upper: differential absorption spectrum (visible and near-IR
region) obtained upon femtosecond flash photolysis (l=387 nm) of 11 in
argon-saturated Triton X-100 with a time delay of 200 ps at room temper-
ature. Lower: time-absorption profile of the spectra shown above l=
475 nm, in which the charge separation and charge recombination were
monitored.
An even more efficient and longer lived charge separation
was observed in pure Triton X-100 (viscosity=240 cP;
Figure 6). Under such experimental conditions, the ZnP sin-
glet excited-state features decay through energy transfer to
afford the corresponding H2P singlet excited state. The
ering the ZnP and H2P singlet excited states, respectively.
Nevertheless, energy transfer outperforms the formation of
the radical-ion-pair state.
The low C60 reduction potential changes the deactivation
pathways drastically in 6. Taking, for example, the driving
force for charge separation, which increases in THF to
0.30 eV, the charge recombination is strongly exothermic
with 1.59 eV. Here, the recombination to form the triplet ex-
cited state is 0.19 eV more favorable than the direct recov-
ery of the ground state. When increasing the solvent polarity
to benzonitrile, the energy of the radical-ion-pair state
(1.43 eV) approached that of the H2P triplet excited state.
As a matter of fact, deactivation proceeds predominantly to
the ground state. Even in toluene, there was evidence for
charge separation, although the dielectric continuum model
suggests that this process should be endothermic, namely,
energetically uphill by 0.17 eV. It is likely that the calculated
value is an overestimation, especially considering that the
charge transfer proceeds through a nonlinear assembly and
that, in turn, a correct distance is difficult to approximate.
Turning to 11, several deactivation scenarios emerge.
Firstly, a unidirectional energy transfer by 0.15 eV, with
close resemblance to the aforementioned 12. Secondly,
charge separation occurred in the triad to form H2P +-C60
C
Cꢀ
out of the H2P singlet excited state within 107 ps. The recov-
ery of the ground state evolving from the intermediate
ZnP +-C60 charge-separated state takes place with k=2.1ꢄ
106 sꢀ1. Again, the charge migration was not observable due
to our instrumental response.
C
Cꢀ
Energetics: The photoexcited 12 deactivates through the fol-
lowing sequence of processes: the initially-formed ZnP
(2.04 eV) singlet excited state deactivates by an exothermic
energy transfer (0.15 eV) to the H2P singlet excited state
(1.9 eV). This process is followed by intersystem crossing to
the H2P triplet excited state (1.40 eV) and the subsequent
recovery of the ground state.
+
Cꢀ
C
The energy of the (H2P) –(ZnP)3 radical-ion-pair state
is 2.09 eV in THF, which makes it impossible to be formed
from any of the singlet excited states. Increasing the solvent
polarity also (e.g., with benzonitrile) increases the driving
force for charge separation to 0.19 and 0.04 eV when consid-
12230
www.chemeurj.org
ꢂ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 12223 – 12233

Charge Transfer in Porphyrin–Fullerene Dendrons
FULL PAPER
+
Cꢀ
C
,
direct charge separation to afford (C60) -H2P-(ZnP)3
cence lifetime measurements confirm the formation of
+
Cꢀ
C
which is exothermic by 0.57 eV. Both processes are fairly ef-
ficient and, thus, fluorescence or intersystem crossing, as in-
trinsic deactivation processes, are nearly quantitatively shut
down. The flexibility of the linker enables a competitive sce-
nario, in which the charge recombination surpasses the
energy transfer (Figure 7). From this point, the only feasible
(C60) -H2P-(ZnP)3 with lifetimes of 100 and 460 ns, re-
spectively. On the other hand, investigations in organic
media (i.e., toluene, THF, and benzonitrile) trigger substan-
tial configurational rearrangements, which place ZnP and
C60 units in proximity to each other. Consequently, the life-
+
Cꢀ
C
time of (C60) -H2P-(ZnP)3 is as short as 100 ps in benzoni-
trile. Implicit is that in Triton X-100/agar matrix a stretched
configuration seems to be favored for C60-H2P-(ZnP)3 by al-
tering the conformational equilibrium of the system, where-
as in organic media coiling up of C60-H2P-(ZnP)3 results
through space p–p/charge-transfer interactions between
ZnP and C60. An alternative rationale implies that in Triton
X-100/agar matrix the motions needed for ZnP quenching
by charge transfer are not competitive with those for the
corresponding energy-transfer reaction. Our current efforts
are directed to the implementation of a rigid spacer that
1) prevents structural rearrangement; 2) provides the neces-
sary electronic coupling to power light harvesting, unidirec-
tional energy transfer, charge transfer, and charge-shift reac-
tions; and 3) slows down the energy-wasting charge recom-
bination.
Experimental Section
General: Fullerene C60 was obtained from Hoechst AG/Aventis and sepa-
rated from higher fullerenes by plug filtration.[37] Solvents were purified
by distillation, or in the case of HPLC grade solvents, used as received.
All the solvents employed for the electrochemical, photochemical, and
photophysical investigations were purchased from chemical suppliers in
dry, spectroscopic grade and were used without further purification. TLC
analysis was carried out on Merck silica gel 60 F254. Column chromatogra-
phy was performed on Macherey–Nagel silica gel 60m (230–400 mesh,
0.04–0.063 mm). The FAB mass spectra were recorded on a Micromass
Zabspec FAB+ mode with 3-nitrobenzyl alcohol as the matrix. The
MALDI-TOF mass spectra were carried out on a Shimadzu Axima Con-
fidence (DCTB: trans-2-[3-(4-tert-butylphenyl)-2-methyl-2-propenylide-
ne]malononitrile, DITH: dithranol, SIN: sinaptic acid). The ESI mass
spectra were obtained on a Bruker Esquire 6000 spectrometer and the EI
mass spectra were recorded on a Varian MAT 311A spectrometer. The
NMR spectra were recorded on Jeol (Jeol JNM EX 400, Jeol JNM GX
400) and Bruker (Bruker Avance 300, Bruker Avance 400) spectrometers.
The measurements were carried out in CDCl3 and [D8]THF, which were
Figure 7. Energy diagram that illustrates the deactivation pathways in
photoexcited 11 in agar/Triton X-100 (i.e., energy transfer from C60-H2P-
Cꢀ
C
*ACHTUNGTRENNUNG
(ZnP)3 to C60-*H2P-(ZnP)3 (ET), formation of (C60) -(H2P) +-(ZnP)3
+
Cꢀ
(CS-I), followed by a charge-shift reaction to form (C60) -H2P-(ZnP)3
C
(CS), and charge recombination to the ground state (CR-2)) and in, for
+
Cꢀ
C
example, THF (i.e., direct formation of (C60) -H2P-(ZnP)3 (CS-II) and
charge recombination to the ground state (CR-II)). Charge recombina-
Cꢀ
C
tion (CR-I) from the intermediately formed (C60) -(H2P) +-(ZnP)3 has
also been added.
process is the recovery of the ground state. By enforcing a
more linear configuration of 11—through the choice of a
highly viscous solvent or a matrix—eliminates the charge
transfer, and instead we note the transduction of singlet ex-
cited-state energy as the major deactivation. The corre-
spondingly formed C60-*ACHTNUTRGNE(UNG H2P)-(ZnP)3 is subject to charge
Cꢀ
C
separation and charge shift to yield (C60) -(H2P) +-(ZnP)3
1
used as standards (solvent peaks in H and 13C NMR spectra: CDCl3: d=
Cꢀ
C
and (C60) -H2P-(ZnP)3 +, respectively.
7.24 and 77.00 ppm, [D8]THF: d=3.58 and 67.57 ppm). The elemental
analysis was obtained by using a CE Instruments Elemental Analyzer
1110 CHNS machine. UV/Vis spectroscopic analysis was performed on
SPECORD S600, Analytik Jena AG, and Varian Cary 5000 UV/Vis/NIR
spectrophotometers. The absorption maxima lmax are given in nm. IR
spectroscopic analysis was carried out on Reakt-IR 1000 Asi Applied
Systems, ATR DiComp detector, and ThermoScientific Nicolet IR100
FT-IR spectrophotometers. The electrochemical experiments were car-
ried out using a BAS-CV50W electrochemical workstation with positive
feedback compensation. Cyclic, differential-pulse, and square-wave vol-
tammetry were performed in a three-electrode cell with a platinum wire
as the counterelectrode, a glassy carbon disk as the working electrode
(Ø=2 mm) versus a silver wire as the reference electrode; ferrocene was
applied as an internal standard and a 0.1m solution of (tBu)4NBF4 as the
supporting electrolyte. All the experiments were performed in an inert
gas atmosphere and at a compound concentration of 0.2 mm. Steady-state
fluorescence spectroscopic analysis was performed on a Horiba Jobin
Yvon Fluoromax 3 spectrophotometer. Deaerated solutions at room tem-
Conclusions
In summary, we have demonstrated unequivocally herein
light harvesting, unidirectional energy transfer, charge trans-
fer, and charge shift in a single donor–acceptor conjugate,
C60-H2P-(ZnP)3. The dendritic architecture built with driving
forces for all of the aforementioned processes is important
in this context. Owing, however, to the flexibility of the link-
ers, which connect the C60, H2P, and ZnP, the outcome of
photoexcitation depends strongly on environmental influen-
ces on the conformation. In Triton X-100 or an agar matrix,
time-resolved transient absorption spectroscopy and fluores-
Chem. Eur. J. 2009, 15, 12223 – 12233
ꢂ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
12231

D. M. Guldi, A. Hirsch, M. A. Garcia-Garibay et al.
perature (298 K) in a 1-cm quartz cuvette were used. All the spectra
were corrected for the instrument response. The monitoring wavelength
corresponded to the maximum of the emission band. For excitation wave-
lengths below l=450 nm a cut-off filter (450 nm) was inserted. In the
time-resolved fluorescence experiments, the fluorescence lifetime meas-
urements were performed on a SPEX Fluorolog-3 (HORIBA-JOBIN
YVON) machine supplied with an integrated TCSPC software. The
wavelength of l=403 nm was selected as the excitation wavelength for
the lifetime measurements (nano-LED-403L; pulse width: <100 ps). The
fluorescence lifetimes were measured at the emission maximum at room
temperature in deaerated solutions in a 1 to 1-cm quartz cuvette. Femto-
second transient absorption studies were performed with 387-nm laser
pulses (1 kHz, pulse width: 150 fs, 200 nJ) from an amplified Ti/sapphire
laser system (Model CPA 2010, Clark-MXR Inc.; output: 775 nm). For
an excitation wavelength of 420 nm, a nonlinear optical parametric con-
verter (NOPA) was used to generate ultrashort tunable visible pulses out
of the pump pulses. The transient absorption pump probe spectrometer
(TAPPS) is referred to as a two-beam setup, in which the pump pulse is
used as an excitation source for transient species and the delay of the
probe pulse is exactly controlled by an optical delay rail. As a probe
(white-light continuum), a small fraction of pulses that stem from the
CPA laser system was focused by a 50-mm lens into a sapphire disc
(thickness: 5 mm). The transient spectra were recorded using fresh
oxygen-free solutions in each laser excitation. All the experiments were
performed at 298 K in a 2-mm quartz cuvette. Transient absorption ex-
periments, based on nanosecond laser photolysis, were performed with
the output of the third harmonics (355 nm) coming from a Nd/YAG laser
(Brilliant, Quantel). Moreover, pulse widths of <5 ns with an energy of
10 mJ were selected. The optical detection is based on a pulsed Xenon
lamp (XBO 450, Osram), a monochromator (Spectra Pro 2300i, Acton
Research), a R928 photomultiplier tube (Hamamatsu Photonics), or a
fast InGaAs photodiode (Nano 5, Coherent) with amplification of
500 MHz and a digital oscilloscope (1 GHz; WavePro7100, LeCroy). The
laser power of every laser pulse was registered by using a bypath with a
fast silicon photodiode. The nanosecond laser photolysis experiments
were performed with 1-cm quartz cells and the solutions were saturated
with argon if no other gas saturation is indicated.
C, C60), 131.96 (b-pyrr-C), 129.07, 128.75, 128.66, 128.56, 128.22 (Ar-C),
121.80, 121.66, 121.50, 121.35, 121.20, 121.15, 121.08, 120.77, 120.74 (Ar-
C, meso-C), 115.94, 115.86, 115.79, 115.73, 115.54 (Ar-C), 72.11 (C60 sp3),
66.16, 65.74, 65.67, 65.58, 65.52 (CH2), 54.37 (OCH3), 53.04 ppm
(COCCO); IR: n˜ =2359, 2340, 1726, 1574, 1471, 1426, 1229, 1163, 1058,
976, 912, 799, 775, 726, 695, 637, 578, 551, 524, 462, 428 cmꢀ1; UV/Vis
(THF): lmax (e)=257 (120000), 327 (46800), 427 (195000), 517 (14200),
551 (7200), 595 (4900), 650 nm (3000mꢀ1 cmꢀ1); MALDI-TOF-MS
(DCTB): m/z: 1716 [M++H].
Dendritic porphyrin–fullerene hybrid C60-H2P-(ZnP)3 (11): Dyad
7
(91 mg, 53 mmol), DMAP (8 mg, 64 mmol), and HOBT (28 mg, 207 mmol)
were dissolved in dry THF (15 mL) under nitrogen. The solution was
cooled to 08C and light was excluded. DCC (43 mg, 207 mmol) was added
with stirring. After 10 min at 08C, porphyrin 10 (117 mg, 159 mmol) was
added to the reaction mixture, which was stirred for 20 min at 08C and
6 days at ambient temperature. At 08C, DCC (62 mg, 300 mmol) was
added and stirred for 1 day at ambient temperature. The solvent was re-
moved and the residue was redissolved in CH2Cl2 then filtered through
Celite 500. The product was purified by column chromatography (silica
gel, CH2Cl2/EtOAc 20:0.75, 1% NEt3, without sand on top) and recrys-
tallized from CH2Cl2/pentane (86 mg, 22 mmol, 42% relative to 10).
1H NMR (400 MHz, CDCl3, RT): d=9.00–8.50 (m, 32H; b-pyrr-H), 8.30–
6.80 (m, 73H; Ar-H), 4.75–3.60 (m, 25H; CH2, CH3), ꢀ3.17 ppm (s, 2H;
NH); 1H NMR (400 MHz, C2D2Cl4, 1008C): d=9.00–8.55 (m, 32H; b-
pyrr-H), 8.30–7.10 (m, 73H; Ar-H), 4.90–4.10 (m, 22H; CH2), 4.00–3.75
(m, 3H; OCH3), ꢀ2.88 ppm (s, 2H; NH); 13C NMR (100 MHz, CDCl3,
RT): d=168.63, 168.55, 168.49, 168.42, 168.33, 168.31, 168.25, 168.21, 168.
10, 168.00 (CO), 163.15, 163.00, 162.96, 162.87, 162.79, 162.72 (CO malon-
yl), 156.44, 156.39, 156.31, 156.19, 156.00, 155.92, 155.85, 155.73, 155.65,
155.56 (Ar-C), 150.13, 149.90 (a-pyrr-C), 144.25, 144.22, 144.19, 144.12,
144.06, 143.43, 143.40, 143.29, 142.75, 142.69 (Ar-C), 142.55, 142.36,
142.30, 142.22, 142.11, 142.05, 141.95, 141.88, 141.80, 141.70, 141.65,
141.56, 141.52, 141.41, 141.34, 141.19, 141.14, 140.79, 140.62, 140.46,
140.44, 140.33, 140.27, 140.14, 140.01, 139.87, 139.83, 139.52, 139.27,
139.19, 139.14, 139.04, 138.86, 138.78, 138.57, 138.52, 138.46, 138.32,
138.14, 138.04, 137.92, 137.80, 137.70, 137.64, 137.54, 137.45, 136.46,
136.33, 136.25 (C60, Ar-C), 134.33 (Ar-C), 131.97, 131.80, 131.00 (b-pyrr-
C), 128.38, 128.17, 127.94, 127.75, 127.65, 127.60, 127.39, 126.48 (Ar-C),
123.62, 121.18, 121.12, 120.91, 120.76, 120.39, 119.97, 119.89, 119.72,
119.52, 119.49, 119.40 (Ar-C, meso-C), 115.32, 115.28, 114.89, 114.80,
114.48, 114.40, 114.26, 113.95, 113.82 (Ar-C), 69.87, 69.75 (C60 sp3), 67.18,
67.10, 65.66, 65.54, 65.48, 65.41, 65.21, 65.07, 65.00, 64.76, 63.45, 63.30,
63.23, 62.98 (CH2), 53.74, 53.72, 53.68 (OCH3), 50.83, 50.78, 50.74,
50.70 ppm (Cquart. malonyl); 13C NMR (100 MHz, C2D2Cl4, 1008C): d=
168.82, 168.78 (CO), 163.43, 163.17 (CO malonyl), 157.11, 157.02, 156.73,
156.66 (Ar-C), 150.63, 150.42, 146.96, 146.90 (a-pyrr-C), 144.64, 144.26,
143.93, 143.85, 143.72, 143.13, 142.63, 142.61, 142.55, 142.40, 141.72,
141.67, 141.61, 141.57, 141.50, 141.27, 141.22, 140.60, 140.34, 139.92,
139.63, 139.16, 138.91, 138.84, 138.76, 138.66, 138.14, 138.07, 137.32,
137.20, 137.16 (C60, Ar-C), 134.75 (Ar-C), 132.32, 132.24, 132.11, 131.31
(b-pyrr-C), 128.82, 128.47, 128.29, 128.04, 127.95, 127.74, 126.79 (Ar-C),
121.87, 121.74, 121.57, 121.51, 120.77, 119.91, 119.88, 119.81 (Ar-C, meso-
C), 115.80, 114.99 (Ar-C), 70.77 (C60 sp3), 67.77, 66.50, 66.09, 66.04, 65.25,
63.71 (CH2), 53.84 ppm (OCH3); IR: n˜ =1749, 1596, 1576, 1481, 1438,
1339, 1286, 1233, 1174, 1069, 1002, 995, 913, 795, 754, 729, 718, 700, 660,
580, 525, 460, 433, 406 cmꢀ1; UV/Vis (CH2Cl2): lmax (e)=259 (160400),
420 (1053000), 517 (20700), 549 (54500), 589 nm (13100mꢀ1 cmꢀ1);
MALDI-TOF-MS (DCTB): m/z: 3875 [M++H]; ESI-MS (CH2Cl2): m/z:
1937.3 [M2+], 1291.5 [M3+], 968.9 [M4+].
tert-Butyl 2-(3-formylphenoxy)acetate (3): 3-(2-Hydroxyethoxy)benzalde-
hyde (16.00 g, 130.0 mmol), tert-butyl bromoacetate (20 mL, 137.2 mmol),
and tetrabutylammonium bromide (4.40 g, 14.0 mmol) dissolved in
CH2Cl2 (100 mL) were added to NaOH (5.20 g, 130.0 mmol) in H2O
(100 mL). The reaction mixture was stirred for 16 h at room temperature.
The organic phase was separated, the aqueous phase was washed four
times with CH2Cl2, and the solvent was removed from the combined or-
ganic phases. The resulting yellow oil was mixed with water and subse-
quently extracted with diethyl ether. This organic phase was washed
twice with 2n NaOH, dried over MgSO4, and the solvent was distilled off
(30.08 g, 127.00 mmol, 97%). 1H NMR (400 MHz, CDCl3, RT): d=9.94
(s, 1H; CHO), 7.49–7.47, 7.47–7.44, 7.44–7.42, 7.42–7.41, 7.33–7.30, 7.21–
7.19, 7.19–7.17 (m, 4H; Ar-H), 4.56 (s, 2H; CH2), 1.46 ppm (s, 9H; CH3);
13C NMR (100 MHz, CDCl3, RT): d=191.79 (CHO), 167.46 (COO),
158.44, 137.72, 130.18, 124.26, 122.01, 112.68 (Ar-C), 82.70 (CCH3), 65.59
(CH2), 27.98 ppm (CCH3); IR: n˜ =3011, 2984, 2945, 2833, 2744, 1741,
1695, 1594, 1482, 1459, 1393, 1370, 1328, 1282, 1227, 1146, 1073, 1038,
996, 977, 953, 923, 869, 845, 787, 733, 679 cmꢀ1; EI-MS: m/z: 236 [M+],
180 [M+ꢀtBu].
1’-Methoxycarbonyl-1’-{2-[3-(5-{10,15,20-tris[3-(hydrogencarboxymethox-
y)phenyl]porphyrin})]phenoxy}ethoxycarbonyl-1,2-methano-[60]fullerene
(7): Dyad 6 (205 mg, 0.12 mmol) was added to formic acid (40 mL). After
stirring overnight, the formic acid was distilled off. Toluene was added
and evaporated several times for purification (quantitative yield).
1H NMR (300 MHz, [D8]THF, RT): d=8.95–8.65 (m, 8H; b-pyrr-H),
8.10–7.53, 7.45–7.28 (m, 16H; Ar-H), 4.86 (s, 2H; CH2), 4.81 (s, 4H;
CH2), 4.71 (s, 2H; CH2), 4.46 (s, 2H; CH2), 3.93 (s, 3H; OCH3),
ꢀ2.91 ppm (s, 2H; NH); 13C NMR (75 MHz, [D8]THF, RT): d=170.57,
170.52, 170.43, 170.34 (COOH), 164.07, 163.67 (COCCO), 158.21, 158.07,
158.01, 157.91 (Ar-C), 145.88, 145.34, 145.03, 144.82, 144.79, 144.55,
144.50, 144.46, 143.96, 143.66, 142.90, 142.64, 142.39, 142.02, 141.92,
141.64, 141.55, 141.27, 141.15, 140.36, 140.26, 139.72, 138.59, 138.49 (Ar-
Dendritic porphyrin H2P-(ZnP)3 (12): Porphyrin 13 (59 mg, 59 mmol),
DMAP (9 mg, 70 mmol), and HOBT (31 mg, 229 mmol) were dissolved in
dry THF (15 mL) under nitrogen. The solution was cooled to 08C and
light was excluded. DCC (47 mg, 229 mmol) was added to the reaction
mixture with stirring. After 5 min, porphyrin 10 (130 mg, 176 mmol) was
added to the reaction mixture, which was stirred for 20 min at 08C and
5 days at ambient temperature. The solvent was removed and the residue
was redissolved in CH2Cl2 then filtered. The product was purified by
column chromatography (silica, CH2Cl2/EtOAc 20:0.75, 1% NEt3, with-
out sand on top) and recrystallized from CH2Cl2/pentane (81 mg,
12232
www.chemeurj.org
ꢂ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 12223 – 12233

Charge Transfer in Porphyrin–Fullerene Dendrons
FULL PAPER
25.65 mmol, 44% relative 13). 1H NMR (400 MHz, CDCl3, RT): d=9.10–
8.75 (m, 32H; b-pyrr-H), 8.40–8.10 (m, 18H; Ar-H), 7.90–6.65 (m, 55H;
Ar-H), 4.80–3.80 (m, 22H; CH2, CH3), 3.65–3.35 (m, 3H; CH2, CH3),
3.25–2.80 (m, 2H; CH2, CH3), ꢀ2.74 ppm (s, 2H; NH); 1H NMR
(400 MHz, C2D2Cl4, 1008C): d=9.20–8.80 (m, 32H; b-pyrr-H), 8.50–8.00
(m, 18H; Ar-H), 8.00–6.70 (m, 55H; Ar-H), 5.00–4.10 (m, 22H; CH2),
3.80–3.50 (s, 3H; OCH3), 3.40–3.20 (m, 2H; CH2), ꢀ2.59 ppm (s, 2H;
NH); 13C NMR (100 MHz, CDCl3, RT): d=168.64, 168.53, 168.47, 168.39
(CO), 166.56, 166.42, 166.30, 166.14, 166.04 (CO malonyl), 156.56, 156.39,
156.35, 156.25, 155.99, 155.93, 155.87 (Ar-C), 150.17, 149.93 (a-pyrr-C),
144.12, 144.07, 143.99, 143.41, 143.32, 142.75, 142.68, 134.38, 134.33 (Ar-
C), 132.01, 131.84, 131.20 (b-pyrr-C), 128.56, 128.10, 127.96, 127.90,
127.56, 127.51, 127.42, 127.37, 127.18, 126.49, 126.44 (Ar-C), 121.22,
121.13, 120.89, 120.40, 119.67, 119.50 (Ar-C, meso-C), 114.15, 113.79 (Ar-
C), 65.67, 65.57, 65.11, 65.02, 63.62, 63.41, 63.28, 63.16 (CH2), 52.37,
52.30, 52.23 (OCH3), 40.83, 40.77, 40.62, 40.58, 40.44, 40.40 ppm (Cquart.
malonyl); 13C NMR (100 MHz, C2D2Cl4, 1008C): d=168.98, 168.94 (CO),
166.71, 166.44 (CO malonyl), 157.36, 157.20, 156.88 (Ar-C), 150.69,
150.47, 147.03 (a-pyrr-C), 144.63, 143.94, 143.13, 134.78 (Ar-C), 132.30,
132.15, 131.41 (b-pyrr-C), 129.07, 128.64, 127.90, 127.77, 127.66, 126.78
(Ar-C), 122.10, 121.59, 121.51, 120.78, 120.06, 119.86 (Ar-C, meso-C),
114.98, 114.87, 114.79 (Ar-C), 66.69, 66.26, 63.92, 63.84 (CH2), 52.39
(OCH3), 41.17 ppm (Cquart. malonyl); IR: n˜ =1595, 1574, 1478, 1437, 1337,
1277, 1159, 1067, 993, 908, 793, 752, 717, 698, 658, 587, 566, 514, 503, 484,
[13] H. Imahori, D. M. Guldi, K. Tamaki, Y. Yoshida, C. Luo, Y. Sakata,
S. Fukuzumi, J. Am. Chem. Soc. 2001, 123, 6617–6628.
[14] H. Imahori, N. V. Tkachenko, V. Vehmanen, K. Tamaki, H. Lemme-
tyinen, Y. Sakata, S. Fukuzumi, J. Phys. Chem. A 2001, 105, 1750–
1756.
[15] V. Chukharev, N. Tkachenko, A. Efimov, D. M. Guldi, A. Hirsch, M.
Scheloske, H. Lemmetyinen, J. Phys. Chem. B 2004, 108, 16377–
16385.
[16] D. Gust, T. A. Moore, A. L. Moore, Acc. Chem. Res. 2001, 34, 40–
48.
[17] L. Echegoyen, L. E. Echegoyen, Acc. Chem. Res. 1998, 31, 593.
[18] Q. Xie, E. Perez-Cordero, L. Echegoyen, J. Am. Chem. Soc. 1992,
114, 3978.
[19] H. Imahori, K. Hagiwara, T. Akiyama, M. Aoki, S. Taniguchi, T.
Okada, M. Shirakawa, Y. Sakata, Chem. Phys. Lett. 1996, 263, 545.
[20] G. McDermott, S. M. Prince, A. A. Freer, A. M. Hawthornthwaite-
Lawiess, M. Z. Papiz, R. J. Cogdell, N. W. Isaacs, Nature 1995, 374,
517–521.
[21] F. Spꢀnig, C. Kovacs, F. Hauke, K. Ohkubo, S. Fukuzumi, D. M.
Guldi, A. Hirsch, J. Am. Chem. Soc. 2009, 131, 8180–8195.
[22] A. McKillop, J.-C. Fiaud, R. P. Hug, Tetrahedron 1974, 30, 1379–
1382.
[23] D. M. Guldi, A. Hirsch, M. Scheloske, E. Dietel, A. Troisi, F. Zer-
betto, M. Prato, Chem. Eur. J. 2003, 9, 4968–4979.
466, 433, 413 cmꢀ1
; UV/Vis (CH2Cl2): lmax (e)=418 (1419000), 515
[24] a) G. R. Geier III, B. J. Littler, J. S. Lindsey, J. Chem. Soc. Perkin
Trans. 2 2001, 712–718; b) J. S. Lindsey, I. C. Schreiman, H. C. Hsu,
P. C. Kearney, A. M. Maguerettaz, J. Org. Chem. 1987, 52, 827–836;
c) J. S. Lindsey, R. W. Wagner, J. Org. Chem. 1989, 54, 828–836;
d) J. S. Lindsey, K. A. MacCrum, J. S. Tyhonas, Y.-Y. Chuang, J. Org.
Chem. 1994, 59, 579–587; e) F. Li, K. Yang, J. S. Tyhonas, K. A.
MacCrum, J. S. Lindsey, Tetrahedron 1997, 53, 12339–12360.
[25] a) C. Bingel, Chem. Ber. 1993, 126, 1957–1959; b) J.-F. Nierengarten,
V. Gramlich, F. Cardullo, F. Diederich, Angew. Chem. 1996, 108,
2242–2244; Angew. Chem. Int. Ed. Engl. 1996, 35, 2101–2103.
[26] C. Kovacs, A. Hirsch, Eur. J. Org. Chem. 2006, 3348–3357.
[27] M. Brettreich, S. Burghardt, C. Bçttcher, T. Bayerl, S. Bayerl, A.
Hirsch, Angew. Chem. 2000, 112, 1915; Angew. Chem. Int. Ed. 2000,
39, 1845.
(26000), 549 (68000), 588 nm (14400mꢀ1 cmꢀ1); MALDI-TOF-MS
(DCTB): m/z: 3156 [M++H]; ESI-MS (CH2Cl2): m/z: 1577.9 [M 2+],
1052.0 [M 3+], 789.37 [M 4+].
Acknowledgements
Financial support from the German Science Foundation (DFG) through
SFB 583 is gratefully acknowledged. G.K. acknowledges NSF IGERT:
Materials Creation Training Program (MCTP) (DGE-0654431).
[1] J. Deisenhofer, H. Michel, Angew. Chem. 1989, 101, 872; Angew.
Chem. Int. Ed. Engl. 1989, 28, 829.
[2] J. Deisenhofer, J. R. Norris in The Photosynthetic Reaction Center,
Academic Press, New York, 1993.
[3] D. Gust, T. A. Moore, A. L. Moore, Acc. Chem. Res. 1993, 26, 198–
205.
[4] T. Asahi, M. Ohkouchi, R. Matsusaka, N. Mataga, R. P. Zhang, A.
Osuka, K. J. Maruyama, J. Am. Chem. Soc. 1993, 115, 5665–5674.
[5] Y. Sakata, H. Imahori, H. Tsue, S. Higashida, T. Akiyama, E. Yoshi-
zawa, M. Aoki, K. Yamada, K. Hagiwara, S. Taniguchi, T. Okada,
Pure Appl. Chem. 1997, 69, 1951–1956.
[28] D. Kuciauskas, S. Lin, G. R. Seely, A. L. Moore, T. A. Moore, D.
Gust, T. Drovetskaya, C. A. Reed, P. D. W. Boyd, J. Phys. Chem.
1996, 100, 15926.
[29] W. A. Lee, M. Graetzel, K. Kalyanasundaram, Chem. Phys. Lett.
1984, 107, 308.
[30] K. Tamaki, H. Imahori, Y. Sakata, Y. Nishimura, I. Yamazaki,
Chem. Commun. 1999, 625.
[31] C. Luo, D. M. Guldi, H. Imahori, K. Tamaki, Y. Sakata, J. Am.
Chem. Soc. 2000, 122, 6535–6551.
[32] D. M. Guldi, K.-D. Asmus, J. Phys. Chem. A 1997, 101, 1472.
[33] C. S. Foote, Top. Curr. Chem. 1994, 169, 347.
[6] M. R. Wasielewski, Chem. Rev. 1992, 92, 435–461.
[7] E. Dietel, A. Hirsch, E. Eichhorn, A. Rieker, S. Hackbarth, B.
Roeder, Chem. Commun. 1998, 1981–1982.
[8] D. M. Guldi, C. Luo, M. Prato, E. Dietel, A. Hirsch, Chem.
Commun. 2000, 373–374.
[9] D. M. Guldi, C. Luo, T. da Ros, M. Prato, E. Dietel, A. Hirsch,
Chem. Commun. 2000, 375–376.
[10] D. M. Guldi, C. Luo, M. Prato, A. Troisi, F. Zerbetto, M. Scheloske,
E. Dietel, W. Bauer, A. Hirsch, J. Am. Chem. Soc. 2001, 123, 9166–
9167.
[34] H. Imahori, K. Tamaki, D. M. Guldi, C. Luo, M. Fujitsuka, O. Ito, Y.
Sakata, S. Fukuzumi, J. Am. Chem. Soc. 2001, 123, 2607–2617.
[35] D. M. Guldi, M. Prato, Acc. Chem. Res. 2000, 33, 695.
[36] Agar (0.75 vol%) was added to an aqueous buffer solution (pH 7)
and heated till the agar dissolved completely, the compound was dis-
solved in Triton X-100 and added to the solution at 608C, and a
transparent jelly was formed after cooling to room temperature.
[37] a) U. Reuther, Ph.D. Thesis, University of Erlangen–Nꢁrnberg (Ger-
many), 2002; b) L. Isaacs, A. Wehrsig, F. Diederich, Helv. Chim.
Acta 1993, 76, 1231–1250.
[11] D. M. Guldi, C. Luo, A. Swartz, M. Scheloske, A. Hirsch, Chem.
Commun. 2001, 1066–1067.
[12] D. M. Guldi, A. Hirsch, M. Scheloske, E. Dietel, A. Troisi, F. Zer-
betto, M. Prato, Chem. Eur. J. 2003, 9, 4968–4979.
Received: August 3, 2009
Published online: October 30, 2009
Chem. Eur. J. 2009, 15, 12223 – 12233
ꢂ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
12233