Photoinduced electron transfer in an amine-corrole-perylene bisimide assembly: Charge separation over terminal components favoured by solvent polarity

Article abstract of DOI:10.1002/chem.201200744

An assembly has been synthesised that consists of four units: a meso-substituted corrole (C3), perylene bisimide (PI), and two electron-rich triphenylamine (DPA) units. PI is connected through a 1,4-phenylene bridge to C3, whereas the two DPA units are linked to C3 through a diphenyl ether linkage, which is used for the first time to connect the various moieties. Various synthetic strategies were elaborated, and the chosen one afforded the final system in six steps in an overall yield of 6 %. The resulting assembly, made of three different units, was named a triad . Excitation of the corrole (C3) or perylene bisimide (PI) units led to the charge-separated state DPA-C3⁺-PI^- with a rate k>10¹¹κ ^-1 in benzonitrile and dichloromethane (CH₂Cl₂) or with k of the order of 10¹⁰κ^-1 in toluene. The latter charge-separated state decayed to the ground state with a rate k=1.8×10⁹κ^-1 in toluene. In the polar solvents benzonitrile and dichloromethane, recombination to the ground state competes with a charge shift to form the distal charge-separated state, DPA ⁺-C3-PI^-, the formation of which occurs with a yield of 50 %. Recombination to the ground state of DPA⁺-C3-PI^- occurs with a rate k=5×10⁷κ^-1 in CH ₂Cl₂ and k=2×10⁷κ^-1 in benzonitrile. Copyright

Full text of DOI:10.1002/chem.201200744

FULL PAPER
DOI: 10.1002/chem.201200744
Photoinduced Electron Transfer in an Amine-Corrole-Perylene Bisimide
Assembly: Charge Separation over Terminal Components Favoured by
Solvent Polarity
Roman Voloshchuk,[a] Daniel T. Gryko,*[a, b] Maciej Chotkowski,[c] Adina I. Ciuciu,[d] and
Lucia Flamigni*[d]
Abstract: An assembly has been syn-
thesised that consists of four units: a
meso-substituted corrole (C3), perylene
bisimide (PI), and two electron-rich tri-
phenylamine (DPA) units. PI is con-
nected through a 1,4-phenylene bridge
to C3, whereas the two DPA units are
linked to C3 through a diphenyl ether
linkage, which is used for the first time
to connect the various moieties. Vari-
ous synthetic strategies were elaborat-
ed, and the chosen one afforded the
final system in six steps in an overall
yield of 6%. The resulting assembly,
made of three different units, was
named a “triad”. Excitation of the cor-
role (C3) or perylene bisimide (PI)
units led to the charge-separated state
DPA-C3+-PIꢀ with a rate k>1011 sꢀ1 in
1010 sꢀ1 in toluene. The latter charge-
separated state decayed to the ground
state with a rate k=1.8ꢀ109 sꢀ1 in tol-
uene. In the polar solvents benzonitrile
and dichloromethane, recombination to
the ground state competes with
a
benzonitrile
and
dichloromethane
charge shift to form the distal charge-
separated state, DPA+-C3-PIꢀ, the for-
mation of which occurs with a yield of
50%. Recombination to the ground
state of DPA+-C3-PIꢀ occurs with a
rate k=5ꢀ107 sꢀ1 in CH2Cl2 and k=2ꢀ
107 sꢀ1 in benzonitrile.
(CH2Cl2) or with k of the order of
Keywords: charge separation · cor-
role · electron transfer · energy
transfer · porphyrinoids
Introduction
occurs in the reaction centre, is the elementary event of
electron flow between two molecules. The first widely ac-
cepted theory of this process was developed by Rudolph A.
Marcus.[4] Energy transfer dominates in antennae, and plays
the role of harvesting photons from the entire visible spec-
trum. Photosynthetic systems occurring in nature are ex-
tremely complex, and have arisen through billions of years
of evolution. For this reason, so-called “artificial photosyn-
thesis” is developing much more slowly than one might
expect. There have been, however, a great number of suc-
cessful attempts to mimic the behaviour of separate stages
of natural systems, which in the future may finally lead to
the mastery of one of the most promising energy sources.
Owing to the excellent properties and availability of por-
phyrins, they have constituted for several decades the most
commonly used components in models for light-energy col-
lection and conversion.[5] However, the need to improve the
performance and increase the variety of such molecular sys-
tems has encouraged the use of other porphyrinoids. Among
these, the corroles,[6] analogues of porphyrins but shorter by
one carbon atom, have seen some success, because they pos-
sess the right properties, namely intense light absorption
throughout the visible spectral range, singlet excited-state
lifetimes in the nanosecond range, high luminescence yields
and high radiative rate constants, strong absorption features
of the excited state, good photostability in most solvents,
and relative ease of oxidation.[7] Energy transfer, and in
many cases also electron transfer, leading to charge-separat-
ed (CS) states with high efficiency and long lifetimes have
In nature, photons (sunlight) are exploited by living organ-
isms as energy in photosynthetic processes.[1] Chemistry can
contribute to this field of research in two ways: by the syn-
thesis and spectroscopic/photophysical characterisation of
model systems for natural counterparts,[2] and by the appli-
cation of acquired knowledge for the derivation of photovol-
taic devices.[3] In nature, two subsystems are required for ef-
ficient photosynthesis: the well-characterised reaction cen-
tres, and the much larger light-harvesting systems that incor-
porate up to hundreds of dyes. Electron transfer, which
[a] Dr. R. Voloshchuk, Prof. Dr. D. T. Gryko
Institute of Organic Chemistry of the Polish Academy of Sciences
Kasprzaka 44/52, 01-224 Warsaw (Poland)
E-mail: dtgryko@icho.edu.pl
[b] Prof. Dr. D. T. Gryko
Faculty of Chemistry, Warsaw University of Technology
Noakowskiego 3, 00-664 Warsaw (Poland)
[c] Dr. M. Chotkowski
Department of Chemistry, Warsaw University
Pasteura 1, 02-093 Warsaw (Poland)
[d] A. I. Ciuciu, Dr. L. Flamigni
Istituto per la Sintesi Organica e Fotoreattivitꢁ (ISOF)
CNR, Via P. Gobetti 101, 40129 Bologna (Italy)
E-mail: flamigni@isof.cnr.it
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201200744.
Chem. Eur. J. 2012, 00, 0 – 0
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
1
&
ÞÞ
These are not the final page numbers!

been observed in corrole-con-
taining arrays.[8] Recently, we
have been able to improve the
performance of a dyad based
on corrole and perylene bisi-
mides (PIs)[8g] by replacing PI
with recently synthesised pery-
lene imides with a different
substitution pattern.[9] The new
PIs maintain the characteristics
of this class of dyes (high ex-
tinction coefficients, ease of re-
duction,
well-characterised
anion radical spectra, and
strong luminescence), but dis-
play differences in their spec-
troscopic properties and, to
some extent, their electro-
chemical properties.[10] This
allows an increase in the driv-
ing forces for charge recombi-
nation, while keeping the proc-
ess highly exergonic. A high
driving force for the charge-re-
combination process places it
in the so-called “Marcus in-
verted” region, and greatly de-
presses the recombination
Figure 1. Structures of C3, PI, DPA, C3-PI, DPA-C3-PI, and DPA-C3.
rate.[4] The change in thermodynamic parameters of the
electron-transfer processes still results in remarkably effi-
cient charge separation (75%) and in an extended CS life-
time of 2.5 ms.[9]
the CS lifetime. The strategy of taking advantage of multi-
step electron transfer is well known and has led to remarka-
ble results.[11] We selected as the new component bis(4-me-
thoxyphenyl)aniline (DPA, Figure 1), which was symmetri-
cally appended to the corrole unit. DPA is a rather strong
reductant, with Eox of the order of 0.6 V versus SCE.[12]
Therefore, ground-state DPA should be able to reduce the
corrole cation formed upon primary electron transfer, pro-
vided that a corrole with a higher oxidation potential is
used. Meso-substituted corrole C3 (Figure 1), possessing two
pentafluorophenyl groups, which are known to increase the
oxidation potential of corroles significantly (for analogous
5,10,15-tris(pentafluorophenyl)corrole, Eox has been report-
ed as +0.86 V in benzonitrile),[13] should provide the right
conditions for the desired charge shift to occur.
An alternative approach to increasing the CS lifetime is
presented in the this paper. Here, we report our work,
which takes advantage of the addition of a further compo-
nent to the corrole-perylene bisimide dyad, resulting in a
triad, which might open the way to a further electron-trans-
fer step to the new component. This process, with further
spatial separation between the electron and the hole, de-
creases the coupling between opposite charges and increases
Abstract in Polish: Barwnik funkcjonalny składaja˛cy sie˛ z
czterech podjednostek: mezo-podstawionego korolu, peryle-
nobisimidu oraz dwꢀch trifenyloamin został zsyntetyzowany
i poddany badaniom spektroskopowym i fotofizycznym.
Analiza rꢀz˙nych strategii doprowadziła do opracowania
Results and Discussion
´
Design and synthesis: We chose to decorate meso-substitut-
ed corrole with two units known for their excellent perform-
ance in electron-transfer model systems. Perylene bisimides
are very popular building blocks in multichromophoric sys-
tems because of their outstanding photophysical properties
such as their intense absorption in the green region of the
visible spectrum, high photostability, fluorescence quantum
yield of approximately 1, and characteristic spectral signa-
ture of anion radicals.[14] The only feature complicating their
use is their very low solubility, which is incompatible with
efektywnej, szescioetapowej syntezy finalnego układu wielo-
chromoforowego. Wykazano, z˙e wzbudzenie korolu lub pery-
lenobisimidu prowadzi do wygenerowania stanu o rozdzielo-
nych ładunkach DPA-C3+-PIꢀ, o czasie z˙ycia k=1.8ꢁ
109 sꢀ1 w toluenie. W rozpuszczalnikach polarnych rekombi-
nacja do stanu podstawowego efektywnie wspꢀłzawodniczy z
utworzeniem drugiego stanu o rozdzielonych ładunkach
DPA+-C3-PIꢀ, podczas gdy w rozpuszczalnikach polarnych
ten drugi proces nie jest obserwowany.
&
2
&
www.chemeurj.org
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Photoinduced Electron Transfer
FULL PAPER
the typical conditions for corrole synthesis.[15] Fortunately, it
is possible to increase the solubility of perylene bisimides by
introducing solubilising groups at the nitrogen atoms with-
out alteration of their redox and optical properties (the
HOMO and LUMO of perylene bisimide have nodes on the
nitrogen atoms that make the electronic interaction of sub-
stituents with the perylene core unfeasible). Among several
different possibilities such as 2,6-diisopropylphenyl,[16] 2,5-
di-tert-butylphenyl[17,18] and various alkyl groups, we turned
our attention to a more efficient “swallow-tail” aliphatic
chain introduced by Langhals.[15,18] As a donor counterpart,
the electron-rich triarylamine-bearing alkoxy substituent
was chosen. Its nitrogen centre can be oxidised easily to a
highly stable radical cation species.[12] The synthesis of the
designed triad DPA-C3-PI was planned to be completed
through the condensation of a complex, acceptor-bearing al-
dehyde with two molecules of an appropriate dipyrrane.
There are two main synthetic strategies for perylene bisi-
mide-corrole dyads: 1) the synthesis of the tetrapyrrolic core
molecule first, followed by the modification of the peripher-
al groups; and 2) the preparation of the elaborated alde-
hyde, which is then used in a condensation reaction to pro-
duce the substituted corrole. On the basis of the results of
our previous endeavours,[8e,h] we decided to follow the
second strategy because of the moderate stability of cor-
roles.[19–22] The corrole-based triad was designed to have
redox-active substituents at all three meso-positions: the
perylene bisimide acceptor and the two triarylamine donor
units.
Scheme 1.
The whole endeavour started with the synthesis of the
corresponding aldehyde, which cannot be synthesised direct-
ly by mixed condensation of aliphatic and aromatic amines
with perylenetetracarboxylic acid dianhydride because of
the vast difference in their nucleophilicities. Therefore, this
condensation had to be divided into two separate condensa-
tion steps. We used monoimide-monoanhydride 1[23] as the
crucial building block, which had to be combined with a unit
containing free amino and protected aldehyde units. The
suitable building block 2 was synthesised in two steps start-
ing from 4-cyanobenzaldehyde.[24] The final condensation
was performed in melted imidazole[25] in the presence of
zinc acetate, and, after the deprotection of the acetal 3, af-
forded the aldehyde 4 in 77% yield (Scheme 1).
The synthesis of the second pivotal building block, the
phenol 7, can, in principle, be achieved through various
strategies. After several unsuccessful attempts (see the Sup-
porting Information), we found that the desired phenol 7
can be prepared easily by the direct double amination of 1-
bromo-4-ethoxybenzene (5) with 4-aminophenol (6) under
Buchwald–Hartwig conditions (Scheme 2).
Phenol 7 was then reacted with pentafluorobenzaldehyde
(8) in the presence of caesium fluoride,[26] and the resulting
aldehyde was efficiently condensed with an excess of pyrrole
under InCl3 catalysis[27] to give dipyrromethane 10
(Scheme 2).
Scheme 2.
forming step following the previously established general
procedures.[8h,19] The concentration of trifluoroacetic acid
(TFA) in the starting solution was approximately 13 mm, but
after 20 min, no signs of reaction were observed by TLC.
After increasing the TFA concentration to 27 mm and stir-
ring for one hour, we noticed the partial disappearance of
the aldehyde along with the initiation of spot formation
(TLC). To diminish any side reactions, we decided to
quench the reaction at this point, and after oxidising with
Having both the required building blocks in hand (alde-
hyde 4 and dipyrrane 10), we attempted the final corrole-
Chem. Eur. J. 2012, 00, 0 – 0
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
3
&
ÞÞ
These are not the final page numbers!

D. T. Gryko, L. Flamigni et al.
2,3-dichloro-5,6-dicyano-1,4-benzoquinone (DDQ), we were
able to isolate the desired DPA-C3-PI in 13% yield after a
straightforward purification process (Scheme 3).
Scheme 4.
Finally, corrole C3, lacking additional chromophores, was
prepared following a published procedure.[16] The symmetric
perylene bisimide PI, reported earlier by Langhals,[28] was
used as a model compound for the acceptor counterpart.
As a model of the donor part, the product of fluorine
atom substitution in pentafluorobenzonitrile with phenol 7
was an obvious choice (a cyano group was preferred to an
aldehyde owing to its greater stability), but the resulting
amine had extremely low fluorescence, which made photo-
physical measurements impossible. Consequently, the alter-
native model DPA was prepared by the alkylation of phenol
7 with tert-butyl chloroacetate (Figure 1).
Scheme 3.
In-depth photophysical measurements require model com-
pounds, both dyads and single units. The dyad C3-PI was
synthesised by the condensation of aldehyde 4 with 5-(pen-
tafluorophenyl)dipyrrane (11)[27] catalysed by TFA
(Scheme 4). Despite the relative simplicity of the substrates,
the reaction appeared to be capricious. When the reaction
was allowed to proceed for one hour without changing the
TFA concentration (ꢁ13 mm), corrole C3-PI was isolated in
2.6% yield. Doubling the TFA concentration allowed an in-
crease in yield up to 7%, but many spots appeared in TLC,
indicating the occurrence of side reactions.
The corrole-triarylamine model compound DPA-C3 was,
in turn, synthesised by condensation of the previously pre-
pared dipyrrane 10 with p-tolualdehyde (12) in the presence
of TFA. After oxidation and a straightforward workup, cor-
role DPA-C3 was isolated in good yield (Scheme 5). It is
worth mentioning that mainly porphyrin was formed when a
higher concentration of TFA was used in this reaction.
Electrochemical studies: Electrochemical studies of DPA,
PI, C3, DPA-C3, C3-PI, and DPA-C3-PI were performed in
order to confirm the magnitude of the coupling in all assem-
blies. The reduction and oxidation potentials are summar-
ised in Table 1 along with assignments of the electron-trans-
fer site. The first irreversible oxidation potential of corrole
C3 (+0.69 V) is in agreement with the reported values for
5,10,15-tris(pentafluorophenyl)corrole and 5,10,15-tris(4-
methylphenyl)corrole of +0.86 V and +0.38 V, respective-
ly.[13] It is rather critical for the interpretation of the photo-
physical results to assign oxidation potentials to both the
&
4
&
www.chemeurj.org
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Photoinduced Electron Transfer
FULL PAPER
model C3, oxidation of the corrole core occurs at +0.69 V
in dichloromethane (CH2Cl2). The lack of a visible shoulder
in the range from +0.55 to +0.75 V suggests that the oxida-
tion wave of the corrole core occurs at a potential not lower
than +0.79 V. On the other hand, knowing that the chromo-
phores are electronically decoupled, we have no better
choice than to assume that corrole oxidation does not occur
at a potential significantly higher than +0.79 V. The next
two, reversible, weakly developed oxidation waves can be
ascribed to the already known[13] second (+0.86 V) and
third (+1.05 V) oxidation processes of meso-substituted cor-
roles. Consequently, we decided to use the value +0.79 V in
the subsequent calculations. The same reasoning also has to
be applied to DPA-C3-PI. In the case of C3-PI, no ambigui-
ty exists.
We noticed a peak at approximately +0.2 V, and after
careful examination of its origin, it was assigned to the prod-
ucts of corrole electrochemistry. Kadish et al. examined a
series of corroles in PhCN, and their behaviour was similar
in all cases.[13] They noted that, in the initial scans in the pos-
itive direction, there were no redox processes between 0 and
+0.35 V, but a well-defined oxidation/reduction couple was
seen on the second positive potential sweep after scanning
through the first reduction.
Neither the oxidation potential of the donors nor the re-
duction potential of the acceptor used to construct the
dyads/triad was significantly affected by the presence of co-
valent linkages between the two redox-active centres. The
electrochemical data thus strongly suggest that ground-state
electronic interactions between the components are weak.
Spectroscopy and photophysics:
Scheme 5.
Steady-state absorption and emission spectroscopy in toluene:
The absorption spectra in toluene (TL) of the models C3,
PI, and DPA and of the arrays C3-PI, DPA-C3, and DPA-
C3-PI are given in Figure 2. In all cases, the absorption fea-
tures are an almost perfect superposition within the experi-
mental error, compared to those of the constituent units
with only minor shifts of ꢂ2 nm in the band maxima. A
very weak electronic interaction between the component
moieties can therefore be assumed.
The arrays can be excited selectively on the C3 moiety at
wavelengths above 550 nm, whereas excitation of the PI and
DPA units can be performed only predominantly: 80% at
490 nm on PI and 70% at 301 nm on DPA. The lumines-
cence spectra of the component units C3 and PI have been
reported previously,[8g] and their properties are collected in
Table 2, in which they are compared to those of the DPA
model and of the arrays in TL. The detected luminescence
at 298 K was due to fluorescence from the lowest singlet ex-
cited states of the various components. At 298 K, DPA dis-
played fluorescence with a maximum at 396 nm, an emission
quantum yield (Ffl) of 0.038 and a lifetime (t) of 2.1 ns; at
77 K, remarkably intense phosphorescence with a maximum
at 440 nm, and a long lifetime of 0.7 s was observed.
Table 1. Oxidation and reduction half-wave potentials (V vs. SCE) of all
studied compounds in CH2Cl2, 0.1m TBAP.
PI (red)
DPA (ox)
ACHTUNGTERNN(UNG Cor)H3 (ox)
DPA
C3
PI
DPA-C3
C3-PI
–
–
+0.62
–
–
+0.70
–
–
+0.69
–
+0.79
+0.75
+0.79
ꢀ0.54[a]
ꢀ0.35
ꢀ0.57
ꢀ0.56
DPA-C3-PI
+0.72
[a] Data from Ref. [14].
corrole and amine units in these arrays. In DPA-C3, the oxi-
dation of the bis(4-methoxyphenyl)aniline unit was easily
visible (and reversible), and the half-wave potential (E1/2)
was +0.70 V (Table 1). Unfortunately, the first, non-reversi-
ble oxidation of the corrole core overlapped with the oxida-
tion wave of triarylamine, and was clearly hidden under the
main wave of oxidation of the amine (+0.79 V, Fig-
ure S1).[29] This assumption is based on the fact that in the
Chem. Eur. J. 2012, 00, 0 – 0
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
5
&
ÞÞ
These are not the final page numbers!

D. T. Gryko, L. Flamigni et al.
The results of the comparative luminescence experiments
in TL on the dyads and the triad at the wavelengths speci-
fied above are reported in Figures 3–5. The experiments
were performed using dilute array solutions and compared
to each model component adjusted to an absorbance corre-
sponding to the absorbance in the dyad; see the insets in the
corresponding figures (Figures 3a, 4a, and 5a).
Selective excitation of the corrole unit in C3-PI was per-
formed at 565 nm; this led to complete quenching of the
corrole unit in the dyad, as shown in Figure 3a. Preferential
excitation of the PI unit at 490 nm (Figure 3b) led to almost
Figure 3. Room-temperature emission spectra of TL solutions of C3, PI,
and C3-PI. Excitation at a) 565 nm and b) 490 nm. The absorption spec-
trum is shown in the inset.
Figure 2. Absorption spectra in the molar absorption coefficient unit of
the triad, model dyads, and components in TL. The molar absorption co-
efficient of DPA is multiplied by two.
Table 2. Luminescence parameters of models and arrays in TL at room temperature and at 77 K.
295 K
77 K
t [ns][e]
[b]
[c]
[d]
Excited state
lmax [nm][a]
Ffl
Ffl
Ffl
t [ns][e]
lmax [nm][a]
E [eV]
DPA
1DPA
396
0.038
2.1
396
440
2.0
3.1
2.8
2.3
1.9
1.9
2.3
1.9
3DPA
70ꢀ107
4.0
PI[f]
1PI
534, 574, 622
656, 716
654,716
535, 574
658, 721
0.92
4.0
3.8
543, 587, 638
654,718
650,712
546, 587, 638
654,719
–
C3[f]
1C3
0.14
<0.01
5.2
5.4
C3-PI
1C3-PI
C3-1PI
DPA-1C3
1DPA-C3
DPA-1C3-PI
0.028
0.0009
0.034[g] 0.086
0.037
3.6
DPA-C3
0.14
0.42
<0.0004
<0.01
4.9
–
–
<0.01
0.030[g]
0.060
<0.01
0.024
DPA-C3-PI
<0.01
0.015
–
1DPA-C3-PI
DPA-C3-1PI
<0.001
–
–
–
–
0.0009
[a] Data from uncorrected spectra. [b] Luminescence quantum yield after excitation of corrole at 565 nm. [c] Luminescence quantum yield after excita-
tion of DPA at 301 nm. [d] Luminescence quantum yield after excitation of perylene at 490 nm. [e] Lifetimes detected after excitation at 301 nm for the
triplet lifetime, 465 nm or 331 nm for nanosecond resolution, and 355 nm or 532 nm for picosecond resolution. [f] Data from Ref. [8g]. [g] Rise.
&
6
&
www.chemeurj.org
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Photoinduced Electron Transfer
FULL PAPER
Figure 4. Room-temperature emission spectra of TL solutions of DPA,
C3, and DPA-C3. Excitation at a) 565 nm and b) 301 nm. The absorption
spectrum is shown in the inset.
complete luminescence quenching (99.9%) in the perylene
unit, whereas the fluorescence of the corrole unit was
quenched by 80%.
In DPA-C3, selective excitation of the corrole unit at
565 nm displayed no quenching in the corrole emission (Fig-
ure 4a), whereas after excitation at 301 nm on the DPA unit,
complete quenching of the luminescence of DPA and sensi-
tisation of the emission in the C3 band (Figure 4b) by a
factor of 3, and Ffl =0.42 were detected. This indicates (see
the absorption spectra) that quantitative energy transfer
takes place from DPA to the C3 unit.
Figure 5. Room-temperature emission spectra of TL solutions of C3, PI,
DPA, and DPA-C3-PI. Excitation at a) 565 nm, b) 490 nm and
c) 301 nm. The absorption spectrum is shown in the inset.
Finally, in the triad DPA-C3-PI, upon excitation of C3 at
565 nm and DPA at 301 nm, the fluorescence of each com-
ponent was quenched almost completely in comparison with
the reference model (Figure 5a and c). Following the excita-
tion of PI at 490 nm, quenching of 99.9% of PI and of 88%
of C3 could be detected, similar to the dyad C3-PI (Fig-
ure 5b).
The excitation spectra read on C3 fluorescence (l=
665 nm) in the dyads C3-PI and DPA-C3 indicate, by com-
parison with Figure 2, that in the luminescence of the cor-
role unit, a contribution of the PI and DPA absorption
bands can be detected, as shown in Figure 6a and b, respec-
tively. Similar results were obtained with the triad DPA-C3-
PI (data not shown). This is an unquestionable sign of
energy-transfer processes from both DPA and the PI unit to
the corrole in the arrays.
Luminescence spectra were also measured at 77 K in TL
glass, but solid TL samples are not a perfect matrix and the
data must be considered with some caution. However, the
following quantitative indication can be derived (data not
shown): 1) the occurrence of a complete energy-transfer
process from DPA to C3 in the DPA-C3 dyad; 2) the occur-
rence of a more efficient energy transfer than at room tem-
perature from PI to the C3 component in the C3-PI dyad, as
can be inferred from the relative values of the PI and C3 lu-
minescence bands after excitation at 490 nm of the PI unit;
Chem. Eur. J. 2012, 00, 0 – 0
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
7
&
ÞÞ
These are not the final page numbers!

D. T. Gryko, L. Flamigni et al.
Figure 7. Fluorescence time profiles in a TL solution of C3-PI after exci-
tation with a 532 nm picosecond pulse. Emission collected a) on the PI
band and b) on the C3 band. The exciting flash is shown as a full black
line and the fitting of the data as a grey line.
Figure 6. Excitation spectra of the dyads, lem =665 nm (C3 fluorescence
band).
and 3) still complete quenching of luminescence from vari-
ous components in the DPA-C3-PI triad. From this, it can
be inferred that, as expected, energy transfer is not affected
by solvent rigidity,[30] whereas the competing electron trans-
fer is. Nonetheless, the postulated electron transfer also
occurs efficiently in a solid matrix at 77 K. The lumines-
cence data are collected in Table 2.
tial function with a rise of 34 ps and a decay of 86 ps
(Figure 7). The rise of the corrole band matches the decay
of the PI band, indicating that the excited state of PI is a
precursor to the corrole excited state, compatible with an
energy-transfer process. It should be noted that on both
bands, some unquenched lifetime, appearing as a non-zero
baseline, can be detected. This is probably due to minimum
traces of contaminants.
Time-resolved absorption and emission spectroscopy in tol-
uene: Time-resolved luminescence on the nanosecond time-
scale was performed with the use of time-correlated single-
photon counting (TCSPC) apparatus with excitation at
465 nm or 331 nm. In order to increase the time resolution,
luminescence experiments with a picosecond laser and a
streak camera were performed upon excitation at 355 nm or
532 nm. The measured lifetimes are collected in Table 2.
The lifetime measured at the maximum of C3 emission in
DPA-C3 after excitation at 532 nm was 3.6 ns, which was es-
sentially identical to that of the model C3 (3.8 ns), confirm-
ing that no quenching occurs in DPA-C3 upon excitation of
the corrole unit. Following excitation at 355 nm on the DPA
moiety, the luminescence of DPA could not be detected in
the dyad, indicating a quenching process faster than the res-
olution of the apparatus (10 ps). The lifetime measured on
the corrole band was 3.6 ns, unquenched with respect to the
model. In the dyad C3-PI, excitation was carried out at
532 nm. The lifetime of PI was reduced to 37 ps, whereas
the fluorescence of C3 could be fitted by a double exponen-
In the triad DPA-C3-PI, the fluorescence lifetime of DPA
was still below the instrument resolution, the fluorescence
lifetime measured on the PI band was slightly faster than in
the corresponding dyad (24 ps versus 37 ps), and the time
profile of the C3 band could be fitted by a double exponen-
tial function with a rise time of 30 ps and a decay of 60 ps
(Figure 8). All processes were slightly faster than in the
dyad, probably because of increased electronic coupling be-
tween the components. The measured lifetimes are collected
in Table 2.
Experiments to determine the time-resolved absorbance
in the picosecond range were performed on the samples in
order to detect all the reaction intermediates. The singlet ex-
cited state of DPA in TL solution displayed a band with a
maximum at 715 nm and a decay lifetime of approximately
2 ns (data not shown). Upon excitation at 355 nm of the
dyad DPA-C3, the end-of-pulse spectrum showed the pres-
ence of the singlet excited state of the corrole, DPA-1C3,
with the typical bleaching and stimulated emission bands in
&
8
&
www.chemeurj.org
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Photoinduced Electron Transfer
FULL PAPER
formation (t=75 ps in C3-PI and t=55 ps in DPA-C3-PI)
and subsequent decay (t=1.1 ns in C3-PI and t=0.54 ns in
DPA-C3-PI) of bands at 710, 800, and 960 nm, previously
ascribed to the reduced PI unit, PIꢀ.[8g]
The lifetime of formation of the bands is in good agree-
ment with the fluorescence decay of 1C3-PI (86 ps) and
DPA-1C3-PI (60 ps) (Table 2). The presence of the strongly
absorbing intermediate indicates the formation and subse-
quent decay of the charge-separated state C3+-PIꢀ in the
dyad. The features of the oxidised corrole C3+, rather weak
and localised around 650 nm,[13] are probably hidden by the
intense PIꢀ bands. In the triad DPA-C3-PI, a spectrum iden-
tical to that of the dyad was registered, as apparent from the
spectral overlap shown in Figure 9. This leads us to assume
that in the triad the “vicinal” CS, DPA-C3+-PIꢀ, is formed
with the hole localised on the corrole, rather than the distal
CS, DPA+-C3-PIꢀ. Oxidised DPA was expected, similarly to
the analogous amine,[12] to display a band around 765 nm,
which was not observed in the present case. Giving further
support to the formation of the DPA-C3+-PIꢀ rather than
the DPA+-C3-PIꢀ state, it was also observed that the life-
time of the charge-separated state in the triad was even
shorter (0.54 ns) than in the dyad (1.1 ns). A much longer
lifetime would be expected for a recombination involving
the two remote sites in DPA+-C3-PIꢀ, owing to the larger
through-bond distance and decreased electronic coupling. It
is worth mentioning that molecular modelling excludes the
possibility of contact through space of the terminal DPA
and PI units. In spite of the rotational freedom around the
ether bond, contact between the extreme components is im-
possible in any possible configuration (see the Supporting
Information). This excludes in a most positive way the oc-
currence of a through-space electron transfer.
Figure 8. Fluorescence time profiles in a TL solution of DPA-C3-PI after
excitation with a 532 nm picosecond pulse. Emission collected a) on the
PI band and b) on the C3 band. The exciting flash is shown as a full
black line and the fitting of the data as a grey line.
the range 570–660 nm and
a weak absorbance above
1
690 nm.[7a] No trace of the DPA band at 715 nm appeared
in the spectrum. This, in agreement with the luminescence
In a nanosecond laser flash photolysis experiment on C3-
PI and DPA-C3-PI upon excitation at 532 nm, only the
spectrum of the CS state rise and decay could be observed
during the 18 ns laser pulse. No residual absorbance ascriba-
ble to the triplet state or other intermediates was detected
after the decay of the CS state, and there was complete re-
covery to the baseline. After nanosecond laser flash photoly-
sis at 355 nm of a DPA-C3 solution, the triplet excited state
of C3 was detected, as could be confirmed by comparison
with a similar experiment on the model C3.
data, confirms
a very fast energy-transfer process to
1
DPA-1C3 from DPA-C3, with a lifetime above our resolu-
tion (10 ps). The results of excitation at 532 nm of the dyad
C3-PI and the triad DPA-C3-PI are reported in Figure 9.
The spectra are almost identical, and the data indicate the
Photophysics in polar solvents: Because of the frustrating
outcome from TL solutions, in which the final charge shift
leading to the distal charge-separated state did not take
place, we decided to explore solvents with higher polarities:
dichloromethane (CH2Cl2) and benzonitrile (BN). The spec-
troscopic properties of the dyads and the triad in these sol-
vents were essentially identical to those in TL; however, the
dynamics of the intramolecular electron-transfer processes
were strongly affected. Time-resolved luminescence re-
vealed that quenching of C3 and PI luminescence in C3-PI
was faster than the resolution of the apparatus (10 ps) in
both solvents. Transient absorbance experiments with pico-
second resolution on the dyad C3-PI yielded the usual C3+
Figure 9. End-of-pulse absorption spectra detected after a 35 ps pulse
(532 nm, 3 mJ/pulse) in optically matched DPA-C3-PI (black) and C3-PI
(grey) in TL. In the insets, the time profiles at 710 nm are fitted by a
biexponential equation.
Chem. Eur. J. 2012, 00, 0 – 0
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
9
&
ÞÞ
These are not the final page numbers!

D. T. Gryko, L. Flamigni et al.
-PIꢀ spectrum (slightly hypsochromically shifted) dominated
by the bands of PIꢀ rising and decaying during the pulse,
with a lifetime of the order of 30–40 ps in both solvents,
which was much shorter than the 1.1 ns measured for the
same CS state in the TL solution. The spectra of the CS
state measured in BN and CH2Cl2 solutions of the triad
DPA-C3-PI were reduced in intensity by approximately
50% with respect to that in TL, as shown in Figure 10. As
Figure 10. End-of-pulse absorption spectra detected after a 35 ps pulse
(532 nm, 3 mJ/pulse) in optically matched DPA-C3-PI in TL (black line),
CH2Cl2 (grey line), BN (dots).
Figure 11. End-of-pulse absorption spectra detected after an 18 ns pulse
(532 nm, 3.5 mJ/pulse) in optically matched DPA-C3-PI in a) BN and
b) CH2Cl2. The insets show the decay at 710 nm and the exponential fit-
ting.
far as the shape is concerned, a shoulder around 765 nm
could be detected. Quite surprisingly, the CS bands, present
at the end of the pulse, did not decay appreciably during the
3.3 ns time window of the picosecond experiment. This
prompted us to perform nanosecond laser flash photolysis
experiments in polar solvents. Whereas no signal was detect-
ed in the dyad C3-PI, the spectra reported in Figure 11 were
registered in solutions of the triad DPA-C3-PI at the end of
the 18 ns laser pulse. The spectral features were those al-
ready detected in the picosecond experiment at the end of
the pulse, with PIꢀ radical bands at 715–720, 805, and
960 nm plus a clear shoulder/band at 765 nm ascribed to
DPA+.
The CS state that formed in these higher-polarity solvents
was, at variance with the TL case, DPA+-C3-PIꢀ. The in-
creased lifetime, to 50 ns for BN and 20 ns for CH2Cl2 com-
pared to 0.54 ns in TL, confirmed the formation of the distal
CS state.
It should be stressed that the high peak power of the pico-
second experiments led to some degradation of the sample
in BN, whereas in the nanosecond laser flash photolysis ex-
periments, the solutions were perfectly stable.
uum model.[31] The latter accounts for the difference in the
solvents in the electrochemical (CH2Cl2) and photochemical
(TL, CH2Cl2, and BN) determinations.[32] In the dyad in TL,
the CS state with the hole localised on the corrole and the
extra electron localised on the perylene derivative, C3+-PIꢀ,
was calculated to be at 1.64 eV. In the triad there are two
CS states, one similar to that of the dyad, DPA-C3+-PIꢀ,
and the other with charges on the two remote sites, DPA+
-C3-PIꢀ. For the former and latter CS states, energy levels
of 1.67 and 1.77 eV, respectively, were determined. In spite
of the fact that the model DPA can be oxidised more easily
than C3, the energy level of the state is opposite to expecta-
tions. This is due to the smaller Coulombic stabilisation
term and larger correction for the ion solvation term in
DPA+-C3-PIꢀ with respect to DPA-C3+-PIꢀ (see Figure 12).
In polar solvents, the situation changes, and the energy
levels of the CS states were calculated as 1.2 eV in CH2Cl2
and 1.1 eV in BN in the dyad C3-PI, for the CS state C3+
-PIꢀ.[32] This value is much lower than in TL, as expected on
the basis of the stabilisation of polar solvents able to reor-
ient around the charges. As far as the triad DPA-C3-PI is
concerned, the estimation led to 1.24 and 1.19 eV for the vi-
cinal and distal states, respectively, in CH2Cl2, and 1.13 and
1.06 eV for the vicinal and distal states, respectively, in BN.
In addition to a general stabilisation of the CS states, in
polar solvents, a smaller stabilisation of the vicinal CS state
was detected with respect to the distal one; this opens the
Photoinduced processes: The luminescence data of Table 2
allow the excited-state energy levels to be placed in the
arrays using the energy calculated from the maxima of the
emission at 77 K (Figure 12). The energy-level scheme can
be completed by adding the states corresponding to the CS
state. This can be done by using the electrochemical data
(Table 1) after correction for the Coulombic interaction and
the ion solvation energy based on the Born dielectric contin-
&
10
&
www.chemeurj.org
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Photoinduced Electron Transfer
FULL PAPER
Figure 12. Energy-level diagram of the arrays in TL (apolar solvent) and of the triad DPA-C3-PI in BN and CH2Cl2 (polar solvents). The CS energy
levels in BN are in grey.
possibility of a charge shift from DPA-C3+-PIꢀ to DPA+
-C3-PIꢀ, as shown in Figure 12, in perfect agreement with
the experimental observations. It is remarkable that the
theory can precisely predict, if not the absolute value, the
relative ordering of the CS states in different solvents.[32]
We first examined the different arrays in TL. Excitation
of the DPA-C3 dyad in the DPA band led to complete
quenching of the DPA emission, which decayed with a life-
time t ꢂ 10 ps and a concomitant quantitative sensitisation
of DPA-1C3. This was ascribable to an efficient energy-
transfer reaction with k ꢃ 1011 sꢀ1. DPA-1C3, on the other
hand, decayed in an unperturbed manner, with a lifetime
identical to that of the corrole model, and k=2.6ꢀ108 sꢀ1.
Excitation of C3-PI in the PI unit led to almost complete
quenching of PI and partial quenching of the C3 unit.
formed with the rates k=4.1ꢀ1010 and 1.6ꢀ1010 sꢀ1 from
DPA-C3-1PI and DPA-1C3-PI, respectively, corresponding to
luminescence lifetimes of 24 and 60 ps. These rates are
faster than the corresponding rates in the dyad. Once
formed, this CS state decays to the ground state with a rate
of k=1.8ꢀ109 sꢀ1, corresponding to a measured lifetime of
0.54 ns. As is evident from the energy-level diagram, a
charge shift to produce the distal CS state DPA+-C3-PIꢀ is
slightly endothermic and does not occur.
The triplet state of corrole is at a lower energy than the
CS state, at 1.52 eV,[8c] and, in principle, might be a product
of CS recombination, as formerly reported for another case
involving a corrole.[8c] However, we were unable to observe
any triplet yield, as demonstrated by the nanosecond laser
flash photolysis experiments, which did not show any residu-
al absorption after the decay of the CS state (see above).
The energy-level diagram of the triad is different in polar
C3-1PI was depopulated both by electron transfer,
a
HOMO–HOMO reaction from the ground-state C3 to the
1
1
excited-state PI, and by energy transfer to C3-PI, as can be
solvents. The initial processes (energy transfer from DPA-
inferred from the excitation spectrum (Figure 6a). In fact,
the presence of the absorption bands of PI in the excitation
spectrum measured on the corrole emission at 665 nm sup-
port this mechanism. The overall rate due to energy and
electron transfer depleting C3-1PI can be calculated via the
general equation k=1/tꢀ1/t0, in which t and t0 refer to the
lifetimes of the excited state in the dyad and in the model,
respectively. The calculated rate is k=2.7ꢀ1010 sꢀ1 on the
basis of t=37 ps for C3-1PI in TL. Excitation of the corrole
unit to 1C3-PI results in the quenching of this state (Fig-
ure 3a) through a LUMO–LUMO electron transfer from
the excited corrole to the PI unit. The relative rate can be
calculated as k=1.1ꢀ1010 sꢀ1 on the basis of a measured life-
C3-PI to DPA-1C3-PI and electron transfer from DPA-1C3-
PI and DPA-C3-1PI to form DPA-C3+-PIꢀ) are analogous
to those in TL. The rate of the latter (charge separation) is
faster than in TL, as expected for an electron-transfer reac-
tion with a higher driving force in the “Marcus normal
region”. DG0 for the charge-separation process, which in TL
is ꢀ0.63 eV, or ꢀ0.23 eV if we consider the excited state lo-
calised on PI or C3, changes to ꢀ1.06 and ꢀ0.66 eV in
CH2Cl2 and ꢀ1.17 and ꢀ0.77 eV in BN, respectively. The in-
crease in the driving force in the “Marcus normal region”
for charge separation explains why the quenching of excited
states in polar solvents is faster than in TL, with k> 1011 sꢀ1,
which is above our resolution. Once DPA-C3+-PIꢀ is
formed with unitary yield, it may undergo either charge re-
combination to the ground state, as in TL, or a charge shift
to produce DPA+-C3-PIꢀ. In polar solvents, the latter step
is permitted by the fact that the charge-shift reaction is ther-
modynamically allowed, though only by 0.05 eV (CH2Cl2) or
0.07 eV (BN). As can be inferred from Figure 10, the yield
1
time of 86 ps for C3-PI. Therefore, in C3-PI, excitation of
both units leads to the CS state C3+-PIꢀ with an almost uni-
tary efficiency, and this state decays to the ground state at a
rate of 9.1ꢀ108 sꢀ1, corresponding to a lifetime of 1.1 ns.
In the triad, upon electron transfer from both excited
states of PI and C3, the vicinal CS state DPA-C3+-PIꢀ is
Chem. Eur. J. 2012, 00, 0 – 0
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
11
&
ÞÞ
These are not the final page numbers!

D. T. Gryko, L. Flamigni et al.
of the distal CS state, DPA+-C3-PIꢀ, is of the order of 50%;
this allows us to deduce that the charge-shift reaction rate is
similar to the rate of charge recombination to the ground
state of DPA-C3+-PIꢀ. Assuming that charge recombination
to the ground state is similar in DPA-C3+-PIꢀ to that meas-
ured in the dyad C3+-PIꢀ in the same solvent (kꢁ3ꢀ
1010 sꢀ1, corresponding to a lifetime of 30–40 ps in the dyad),
we can assign to the charge-shift reaction a similar value of
approximately 3ꢀ1010 sꢀ1.
used. Further improvement is expected to be made through
a more judicious choice of the electron donor appended to
the corrole component.
Experimental Section
General synthetic section: All chemicals were used as received unless
otherwise noted. Reagent-grade solvents (CH2Cl2, hexanes, cyclohexane)
were distilled prior to use. All reported 1H NMR and 13C NMR spectra
were recorded on an AM 500 MHz or 400 MHz spectrometer. Chemical
shifts d (ppm) were determined with TMS as the internal reference; J
values are given in Hz. UV/Vis spectra were recorded in toluene. Chro-
matography was performed on silica (200–400 mesh), or dry-column
vacuum chromatography (DCVC)[33] was performed on preparative thin-
layer chromatography silica. Mass spectra were obtained through EI,
field desorption (FD), or electrospray MS (ESI-MS). The purity of all
new corroles was established on the basis of 1H NMR spectra and ele-
mental analyses. The following compounds were synthesised according to
literature procedures:
The distal CS state DPA+-C3-PIꢀ decays to the ground
state with rates of k=2.0ꢀ107 sꢀ1 in BN and k=5.0ꢀ107 sꢀ1
in CH2Cl2. These rates increase as the driving force of the
process in the two solvents increases, which is typical of a
reaction occurring in the “normal” Marcus region. The CS
lifetime values are not impressive for a triad displaying two-
step electron transfer, and certainly much longer lifetimes
have been reported in the literature for other triads,[11] and
even for corrole-based dyads.[8c,g,9] However, we have been
successful in showing that the design of such a triad can
indeed lead to an improvement in the CS lifetime with re-
spect to the corresponding dyad. Most importantly, we have
shown that solvent polarity can play an important role in
stabilising the vicinal and distal CS states to a different
extent, favouring a charge shift over the terminal units. Re-
markably, this can be predicted with a good level of preci-
sion by the conventional Weller treatment of the electro-
chemical data.[31]
Aldehyde 4:
1.32 mmol), acetal 2 (410 mg, 1.98 mmol) and a catalytic amount of Zn-
AHCTUNGRTEG(NNNU OAc)2 were heated in melted imidazole (7.2 g) at 1608C for 4 h. After
A mixture of perylene monoanhydride 1 (756 mg,
cooling to 808C, water (5 mL) was added to the reaction mixture, fol-
lowed by 2 N HCl, and the suspension formed was allowed to stay over-
night for aggregation. The resulting mixture was filtered through a sinter
glass filter (G4), and the solid was washed with water. After drying
under vacuum at 708C, the crude product was chromatographed (SiO2,
chloroform) to give acetal
3
(790 mg). 1H NMR (400 MHz, CDCl3,
TMS): d=0.84 (12H, m; CH3), 1.20–1.50 (16H, m; Alk), 1.88 (2H, m;
Alk), 2.25 (2H, m; Alk), 3.70 (2H, d, J=10.8 Hz; ac.-CH2), 3.82 (2H, d,
J=11.2 Hz; ac.-CH2), 5.19 (1H, m; N-CH), 5.50 (1H, s; ac.-N-CH), 7.37
(2H, m; C6H4), 7.72 (2H, m; C6H4), 8.57–8.80 ppm (8H, m; Ar). Depro-
tection of acetal 3 to aldehyde 4 was performed by stirring the solution
in a mixture of CH2Cl2/TFA/H2O (40:10:1) overnight. The yield after two
steps was 687 mg (77%). Rf 0.45 (2% MeOH in CH2Cl2, identical for al-
dehyde and acetal); 1H NMR (500 MHz, CDCl3, TMS): d=0.83 (6H, t,
J=7.1 Hz; CH3), 1.20–1.45 (16H, m; Alk), 1.88 (2H, m; Alk), 2.25 (2H,
m; Alk), 5.19 (1H, m; N-CH), 7.57 (2H, d, J=8.2 Hz; C6H4), 8.11 (2H,
d, J=8.6 Hz; C6H4) 8.60–8.80 (8H, m; Ar), 10.14 ppm (1H, s; CHO);
13C NMR (125 MHz, CDCl3, TMS): d=14.0, 22.6, 26.9, 29.2, 31.7, 32.4,
54.9, 122.9, 123.1, 123.5, 126.4, 126.7, 129.5, 129.8, 129.9, 130.7, 132.0,
134.1, 135.5, 136.4, 140.5, 163.3, 191.3 ppm; HRMS-FD (m/z): [M]+· calcd
for C44H40N2O5: 676.2937; found: 676.2928.
Conclusion
A complex multichromophoric system can be synthesised
using an approach consisting of the preparation of two sepa-
rate units (aldehyde and dipyrrane) followed by a final cycli-
sation step. This strategy leads to a stable meso-substituted
corrole in a fairly efficient manner. The nucleophilic aro-
matic substitution of the fluorine atom in C6F5CHO plays a
vital role in this strategy. We have characterised the photo-
induced processes occurring in TL solutions in a triad based
on meso-substituted corrole with an appended perylene bisi-
mide electron acceptor and a bis(4-methoxyphenyl)aniline
electron donor. The triad was designed to achieve a two-
step electron transfer, which could provide charge separa-
tion over the extreme DPA and PI components. The excita-
tion of DPA in the dyad and triad results in quantitative
energy transfer to the corrole component, DPA-1C3-PI.
However, the excitation of both the PI and C3 (or DPA)
components quantitatively yields DPA-C3+-PIꢀ, but a fur-
ther charge shift in the ground state leading to DPA+-C3-
PIꢀ is inhibited in TL solutions. We can rationalise this on
the basis of the negative thermodynamic parameters for the
charge-shift reaction. The situation is reversed in polar sol-
vents, in which almost identical initial steps are followed by
a charge shift leading to the distal CS state DPA+-C3-PIꢀ
with an efficiency of 50% and a lifetime of several tens of
nanoseconds in polar solvents. Although not impressive, the
increase in CS lifetime confirms the success of the strategy
Corrole DPA-C3-PI: Aldehyde 4 (366 mg, 0.54 mmol) and dipyrrane 10
(694 mg, 1.08 mmol) were dissolved in CH2Cl2 (8 mL), and a TFA stock
solution (180 mL; 10 mL of TFA dissolved in 100 mL of CH2Cl2) was
added. After 1 h, the reaction was quenched with triethylamine (30 mL)
and diluted to 100 mL with CH2Cl2.
A solution of DDQ (319 mg,
1.4 mmol) in toluene was added while stirring, and after 1 h the reaction
mixture was filtered through a silica pad. Product-containing fractions
were collected and evaporated on Celite. Chromatography (SiO2, tol-
uene) afforded 86 mg of DPA-C3-PI (13%). A crystalline sample was
obtained by reprecipitation in MeOH from the CH2Cl2 solution. Rf 0.6
(toluene); UV/Vis (CH2Cl2): lmax (eꢀ10ꢀ3) 302 (71.2), 414 (139.2), 457
(30.3), 490 (66.3), 527 (108.5), 564 (23.7), 613 nm (13.6); 1H NMR
(400 MHz, CDCl3, TMS): d=0.87 (6H, m; CH3), 1.20–1.45 (26H, m;
Alk), 1.82 (2H, m; Alk), 1.95 (2H, m; Alk), 2.33 (2H, m; Alk), 4.04
(8H, m; O-CH2), 5.24 (1H, m; N-CH), 6.80–7.20 (24H, m; Ar) 7.72 (2H,
d, J=7.8 Hz; C6H4), 8.10–8.65 (12H, m; C6H4+Ar+b-H), 8.80 (2H, d,
J=4.6 Hz; bꢀH), 8.86 (2H, d, J=4.8 Hz; bꢀH), 8.99 ppm (2H, d, J=
4.1 Hz; bꢀH); MS-FD (m/z): [M]+· calcd for C118H94N8O10
: 1934.7;
found: 1934.5.
Compound 7: An oven-dried Schlenk flask was charged with p-amino-
phenol (5, 3.27 g, 30 mmol), Pd2A(dba)3 (275 mg, 0.3 mmol), tBuONa (10 g,
0.1 mol), dry dioxane (45 mL), 1-bromo-4-ethoxybenzene (15.08 g,
75 mmol), and P(tBu)3 (4.8 mL, 0.25m solution in dioxane, 1.2 mmol)
CTHUNGTRENNUNG
AHCTUNGTRENNUNG
&
12
&
www.chemeurj.org
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Photoinduced Electron Transfer
FULL PAPER
under argon. The reaction mixture was stirred at 908C overnight. Upon
cooling, the reaction mixture was diluted with diethyl ether (200 mL),
washed with brine (3ꢀ50 mL), and then dried over Na2SO4. After solvent
evaporation, the oily residue was purified by chromatography (AcOEt/
hexanes 1:4) to afford the desired triarylamine 7 (10.4 g, 99%) as a col-
ourless oil (became bluish on standing). Rf 0.5 (EtOAc/hexanes 1:4);
1H NMR (400 MHz, [D6]DMSO, TMS): d=1.29 (6H, t, J=6.8 Hz;
CH2CH3), 3.94 (4H, q, J=7.2 Hz; CH2CH3), 6.67 (2H, d, J=8.8 Hz; Ar),
6.78 (10H, m; Ar), 9.19 ppm (1H, brs; OH); 13C NMR (100 MHz,
[D6]DMSO, TMS): d=15.4, 63.8, 115.8, 116.7, 124.5, 126.2, 140.3, 142.2,
153.8, 154.2 ppm; HRMS-ESI (m/z): [M]+· calcd for C22H23NO3:
349.1672; found: 349.1675.
3.12 mmol; dissolved in a minimal amount of toluene) was added under
vigorous stirring. After 2 h, the reaction mixture was passed through a
silica pad and concentrated. Flash column chromatography (toluene) af-
forded 251 mg of corrole DPA-C3 (15%). It was difficult to crystallise
the product in typical solvents, but it precipitated smoothly when evapo-
rating the solution in CH2Cl2/hexanes 1:1. Rf 0.5 (toluene); UV/Vis
(CH2Cl2): lmax (eꢀ10ꢀ3) 301 (64.0), 414 (142.7) 563 (21.3), 615 nm (12.4);
1H NMR (600 MHz, CDCl3, TMS): d=(ꢀ4)–(ꢀ1.5) (3H, brs; NH), 1.42
(12H, t, J=7.2 Hz; CH2CH3), 2.69 (3H, s; tol.-CH3), 4.02 (8H, q, J=
7.0 Hz; CH2CH3), 6.84 (8H, m; Ar), 7.07 (12H, m; Ar), 7.12 (4H, m;
Ar), 7.57 (2H, d, J=7.3 Hz; tol.), 7.87 (2H, d, J=7.3 Hz; tol.), 8.60 (2H,
brs; bꢀH), 8.73 (4H, m; bꢀH), 9.95 ppm (2H, m; bꢀH); 13C NMR
(150 MHz, CDCl3, TMS): d=14.9, 26.9, 63.7, 113.3, 115.3, 116.9, 117.4,
122.8, 125.5, 125.9, 127.7, 128.1, 134.6, 135.0, 137.5, 138.4, 140.9, 141.2,
142.7, 145.3, 145.4, 147.1, 151.7, 155.0 ppm; MS-FD (m/z): [M]+· calcd for
C82H62F8N6O6: 1378.5; found: 1378.5.
Compound 9: Phenol 7 (4.2 g, 12 mmol) and cesium fluoride (3.65 g,
24 mmol) were suspended in dry DMF (20 mL). Pentafluorobenzalde-
hyde (8, 2.35 g, 12 mmol) was added and the reaction was stirred at RT
for 3 h (under argon). The reaction mixture was diluted with water and
extracted with CH2Cl2 (3ꢀ100 mL). The combined organic layers were
dried over MgSO4, concentrated, and passed through a silica pad. Crys-
tallisation from chloroform/hexanes afforded 4.31 g (68%) of aldehyde 9
DPA: Phenol
7 (1.82 g, 5.2 mmol), tert-butyl chloroacetate (1.17 g,
7.8 mmol), and Cs2CO3 (2.54 g, 7.8 mmol) were heated in dry DMF
(15 mL) at 708C for 2 h. The reaction was diluted with water and extract-
ed with CH2Cl2 (3ꢀ50 mL). The organic layers were collected and dried
over MgSO4, then chromatographed (EtOAc/hexanes 1:9). Crystallisation
from chloroform/hexanes afforded 2.36 g of the title compound (98%).
Rf 0.5 (EtOAc/hexanes 1:3); 1H NMR (400 MHz, CDCl3, TMS): d=1.40
(6H, t, J=6.8 Hz, CH2CH3), 1.49 (9H, s; CH3), 3.98 (4H, q, J=6.8 Hz;
CH2CH3), 4.46 (2H, s; CH2), 6.76 (6H, m; Ar), 6.93 ppm (6H, m; Ar);
13C NMR (100 MHz, CDCl3, TMS): d=14.9, 27.9, 63.6, 66.2, 82.2, 115.1,
115.3, 124.1, 125.1, 141.7, 142.9, 153.0, 154.4, 168.3 ppm; HRMS-EI (m/z):
[M]+· calcd for C28H33NO5: 463.2359; found: 463.2344.
1
as an off-white solid. Rf 0.45 (EtOAc/hexanes 1:4); m.p. 1568C; H NMR
(500 MHz, CDCl3, TMS): d=1.40 (6H, t, J=7.0 Hz; CH3), 4.00 (4H, q,
J=6.9 Hz; ac.-CH2), 6.78–7.03 (12H, m; Ar), 10.28 ppm (1H, s; CHO);
13C NMR (125 MHz, CDCl3, TMS): d=14.9, 63.7, 110.8 (t, J=40 Hz),
115.3, 117.1, 122.2, 126.1, 140.0, 140.9, 141.2 (dm, J=255 Hz), 145.9, 147.6
(dm, J=260 Hz), 150.6, 155.2, 182.0 ppm; elemental analysis calcd (%)
for C29H23F4NO4·0.25H2O: C 65.72, H 4.47, N 2.64; found: C 65.93, H
4.45, N 2.65; HRMS-ESI (m/z): [M]+· calcd for C29H23F4NO4: 525.1558;
found: 525.1554.
Compound 10: A solution of aldehyde 9 (4.82 g, 9.18 mmol) in freshly
distilled pyrrole (64 mL) was degassed with argon for 15 min. InCl3
(203 mg, 0.92 mmol) was added and the reaction mixture was stirred at
RT for 1.5 h under an argon atmosphere. Next, powdered NaOH
(734 mg, 18.4 mmol) was added and the reaction mixture was stirred for
a further 1 h. The reaction mixture was filtered through Celite, the pyr-
role was distilled off under reduced pressure, and the oily residue was pu-
rified by flash chromatography on a short column (EtOAc/hexanes 1:3!
1:1) affording 5.48 g of 10 (93%). All attempts to crystallise this dipyr-
rane failed, and it was stored in the form of a dry foam. 1H NMR
(500 MHz, CDCl3, TMS): d=1.39 (6H, t, J=7.0 Hz; CH3), 3.99 (4H, q,
J=7.0 Hz; CH2), 5.92 (1H, s; meso-CH), 6.05 (2H, brs; pyrr.-CH), 6.16
(2H, dd, 3J=5.6 Hz, 4J=2.7 Hz; pyrr.-CH), 6.73 (2H, m; pyrr.-CH),
6.76–6.84 (6H, m; Ar), 6.90 (2H, m; Ar), 6.98 (4H, m; Ar), 8.15 ppm
(2H, brs; NH); 13C NMR (125 MHz, CDCl3, TMS): d=14.9, 33.2, 63.7,
107.6, 108.7, 115.3, 116.2 (m), 116.7, 118.0, 122.8, 125.8, 128.5, 133.8 (m),
141.2, 141.8 (dm, J=251 Hz), 145.1, 145.2 (dm, J=246 Hz), 151.5,
154.9 ppm; HRMS-FD (m/z): [M]+· calcd for C37H31F4N3O3: 641.2302;
found: 641.2307.
Electrochemistry: Cyclic voltammetry was carried out with a CH Instru-
ments 604C potentiostat. A homemade three-electrode electrochemistry
cell was used, which consisted of a platinum button, a platinum wire
counter electrode, and a saturated calomel reference electrode (SCE).
The SCE was separated from the bulk of the solution by a fritted-glass
bridge of low porosity, which contained the solvent/supporting electrolyte
mixture. All potentials are referenced to SCE.
Spectroscopy and photophysics: The solvents used were spectroscopic-
grade toluene and dichloromethane (C. Erba) and HPLC-grade benzoni-
trile (SIGMA-Aldrich). Absorption spectra were recorded with
a
Perkin–Elmer Lambda 650 spectrophotometer, and emission spectra, un-
corrected for the photomultiplier response unless otherwise stated, were
recorded by a Spex Fluorolog II spectrofluorimeter equipped with a Ha-
mamatsu R3896 photomultiplier. The phosphorescence spectrum of DPA
was determined both with a Spex Fluorolog II spectrofluorimeter equip-
ped with a phosphorimeter accessory (1934D, Spex) and with a Perkin–
Elmer LS-50 spectrofluorimeter. The latter was equipped with a pulsed
xenon lamp and gated photomultiplier with a modified response for oper-
ation to about 650 nm. The phosphorescence lifetime on the millisecond
scale was measured using the same Perkin–Elmer LS-50 spectrofluorime-
ter. Fluorescence lifetimes in the nanosecond region were detected by a
Time-Correlated Single-Photon Counting apparatus (IBH) with excita-
tion at 465 nm or, in the case of DPA, with an FLS920 spectrofluorimeter
(Edinburgh) equipped with an S900 single-photon photomultiplier detec-
tion system, with excitation at 331 nm. Emission quantum yields where
determined against PI in TL (Ffl =0.92)[8g] or against C3 in TL (Ffl =
0.14).[7a] Air-equilibrated quinine sulfate in 1N H2SO4 was used as a
standard for the determination of the DPA emission quantum yield with
Ffl =0.546.[34]
Corrole C3-PI: Aldehyde 4 (220 mg, 0.33 mmol) and 5-(pentafluorophe-
nyl)dipyrrane (11, 203 mg, 0.65 mmol) were dissolved in CH2Cl2 (5 mL),
and the TFA stock solution (55 mL) was added. After 20 min, the reaction
was quenched with triethylamine (10 mL), and was diluted to 50 mL with
CH2Cl2. A solution of DDQ (195 mg, 0.83 mmol) in toluene was added
while stirring, and after 1 h, the reaction mixture was filtered through a
silica pad. Product-containing fractions were collected and evaporated
with Celite. Purification by DCVC (CH2Cl2/hexanes 4:1) afforded 29 mg
of C3-PI (7%). Rf 0.6 (CH2Cl2/hexanes 3:2); UV/Vis (CH2Cl2): lmax (eꢀ
10ꢀ3) 410 (163.1), 458 (35.2), 490 (83.4), 527 (138.7), 561 (28.0), 611 nm
(16.0); 1H NMR (500 MHz, [D8]THF, TMS): d=ꢀ1.71 (3H, brs; NH),
0.85 (6H, m; CH3), 1.20–1.50 (16H, m; Alk), 1.93 (2H, m; Alk), 2.39
(2H, m; Alk), 5.25 (1H, m; Alk), 7.89 (2H, m; C6H4), 8.40 (2H, m;
C6H4), 8.59 (10H, m; bꢀH+Ar), 8.81 (4H, brs; bꢀH), 9.09 ppm (2H, s,
bꢀH); MS-FD (m/z): [M]+· calcd for C74H50F10N6O4: 1276.4; found:
1276.3.
For picosecond time-resolved luminescence, a Streak Camera Hamamat-
su C1587 equipped with an M1952 unit was used as the detector after ex-
citation with a Nd-YAG laser at 355 nm and 532 nm (Continuum PY62/
10, 35 ps pulse, 1 mJ).[35] Transient absorbance in the picosecond range
made use of a pump-and-probe system based on a Nd-YAG laser (Con-
tinuum PY62/10, 35 ps pulse). The second (532 nm) and third (355 nm)
harmonics with an energy of around 3 mJ/pulse were used to excite the
samples with an absorbance around 0.5. More details are reported else-
where.[9] Transient absorbance measurements in the nano- and microsec-
ond ranges made use of a laser flash photolysis apparatus based on a
Nd:YAG laser (JK Lasers) delivering pulses of 18 ns at 355 nm and
Corrole DPA-C3: p-Tolualdehyde (12, 144 mg, 1.2 mmol) and dipyrrane
10 (1.54 g, 2.4 mmol) were dissolved in 18 mL of CH2Cl2, and the TFA
stock solution was added (200 mL). The reaction mixture was stirred at
RT for 20 min (monitored by TLC). It was then quenched with triethyla-
mine (35 mL), diluted to 200 mL with CH2Cl2, and DDQ (708 mg,
Chem. Eur. J. 2012, 00, 0 – 0
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
13
&
ÞÞ
These are not the final page numbers!

D. T. Gryko, L. Flamigni et al.
532 nm. The energy used was 3 mJ/pulse, and the absorbance of the solu-
tions was around 1. More details can be found elsewhere.[11f]
D. T. Gryko, S. Fukuzumi, J. Am. Chem. Soc. 2008, 130, 14263–
14272; d) M. Tasior, D. T. Gryko, M. Cembor, J. S. Jaworski, B. Ven-
tura, L. Flamigni, New J. Chem. 2007, 31, 247–259; e) L. Flamigni,
D. Wyrostek, R. Voloshchuk, D. T. Gryko, Phys. Chem. Chem. Phys.
The estimated errors are 10% on lifetimes, molar extinction coefficients,
and quantum yields. The working temperature, if not otherwise specified,
was 295ꢄ2 K. Molecular dimensions were estimated after MM2 minimi-
zation by CS Chem3D Ultra 6.0 software.
´
2010, 12, 474–483; f) M. Tasior, D. T. Gryko, D. Pielacinska, A. Za-
nelli, L. Flamigni, Chem. Asian J. 2010, 5, 130–140; g) L. Flamigni,
B. Ventura, M. Tasior, T. Becherer, H. Langhals, D. T. Gryko, Chem.
Eur. J. 2008, 14, 169–183; h) L. Flamigni, D. T. Gryko, Chem. Soc.
Rev. 2009, 38, 1635–1646.
[9] L. Flamigni, A. I. Ciuciu, H. Langhals, B. Bçck, D. T. Gryko, Chem.
Asian J. 2012, 7, 582–592.
[10] L. Flamigni, A. Zanelli, H. Langhals, B. Bçck, J. Phys. Chem. A
2012, 116, 1503–1509.
Acknowledgements
We thank the Polish Ministry of Science and Higher Education (Contract
N204 123837), CNR of Italy (PM.P04.010, Project MACOL) and the Eu-
rocores project “Solarfueltandem” for financial support. The research
leading to these results has received partial funding from the European
Communityꢃs Seventh Framework Programme under the TOPBIO proj-
ect, grant agreement no. 264362.
[11] a) D. Gust, T. A. Moore, A. L. Moore, A. N. MacPherson, A. Lopez,
J. M. DeGraziano, I. Gouni, E. Bittersman, G. R. Seely, F. Gao,
R. A. Nieman, X. C. Ma, L. J. Demanche, S.-C. Hung, D. K. Luttrull,
S.-J. Lee, P. K. Kerrigan, J. Am. Chem. Soc. 1993, 115, 11141–11152;
b) K. A. Jolliffe, T. D. Bell, K. P. Ghiggino, S. J. Langford, M. N.
Paddon Row, Angew. Chem. 1998, 110, 959–964; Angew. Chem. Int.
Ed. 1998, 37, 915–919; c) H. Imahori, D. M. Guldi, K. Tamaki, Y.
Yoshida, C. Luo, Y. Sakata, S. Fukuzumi, J. Am. Chem. Soc. 2001,
123, 6617–6628; d) H. Imahori, Y. Sekiguchi, Y. Kashiwagi, T. Sato,
Y. Araki, O. Ito, H. Yamada, S. Fukuzumi, Chem. Eur. J. 2004, 10,
3184–3196; e) M. Borgstrçm, N. Shaikh, O. Johansson, M. F. Ander-
lund, S. Styring, B. ꢅkermark, A. Magnuson, L. Hammarstrçm, J.
Am. Chem. Soc. 2005, 127, 17504–17515; f) L. Flamigni, E. Baran-
off, J.-P. Collin, J.-P. Sauvage, Chem. Eur. J. 2006, 12, 6592–6606.
[12] E. Baranoff, I. M. Dixon, J.-P. Collin, J.-P. Sauvage, B. Ventura, L.
Flamigni, Inorg. Chem. 2004, 43, 3057–3066.
[13] J. Shen, J. Shao, Z. Ou, W. E, B. Koszarna, D. T. Gryko, K. M.
Kadish, Inorg. Chem. 2006, 45, 2251–2265.
[14] H. Langhals, A. Obermeier, Y. Floredo, A. Zanelli, L. Flamigni,
Chem. Eur. J. 2009, 15, 12733–12744.
[15] H. Langhals, Nachr. Chem. Tech. Lab. 1980, 28, 716–718.
[16] F. Graser, to BASF AG, Ger. Offen. DE 3049215, July 15, 1982
(Chem. Abstr. 1982, 97, 129114).
[17] A. Rademacher, S. Mꢄrkle, H. Langhals, Chem. Ber. 1982, 115, 2927.
[18] H. Langhals, S. Demmig, T. Potrawa, J. Prakt. Chem. 1991, 333,
733–748.
[1] a) A. W. Rutherford, P. Faller, Philos. Trans. R. Soc. London Ser. B
2003, 358, 245–253; b) B. Conlan, Photosynth. Res. 2008, 98, 687–
700; c) P. L. Marek, H. Hahn, T. S. Balaban, Energy Environ. Sci.
2011, 4, 2366–2378.
[2] a) M. R. Wasielewski, Acc. Chem. Res. 2009, 42, 1910–1921; b) A.
Magnuson, M. Anderlund, O. Johansson, P. Lindblad, R. Lomoth, T.
Polivka, S. Ott, K. Stensjç, S. Styring, V. Sundstrçm, L. Hammar-
strçm, Acc. Chem. Res. 2009, 42, 1899–1909; c) L. Flamigni, J.-P.
Collin, J.-P. Sauvage, Acc. Chem. Res. 2008, 41, 857–871; d) D. Gust,
T. A. Moore, A. L. Moore, Acc. Chem. Res. 2009, 42, 1890–1898.
[3] A. Hagfeldt, M. Grꢄtzel, Acc. Chem. Res. 2000, 33, 269–277.
[4] R. A. Marcus, Angew. Chem. 1993, 105, 1161–1172; Angew. Chem.
Int. Ed. Engl. 1993, 32, 1111–1121, and references therein.
[5] a) D. Holten, D. F. Bocian, J. L. Lindsey, Acc. Chem. Res. 2002, 35,
57–69; b) L. Flamigni, V. Heitz, J.-P. Sauvage, Struct. Bonding 2006,
121, 217–261; c) M. R. Wasielewski, Chem. Rev. 1992, 92, 435–461;
d) D. Gust, T. A. Moore, A. L. Moore, Acc. Chem. Res. 2001, 34,
40–48; e) B. Du, D. Fortin, P. D. Harvey, Inorg. Chem. 2011, 50,
11493–11505.
[19] D. T. Gryko, B. Koszarna, Synthesis 2004, 2205–2209.
[20] D. T. Gryko, K. Jadach, J. Org. Chem. 2001, 66, 4267–4275.
[21] a) D. T. Gryko, B. Koszarna, Org. Biomol. Chem. 2003, 1, 350–357;
b) D. T. Gryko, K. E. Piechota, J. Porphyrins Phthalocyanines 2002,
6, 81–97.
[22] a) D. T. Gryko, J. Porphyrins Phthalocyanines 2008, 12, 906–917;
b) M. Tasior, R. Voloshchuk, Y. M. Poronik, T. Rowicki, D. T.
Gryko, J. Porphyrins Phthalocyanines 2011, 15, 1011–1023.
[23] H. Kaiser, J. Lindner, H. Langhals, Chem. Ber. 1991, 124, 529–535.
[24] A. Marcos Ramos, S. C. J. Meskers, E. H. A. Beckers, R. B. Prince,
L. Brunsveld, R. A. J. Janssen, J. Am. Chem. Soc. 2004, 126, 9630–
9644.
[25] H. Langhals, Chem. Ber. 1985, 118, 4641–4645.
[26] D. T. Gryko, D. Wyrostek, A. Nowak-Krꢆl, K. Abramczyk, M. Ro-
gacki, Synthesis 2008, 4028–4032.
[27] J. K. Laha, S. Dhanalekshmi, M. Taniguchi, A. Ambroise, J. S. Lind-
sey, Org. Process Res. Dev. 2003, 7, 799–812.
[28] S. Demmig, H. Langhals, Chem. Ber. 1988, 121, 225–230.
[29] A. J. Bard, L. R. Faulkner, Electrochemical methods. Fundamentals
and applications, 2nd ed., Wiley, New York, 2001.
[30] L. Hammarstrçm, F. Barigelletti, L. Flamigni, N. Armaroli, A. Sour,
J.-P. Collin, J.-P. Sauvage, J. Am. Chem. Soc. 1996, 118, 11972–
11973.
[31] A. Weller, Z. Phys. Chem. (Muenchen Ger.) 1982, 133, 93–98.
[32] The correction of the electrochemical potential can be done accord-
ing to the following equation, which provides the energy for the
charge-separation reaction (relative to the neutral ground state):
DG0 =EoxꢀEred ꢀ (14.32/RDAeS) + 14.32 (1/2rD +1/2rA) (1/eSꢀ1/eP)
(1), in which Eox and Ered are the potentials corresponding to the
first reduction of the electron acceptor and the first oxidation of the
[6] a) R. Paolesse in The Porphyrin Handbook, Vol. 2 (Eds.: K. M.
Kadish, K. M. Smith, R. Guilard), Academic Press, New York, 2000,
pp. 201–232; b) J. L. Sessler, S. J. Weghorn in Expanded, Contracted
& Isomeric Porphyrins, Pergamon, Oxford, 1997, p. 11; c) C. Erben,
S. Will, K. M. Kadish in The Porphyrin Handbook, Vol. 2, (Eds.:
K. M. Kadish, K. M. Smith, R. Guilard), Academic Press, San
Diego, 2000, p. 233; d) D. T. Gryko, Eur. J. Org. Chem. 2002, 1735–
1743; e) R. Guilard, J.-M. Barbe, C. Stern, K. M. Kadish in The Por-
phyrin Handbook Vol. 18 (Eds.: K. M. Kadish, K. M. Smith, R. Gui-
lard), Academic Press, New York, 2003, pp. 303–351; f) D. T. Gryko,
J. P. Fox, D. P. Goldberg, J. Porphyrins Phthalocyanines 2004, 8,
1091–1105; g) S. Nardis, D. Monti, R. Paolesse, Mini-Rev. Org.
Chem. 2005, 2, 355–372; h) D. P. Goldberg, Acc. Chem. Res. 2007,
40, 626–634; i) I. Aviv, Z. Gross, Chem. Commun. 2007, 1987–1999.
[7] a) B. Ventura, A. Degli Esposti, B. Koszarna, D. T. Gryko, L. Fla-
migni, New J. Chem. 2005, 29, 1559–1566; b) T. Ding, E. A.
Alemꢁn, D. A. Mordarelli, C. J. Ziegler, J. Phys. Chem. A 2005, 109,
7411–7417; c) R. Paolesse, A. Marini, S. Nardis, A. Froiio, F.
Mandoj, D. J. Nurco, L. Prodi, M. Montalti, K. M. Smith, J. Porphyr-
ins Phthalocyanines 2003, 7, 25–36; d) J. H. Palmer, M. W. Day,
A. D. Wilson, L. M. Henling, Z. Gross, H. B. Gray, J. Am. Chem.
Soc. 2008, 130, 7786–7787; e) F. Nastasi, S. Campagna, T. H. Ngo,
W. Dehaen, W. Maes, M. Kruk, Photochem. Photobiol. Sci. 2011, 10,
143–150; f) D. Kowalska, X. Liu, U. Tripathy, A. Mahammed, Z.
Gross, S. Hirayama, R. P. Steer, Inorg. Chem. 2009, 48, 2670–2676.
[8] a) L. Flamigni, B. Ventura, M. Tasior, D. T. Gryko, Inorg. Chim.
Acta 2007, 360, 803–813; b) M. Tasior, D. T. Gryko, J. Shen, K. M.
Kadish, T. Becherer, H. Langhals, B. Ventura, L. Flamigni, J. Phys.
Chem. C 2008, 112, 19699–19709; c) F. D’Souza, R. Chitta, K.
Ohkubo, M. Tasior, N. K. Subbaiyan, M. E. Zandler, M. K. Rogacki,
&
14
&
www.chemeurj.org
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Photoinduced Electron Transfer
FULL PAPER
electron donors reported in Table 1. eS is the dielectric constant;
eS =2.4 for TL, eS =8.9 for CH2Cl2, eS =25.9 for BN, and ep =8.9 for
CH2Cl2. The centre-to-centre donor–acceptor distance calculated
after energy minimisation, RDA, is 14.7 ꢅ in C3+-PIꢀ and 20 ꢅ in
DPA+-C3-PIꢀ (average value), whereas the radii are taken as rD =
6 ꢅ (C3), rD =5 ꢅ (DPA), and rA =6 ꢅ (PI).
[34] D. F. Eaton, Pure Appl. Chem. 1988, 60, 1107–1114.
[35] L. Flamigni, A. M. Talarico, B. Ventura, R. Rein, N. Solladiꢇ, Chem.
Eur. J. 2006, 12, 701–712, and references therein.
Received: March 5, 2012
Revised: July 27, 2012
[33] D. S. Pedersen, C. Rosenbohm, Synthesis 2001, 2431–2434.
Published online: && &&, 0000
Chem. Eur. J. 2012, 00, 0 – 0
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
15
&
ÞÞ
These are not the final page numbers!

D. T. Gryko, L. Flamigni et al.
Double electron transfer: An assembly
consisting of four units, a meso-substi-
tuted corrole (C3), perylene bisimide
(PI), and two electron-rich triphenyla-
mine (DPA) units (see figure), is syn-
thesized and its photophysical charac-
teristics are determined. Excitation of
the corrole (C3) or perylene bisimide
(PI) units leads to the charge-sepa-
rated state DPA-C3+-PIꢀ with a rate
k>1011 sꢀ1 in benzonitrile and
Charge Separation
R. Voloshchuk, D. T. Gryko,*
M. Chotkowski, A. I. Ciuciu,
L. Flamigni* .................. &&&& — &&&&
Photoinduced Electron Transfer in an
Amine-Corrole-Perylene Bisimide
Assembly: Charge Separation over
Terminal Components Favoured by
Solvent Polarity
dichloromethane and with k of the
order of 1010 sꢀ1 in toluene.
&
16
&
www.chemeurj.org
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!