Fullerene C60-perylene-3,4:9,10-bis(dicarboximide) light-harvesting dyads: spacer-length and bay-substituent effects on intramolecular singlet and triplet energy transfer

DOI: 10.1002/chem.200800156

Fullerene C₆₀–Perylene-3,4:9,10-bis(dicarboximide) Light-Harvesting Dyads:

Spacer-Length and Bay-Substituent Effects on Intramolecular Singlet and

Triplet Energy Transfer

JØrôme Baffreau,^[a]StØphanie Leroy-Lhez,^[a]NguyÞn Vân Anh,^[b]

RenØ M. Williams,*^[b]and PiØtrick Hudhomme*^[a]

Abstract: Novel covalent fullerene C₆₀–

perylene-3,4:9,10-bis(dicarboximide)

(C₆₀–PDI) dyads (1–4) were synthe-

sized and characterized. Their electro-

chemical and photophysical properties

were investigated. Electrochemical

studies show that the reduction poten-

tial of PDI can be tuned relative to C₆₀

by molecular engineering through al-

tering the substituents on the PDI bay

region. It was demonstrated using

steady-state and time-resolved spec-

troscopy that a quantitative, photoin-

duced energy transfer takes place from

the PDI moiety, acting as a light-har-

vesting antenna, to the C₆₀unit, playing

the role of energy acceptor. The bay-

substitution (tetrachloro [1 and 2] or

tetra-tert-butylphenoxy [3 and 4]) of

the PDI antenna and the linkage

length (C₂[1 and 3] or C₅[2 and 4]) to

the C₆₀acceptor are important parame-

ters in the kinetics of energy transfer.

to two different conformers for each

C₆₀–PDI system. Subsequent triplet–

triplet energy-transfer times (from the

C₆₀unit to the PDI) are slower and in

the order of 0.8 ns (1), 6.2 ns (2), 2.7 ns

(3), and 9 ns (4). Nanosecond transient

absorption spectroscopy of final PDI

triplet states show a marked influence

of the bay substitution (tetrachloro- or

tetra-tert-butylphenoxy), and triplet-

state lifetimes (10–20 ms) and the PDI

triplet quantum yields (0.75–0.52) were

estimated. The spectroscopy showed no

substantial solvent effect upon compar-

ing toluene (non-polar) to benzonitrile

(polar), indicating that no electron

transfer is occurring in these systems.

Femtosecond

transient

absorption

spectroscopy indicates singlet–singlet

energy-transfer times (from the PDI to

the C₆₀unit) of 0.4 and 5 ps (1), 4.5 and

27 ps (2), 0.8 and 12 ps (3), and 7 and

50 ps (4), these values being ascribed

Keywords:

energy transfer

dyes/pigments

fullerenes

·

light-harvesting · perylenediimide

Introduction

Photoinduced energy- and electron-transfer processes are

crucial phenomena occurring in natural photosynthesis,^[1]

and the development of artificial photosynthetic systems is

still of considerable interest for the elucidation of the mech-

anisms that convert sunlight into chemical energy.^[2]Several

strategies have been employed to mimic this natural photo-

synthetic system.^[3]Among these, artificial light-harvesting

antennas incorporated into molecular electron-donor–ac-

ceptor systems are prone to convert excitation energy into

an electrochemical potential or chemical energy in the form

of a long-lived charge-separation state. Fullerene C₆₀is now

recognized to be a superior electron acceptor with regard to

its inherent redox properties, low reorganization energy,^[4]

and highly symmetrical 3D structure,^[5]and the porphyrins

are the most frequently used electron donors in photosyn-

thetic model systems.^[6]In these molecular donor–acceptor

dyads, intramolecular electronic interactions almost always

[a] Dr. J. Baffreau, Dr. S. Leroy-Lhez, Prof. P. Hudhomme

UniversitØ dꢀAngers, CNRS

Laboratoire de Chimie et IngØnierie MolØculaire dꢀAngers

CIMA UMR 6200, 2 Boulevard Lavoisier, 49045 Angers (France)

Fax : (+33)241-735-405

E-mail: pietrick.hudhomme@univ-angers.fr

[b] N. Vân Anh, Dr. R. M. Williams

Molecular Photonics Group

Vanꢀt Hoff Institute for Molecular Sciences

Universiteit van Amsterdam

Nieuwe Achtergracht 129, 1018 WS Amsterdam (The Netherlands)

Fax : (+31)205-256-456

E-mail: williams@science.uva.nl

Supporting information for this article is available on the WWW

under http://www.chemeurj.org/ or from the author: Color versions of

Figures 6–9, nano- and femtosecond transient absorption spectra of

dyads 1–4, ¹H NMR of compounds 1, 3, 12a, 13a, ¹³C and HMBC,

HMQC 2D NMR spectra of compounds 1 and 12a.

4974

ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Chem. Eur. J. 2008, 14, 4974 –4992

FULL PAPER

dominate, generating (long- or short-lived) charge-separated

states. Moreover, porphyrins have the advantage of being

excellent light-harvesting antennas.^[7]It was recently shown

that such architectures can be tailored to construct molecu-

lar photonic devices as well as artificial photosynthetic sys-

tems. This kind of photoinduced electron transfer was re-

cently extended to the production of photovoltaic cells that

convert light into electric energy.^[8]Attractive systems are

now developed for the construction of organic solar cells, es-

pecially incorporating fullerene C₆₀in the photoactive

layer.^[9]In this area, the direct covalent bonding of various p

donors to C₆₀has emerged as an active field of research.^[10]

However, for such application, back electron transfer and

energy transfer from the donor to C₆₀are events to be mini-

mized to increase the quantum yield and lifetime of charge

separation.^[11]Consequently, competition between energy-

and electron-transfer in a functionalized electron-donor–

fullerene dyad has to be carefully considered in designing

new systems for photovoltaic applications. In many exam-

ples it was shown that efficient photoinduced electron trans-

fer from a donor such as oligophenylenevinylene,^[12]oligo-

thiophene^[13]or perylene^[14]to fullerene C₆₀was the main

pathway, whereas energy transfer could appear as a compet-

itive process.^[15]

Following the observation of the very fast photoinduced

electron transfer occurring on the subpicosecond timescale

from conducting polymers to C₆₀,^[16]a major breakthrough

towards efficient organic photovoltaic devices was realized

with the development of the bulk-heterojunction concept.^[17]

This approach consists of generating an interpenetrating net-

work by blending the p-type electron-donating conjugated

polymer and C₆₀or another fullerene derivative as n-type

acceptor material.^[18]In this field, intensive efforts on organ-

ic solar cells are still focused on interpenetrating networks

using the soluble acceptor [60]PCBM.^[19]However, the dis-

advantage of this fullerene derivative is its small molar ab-

sorption coefficient in the visible region. Consequently, one

of the biggest challenges is now the access to materials that

present the optimal absorption range, matching as well as

possible the solar irradiation spectrum. This can be achieved

with the development of p-type low band-gap polymers^[20]

that should increase the absorption in the visible and near-

infrared (NIR) region of the solar spectrum. Another strat-

egy concerns dendrimer-based light-harvesting structures

that have attracted attention in the past decade.^[21]The func-

tionalization of C₆₀with oligomeric dendrons has given in-

teresting light-harvesting systems in which peripheral chro-

mophores are able to transfer the collected energy to the

central core of the dendrimer.^[22]Considering the n-type ma-

terial incorporated in the blend, the improvement of the

light absorption of fullerene derivatives and its relationship

with the efficiency of photovoltaic cells was demonstrated

with the replacement of the C₆₀derivative by the C₇₀ana-

logue.^[23]

Abstract in French: De nouveaux assemblages covalents ful-

lerne C₆₀–pØrylne-3,4:9,10-bis(dicarboximide) (C₆₀–PDI)

(1–4) ont ØtØ synthØtisØs et caractØrisØs. Les propriØtØs Ølec-

trochimiques et photophysiques ont Øgalement ØtØ Øtablies.

Par ingØnierie molØculaire et en jouant sur la nature des sub-

stituants prØsents sur le noyau PDI, le premier potentiel de

rØduction du PDI peut Þtre modulØ par rapport à celui de

C₆₀. Les Øtudes spectroscopiques à lꢀØtat stationnaire et en

temps rØsolu ont permis de mettre en Øvidence un transfert

dꢀØnergie photoinduit quantitatif intervenant de la partie PDI

jouant le rôle dꢀantenne vers lꢀaccepteur C₆₀. Par ailleurs, la

nature des substituants (tØtrachloro pour 1 et 2, ou tØtra-tert-

butylphØnoxy pour 3 et 4) sur lꢀunitØ PDI et la longueur de

lꢀespaceur (C₂pour 1 et 3 ou C₅pour 2 et 4) entre le PDI et

le fullerne C₆₀sont des paramtres importants sur les cinØti-

ques de transfert dꢀØnergie. La spectroscopie dꢀabsorption

transitoire femtoseconde indique des temps de transfert dꢀØ-

nergie singulet–singulet du PDI vers le C₆₀de 0.4 et 5 ps

(pour 1), 4.5 et 27 ps (pour 2), 0.8 et 12 ps (pour 3), 7 et

50 ps (pour 4), attribuØs à la prØsence de deux conformations

privilØgiØes pour chaque systme C₆₀–PDI. Par ailleurs, les

temps de transfert dꢀØnergie triplet–triplet de C₆₀vers le PDI

sont plus lents et de lꢀordre de 0.8 ns (pour 1), 6.2 ns (pour

2), 2.7 ns (pour 3) et 9 ns (pour 4). La spectroscopie dꢀab-

sorption transitoire nanoseconde montre que lꢀØtat triplet

final du PDI dØpend de la nature de la substitution. Des

durØes de vie de lꢀØtat triplet de 10–20 ms et des rendements

quantiques de 0.75–0.52 ont ØtØ dØterminØs. Aucun effet signi-

ficatif en fonction de la nature non-polaire (tolune) ou po-

laire (benzonitrile) du solvant nꢀa ØtØ mis en Øvidence, indi-

quant quꢀaucun transfert dꢀØlectron ne semble intervenir au

sein de ces systmes.

We are interested in the concept of linking a dye molecule

to fullerene C₆₀as a potential system presenting efficient

light-harvesting properties. The idea behind this is that the

dye could act as an antenna by absorbing sunlight, thereby

inducing an intramolecular energy transfer towards the full-

erene. Perylene-3,4:9,10-bis(dicarboximide) (PDI) dyes were

chosen as potential antennas because of their high chemical

stability, high photoluminescence quantum yield, and the

possibility for tuning the absorption range by variation of

the substitution pattern on the perylene core. During the

last few years, PDI derivatives have been developed as one

of the most useful classes of chromophores.^[24]One of the

most active areas of research involving PDI systems is relat-

ed to their utilization in photoinduced electron- and/or

energy-transfer processes. In this regard, electroactive units

such as fluorenone and anthraquinone,^[25]tetrathiafulva-

lene,^[26]pyrene,^[27]oligothiophene,^[28]zinc phthalocyanine,^[29]

3,4-ethylenedioxythiophene,^[30]zinc porphyrin,^[31]hexaazatri-

phenylenes,^[32]anthracene dendrimers,^[33]perylenemonoi-

mide,^[34]tetraboron-dipyrrin,^[35]and poly(fluorene-alt-phen-

ylene)^[36]were connected to the PDI dye to reach molecular

ensembles in which an electron- and/or energy-transfer

event occurs. In the search for such electronic interaction in-

volving PDI, supramolecular architectures were also recent-

ly described involving an oligo(p-phenylenevinylene),^[37]

a

Chem. Eur. J. 2008, 14, 4974 –4992

ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemeurj.org

4975

P. Hudhomme, R. M. Williams et al.

bis(ruthenium phthalocyanine)

moiety,^[38]or zeolite crystals.^[39]

In the last three years, and in

parallel to our first investiga-

tion,^[40]new electroactive PDI-

based systems involving the at-

tractive fullerene C₆₀electron

acceptor were considered and

the synthesis of many C₆₀–PDI

dyad or triad systems have now

been reported.^[41]Nevertheless,

the spectral and kinetic charac-

terization of the photoinduced

processes that are inherent to

the combination of these elec-

troactive units have received

little attention. It was only re-

cently shown that an efficient

photoinduced electron transfer

occurs in the C₆₀–PDI dyad in

which electron-donating pyrro-

Scheme 1. Representation of the photoactive layer of an organic solar cell incorporating the p-type P3HT

polymer donor and the C₆₀–PDI dyad.

to reach unsymmetrical 3,4:9,10-perylenetetracarboxylic

lidino groups are introduced on the bay region of the PDI

moiety.^[41g,h]The consequence of such substitution is the cor-

responding low oxidation potential of the PDI moiety, thus

facilitating the electron transfer from PDI to C₆₀.

monoanhydride monoimide derivatives non-substituted at

the bay region have been reported previously.^[42]Application

of these strategies on 1,6,7,12-tetrachloro-3,4:9,10-perylene

tetracarboxylic dianhydride 5 as starting material proved to

be fastidious. As an alternative method, we achieved the

direct condensation of PDI 5 with pentan-1-amine and 2-

aminoethanol or 5-aminopentan-1-ol in stoichiometric ratio

(Scheme 2). Unsymmetrical (6a or 6b) and symmetrical (7

and 8a or 8b) PDI compounds were efficiently separated by

silica-gel column chromatography thanks to the difference

in their respective polarity. This method constitutes an effi-

cient synthesis of unsymmetrical PDI 6a or 6b bearing the

alcohol functionality after its separation from N,N’-dipen-

tylPDI derivative 7 and symmetrical PDI dialcohols 8a or

8b.

The introduction of tert-butylphenoxy groups at the bay

region to achieve tert-butylphenoxyPDI 9a and 9b was car-

ried out using a nucleophilic substitution of chlorine atoms

with 4-tert-butylphenol in the presence of potassium carbon-

ate according to a well-known strategy.^[43]Another route

was attempted to synthesize compound 9a starting from

N,N’-dipentylPDI derivative 7 previously obtained as a by-

product. The latter was firstly transformed into the corre-

sponding tert-butylphenoxyPDI 10 in 69% yield following a

similar nucleophilic-substitution reaction. Subsequent treat-

ment with potassium hydroxide in tert-butyl alcohol^[42c,44]

gave the corresponding dianhydride 11 in 93% yield. Un-

symmetrical PDI 9a was obtained in 34% yield by conden-

sation of pentan-1-amine and 2-aminoethanol in stoichio-

metric ratio followed by separation from both symmetrical

PDI derivatives (Scheme 3).

In this paper we describe the synthesis as well as the elec-

trochemical and photophysical properties of C₆₀–PDI dyads

1–4 in the search for an efficient energy transfer from PDI

to C₆₀. In designing these systems it was assumed that the

distance between PDI and C₆₀moieties and their mutual ori-

entation could be crucial parameters for the rate of energy

transfer and could play an important role in the electronic

interaction between both partners. Another objective was to

demonstrate that the nature of the substituents on the PDI

bay region strongly influence the electronic properties of

these dyads for their further utilization in photovoltaic devi-

ces. In designing these solar cells it is considered that inside

the photoactive layer the following events occur: 1) self-as-

sembly into an interpenetrating nanoscopic network, 2) an

energy transfer from PDI towards fullerene C₆₀with the dye

acting exclusively as a light-harvesting antenna, and 3) a se-

lective electron transfer between the p-type polymer donor

such as poly3-hexylthiophene (P3HT) to the C₆₀unit

(Scheme 1). We report here on the second point, that is,

energy-transfer phenomena occurring from the PDI unit to-

wards fullerene C₆₀and on the influence of the spacer

length and bay substitution of the PDI on the kinetics and

spectral features.

Results and Discussion

Synthesis: The strategy employed for the preparation of

C₆₀–PDI dyads 1–4 is based upon the cyclopropanation reac-

tion, which has proven to be very efficient for the function-

alization of C₆₀. Consequently, initial efforts were focused

on the synthesis of unsymmetrical PDI precursors. Methods

Subsequent transformation of the alcohol into the malo-

nate functionality was performed using ethyl malonyl chlo-

ride in the presence of pyridine and corresponding deriva-

tives 12 and 13 were isolated in satisfactory yields

(Scheme 4). Reaction of cyclopropanation was carried out

4976

www.chemeurj.org

ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Chem. Eur. J. 2008, 14, 4974 –4992

Light-Harvesting Dyads

FULL PAPER

Scheme 2. i) C₅H₁₁NH₂/HO(CH₂)_nNH₂1:1, toluene, reflux, then column chromatography (6a: 28%, 6b: 17%).

N

Scheme 3. i) p-tert-Butylphenol, K₂CO₃, N-methylpyrrolidone, 1308C, (9a: 60%, 9b: 58%; 10: 69%); ii) KOH, tBuOH, reflux, then 1n HCl, 93%;

iii) C₅H₁₁NH₂/HO(CH₂)₂NH₂1:1, toluene, reflux, then column chromatography, 34%.

ACHTREUNG

by reaction with C₆₀in the presence of iodine and 1,8-

diazabicyclo

[5.4.0]undec-7-ene (DBU).^[45]Methano[60]fuller-

ene dyads 1–4 were purified by silica-gel column chromatog-

raphy.

With the aim of simplifying the synthetic scheme of dyads

3 and 4, we considered the straightforward post-functionali-

zation of dyads 1 and 2, respectively. Attempts to replace

chlorine atoms of dyad 1 by tert-butylphenoxy groups to

yield dyad 3 according to the reaction as described above

were unsuccessful. Using such experimental conditions on

dyad 1, the nucleophilic substitution of chlorine atoms by

tert-butylphenoxy groups was accompanied by the concomi-

tant cleavage of the malonate functionality and the unique

alcohol 9a was isolated in 66% yield.

ACHTREUNG

Chem. Eur. J. 2008, 14, 4974 –4992

ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemeurj.org

4977

P. Hudhomme, R. M. Williams et al.

on temperature was noted for

each proton, with a shielding

effect for A and B protons, and

on the contrary a deshielding

effect for A’ and B’ protons

(Figure 1d).

The ¹H NMR spectrum of

dyad 3 recorded in CDCl₃at

293 K presents a similar behav-

ior, but the signals correspond-

ing to the -NCH₂-CH₂O- link-

age appear broader than in the

case of dyad 1. We assign this

difference to spin relaxation re-

sulting from the strong steric

hindrance between the fuller-

ene and the tert-butylphenoxy

groups on the bay region of the

PDI. Here again, the -NCH₂-

CH₂O- linkage appears as a

series of four signals (4.90, 4.82,

4.72, and 4.64 ppm) that are de-

Scheme 4. i) ClCOCH₂CO₂Et, pyridine, CH₂Cl₂, (12a: 93%, 12b: 94%, 13a: 91%, 13b: 88%); ii) C₆₀, DBU,

I₂, toluene, (1: 53%, 2: 42%, 3: 47%, 4: 51%).

¹H NMR spectroscopy: To evaluate the extent of the inter-

action between C₆₀and PDI units in solution, ¹H NMR spec-

shielded relative to precursor 13a (up to Dd=0.45 ppm). A

variable-temperature NMR study showed a clear narrowing

of all the peaks with a coalescence phenomenon arising at

313 K for both -NCH₂and -OCH₂groups (Figure 1e).

tra of dyads 1 and 3 were compared with those of their mal-

1

onate precursors 12a and 13a, respectively. H NMR spectra

of compounds 12a and 13a (CDCl₃, 293 K) show that the

four protons of the -NCH₂-CH₂O- linkage fortuitously pres-

ent a single signal at 4.55 ppm for PDI 12a and at 4.45 ppm

for PDI 13a. This observation was confirmed by HMBC and

HMQC 2D NMR correlated spectra carried out on com-

pound 12a (Supporting Information). It is worth noting that

the case of dyad 1 appears very different from its precursor

1

These H NMR studies clearly demonstrate the presence

of a restricted rotation inducing conformational effects

around the -NCH₂CH₂O- linkage in dyads 1 and 3. This phe-

nomenon as well as the temperature-dependent dynamic

effect observed for dyad 3 can be unambiguously attributed

to the presence of C₆₀.^[49]

As a high purity of the compounds is crucial for further

photophysical studies, an efficient way to purify dyads 1–4

and reference compounds was needed. For this purpose, an-

alytical and semi-preparative HPLC were carried out on a

silica column using toluene/CH₂Cl₂3:1 for dyad 1–4 and tet-

rachloro-substituted PDI 7. For tetraphenoxy-substituted

PDI 10 and 1,1-di(ethoxycarbonyl)-[6,6]-methanofullerene

14, the latter being prepared starting from a-bromoethyl

malonate as reported in the literature (Scheme 5),^[50]toluene

was used as the eluent for analytical and semipreparative

HPLC. The purity of dyads 1–4 was estimated to be superior

to 99.9%.

1

12a. In the H NMR spectrum of dyad 1 recorded in CDCl₃

at 293 K, the ethylene linkage appears as a four-proton spin

system, demonstrating the non-equivalence of the four pro-

tons (Figure 1a). At this stage it is important to note that

the C₆₀unit in dyad 1 is also responsible for an important

deshielding effect on these four -NCH₂-CH₂O- protons (up

to Dd=0.54 ppm) compared to 12a, as well as for the -CH₂-

protons of the ethyl ester functionality (Dd=0.44 ppm). In

agreement with HMQC correlation technique, both signals

centered at 5.09 and 4.78 ppm were assigned to the -OCH₂

group (noted H_Aand H_A’, respectively), and signals at 4.92

and 4.62 ppm to the -NCH₂group (noted H_Band H_B’, re-

spectively) (Figure 1a). The corresponding coupling con-

stants are of ²J=14 Hz (geminal protons), ³J=3.5 and

8.5 Hz. According to the Karplus equation^[46]and its applica-

tion in determining the conformations of ethane deriva-

tives,^[47]dihedral torsion angles of approximately 458 and

Electrochemical properties: Cyclic voltammetry was used to

electrochemically characterize dyads 1–4 (Table 1). Com-

pounds 7, 10, and 14 were used as references allowing the

determination of the nature of reduced species. Dyad 1

showed three reversible reduction waves and the first one-

electron process at E⁰_red1=ꢀ0.84 V (vs Fc⁺/Fc) was assigned

to the formation of the anion radical of the PDI moiety

3

1658 were estimated for dyad 1 using these J coupling con-

stants. Consequently, two preferential conformations of the

-NCH₂-CH₂O- linkage can be considered from these param-

eters (Figure 1b), as was also predicted by theoretical calcu-

lations.^[48]Variable-temperature NMR experiments were

conducted at 313 K and 343 K, but no evolution in the dif-

ferent coupling constants was observed (Figure 1c). Interest-

ingly, a significant linear dependence of the chemical shift

^ꢀC

(C₆₀–PDI ). The following two one-electron reduction pro-

cess at E⁰_red2=ꢀ1.08 V corresponding to the generation of

^ꢀC

2ꢀ

the C₆₀–PDI species suggested that the first reduction

wave of fullerene and the second reduction wave of PDI

were overlapping. The third one-electron wave appearing at

4978

www.chemeurj.org

ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Chem. Eur. J. 2008, 14, 4974 –4992

Light-Harvesting Dyads

FULL PAPER

Figure 1. a) ¹H NMR spectra of the -NCH₂-CH₂O- linkage in dyad 1 at 293 K; b) folded (left) and extended (right) conformations in dyad 1 derived from

the coupling constants using the Karplus equation; c) ¹H NMR spectra of dyad 1 at 293 K (left), 313 K (center), and 343 K (right); d) linear temperature

1

~

!

&

*

dependence of protons A ( ), B ( ), A’ ( ), and B’ ( ) chemical shifts in dyad 1; e) H NMR spectra of dyad 3 at 293 K (left), 313 K (center), and

343 K (right).

Chem. Eur. J. 2008, 14, 4974 –4992

ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemeurj.org

4979

P. Hudhomme, R. M. Williams et al.

tert-butylphenoxy groups at the bay region, the C₆₀moiety

appears to be the favored electron acceptor. In the latter

case, the electron-donating mesomeric effect of phenoxy

groups appeared predominant with less consideration for

their possible electron-withdrawing inductive effect. This

demonstrates that electronic properties of C₆₀–PDI dyads

can be perfectly tuned thanks to a well-defined substitution

of the PDI bay region.

Scheme 5. i) EtO₂C-CHBr-CO₂Et, NaH, toluene, 57%.

Electronic absorption: Elec-

tronic absorption spectra of

Table 1. Redox-potential values (vs Fc⁺/Fc) of dyads 1–4 and reference compounds 7, 10, and 14 recorded in

a CH₂Cl₂solution using Bu₄NPF₆0.1m as the supporting electrolyte, Ag wire as the reference, platinum wires

dyads 1–4 and reference com-

pounds 7, 10, 14 were recorded

in toluene and dichloromethane

at room temperature (Figure 2

and Table 2). No significant sol-

vent effect was observed, but

further insight into solvato-

chromism was prevented by the

lack of solubility of the dyads.

All compounds containing the

PDI unit exhibit the three typi-

cal absorption peaks at approxi-

mately 430, 485, and 520 nm for

as counter and working electrodes. Scan rate: 100 mVs^ꢀ1

.

Compound

E⁰[V]

red4

red3

red2

red1

ox1

ox2

1

2

3

4

7

10

14

–

ꢀ1.47

ꢀ1.45

ꢀ1.35

ꢀ1.37

ꢀ1.08^[a]

ꢀ1.07^[a]

ꢀ1.24

ꢀ1.23

ꢀ1.06

ꢀ1.37

ꢀ1.47

ꢀ0.84

ꢀ0.86

ꢀ1.10

ꢀ1.09

ꢀ0.87

ꢀ1.21

ꢀ1.08

+1.24^[b]

–

+0.83

+0.82

––

ꢀ1.87^[b]

ꢀ1.46

ꢀ1.44

+1.23^[b]

–

+0.80

+1.14

+1.22^[b]

–

[a] Two one-electron process. [b] Irreversible process.

E⁰_red3=ꢀ1.47 V resulted from the formation of C₆₀^2ꢀ–PDI^2ꢀ

species. One irreversible oxidation process was observed at

E⁰_ox1=+1.24 V.

1, 2, and 7 or 450, 540, and 580 nm for 3, 4, and 10. The

values reported for a PDI derivative bearing hydrogen

atoms at the bay region are intermediate (434, 460, 491, and

528 nm).^[52,53]In agreement with information issued from

electrochemical analyses, the hypsochromic effect observed

for 1, 2, and 7 could be ascribed to the bay-substitution pat-

tern of the PDI moiety and, more precisely, of the electron-

withdrawing inductive effect of chlorine atoms. On the

other hand, the bathochromic effect for 3, 4, and 10 could

be related to the electron-donating mesomeric effect of

phenoxy groups. It appears that electronic effects are pre-

dominant.^[24]The expected hypsochromic effect due to the

decrease of conjugation because of the torsion induced by

the bay substitution is negligible. According to litera-

ture,^[52,53]these bands could be attributed to the transitions

3–0, 1–0, 0–0, with a weak transition 2–0 that could be only

Cyclic voltammogram of dyad 3 showed four reversible

one-electron reduction waves. The first process arising at

E⁰_red1=ꢀ1.10 V (vs Fc⁺/Fc) was assigned to the formation

^ꢀC

of the anion radical of the C₆₀moiety (C₆₀–PDI). This was

followed by both one-electron reduction processes of the

^ꢀC

PDI moiety leading successively to C₆₀–PDI (E⁰

=

red2

^ꢀC

ꢀ1.24 V) then C₆₀–PDI species (E⁰_red3=ꢀ1.35 V). The

fourth one-electron reduction wave was corresponding to

the generation of the C₆₀^2ꢀ–PDI^2ꢀspecies at E⁰_red4=ꢀ1.46 V.

A first reversible one-electron oxidation process arising at

E⁰_ox1=+0.83 V was followed by an irreversible wave ap-

pearing at E⁰_ox2=+1.23 V.

2ꢀ

Considering the cyclic voltammogram of dyads 2 and 4, it

should be noted that the influence of the length of the alkyl

chain on the imide position is not characteristic. The reason

is that the imide nitrogen atoms correspond to nodes in the

LUMO of PDI derivatives, thus indicating that substituents

at those positions do not really influence the corresponding

energy levels.^[51]

It was reported in the case of C₆₀–PDI dyads with PDI

bearing only hydrogen atoms at the bay region^[41d]that the

first reduction potential corresponds to a reversible two

^ꢀC

one-electron process leading directly to the C₆₀–PDI spe-

cies. In the case of dyads 1 and 2 with the PDI substituted

by four chlorine atoms at the bay region, the first reduction

occurs on the PDI moiety. The enhancement of the electron

affinity is explained by the electron-withdrawing inductive

effect of the chlorine atoms stabilizing the anion radical

Figure 2. Absorption spectra of dyads 1 (c) and 3 (c) and reference

^ꢀC

PDI , this minimizing their possible electron-donating

mesomeric effect. For dyads 3 and 4 with the PDI bearing

compounds

7 (g), 10 (a), and 14 (b) in toluene at 298 K

(c<10^ꢀ6m).

4980

www.chemeurj.org

ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Chem. Eur. J. 2008, 14, 4974 –4992

Light-Harvesting Dyads

FULL PAPER

Table 2. Electronic absorption data for dyads 1–4 and reference compounds 7, 10, 14.^[a]

chromic shift of the emission

maxima was noticed in agree-

ment with the absorption spec-

trum. The consequence is an

overlapping of emissions cen-

tered on the PDI and C₆₀moi-

eties.

The Stokesꢀ shift values in

both toluene and dichlorome-

thane are in the 700–900 cm^ꢀ1

range, except for 14 for which

the Stokesꢀ shift value is smaller

Toluene

Dichloromethane

l_max[nm]

e_max[m^ꢀ1cm^ꢀ1

]

l_max[nm]

e_max[m^ꢀ1cm^ꢀ1

]

1

2

3

4

7

10

14

524

522

579

576

522

575

29800

31350

36800

37300

33900

42000

522

520

586

582

518

582

32800

38500

40300

42700

689 (490, 428)

250 (1700, 3000)

693 (496, 431)

105 (1400, 2500)

[a] Other absorption bands of compound 14 and their respective molar absorption coefficients.

barely distinguished around 460 nm for 7 and 500 nm for 10.

The extinction coefficient values found are in agreement

with those reported in literature and are within the range

30000–40000 cm^ꢀ1for these compounds.^[26b,54]The strong

absorption band at 330 nm observed for 1–4 and 14 was as-

cribed to fullerene C₆₀absorption. The sharp peak at 430 nm

observed for 14 is characteristic of [6,6]-methanofullerene

derivatives, as well as the peak at approximately 690 nm

(Table 3), all these values being coherent with a weak geo-

metrical reorganization of the considered fluorophore at the

excited state (PDI for 1–4, 7, 10 and C₆₀for 14). Moreover,

!

that corresponds to the S₁S₀transition.^[55]Rather similar

spectral features are also observed for pyrrolidine C₆₀deriv-

atives^[55b]and for C₆₀Diels–Alder adducts.^[55c]Moreover, the

UV-visible absorption spectra of dyads 1–4 match the profile

obtained by superposition of the spectra of corresponding

reference compounds 7 or 10 and 14. Independently of the

bay-substitution pattern or spacer length between PDI and

C₆₀moieties, this characterizes the absence of significant in-

teraction in the ground state between these two partners.

Fluorescence emission: Fluorescence emission spectra of

reference compounds 7, 10, and 14 were recorded in toluene

and dichloromethane and were found to be independent of

the excitation wavelength (Figure 3A and Table 3).

Fluorescence excitation spectra matched the absorption

profile over the entire wavelength range. As expected, PDI-

based compounds 7 and 10 are strongly emissive in the visi-

ble range with fluorescence quantum yields being close to

unity at room temperature. On the other hand, compound

14 presented a typical weak emission centered at 697 nm

with a fluorescence quantum yield of approximately 510^ꢀ4

at room temperature, characteristic of [6,6]-methanofuller-

ene derivatives.^[55–57]All the fluorescence emission spectra of

dyads 1–4 are composed of two bands, one characteristic of

the PDI moiety in the 500–650 nm range, the second corre-

sponding to the methanofuller-

Figure 3. Emission spectra of A) reference compounds 7 (c, l_exc

=

485 nm), 10 (a, l_exc=550 nm), and 14 (c, l_exc=450 nm) and

B) dyads 1 (c, l_exc=520 nm) and 3 (c, l_exc=550 nm) in toluene at

298 K (c<10^ꢀ6m).

ene in the 670–750 nm region

Table 3. Photophysical properties (emission) of dyads 1–4 and reference compounds 7, 10, 14 at 298 K.

(Figure 3B and Table 3). The

very weak importance of C₆₀

emission in dyad 2 relative to

dyad 1 should be noticed. The

spacer length between PDI and

C₆₀seems to have an influence

on the relative intensity of

these two bands. For dyads 3

and 4 bearing tert-butylphenoxy

Toluene

F_f

Dichloromethane

l_max[nm]

Dn˜ [cm^ꢀ1

]

t [ps]

k_en10¹⁰[s^ꢀ1

]

l_max[nm]

Dn˜ [cm^ꢀ1

]

F_f

1

2

3

4

7

10

14

549 (696)

548 (696)

602 (698)

600 (697)

548

909

923

808

694

923

167

3.310^ꢀ3

3.810^ꢀ3

3.410^ꢀ3

4.610^ꢀ3

1

9

30

25

11.1

3.3

3.9

545 (694)

545 (696)

609 (694)

608 (694)

923

808

882

808

923

1

2.310^ꢀ3

3.110^ꢀ3

2.510^ꢀ3

3.210^ꢀ3

54

1.8

5580

5950

1500^[a]

–544

–607

–696

601

697

1

882

62

1

5.010^ꢀ4

5.010

ꢀ4

groups at the bay area, a batho- [a] In good agreement with ref. [55a].

Chem. Eur. J. 2008, 14, 4974 –4992

ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemeurj.org

4981

P. Hudhomme, R. M. Williams et al.

there is a dramatic quenching of the PDI fluorescence emis-

sion for all the dyads. At room temperature, the fluores-

cence quantum yields are in the range 2–310^ꢀ3, this being

also dependent on the spacer length. Considering identical

bay substitution, the shorter the alkyl chain between PDI

and C₆₀units, the stronger the fluorescence quenching.

This strong quenching (up to 99.8%, depending on the

dyad) should result from an intramolecular interaction be-

tween PDI and C₆₀in the excited state as the quantum yield

of neither 7 nor 10 was affected by the addition of up to ten

equimolecular amounts of 14 in solution. To get further in-

sights into the nature of this new process, the fluorescence

emission spectrum of 1 was recorded by selectively exciting

the PDI unit in the dyad (Figure 3B). The C₆₀emission fea-

tures are still observed, evidencing a singlet–singlet energy

transfer between PDI and C₆₀. This process is in agreement

with the energy levels of both PDI and C₆₀units, 2.32 eV for

7 and 1.79 eV for 14 in toluene. This was confirmed by the

excitation spectrum recorded at 695 nm in dichloromethane.

Indeed, the characteristic absorption bands of PDI are still

present in this spectrum (Figure 4).

than with tert-butylphenoxy-substituted dyad 3 (Table 3).

This phenomenon is confirmed by the emission spectra of

dyads 1–4 recorded in rigid solvent glass (methylcyclohex-

ane/toluene 7:3) at low temperature (77 K) (Figure 5A and

B). A slight bathochromic shift with decreasing temperature

for fluorescence emission of dyads 1 and 2 was noticed

(Dl_max=11 nm between 77 K and room temperature). These

spectra showed well-structured vibronic bands in agreement

with the lack of solvent relaxation in rigid medium. Conse-

quently, at 77 K a better comparison of the intensity ratio

between the PDI emission band (550 nm for dyads 1 and 2

and 600 nm for dyads 3 and 4) and the C₆₀emission band

(ca. 700 nm) could be performed. Considering the same bay

substitution and PDI fluorescence-emission intensity being

normalized, it was clearly displayed from these spectra that

C₆₀-centered emission is more important for dyads present-

ing the shortest spacer, thus confirming the real influence of

the spacer length on the efficiency of energy transfer be-

tween PDI and C₆₀. Furthermore, no phosphorescence emis-

sion was detected under these experimental conditions.

Figure 4. Excitation spectra of dyad 1 at l_obs=570 nm (g) and at l_obs

695 nm (a) in dichloromethane at 298 K (c<10^ꢀ6m).

=

Similar experiments could not be conducted for dyads 3

or 4 as fluorescence emissions of PDI and C₆₀are overlap-

ping. Nevertheless, by subtracting the contribution of PDI

from the emission spectrum of 3 or 4, the fullerene emission

is still observed, even when the PDI is selectively excited at

wavelengths above 550 nm.^[40]This result associated with the

strong fluorescence quenching of PDI emission corroborates

the hypothesis of an energy transfer from PDI to C₆₀for

these dyads. Energy levels are still in agreement as the first

excited state of 10 is lying at 2.11 eV, that is, 0.32 eV above

the first excited state of com-

Figure 5. Normalized fluorescence emission spectra at 77 K in methylcy-

clohexane/toluene 7:3 matrix: A) for dyads 1 (c) and 2 (a, l_exc

490 nm); B) for dyads 3 (c) and 4 (a, l_exc=450 nm).

=

pound 14 in toluene (Table 4).

Table 4. Energy level of the charge-separated state for dyads 1–4 and parameters used for the calculation.

Altogether these results un-

derline the efficiency of this

process as a function of bay-

area substitution and spacer

length. Indeed, the quenching

of fluorescence appears higher

Energy level [eV]

[a]

Dyad

E_ox[eV]

E_red[eV]

E^0ꢀ0[eV]

DG [eV]

E_PDI

E_CS

E_C60

1

2

3

4

+1.24

+0.83

ꢀ0.84

ꢀ1.10

2.32

2.11

ꢀ0.24

ꢀ0.18

2.33

2.11

2.09

1.93

1.79

for chlorine-substituted dyad 1 [a] vs Fc⁺/Fc in CH₂Cl₂.

4982

www.chemeurj.org

ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Chem. Eur. J. 2008, 14, 4974 –4992

Light-Harvesting Dyads

FULL PAPER

The intramolecular energy transfer from PDI to C₆₀could

be discussed in terms of Coulombic interactions (Fçrster)^[58]

or exchange interactions (Dexter).^[59]

vent was also used to check the presence of electron transfer

in polar solvents in some femto- and nanosecond transient

absorption experiments.

In order to know if this energy transfer could be due to

Coulombic interaction, Fçrster radii R₀were calculated for

each dyad (Table 5), according to Equation (1):^[58]

Single photon counting (SPC): SPC was used to determine

the emission decays of the dyads 1–4 and the reference sys-

tems 7, 10, and 14 in toluene. The decay curves recorded for

dyads 1 and 2 (at 550 nm) and dyads 3 and 4 (at 610 nm)

monitoring the PDI emission gave bi-exponential decays.

The major (95–99%) short components (9–54 ps) as well as

the energy-transfer rates (k_en) derived are given in Table 3.

Figure 6 shows the decay traces of the four C₆₀–PDI sys-

tems. Interestingly, a minor long component (0.2–4%, 3–

5 ns) is observed in the time-resolved emission.^[61]Table 3

shows the mono-exponential fluorescence lifetimes of the

reference compounds 7, 10, and 14. The lifetimes are virtual-

ly independent of deaerating and are in accordance with re-

ported lifetimes for similar systems.^{[27,53,55–57,62]}

1

Z

R₀¼ 0:2108½k²ꢀ_D⁰n^ꢀ4

I_DðlÞe_AðlÞl⁴dlꢁ

1=6

ð1Þ

0

where k is the orientation factor, ꢀ⁰_Dis the emission quan-

tum yield of the reference donor chromophore, I_D(l) is the

corrected fluorescence intensity of the donor with the total

intensity normalized to unity and n is the refractive index of

2

the solvent. For dyads 1–4, ꢀ⁰_Dis equal to one and k²to =

3

because of the random orientation of the dipole between

molecules. All the values obtained for R₀in both toluene

and dichloromethane are superior to the distance calculated

by molecular optimization between PDI and C₆₀for each

dyad.^[48]According to this result, the nature of the energy

transfer to dyads 1–4 should be of Fçrster type. Using these

R₀values and the experimental rates, center-to-center dis-

tances could be estimated (Supporting Information).

Table 5. Calculated Fçrster radii R₀[] for dyads 1–4.

Toluene

Dichloromethane

1

2

3

4

32.3

28.1

31.5

30.0

To study the possibility of a competitive reductive (for

dyads 1 and 2) or oxidative (for dyads 3 and 4) electron-

transfer process, the “Gibbs energy of photoinduced elec-

tron transfer” was estimated for each dyad.^[60]For all the

dyads, the corresponding charge-separated state lies at an

intermediate energy level between the fullerene and PDI

singlet-excited states (Table 4). If the Coulomb and solvent-

correction terms are taken into account, the energy level

will be lowered, making it energetically more accessible in

dichloromethane and benzonitrile. In toluene, however,

charge separation would be an endergonic process (using

“average” center-to-center distances, listed in Supporting In-

formation, and average ionic radii of 4 ). Apparently the

excited-state processes in polar solvents are governed by ki-

netic factors: charge separation is energetically possible, but

it is not observed (see below).

Figure 6. Emission decay traces obtained with SPC of compounds 1–4 in

toluene. Pulse is also shown (l_ex=324 nm, 17 ps FWHM, detection wave-

length is indicated).

By using the quenched lifetimes and applying k_en=1/tꢀ

1/t₀, singlet energy-transfer rates of 11.110¹⁰s^ꢀ1(for 1),

3.310¹⁰s^ꢀ1(for 2), 3.910¹⁰s^ꢀ1(for 3), and 1.810¹⁰s^ꢀ1

(for 4) were obtained (Table 3). This correlates reasonably

well with the rates that can be deduced from the steady-

state data (Table 3). From these data it is already clear that

the rate of energy transfer is modulated both by the bay

substituent and by the spacer length. The shortest spacers

display the faster rates. Considering the same spacer length,

energy transfer is more efficient for the chlorinated than for

the tert-butylphenoxy-substituted dyads. It can be noted that

the energy-transfer rates obtained for dyads 1–4 are higher

than those reported in the literature for other C₆₀fullerene-

based dyads.^{[14,15,41,63]}

Time-resolved spectroscopy: To get more information on

the effects of the bay substitution and the spacer length on

the interaction between the fullerene unit and the PDI chro-

mophore, various types of time-resolved spectroscopy were

applied. The photophysical properties of the systems (1–4)

containing fullerene C₆₀covalently linked to the perylenedi-

imide as well as those of the reference compounds 7, 10,

and 14 in toluene were investigated. Benzonitrile as a sol-

Femtosecond transient absorption:^[64]Femtosecond transient

absorption was used to get more accurate information on

Chem. Eur. J. 2008, 14, 4974 –4992

ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemeurj.org

4983

P. Hudhomme, R. M. Williams et al.

the singlet and triplet energy-transfer processes in the dyads.

The spectral data obtained for 2 in toluene are shown in

Figure 7. The typical bands of the PDI moiety, such as the

ground-state bleaching (at 500 and 525 nm), the singlet-state

emission (at ca. 600 nm), the singlet-excited-state absorption

(at ca. 800 nm), and the triplet-excited-state absorption (at

ca. 560 nm) are observed, similar to those reported before

for 1.^[65]The kinetics of 2, however, are different, showing a

slower decay of the PDI excited-singlet state and a slower

rise of the 560 nm band. For all dyads (1–4) the singlet excit-

ed state of the fullerene adduct is observed as an intermedi-

ate state (at 80 ps for 2, see Figure 7A). Its absorption is

rather weak and covers the whole visible range, with a mini-

mum at around 600 nm.^[65,66]

slower singlet and triplet energy transfer. Table 6 summariz-

es the characteristic times observed with femtosecond transi-

ent absorption for the dyads 1 to 4 in toluene. See Support-

ing Information for corresponding rates.

Table 6. Decay and rise times observed for dyads 1–4 in toluene with

femtosecond transient absorption spectroscopy, corresponding to singlet–

singlet energy transfer (t₁, t₂) and triplet–triplet energy transfer (t₃).

t₁[ps]

t₂[ps]

t₃[ns]

1

2

3

4

0.4

4.5

0.8

7

5

27

12

50

0.8

6.2

2.7

9

Overall, the phenoxy-substituted compounds 3 and 4 dis-

play similar behavior to 1 and 2, but the spectral features of

the PDI singlet and triplet excited states are different

(Figure 8), especially the modulation by the strong ground-

state bleaching at 580 nm for 3 and 4. The phenoxy-substi-

tuted PDI moiety shows ground-state bleaching (at 540 and

580 nm), the singlet-state emission (at ca. 600 and 650 nm),

and a weak singlet excited-state absorption (at ca. 700 nm).

The final state populated (the triplet of PDI) is character-

ized by maxima at 510, 550 and by the strong bleach at

580 nm and an almost flat absorption band between 600 and

800 nm. Again, the dyad with the longer spacer (4) shows

Interestingly, the ultrafast component observed in the

femtosecond transient absorption spectroscopy is not detect-

ed with SPC, due to its lower time resolution. The slower

components (t₂) are in reasonable agreement with the fluo-

rescence emission lifetimes (Table 3).

Femtosecond transient absorption spectroscopy was also

performed in benzonitrile for compounds 1 and 3 to assess

solvent-dependent photophysical behavior (i.e., the occur-

rence of electron transfer). Very similar spectral features

were observed as well, as slightly different kinetics (see Sup-

porting Information), the latter being attributed to small dif-

ferences in solvent-dependent

conformations

resulting

in

slightly altered interchromo-

phoric distances in the two sol-

vents for the extended and

folded forms. This clearly indi-

cates that

a

charge-transfer

state is not formed (or has no

population build-up). Further-

more, the singlet energy trans-

fer from the PDI to the C₆₀unit

and subsequent triplet energy

transfer from the fullerene trip-

let state to the perylene-based

triplet are relatively insensitive

to the solvent polarity.

The femtosecond spectrosco-

py confirms that light energy

harvested by the PDI antenna

is transferred very efficiently to

the C₆₀moiety on a (sub)pico-

second (for 1) to tens-of-pico-

second timescale and that no

electron transfer occurred be-

tween PDI and C₆₀moieties.

No spectral features attributa-

ble to either the PDI radical

anion or cation,^[67]or the C₆₀

radical cation^[68]or anion^[69]

were observed. Using NIR de-

Figure 7. A) Femtosecond transient absorption spectra of compound

FWHM); kinetic traces B) at 600 nm on a 100-ps timescale and C) at 560 nm on the full timescale with tri-ex-

2 in toluene (l_ex=530 nm, 150 fs

ponential fit (t₁=4.5 ps, t₂=27 ps, and t₃=6.2 ns). Incremental time delays are indicated in A).

tection, compounds 1 and 3

4984

www.chemeurj.org

ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Chem. Eur. J. 2008, 14, 4974 –4992

Light-Harvesting Dyads

FULL PAPER

Figure 8. A) 3D surface plot of the femtosecond transient absorption data obtained for 3 in toluene (l_ex=530 nm, 150 fs FWHM). Signal intensity is also

projected on the x-y plane. The incremental time delay is short (0.04 ps) at the start and longer (50 ps) at later times (total timescale ꢀ10 to 3600 ps).

Note that the time increases to the front and the wavelength increases to the left. B) Femtosecond transient absorption spectra of compound 3. Incre-

mental time delays are indicated in the legend of (B). Kinetic traces at 600 nm (C) and at 500 nm (D) together with tri-exponential fit (t₁=0.8 ps, t₂=

12 ps, and t₃=2.7 ns).

show only singlet–singlet absorption in toluene^[41h]at 990 nm

for 1 and at 960 and 1040 nm for 3 (see Supporting Informa-

tion) with kinetics similar to the singlet excited-state kinetics

observed with the femtosecond transient absorption data

using visible-light detection. Moreover, the final excited

state is a low-energy (1.2 eV) triplet state localized on the

PDI,^[53]normally rarely observed in perylene dyes. Two rates

of fast singlet energy transfer are observed, which were at-

tributed to the co-existence of a folded and an extended

conformer. The two faster rates observed with transient ab-

Chem. Eur. J. 2008, 14, 4974 –4992

ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemeurj.org

4985

P. Hudhomme, R. M. Williams et al.

sorption show that the interchromophoric distance in the

folded conformer of 2 and 4 is longer than for 1 and 3, in

agreement with DFT and Hartree–Fock calculations.^[48]

perfect. Tetrachloro-substituted PDI (1 and 2) are character-

ized by a very strong triplet-excited state absorption at

560 nm. Ground-state bleaching features are observed at

490 and 525 nm. Tetra-tert-butylphenoxy-substituted PDI (3

and 4) show a markedly different spectrum, due to the

strong ground-state bleaching at 580 nm, characterized by

maxima at 510 and 550 nm (for the spectra in benzonitrile,

see Supporting Information). The observed triplet-state life-

times in air and under deaerated conductions are given in

Table 7. Notably, the lifetimes under degassed conditions

Nanosecond transient absorption: The ultrafast spectroscopy

described in the previous section clearly shows the forma-

tion of a longer-lived triplet state that is localized on the

PDI unit in all four dyads. The triplet–triplet absorption

spectra of the PDI chromophores observed with nanosecond

transient absorption spectroscopy are given in Figure 9. The

similarity to the final femtosecond transients (at 3.6 ns) is

Table 7. Triple-state lifetimes of compounds 1–4 in toluene (tol) and ben-

zonitrile (PhCN) under aerated (air) and de-aerated (Ar) conditions as

observed by using nanosecond transient absorption spectroscopy.

t [ns, tol, air]

t [ns, tol, Ar]

t [ns PhCN, air]

t [ns, PhCN, Ar]

1

2

3

4

440

400

330

310

12000

10000

14000

15000

640

590

680

590

15000

14000

20000

are very sensitive to traces of oxygen left in solution and are

influenced by triplet–triplet annihilation (and can thus be

considered lower limits). The effects of de-oxygenation and

comparison of the spectral shape with literature data^[53,70]

clearly show the nature of the PDI-localized triple state.^[71]

The spectroscopy of compound 1 does not reveal any indi-

cation of C₆₀triplet-state formation in the femtosecond and

nanosecond transient absorption, and its intermediacy is in-

ferred from the intrinsic fullerene properties. This is differ-

ent, however, for compound 4. Figure 9C shows the clear ki-

netic difference for the 500 and 710 nm bands in the first

50 ns (Supporting Information Figure S6 for spectra of 1

and 2). From these data an initial rise time of 9 ns is de-

duced (at 500 nm), which corresponds to the triplet energy

transfer from C₆₀to the PDI. Now, a second, slower triplet

energy transfer (in 33 ns) is inferred from the decay at

710 nm. A rise time at 500 nm is not observed due to over-

lapping bands and the small yield of this slower process.

However, in a large part of the population of excited-state

molecules, the fullerene triplet does not transfer its energy

and decays with a lifetime very similar to that of the PDI

triplet (Table 7). Clearly, the two conformations deduced

from two singlet energy-transfer rates in the femtosecond

transient now also give two different triplet energy-transfer

rates.

Similar observations were made for compound 3, for

which the 710 nm band is clear in the nanosecond transient.

Closer inspection of Figure 7 also reveals a spectroscopic

Figure 9. Nanosecond transient absorption spectra of the triplet state in

toluene of A) compound 2, incremental time delay is 2500 ns (argon de-

gassed); B) compound 3, incremental time delay is 120 ns non-degassed;

C) compound 4, incremental time delays is 1 ns. The first spectrum is

dotted in A) and B) (time increases from grey to black). A) and B) pres-

ent the decay of the PDI triplets. In B) some triplet C₆₀is observed at

710 nm. This 710 nm band shows a different decay behavior in (C), in

which also the rise of the PDI triplet can be observed. For C) the time

zero was chosen slightly before the laser pulse.

3

sign for C₆₀formation in compound 2, by the small hump

at 710 nm in the last trace (at 2988 ps). In the nanosecond

transient, however, this appears to be masked by the strong

3

and more red-shifted PDI absorption (Figure 9). The fea-

ture at 710 nm is also not observed for compound 1.^[65]

For comparison and completeness the triplet–triplet ab-

sorption spectra^[72]of the methanofullerene adduct 14 and

4986

www.chemeurj.org

ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Chem. Eur. J. 2008, 14, 4974 –4992

Light-Harvesting Dyads

FULL PAPER

pristine C₆₀are given in the Supporting Information. The

singlet–singlet absorption of 7 was reported previously.^[65]

Quantum yields of triplet PDI formation were estimated

by using the comparative method^[73]and molar-absorption

coefficients.^[53,66]From this analysis we estimate that F_T

(1)=0.75, F_T(2)=0.70 and F_T(3)=0.60, F_T(4)=0.52. The

molar-absorption coefficients of the transient triplet species

were estimated by using the iso-absorptive points^[53](Sup-

porting Information). The observed triplet absorption spec-

trum has a maximum at 510 nm (e_(T1–Tn)=9400m^ꢀ1cm^ꢀ1) for

compounds 3 and 4. The observed triplet absorption spectra

directional energy transfer. The PDI was introduced to play

the role of a visible-light antenna, and after excitation of

this PDI moiety in the dyads, we could prove a quantitative

intramolecular energy transfer (>99%) from the PDI donor

to the C₆₀acceptor. A Fçrster-type mechanism is suggested

to explain this singlet–singlet energy transfer and it was

demonstrated that the final populated excited state is a low-

energy triplet state localized on the PDI moiety.

Interestingly, various intramolecular energy-transfer

rates^[41h,65]between C₆₀and different PDI-containing chloro-,

phenoxy-, and pyrrolidino-substitutents at the bay position

(the orange, red, and green PDI family^[75]) at different mo-

lecular configurations are now available, making an exten-

sive comparison of factors influencing the rates possible (see

Supporting Information). It can already be noted that the

combination of the green PDI with C₆₀through a pyrrolidine

bridge results in much slower

of

1 and 2 have a maximum at 560 nm (e_(T1–Tn)=

51600m^ꢀ1cm^ꢀ1).

We can now assess the energetic scheme reported in

Figure 10 for dyad 1. Excitation at 520 nm, at which the

PDI moiety selectively absorbs, leads to the population of

singlet and triplet energy trans-

fer^[41h](in toluene), not only

due to the restricted conforma-

tional freedom, but also to the

much less favorable Fçrster

overlap. This would imply that

these types of NIR-absorbing

imide dyes (that can harvest

more of the sunlight) should be

designed to display Dexter-type

energy transfer.

It was also shown in these

C₆₀–PDI dyads that the position

of the first reduction potential

of the PDI moiety in solution

relative to this of C₆₀can be

finely tuned by molecular engi-

neering around the PDI bay

region. This electrochemical pa-

Figure 10. Energy-level diagram (on eV scale) showing the processes that take place in the two conformations

of dyad 1 upon selective excitation of PDI moiety (toluene, RT). Lifetimes and efficiencies are also indicated.

1

singlet state centered on PDI. PDI*–C₆₀decays by very fast

energy transfer to form PDI–¹C₆₀* for which two rates are

observed, attributed to a folded and an extended conforma-

tion. Fluorescence emission is more than 99% quenched. A

rameter was correlated with preliminary results obtained

from the conception of photovoltaic devices. These were

tested by blending P3HT with C₆₀–PDI dyad and it was

clearly evidenced in agreement with absorption and electro-

chemical properties that better photovoltaic characteristics

were obtained with dyad 3 than with dyad 1.^[76]Further in-

vestigation concerning the incorporation of these dyads in

photovoltaic devices is underway: firstly, the possibility of

using the C₆₀–PDI dyad associated in variable amounts with

[60]PCBM in the photoactive layer and secondly, the design

of new light-harvesting dyads in which the first reduction

process should be more selective between the two units.

+

charge-separated state PDI^ꢀ–C₆₀is not populated, although

its energy level is intermediate (Table 4). Clearly, singlet

3

energy transfer is much faster. The final triplet state PDI*–

C₆₀is populated in high yield. Although the triplet state

PDI–³C₆₀* is not observed for dyad 1, it is clearly observed

for dyads 2–4. Thus, it is clear that intersystem crossing from

PDI–¹C₆₀* to PDI–³C₆₀* precedes the energy transfer to the

final PDI triplet state. Indeed, intersystem crossing is known

to be very efficient for most mono-functionalized C₆₀deriva-

tives (F_Tclose to unity, t_fꢂ1.5 ns and F_fꢂ510^ꢀ4) such as

14 and other compounds described in the literature.^{[55,57,72,74]}

Experimental Section

Materials and methods: Reagents were obtained commercially and were

used without any purification. Fullerene C₆₀(99.5+ %) was purchased

from MER Corporation Fullerenes. 1,6,7,12-Tetrachloroperylenetetracar-

boxylic dianhydride 5 was furnished by BASF-AG (Ludwigshafen, Ger-

many). Dry solvents were obtained by distillation over suitable dessicants

(THF and toluene from Na/benzophenone, CH₂Cl₂from P₂O₅, CH₃CN

Conclusion

Novel [60]fullerene–perylenediimide (C₆₀–PDI) dyads were

developed as efficient light-harvesting systems designed for

Chem. Eur. J. 2008, 14, 4974 –4992

ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemeurj.org

4987

P. Hudhomme, R. M. Williams et al.

from CaH₂). Thin-layer chromatography (TLC) was performed on alumi-

nium sheets coated with silica gel 60F254. Column chromatography was

carried out on silica gel (Kieselgel 60 Merck 5–40 mm).

also used for the white-light generation in a 2 mm water cell. For femto-

second transient absorption in the NIR region a Control Development

NIR-256L-1.7T1-USB optical spectrometer system, InGaAs detector

with 512 element arrays responding to wavelengths range from 900–

1700 nm was used. Detection light was generated with a sapphire plate.

The HPLC system consisted of a series of a model Waters 501 pump, a

model Waters 2487 UV/Vis detector. Analytical separation was accom-

plished at 258C on a 250 mm4.6 mm ID column packed with a 5 mm

AIT Kromasil Silica stationary phase. Semi-preparative separation was

accomplished at 258C on a 250 mm10 mm column packed with a 5 mm

AIT Kromasil Silica stationary phase. The operating conditions for the

semi-preparative separation were a flow-rate of 1.5 mLmin^ꢀ1(for 10),

2 mLmin^ꢀ1(for 4, 7, 14), or 3 mLmin^ꢀ1(for 1, 2, 3), a mobile phase of

toluene/CH₂Cl₂(3:1, v/v) (for dyads 1–4 and 7) or CH₂Cl₂(for 10 and 14)

(HPLC grade Fisher Scientific). The injected solution was composed of

15 mg of compound in 2 mL of solvent. Detection was achieved at the

absorbance of 335 nm (for 14), 520 nm for 1, 2, 7), or 580 nm (for 3, 4,

10).

Time-resolved fluorescence measurements were performed using a pico-

second single-photon counting (SPC) setup. The frequency-doubled

(300–340 nm, 1 ps, 3.8 MHz) output of a cavity-dumped DCM dye laser

(Coherent model 700) pumped by a mode-locked Ar-ion laser (Coherent

486 AS Mode Locker, Coherent Innova 200 laser) was used as the excita-

tion source. A (Hamamatsu R3809) microchannel-plate photomultiplier

was used as detector. The instrument response (ꢂ17 ps FWHM) was re-

corded using the Raman scattering of a doubly de-ionized water sample.

Time windows (4000 channels) of 5 ns (1.25 ps/channel) to 50 ns (12.5 ps/

channel) could be used, enabling the measurement of lifetimes of 5 ps–

40 ns. The recorded traces were deconvoluted with the system response

and fitted to an exponential function using the Igor Pro program.

Melting points were determined by using a microscope with a Kçfler hot

1

stage and are uncorrected. H (500 MHz) and ¹³C (125 MHz) NMR spec-

In nanosecond pump-probe experiments,

a (Coherent) Infinity Nd:

tra were recorded on a Bruker Avance DRX 500 spectrometer. Chemical

shifts are reported as d values in ppm using CHCl₃as the reference.

Infra-red spectra were obtained on a BIO-RAD FTS 155 spectrometer

and UV/Vis experiments were performed on a Perkin–Elmer Lambda 19

YAG-XPO laser was used for excitation. The laser illuminated a slit of

102 mm. Perpendicular to this, the probe light was provided by an

EG&G (FX504) low-pressure Xenon lamp that irradiated the sample

through a 1 mm pinhole. The overlap of the two beams falls within the

first two millimeters of the cell, after the slit. The probe light from both

the signal and the reference channels is then collected in optical fibers

that are connected to an Acton SpectraPro-150 spectrograph that is cou-

pled to a Princeton Instruments ICCD-576-G/RB-EM gated intensified

CCD camera. Using a 5 ns gate this camera simultaneously records the

spectrally dispersed light from both optical fibers on separate stripes of

the CCD. Deaeration was performed by bubbling with Argon for 20 to

30 min.

NIR spectrometer. Mass spectra were recorded on

a MALDI-TOF

Bruker Biflex III. Elemental microanalyses were performed by the

CNRS Central Service of Microanalysis (Vernaison, France).

Cyclic voltammetry was performed in a three-electrode cell equipped

with a platinum millielectrode and a platinum wire counterelectrode. A

silver wire served as quasi-reference electrode and its potential was

checked against the ferricinium/ferrocene couple (Fc⁺/Fc) before and

after each experiment. The electrolytic media involved CH₂Cl₂(HPLC

grade), o-dichlorobenzene (Aldrich spectroscopic grade), and 0.1m of tet-

rabutylammonium hexafluorophosphate (TBAHP, puriss quality). All ex-

periments were performed in a glove box containing dry, oxygen-free

(<1 ppm) argon, at RT. Electrochemical experiments were carried out

using an EGG PAR 273A potentiostat.

Dyad 1: To a solution of 12a (189 mg, 0.25 mmol) in anhydrous toluene

(80 mL) were added successively C₆₀(360 mg, 0.50 mmol), iodine (94 mg,

0.37 mmol), and DBU (85 mL, 0.57 mmol). The reaction mixture was

stirred at RT under nitrogen for 3 h. After addition of water (50 mL), the

resulting emulsion was extracted with CH₂Cl₂(330 mL). The combined

organic phases were washed with 1n HCl (250 mL), water (250 mL),

dried over MgSO₄and concentrated under reduced pressure. The residue

was purified by column chromatography on silica gel by using CS₂to

remove unreacted C₆₀, then CH₂Cl₂/EtOAc 9:1. A second column chro-

matography on silica gel was performed by using toluene/EtOAc 95:5.

Recrystallization using CH₂Cl₂/petroleum ether afforded pure dyad 1 as a

red powder (195 mg; 53%). ¹H NMR (CDCl₃): d=8.66 (s, 2H), 8.63 (s,

2H), 5.09 (ddd, ³J=3.5, ³J=8.5 Hz, ²J=14 Hz, 1H), 4.92 (ddd, ³J=3.5,

³J=8.5 Hz, ²J=14 Hz, 1H), 4.78 (ddd, ³J=3.5, ³J=8.5 Hz, ²J=14 Hz,

1H), 4.62 (ddd, ³J=3, ³J=3.5, ³J=8.5 Hz, ²J=14 Hz, 1H), 4.57 (q, ³J=

7 Hz, 2H), 4.20 (t, ³J=7 Hz, 2H), 1.75 (m, 2H), 1.50 (t, ³J=7 Hz, 3H),

1.42 (m, 4H), 0.94 ppm (t, ³J=7 Hz, 3H); ¹³C NMR (CDCl₃): d=163.82,

163.21, 162.26, 162.19, 145.15, 145.11, 145.05, 145.00, 144.96, 144.81,

144.75, 144.66, 144.57, 144.55, 144.51, 144.44, 144.42, 144.40, 144.25,

144.16, 143.76, 143.66, 143.15, 142.96, 142.89, 142.84, 142.79, 142.76,

142.74, 142.65, 142.32, 142.06, 141.98, 141.80, 141.74, 141.63, 141.50,

141.00, 140.87, 140.83, 140.58, 139.87, 138.91, 138.54, 138.46, 135.58,

135.45, 133.44, 132.98, 131.39, 131.21, 128.89, 128.30, 123.38, 123.21,

122.54, 71.31, 71.30, 63.75, 63.57, 51.97, 40.98, 39.12, 29.15, 27.76, 22.41,

Electronic absorption spectra were recorded on a Lambda 19 NIR model

from Perkin–Elmer. Fluorescence spectra were recorded in non-deoxy-

genated solvents (spectroscopic grade) at 208C with a QM-4/QuantaMas-

ter fluorometer from PTI equipped with rapid monochannel detection

and continuous excitation source. Quantum yields were determined using

cresyl violet as a standard reference (F_f=0.54 at 208C in MeOH).

For time-resolved spectroscopy, samples were dissolved in spectroscopic

solvents (toluene and benzonitrile from Aldrich) and filtered (0.4 mm

PVDF HPLC filters) to remove particles and potential aggregates. The

samples had an absorbance of ca. 0.7–0.9 (1 cm) for nanosecond transient

measurements and of ca. 0.3–0.7 (2 mm) for femtosecond transient at the

excitation wavelength. The UV/Vis absorption spectra of the samples

were measured before and after the laser experiments and were found to

be virtually identical, thus ruling out any possible degradation or chemi-

cal change in the samples. All photophysical data reported here have an

error limit of 5–10%, unless indicated otherwise. The experiments were

performed at RT.

Femtosecond transient absorption experiments were performed with a

Spectra-Physics Hurricane Titanium:Sapphire regenerative amplifier

system. The full spectrum setup was based on an optical parametric am-

plifier (Spectra-Physics OPA 800C) as the pump. The residual fundamen-

tal light, from the pump OPA, was used for white-light generation, which

was detected with a CCD spectrograph (Ocean Optics). The polarization

of the pump light was controlled by a Berek Polarization Compensator

(New Focus). The Berek Polarizer was always included in the setup to

provide the Magic-Angle conditions. The probe light was double-passed

over a delay line (Physik Instrumente, M-531DD) that provides an exper-

imental time window of 3.6 ns with a maximal resolution of 0.6 fs per

step. The OPA was used to generate excitation pulses at 530 nm. The

laser output was typically 3.5–5 mJpulse^ꢀ1(130 fs FWHM) with a repeti-

tion rate of 1 kHz. The samples were placed into cells of 2 mm path

length (Hellma) and were stirred with a downward projected PTFE shaft

using a direct-drive spectro-stir (SPECTRO-CELL). This stir system was

14.26, 13.99 ppm; IR (KBr): n˜ =1747, 1705, 1667, 1589, 1236, 527 cm^ꢀ1

;

MS (MALDI-TOF, pos. mode, dithranol): m/z: calcd for: 1472.03; found:

1472.07 [M]⁺; elemental analysis calcd (%) for C₉₆H₂₄Cl₄N₂O₈(1475.04):

C

C 78.17, H 1.64; found: C 77.91, H 1.92.

Dyad 2: To a solution of 12b (120 mg, 0.15 mmol) in anhydrous toluene

(50 mL) were added successively C₆₀(216 mg, 0.30 mmol), iodine (58 mg,

0.23 mmol), and DBU (50 mL, 0.33 mmol). The reaction mixture was

stirred at RT under nitrogen for 4 h. After addition of water (50 mL), the

resulting emulsion was extracted with CH₂Cl₂(330 mL). The combined

organic phases were washed with 1n HCl (50 mL), water (250 mL),

dried over MgSO₄and concentrated under reduced pressure. The residue

was purified by column chromatography on silica gel by using CS₂to

remove unreacted C₆₀, then petroleum ether/CH₂Cl₂3:7. Recrystalliza-

tion using CH₂Cl₂/petroleum ether afforded pure dyad 2 as a red powder

4988

www.chemeurj.org

ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Chem. Eur. J. 2008, 14, 4974 –4992

Light-Harvesting Dyads

FULL PAPER

1

(96 mg; 42%). H NMR (CDCl₃): d=8.67 (s, 2H), 8.66 (s, 2H), 4.60–4.50

³J=7.5 Hz, 2H), 4.30–4.10 (brs, 1H), 4.20 (t, ³J=7.5 Hz, 2H), 4.00 (t,

³J=7 Hz, 2H), 1.80 (m, 2H), 1.50 (m, 4H), 0.95 ppm (t, J=7.5 Hz, 3H);

3

(m, 2H), 4.55 (q, J=7 Hz, 2H), 4.26 (t, J=7 Hz, 2H), 4.21 (t, J=7 Hz,

2H), 1.94 (quint, ³J=7.5 Hz, 2H), 1.87 (quint, ³J=7.5 Hz, 2H), 1.75

(quint, ³J=7.5 Hz, 2H), 1.64 (quint, ³J=7.5 Hz, 2H), 1.50 (t, ³J=7 Hz,

3H), 1.45–1.40 (m, 4H), 0.94 ppm (t, ³J=7 Hz, 3H); IR (KBr): n˜ =1744,

1703, 1665, 1587, 525 cm^ꢀ1; MS (MALDI-TOF, pos. mode, dithranol):

m/z: calcd for: 1514.08; found: 1512.95 [MꢀH]⁺; elemental analysis calcd

(%) for C₉₉H₃₀Cl₄N₂O₈(1517.12): C 78.38, H 1.99; found: C 76.98, H

2.38.

IR (KBr): n˜ =3436 (br, OH), 1702, 1658, 1587 cm^ꢀ1; MS (MALDI-TOF,

pos. mode, dithranol): m/z: calcd for: 640.01; found: 640.08 [M]⁺; ele-

C

mental analysis calcd (%) for C₃₁H₂₀Cl₄N₂O₅(642.31): C 57.97, H 3.14, N

4.36; found: C 56.43, H 3.18, N 4.17.

N-2’-Hydroxypentyl-N’-pentyl-1,6,7,12-tetrachloroperylene-3,4:9,10-bis-

(dicarboximide) (6b): To

3,4:9,10-tetracarboxylic dianhydride

a

solution of 1,6,7,12-tetrachloroperylene

(10 g; 18.9 mmol) in toluene

5

(110 mL) were added simultaneously n-pentylamine (2.2 mL, 18.9 mmol)

and 5-aminopentanol (2.0 mL, 18.9 mmol). The reaction mixture was

heated under reflux for 3 d and after cooling to RT the solvent was con-

centrated under reduced pressure. The residue was extracted by using

CH₂Cl₂(200 mL). The filtrate was concentrated under reduced pressure

and the residue was purified by column chromatography on silica gel.

Symmetrical derivative 7 was first isolated using CH₂Cl₂, then unsymmet-

rical compound 6b was obtained as a red powder using CH₂Cl₂/EtOAc

4:1 (2.24 g, 17%). ¹H NMR (CDCl₃): d=8.68 (s, 2H), 8.67 (s, 2H), 4.24

(t, ³J=7 Hz, 2H), 4.21 (t, ³J=7 Hz, 2H), 4.20 (s, 1H), 3.69 (t, ³J=7 Hz,

2H), 1.80 (quint, ³J=7.5 Hz, 2H), 1.75 (quint, ³J=7.5 Hz, 2H), 1.68

(quint, ³J=7.5 Hz, 2H), 1.53 (quint, ³J=7.5 Hz, 2H), 1.42 (m, 4H),

0.93 ppm (t, ³J=7 Hz, 3H); IR (KBr): n˜ =3407 (br, OH), 1703, 1665,

1588 cm^ꢀ1; MS (MALDI-TOF, pos. mode, dithranol): m/z: calcd for:

682.06; found: 682.03; elemental analysis calcd (%) for C₃₄H₂₆Cl₄N₂O₅

(684.39): C 59.67, H 3.83, N 4.09; found: C 59.33, H 3.92, N 4.01.

Dyad 3: To a solution of 13a (363 mg, 0.3 mmol) in anhydrous toluene

(100 mL) were added successively C₆₀(432 mg, 0.6 mmol), iodine

(114 mg, 0.45 mmol), and DBU (102 mL, 0.68 mmol). The reaction mix-

ture was stirred at RT under nitrogen for 4 h. After addition of water

(50 mL), the resulting emulsion was extracted with CH₂Cl₂(360 mL).

The combined extracts were washed with washed with 1n HCl (2

50 mL), water (250 mL), dried over MgSO₄and concentrated under re-

duced pressure. The residue was purified by column chromatography on

silica gel by using CS₂to remove unreacted C₆₀, then CH₂Cl₂/EtOAc 9:1.

A second column chromatography on silica gel using petroleum ether/

CH₂Cl₂3:7 was performed. Recrystallization using CH₂Cl₂/petroleum

ether afforded pure dyad 3 as a purple powder (270 mg; 47%). ¹H NMR

(CDCl₃, 298 K): d=8.19 (s, 2H), 8.15 (s, 2H), 7.18 and 7.17 (2d, ³J=

3

8.6 Hz, 8H), 6.75 and 6.73 (2d, J=8.6 Hz, 8H), 4.90 (brs, 1H), 4.82 (brs,

1H), 4.72 (brs, 1H), 4.64 (brs, 1H), 4.50 (q, ³J=7 Hz, 2H), 4.09 (t, ³J=

7.5 Hz, 2H), 1.66 (brt, 2H), 1.43 (t, ³J=7 Hz, 3H), 1.27 (s, 36H), 1.23–

1.36 (m, 4H), 0.88 ppm (t, ³J=7 Hz, 3H); the ¹H NMR recorded in

CDCl₃at 343 K presents two signals at 4.86 (brs, 2H) and 4.67 ppm (brs,

2H) instead of the four broad signals at 5.09 (1H), 4.92 (1H), 4.78 (1H),

4.62 ppm (1H); ¹³C NMR (CDCl₃): d=163.40, 163.33, 156.00, 155.80,

152.83, 152.71, 147.22, 147.18, 145.11, 144.98, 144.88, 144.75, 144.70,

144.50, 144.43, 143.73, 143.42, 142.77, 142.72, 141.99, 141.69, 139.49,

129.02, 128.21, 126.67, 126.63, 122.56, 121.75, 121.00, 120.22, 120.20,

119.83, 119.35, 119.30, 119.20, 77.57, 71.38, 63.63, 63.52, 52.19, 40.63,

38.37, 34.33, 31.43, 29.69, 29.19, 27.73, 22.35, 21.45, 14.15, 13.95 ppm; IR

N,N’-Dipentyl-1,6,7,12-tetrachloroperylene-3,4:9,10-bis(dicarboximide)

(7): Synthesis of this compound was described previously.^[26b]

N-2’-Hydroxyethyl-N’-pentyl-1,6,7,12-tetra(p-tertiobutyl)phenoxypery-

lene-3,4:9,10-bis(dicarboximide) (9a): A mixture of perylenediimide 6a

(860 mg, 1.4 mmol), p-tertiobutylphenol (2.1 g, 14 mmol), and K₂CO₃

(2 g, 14.5 mmol) in N-methylpyrrolidone (NMP) (50 mL) was stirred at

1308C under nitrogen for 14 h. After being cooled to RT, the reaction

mixture was poured into HCl acid (1n) (50 mL). The precipitate was col-

lected by filtration, washed thoroughly with water until neutrality, then

purified by column chromatography on silica gel using successively

CH₂Cl₂then CH₂Cl₂/acetone 30:1. Recrystallization from CH₂Cl₂/metha-

nol afforded compound 9a as a purple powder, which was washed with

petroleum ether to remove traces of methanol (920 mg, 60%). ¹H NMR

(CDCl₃, 500 MHz): d=8.24 (s, 2H), 8.22 (s, 2H), 7.25 (d, ³J=9 Hz, 8H),

6.80 (d, ³J=9 Hz, 8H), 4.40 (t, ³J=7 Hz, 2H,), 4.40–4.20 (brs, 1H), 4.10

(t, ³J=7 Hz, 2H), 3.93 (t, ³J=7 Hz, 2H), 1.70 (brt, 2H), 1.40–1.35 (m,

4H), 1.30 (s, 36H), 0.95 ppm (t, ³J=7 Hz, 3H); IR (KBr): n˜ =3438 (br,

OH), 1693, 1657, 1586, 1507, 1292 cm^ꢀ1; MS (MALDI-TOF, pos. mode,

(KBr): n˜ =2958, 1748, 1700, 1663, 1588, 1503, 1284, 526 cm^ꢀ1

; MS

(MALDI-TOF, pos. mode, dithranol): m/z: calcd for: 1928.54; found:

1928.52 [M]⁺; elemental analysis calcd (%) for C₁₃₆H₇₆N₂O₁₂(1930.06):

C

C 84.63, H 3.97; found: C 82.90, H 4.25.

Dyad 4: To a solution of 13b (150 mg, 0.12 mmol) in anhydrous toluene

(30 mL) were added successively C₆₀(130 mg, 0.18 mmol), iodine (46 mg,

0.18 mmol), and DBU (54 mL, 0.36 mmol). The reaction mixture was

stirred at RT under nitrogen for 48 h. The solvent was concentrated

under reduced pressure and the residue was purified by column chroma-

tography on silica gel using successively CS₂to remove unreacted C₆₀, tol-

uene/acetone 9:1, then petroleum ether/CH₂Cl₂1:4. Recrystallization

using CH₂Cl₂/petroleum ether afforded pure dyad 4 as a purple powder

dithranol): m/z: calcd for: 1096.52; found: 1096.23 [M]⁺; elemental anal-

C

ysis calcd (%) for C₇₁H₇₂N₂O₉(1097.34): C 77.71, H 6.61, N 2.55; found:

C 76.54, H 6.31, N 2.57.

1

(120 mg; 51%). H NMR (CDCl₃): d=8.22 (s, 2H), 8.19 (s, 2H), 7.22 and

N-2’-Hydroxypentyl-N’-pentyl-1,6,7,12-tetra(p-tertiobutyl)phenoxypery-

lene-3,4:9,10-bis(dicarboximide) (9b): A mixture of perylenediimide 6b

(68.5 mg, 0.1 mmol), p-tertiobutylphenol (150 mg, 1 mmol), and K₂CO₃

(143 mg, 1.05 mmol) in N-methylpyrrolidone (NMP) (5 mL) was stirred

at 1308C under nitrogen for 1 h. After being cooled to RT, the reaction

mixture was poured into HCl acid (1n) (15 mL). The precipitate was col-

lected by filtration, washed thoroughly with water until neutrality, then

purified by column chromatography on silica gel using successively

CH₂Cl₂then CH₂Cl₂/acetone 30:1. Recrystallization from CH₂Cl₂/metha-

nol afforded compound 9b as a purple powder which was washed with

petroleum ether to remove traces of methanol (66 mg, 58%). ¹H NMR

7.21 (2d, ³J=9 Hz, 8H), 6.81 and 6.79 (2d, ³J=9 Hz, 8H), 4.50 (q, ³J=

7 Hz, 2H), 4.49 (t, ³J=7 Hz, 2H), 4.15 (t, ³J=7 Hz, 2H), 4.10 (t, ³J=

7 Hz, 2H), 1.86 (quint, ³J=7 Hz, 2H), 1.78 (quint, ³J=7 Hz, 2H), 1.68

3

(quint, J=7 Hz, 2H), 1.57 (m, 2H), 1.44 (t, J=7 Hz, 3H), 1.36 (m, 4H),

3

1.28 (s, 36H), 0.88 ppm (t, J=7 Hz, 3H); IR (KBr): n˜ =2957, 1746, 1698,

1661, 1588, 1504, 1283, 526 cm^ꢀ1; MS (MALDI-TOF, pos. mode, dithra-

nol): m/z: calcd for: 1970.59; found: 1970.29 [M]⁺; elemental analysis

C

calcd (%) for C₁₃₉H₈₂N₂O₁₂(1972.1): C 84.65, H 4.19; found: C 83.35, H

4.35.

N-2’-Hydroxyethyl-N’-pentyl-1,6,7,12-tetrachloroperylene-3,4:9,10-bis(di-

carboximide) (6a): To a solution of 1,6,7,12-tetrachloroperylene-3,4:9,10-

tetracarboxylic dianhydride 5 (10 g; 18.9 mmol) in toluene (110 mL) were

added simultaneously n-pentylamine (2.2 mL, 18.9 mmol) and ethanol-

amine (1.2 mL, 18.9 mmol). The reaction mixture was heated under reflux

for 24 h and after cooling the solvent was concentrated under reduced

pressure. The residue was extracted by using hot CH₂Cl₂(10100 mL)

and the combined extracts were concentrated under reduced pressure.

The residue was purified by column chromatography on silica gel. Sym-

metrical derivative 7 was first isolated by using CH₂Cl₂, then unsymmetri-

cal compound 6a was obtained as a red powder using CH₂Cl₂/EtOAc 4:1

(3.4 g, 28%). ¹H NMR (CDCl₃): d=8.62 (s, 2H), 8.60 (s, 2H), 4.50 (t,

3

(CDCl₃, 500 MHz): d=8.22 (s, 4H), 7.23 (d, J=9 Hz, 8H), 6.83 and 6.82

3

(2d, J=9 Hz, 8H), 4.13 (t, J=7 Hz, 2H), 4.10 (t, J=7 Hz, 2H), 3.63 (t,

³J=7 Hz, 2H), 1.72 (quint, ³J=7.5 Hz, 2H), 1.67 (brt, ³J=7.5 Hz, 1H),

1.61 (quint, ³J=7.5 Hz, 2H), 1.44 (quint, ³J=7.5 Hz, 2H), 1.38–1.32 (m,

6H), 1.29 (s, 36H), 0.88 ppm (t, ³J=7 Hz, 3H); IR (KBr): n˜ =3500 (br,

OH), 2962, 1697, 1658, 1588, 1505, 1291 cm^ꢀ1; MS (MALDI-TOF, pos.

mode, dithranol): m/z: calcd for: 1138.57; found: 1138.03 [M]⁺; elemen-

C

tal analysis calcd (%) for C₇₄H₇₈N₂O₉(1139.42): C 78.00, H 6.90, N 2.46;

found: C 77.48, H 6.74, N 2.41.

N,N’-Dipentyl-1,6,7,12-tetra(p-tertiobutyl)phenoxyperylene-3,4:9,10-

3,4:9,10-bis(dicarboximide) (10): A mixture of perylenediimide 7 (5 g,

Chem. Eur. J. 2008, 14, 4974 –4992

ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemeurj.org

4989

P. Hudhomme, R. M. Williams et al.

7.5 mmol), p-tertiobuylphenol (11.3 g, 75 mmol), and K₂CO₃(10.4 g,

75 mmol) in N-methylpyrrolidone (NMP) (160 mL) was stirred at 1308C

under nitrogen for 14 h. After being cooled to RT, the reaction mixture

was poured into HCl acid (1n) (40 mL). The precipitate was collected by

filtration, washed thoroughly with water until neutrality, then purified by

column chromatography on silica gel using CH₂Cl₂/acetone 30:1. Recrys-

tallization from CH₂Cl₂/methanol afforded compound 10 as a purple

powder, which was washed with petroleum ether to remove traces of

methanol (5.8 g, 69%). ¹H NMR (CDCl₃, 500 MHz): d=8.22 (s, 4H),

7.23 (d, ³J=9 Hz, 8H), 6.82 (d, ³J=9 Hz, 8H), 4.10 (t, ³J=7.5 Hz, 4H),

1.67 (quint, ³J=7.5 Hz, 4H), 1.34 (m, 8H), 1.29 (s, 36H), 0.88 ppm (t,

³J=7 Hz, 6H); IR (KBr): n˜ =2956, 1706, 1659, 1590, 1497, 1289 cm^ꢀ1; MS

(MALDI-TOF, pos. mode, dithranol): m/z: calcd for: 1122.58; found:

147.32, 147.26, 132.84, 126.67, 122.56, 122.01, 120.91, 120.25, 120.17,

119.75, 119.59, 119.37, 119.20, 77.59, 62.43, 61.49, 41.41, 40.62, 38.93,

34.37, 31.45, 29.18, 27.74, 22.36, 13.98 ppm; IR (KBr): n˜ =2971, 1700,

1657, 1588, 1507, 1291 cm^ꢀ1; MS (MALDI-TOF, pos. mode, dithranol):

m/z: calcd for: 1210.55; found: 1210.49 [M]⁺; elemental analysis calcd

C

(%) for C₇₆H₇₈N₂O₁₂(1211.44): C 75.35, H 6.49, N 2.31; found: C 73.94,

H 6.23, N 2.31.

Ethyl 2-(N,N’-pentyl-1,6,7,12-tetra(p-tertiobutyl)perylene-3,4:9,10-bis(di-

carboximide)pentyl malonate (13b): To a solution of alcohol 9b (300 mg,

0.26 mmol) in anhydrous CH₂Cl₂(40 mL) were added successively at 08C

under nitrogen fresh pyridine (0.17 mL, 2.1 mmol) and ethyl malonyl

chloride 90% (0.2 mL, 1.6 mmol). The reaction mixture was refluxed for

36 h. After cooling to RT, the solvent was concentrated under reduced

pressure. Recrystallization from CH₂Cl₂/methanol afforded compound

13a as a purple powder, which was washed with petroleum ether to

remove traces of methanol (291 mg, 88%). ¹H NMR (CDCl₃, 500 MHz):

d=8.22 (s, 4H), 7.23 (d, ³J=9 Hz, 8H), 6.83 (d, ³J=9 Hz, 4H), 6.81 (d,

³J=9 Hz, 4H), 4.16 (q, ³J=7 Hz, 2H), 4.13–4.08 (m, 6H), 3.33 (s, 2H),

1.74–1.64 (m, 6H), 1.47–1.41 (m, 2H), 1.37–1.33 (m, 4H), 1.29 (s, 36H),

1.24 (t, ³J=7 Hz, 3H), 0.88 ppm (t, ³J=7 Hz, 3H); IR (KBr): n˜ =2972,

2867, 1702, 1652, 1590, 1509, 1289 cm^ꢀ1; MS (MALDI-TOF, pos. mode,

1122.02 [M]⁺; elemental analysis calcd (%) for C₇₄H₇₈N₂O₈(1123.42): C

C

79.11, H 7.00, N 2.49; found: C 78.76, H 7.01, N 2.41.

Ethyl 2-N-(N’-pentyl-1,6,7,12-tetrachloroperylene-3,4:9,10-3,4:9,10-bis(di-

carboximide)ethyl malonate (12a): To a solution of alcohol 6a (193 mg,

0.3 mmol) in anhydrous CH₂Cl₂(50 mL) were added successively at 08C

under nitrogen fresh pyridine (77 mL, 0.94 mmol) and ethyl malonyl chlo-

ride 90% (130 mL, 0.92 mmol). The reaction mixture was refluxed for

1 h. After cooling to RT, the solvent was concentrated under reduced

pressure. The residue was purified by column chromatography on silica

gel using CH₂Cl₂/EtOAc 4:1. Recrystallization from CH₂Cl₂/petroleum

ether afforded compound 12a as a red powder (210 mg, 93%). ¹H NMR

(CDCl₃, 500 MHz): d=8.69 (s, 2H), 8.68 (s, 2H), 4.55 (s, 4H), 4.21 (brt,

³J=7.5 Hz, 2H), 4.13 (q, ³J=7 Hz, 2H), 3.36 (s, 2H), 1.75 (quint, ³J=

dithranol): m/z: calcd for: 1252.60; found: 1251.79 [M]⁺; elemental anal-

C

ysis calcd (%) for C₇₉H₈₄N₂O₁₂(1253.52): C 75.69, H 6.75, N 2.23; found:

C 75.25, H 6.65, N 2.13.

3

7.5 Hz, 2H), 1.44–1.39 (m, 4H), 1.23 (t, J=7.5 Hz, 3H), 0.93 ppm (t, J=

7 Hz, 3H); ¹³C NMR (CDCl₃): d=166.62 (C=O ester), 166.24 (C=O

ester), 162.38 (C=O imide), 162.21 (C=O imide), 135.46, 135.30, 133.08,

132.94, 131.44, 131.40, 128.85, 128.47, 123.36, 123.31, 123.25, 122.84, 62.45,

61.54, 41.36, 40.93, 39.40, 29.13, 27.75, 22.40, 14.02, 13.97 ppm; IR (KBr):

n˜ =1764, 1744, 1706, 1667, 1594 cm^ꢀ1; MS (MALDI-TOF, pos. mode, di-

Acknowledgements

This work was supported by the French National Research Agency

(ANR Nanorgysol) and the Conseil GØnØral du Maine et Loire for the

grant to J.B. We thank Dr. Jacques Delaunay (CIMA, UniversitØ dꢀAn-

gers) for spectroscopic analyses and Dr Dario Bassani (LCOO, Univer-

sitØ Bordeaux 1) for fruitful discussions. BASF-AG (Ludwigshafen) is

also acknowledged for providing 1,6,7,12-tetrachloroperylene tetracar-

boxylic dianhydride. We are grateful to NOW (de Nederlandse Organisa-

tie voor Wetenschappelijk Onderzoek) for the grant for the femtosecond

equipment and to the UvA (Universiteit van Amsterdam) for structural

support. We thank the Vietnamese Oversea Scholarship Program

(VOSP) of the Vietnamese government for the support of N.V.A. within

the UvA-HUT project.

thranol): m/z: calcd for: 754.04; found: 753.97 [M]⁺; elemental analysis

C

calcd (%) for C₃₆H₂₆Cl₄N₂O₈(756.41): C 57.16, H 3.46, N 3.70; found: C

56.34, H 3.59, N 3.55.

Ethyl 2-N-(N’-pentyl-1,6,7,12-tetrachloroperylene-3,4:9,10-bis(dicarbox-

imide)pentyl malonate (12b): To

a solution of alcohol 6b (200 mg,

0.29 mmol) in anhydrous CH₂Cl₂(40 mL) were added successively at 08C

under nitrogen fresh pyridine (0.1 mL, 1.22 mmol) and ethyl malonyl

chloride 90% (0.13 mL, 0.89 mmol). The reaction mixture was refluxed

for 2 h. After cooling to RT, the solvent was concentrated under reduced

pressure. The residue was purified by column chromatography on silica

gel using CH₂Cl₂/EtOAc 95:5. Recrystallization from CH₂Cl₂/methanol

afforded compound 12b as a red powder, which was washed with petrole-

[1] R. Huber, Angew. Chem. 1989 , 101 , 84 9 –871; Angew. Chem. Int.

Ed. Engl. 1989 , 28 , 84 8 –869.

[2] D. Gust, T. A. Moore, A. L. Moore, Acc. Chem. Res. 2001 , 34 , 4 0 –

48.

[3] a) T. J. Meyer, Acc. Chem. Res. 1989 , 22 , 16 3 –170; b) M. R. Wasie-

lewski, Chem. Rev. 1992 , 92 , 43 5 –461; c) D. Gust, T. A. Moore,

A. L. Moore, Acc. Chem. Res. 1993 , 26 , 19 8 –205.

1

um ether to remove traces of methanol (220 mg, 94%). H NMR (CDCl₃,

500 MHz): d=8.69 (s, 2H), 8.68 (s, 2H), 4–24–4.19 (m, 6H), 4.19 (q, ³J=

7 Hz, 2H), 3.38 (s, 2H), 1.82–1.74 (m, 4H), 1.65–1.38 (m, 6H), 1.51 (m,

2H), 1.28 (t, ³J=7 Hz, 3H), 0.93 ppm (t, ³J=7 Hz, 3H); IR (KBr): n˜ =

1734, 1704, 1665, 1588 cm^ꢀ1; MS (MALDI-TOF, pos. mode, dithranol):

m/z: calcd for: 796.09; found: 795.89 [M]⁺; elemental analysis calcd (%)

C

[4] For the small reorganization energy, see: a) H. Imahori, K. Hagi-

wara, T. Akiyama, M. Aoki, S. Taniguchi, T. Okada, M. Shirakawa,

Y. Sakata, Chem. Phys. Lett. 1996, 263, 545–550; b) H. Imahori,

N. V. Tkachenko, V. Vehmanen, K. Tamaki, H. Lemmetyinen, Y.

Sakata, S. Fukuzumi, J. Phys. Chem. A 2001 , 105 , 175 0 –1756; c) H.

Imahori, H. Yamada, D. M. Guldi, Y. Endo, A. Shimomura, S.

Kundu, K. Yamada, T. Okada, Y. Sakata, S. Fukuzumi, Angew.

Chem. 2002 , 114 , 245 0 –2453; Angew. Chem. Int. Ed. 2002 , 41 , 234 4 –

2347; d) S. Fukuzumi, K. Ohkubo, H. Imahori, D. M. Guldi, Chem.

Eur. J. 2003 , 9 , 158 5 –1593; e) The internal reorganization energy

was first estimated in 1995 to be 0.3 eV (see ref. [55b]).

[5] a) Fullerenes: Principles and Applications (Eds.: F. Langa, J.-F. Nier-

engarten), RSC, Cambridge, 2007; b) Fullerene Chemistry, Thematic

Issue C. R. Chimie (Eds.: J.-F. Nierengarten, N. Martín, P. Fowler),

2006, 9, 859–1116.

[6] a) H. Imahori, Y. Sakata, Adv. Mater. 1997 , 9 , 53 7 –546; b) H. Ima-

hori, Org. Biomol. Chem. 2004 , 2 , 142 5 –1433; c) D. I. Schuster, P.

Cheng, S. R. Wilson, V. Prokhorenko, M. Katterle, A. R. Holzwarth,

S. E. Braslavsky, G. Klihm, R. M. Williams, C. Luo, J. Am. Chem.

for C₃₉H₃₂Cl₄N₂O₈(798.49): C 58.66, H 4.04, N 3.51; found: C 57.86, H

3.52, N 3.47.

Ethyl

2-N-(N’-pentyl-1,6,7,12-tetra(p-tertiobutyl)phenoxy

perylene-

3,4:9,10-bis(dicarboximide)ethyl malonate (13a): To a solution of alcohol

9a (134 mg, 0.12 mmol) in anhydrous CH₂Cl₂(20 mL) were added succes-

sively at 08C under nitrogen fresh pyridine (35 mL, 1.22 mmol) and ethyl

malonyl chloride 90% (55 mL, 0.27 mmol). The reaction mixture was re-

fluxed for 2 h. After cooling to RT, the solvent was concentrated under

reduced pressure. The residue was purified by column chromatography

on silica gel using CH₂Cl₂/EtOAc 9:1. Recrystallization from CH₂Cl₂/

methanol afforded compound 13a as a purple powder, which was washed

with petroleum ether to remove traces of methanol (135 mg, 91%).

¹H NMR (CDCl₃, 500 MHz): d=8.23 (s, 2H), 8.21 (s, 2H), 7.23 (d, ³J=

9 Hz, 4H), 7.22 (d, ³J=9 Hz, 4H), 6.83 (d, ³J=9 Hz, 4H), 6.81 (d, ³J=

9 Hz, 4H), 4.45 (s, 4H), 4.10 (d, ³J=7.5 Hz, 2H), 4.08 (q, ³J=7 Hz, 2H),

3.30 (s, 2H), 1.67 (quint, ³J=7.5 Hz, 2H), 1.37–1.33 (m, 4H), 1.29 (s,

36H), 1.17 (t, ³J=7 Hz, 3H), 0.87 ppm (t, ³J=7 Hz, 3H); ¹³C NMR

(CDCl₃): d=166.49, 166.27, 166.47, 163.42, 156.09, 155.78, 152.91, 152.76,

4990

www.chemeurj.org

ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Chem. Eur. J. 2008, 14, 4974 –4992

Light-Harvesting Dyads

FULL PAPER

Soc. 1999 , 121 , 1159 9 –11600; d) D. I. Schuster, P. Cheng, P. D. Jar-

owski, D. M. Guldi, C. Luo, L. Echegoyen, S. Pyo, A. R. Holzwarth,

S. E. Braslavsky, R. M. Williams, G. Klihm, J. Am. Chem. Soc. 2004 ,

126 , 725 7 –7270.

[30] J. L. Segura, R. Gómez, E. Reinold, P. Bäuerle, Org. Lett. 2005 , 7 ,

234 5 –2348.

[31] R. F. Kelley, M. J. Tauber, M. R. Wasielewski, J. Am. Chem. Soc.

2006 , 128 , 477 9 –4791.

[7] a) M.-S. Choi, T. Yamazaki, I. Yamazaki, T. Aida, Angew. Chem.

2004 , 116 , 15 2 –160; Angew. Chem. Int. Ed. 2004 , 43 , 15 0 –158; b) H.

Imahori, J. Phys. Chem. B 2004 , 108 , 613 0 –6143; c) D. I. Schuster,

K. Li, D. M. Guldi, C. R. Chim. 2006 , 9 , 89 2 –908.

[8] J. J. M. Halls, R. H. Friend in Clean Electricityform Photovoltaics

(Eds.: M. D. Archer, R. Hill), Imperial College Press, London, 2001,

377–445.

[9] P. Hudhomme, J. Cousseau in Fullerenes: Principles and Applica-

tions (Eds.: F. Langa, J.-F. Nierengarten), RSC, Cambridge, 2007,

221–265.

[32] T. Ishi-i, K. Murakami, Y. Imai, S. Mataka, J. Org. Chem. 2006 , 71 ,

575 2 –5760.

[33] M. Takahashi, H. Morimoto, K. Miyake, M. Yamashita, H. Kawai,

Y. Sei, K. Yamaguchi, Chem. Commun. 2006 , 308 4 –3086.

[34] S. Leroy-Lhez, L. Perrin, J. Baffreau, P. Hudhomme, C. R. Chim.

2006, 9, 240–246.

[35] M. D. Yilmaz, O. A. Bozdemir, E. U. Akkaya, Org. Lett. 2006 , 8 ,

287 1 –2873.

[36] R. Gómez, D. Veldman, R. Blanco, C. Seoane, J. L. Segura, R. A. J.

Janssen, Macromolecules 2007, 40, 2760–2772.

[10] a) N. Martín, L. Sµnchez, B. Illescas, I. PØrez, Chem. Rev. 1998, 98,

2527–2547; b) J.-F. Nierengarten, Sol. EnergyMater. Sol. Cells 2004 ,

83 , 18 7 –199; c) J. L. Segura, N. Martín, D. M. Guldi, Chem. Soc.

Rev. 2005 , 34 , 3 1 –47; d) P. Hudhomme, C. R. Chim. 2006 , 9 , 88 1 –

891.

[11] R. A. Marcus, Pure Appl. Chem. 1997 , 69 , 1 3 –29.

[12] a) J.-F. Nierengarten, J.-F. Eckert, J.-F. Nicoud, L. Ouali, V. Krasni-

kov, G. Hadziioannou, Chem. Commun. 1999 , 61 7 –618; b) P. A.

van Hal, S. C. J. Meskers, R. A. J. Janssen, Appl. Phys. A 2004 , 79 ,

4 1 –46; c) N. Armaroli, G. Accorsi, J. N. Clifford, J.-F. Eckert, J.-F.

Nierengarten, Chem. Asian J. 2006 , 1 , 56 4 –574.

[13] P. A. van Hal, E. H. A. Beckers, S. C. J. Meskers, R. A. J. Janssen, B.

Jousselme, P. Blanchard, J. Roncali, Chem. Eur. J. 2002 , 8 , 541 5 –

5429.

[14] a) J. L. Segura, R. Gómez, N. Martín, C. Luo, D. M. Guldi, Chem.

Commun. 2000 , 70 1 –702; b) I. B. Martini, B. Ma, T. Da Ros, R. Hel-

geson, F. Wudl, B. J. Schwartz, Chem. Phys. Lett. 2000 , 327 , 25 3 –

262.

[15] a) J.-F. Eckert, J.-F. Nicoud, J.-F. Nierengarten, S.-G. Liu, L. Eche-

goyen, F. Barigelletti, N. Armaroli, L. Ouali, V. Krasnikov, G. Had-

ziioannou, J. Am. Chem. Soc. 2000 , 122 , 746 7 –7479; b) F. Langa,

M. J. Gomez-Escalonilla, J.-M. Rueff, T. M. Figueira Duarte, J.-F.

Nierengarten, V. Palermo, P. Samori, Y. Rio, G. Accorsi, N. Armaro-

li, Chem. Eur. J. 2005 , 11 , 440 5 –4415.

[16] N. S. Sariciftci, L. Smilowitz, A. J. Heeger, F. Wudl, Science 1992 ,

258 , 147 4 –1476.

[17] G. Yu, J. Gao, J. C. Hummelen, F. Wudl, A. J. Heeger, Science 1995 ,

270 , 178 9 –1791.

[37] E. E. Neuteboom, E. H. A. Beckers, S. C. J. Meskers, E. W. Meijer,

R. A. J. Janssen, Org. Biomol. Chem. 2003 , 1 , 19 8 –203.

[38] M. S. Rodriguez-Morgade, T. Torres, C. Atienza-Castellanos, D. M.

Guldi, J. Am. Chem. Soc. 2006 , 128 , 1514 5 –15154.

[39] H. Li, Y. Wang, W. Zhang, B. Liu, G. Calzaferri, Chem. Commun.

2007 , 285 3 –2854.

[40] J. Baffreau, L. Perrin, S. Leroy-Lhez, P. Hudhomme, Tetrahedron

Lett. 2005 , 46 , 459 9 –4603.

[41] a) J. L. Hua, F. S. Meng, F. Ding, H. Tian, Chem. Lett. 2004 , 33 , 43 2 –

433; b) J. L. Hua, F. S. Meng, F. Ding, F. Y. Li, H. Tian, J. Mater.

Chem. 2004 , 14 , 184 9 –1853; c) Y. Liu, N. Wang, Y. J. Li, H. B. Liu,

J. H. Xiao, X. H. Xu, C. S. Huang, S. Cui, D. B. Zhu, Macromole-

cules 2005 , 38 , 488 0 –4887; d) R. Gómez, J. L. Segura, N. Martín,

Org. Lett. 2005, 7, 717–720; e) Y. Li, Y, Li, H. Liu, S. Wang, N.

Wang, J. Zhuang, X. Li, X. He, D. Zhu, Nanotechnology 2005 , 16 ,

189 9 –1904; f) N. Wang, Y. J. Li, X. R. He, H. Y. Gan, Y. L. Li, C. S.

Huang, X. H. Xu, J. C. Xiao, S. Wang, H. B. Liu, D. B. Zhu, Tetrahe-

dron 2006 , 62 , 121 6 –1222; g) Y. Shibano, T. Umeyama, Y. Matano,

N. V. Tkachenko, H. Lemmetyinen, H. Imahori, Org. Lett. 2006 , 8 ,

442 5 –4428; h) Y. Shibano, T. Umeyama, Y. Matano, N. V. Tkachen-

ko, H. Lemmetyinen, Y. Araki, O. Ito, H. Imahori, J. Phys. Chem. C

2007 , 111 , 613 3 –6142.

[42] a) Y. Nagao, T. Misono, Bull. Chem. Soc. Jpn. 1981 , 54 , 119 1 –1194;

b) H. Trçster, Dyes Pigm. 1983 , 4 , 171- 177; c) H. Kaiser, J. Lindner,

H. Langhals, Chem. Ber. 1991 , 124 , 52 9 –535.

[43] a) G. Seybold, A. Stange, BASF AG Patent, DE 3545004, 1987;

b) G. Seybold, G. Wagenblast, Dyes Pigm. 1989 , 11 , 30 3 –317.

[44] F. Würthner, C. Thalaker, A. Sautter, W. Schärtl, W. Ibach, O. Holl-

richer, Chem. Eur. J. 2000 , 6 , 387 1 –3886.

[45] a) J.-F. Nierengarten, V. Gramlich, F. Cardullo, F. Diederich, Angew.

Chem. 1996 , 108 , 224 2 –2244; Angew. Chem. Int. Ed. Engl. 1996 , 35 ,

210 1 –2103; b) J.-F. Nierengarten, A. Herrmann, R. R. Tykwinski,

M. Ruttimann, F. Diederich, Helv. Chim. Acta 1997 , 80 , 29 3 –316;

c) R. Kessinger, J. Crassous, A. Herrmann, M. Ruttimann, L. Eche-

goyen, F. Diederich, Angew. Chem. 1998 , 110 , 202 2 –2025; Angew.

Chem. Int. Ed. 1998 , 37 , 191 9 –1922.

[46] a) M. Karplus, J. Chem. Phys. 1959 , 30 , 1 1 –15; b) M. Karplus, J.

Am. Chem. Soc. 1963 , 85 , 287 0 –2871.

[47] H. Friebolin, Basic One- and Two-Dimensional NMR Spectroscopy,

VCH, Weinheim, 1991, 73–94.

[18] a) S. Günes, H. Neugebauer, N. S. Sariciftci, Chem. Rev. 2007, 107,

1324–1338; b) B. C. Thompson, J. M. J. FrØchet, Angew. Chem.

2007, 120, 62–82; Angew. Chem. Int. Ed. 2007, 47, 58–77.

[19] a) [60]PCBM=1-(3-methoxycarbonyl)propyl-1-phenyl-

A

J. C. Hummelen, B. W. Knight, F. LePeq, F. Wudl, J. Org. Chem.

1995 , 60 , 53 2 –538.

[20] C. Winder, N. S. Sariciftci, J. Mater. Chem. 2004 , 14 , 107 7 –1086.

[21] A. Adronov, J. M. J. Frechet, Chem. Commun. 2000 , 170 1 –1710.

[22] G. Accorsi, N. Armaroli, J.-F. Eckert, J.-F. Nierengarten, Tetrahedron

Lett. 2002 , 43 , 6 5 –68.

[23] M. M. Wienk, J. M. Kroon, W. J. H. Verhees, J. Knol, J. C. Humme-

len, P. A. van Hal, R. A. J. Janssen, Angew. Chem. 2003 , 115 , 349 3 –

3497; Angew. Chem. Int. Ed. 2003 , 42 , 337 1 –3375.

[48] J. Baffreau, S. Leroy-Lhez, N. Gallego-Planas, P. Hudhomme, J.

Mol. Struct. 2007, 815, 145–150.

[24] F. Würthner, Chem. Commun. 2004 , 156 4 –1579.

[49] H. Günther, NMR Spectroscopy, Wiley, New York, 1980, 234–280.

[50] C. Bingel, Chem. Ber. 1993 , 126 , 195 7 –1959.

[51] H. Langhals, S. Demmig, H. Huber, Spectrochim. Acta Part A 1988 ,

44 , 118 9 –1193.

[52] S. Prathapan, S. I. Yang, J. Seth, M. A. Miller, D. F. Bocian, D.

Holten, J. S. Lindsey, J. Phys. Chem. B 2001 , 105 , 823 7 –8248.

[53] W. E. Ford, P. V. Kamat, J. Phys. Chem. 1987 , 91 , 637 3 –6380.

[54] C. Addicott, I. Oesterling, T. Yamamoto, K. Müllen, P. J. Stang, J.

Org. Chem. 2005 , 70 , 79 7 –801.

[55] a) S. M. Anthony, S. M. Bachilo, R. B. Weisman, J. Phys. Chem. A

2003 , 107 , 1067 4 –10679; b) R. M. Williams, J. M. Zwier, J. W. Verho-

even, J. Am. Chem. Soc. 1995 , 117 , 409 3 –4099; c) R. M. Williams,

[25] H. Langhals, S. Saulich, Chem. Eur. J. 2002 , 8 , 563 0 –5643.

[26] a) X. Guo, D. Zhang, H. Zhang, Q. Fan, W. Xu, X. Ai, L. Fan, D.

Zhu, Tetrahedron 2003 , 59 , 484 3 –4850; b) S. Leroy-Lhez, J. Baf-

freau, L. Perrin, E. Levillain, M. Allain, M.-J. Blesa, P. Hudhomme,

J. Org. Chem. 2005 , 70 , 631 3 –6320.

[27] B. K. Kaletas, R. Dobrowa, A. Sautter, F. Würthner, M. Zimine, L.

De Cola, R. M. Williams, J. Phys. Chem. A 2004 , 108 , 190 0 –2909.

[28] a) C.-C. You, C. R. Saha-Mçller, F. Würthner, Chem. Commun.

2004, 2345–2348; b) L. X. Chen, S. Xiao, L. Yu, J. Phys. Chem. A

2006, 110, 11730–11738.

[29] X. Li, L. E. Sinks, B. Rybtchinski, M. R. Wasielewski, J. Am. Chem.

Soc. 2004 , 126 , 1081 0 –10811.

Chem. Eur. J. 2008, 14, 4974 –4992

ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemeurj.org

4991

P. Hudhomme, R. M. Williams et al.

M. Koeberg, J. M. Lawson, Y.-Z. An, Y. Rubin, M. N. Paddon-Row,

J. W. Verhoeven, J. Org. Chem. 1996 , 61 , 505 5 –5062.

[56] D. M. Guldi, K.-D. Asmus, Proc. SPIE-Int. Soc. Opt. Eng. 1996,

2854, 236–245.

[66] T. W. Ebbesen, K. Tanigaki, S. Kuroshima, Chem. Phys. Lett. 1991 ,

181 , 50 1 –504.

[67] W. E. Ford, H. Hiratsuka, P. V. Kamat, J. Phys. Chem. 1989 , 93 ,

669 2 –6696.

[57] B. Ma, C. E. Bunker, R. Guduru, X.-F. Zhang, Y.-P. Sun, J. Phys.

Chem. A 1997 , 101 , 562 6 –5632.

[58] a) T. Z. Fçrster, Naturforsch. A 1949, 4, 321–327; b) T. Z. Fçrster,

Discuss. FaradaySoc. 1959, 27, 7–17.

[68] S. D. Yao, Z. R. Lian, W. F. Wang, J. S. Zhang, N. Y. Lin, H. Q. Hou,

Z. M. Zhang, Q. Z. Qin, Chem. Phys. Lett. 1995, 239, 112–116.

[69] T. Kato, T. Kodama, M. Oyama, S. Okazaki, T. Shida, T. Nakagawa,

Y. Matsui, S. Suzuki, H. Shiromaru, K. Yamauchi, Y. Achiba, Chem.

Phys. Lett. 1991 , 186 , 3 5 –39.

[70] A. Prodi, C. Chiorboli, F. Scandola, E. Iengo, E. Alessio, R. Dobra-

wa, F. Würthner, J. Am. Chem. Soc. 2005 , 127 , 145 4 –1462.

[71] T. van der Boom, R. T. Hayes, Y. Zhao, P. J. Bushard, E. A. Weiss,

M. R. Wasielewski, J. Am. Chem. Soc. 2002, 124, 9582–9590.

[72] R. V. Bensasson, E. Bienvenüe, C. Fabre, J.-M. Janot, E. J. Land, S.

Leach, V. Leboulaire, A. Rassat, S. Roux, P. Seta, Chem. Eur. J.

1998 , 4 , 27 0 –278.

[59] D. L. Dexter, J. Chem. Phys. 1953 , 21 , 83 6 –850.

[60] D. Rehm, A. Weller, Isr. J. Chem. 1970, 8, 259–271.

[61] We attribute this to the presence of a PDI derivative impurity in

spite of repetitive efforts to purify dyads by preparative HPLC.

Indeed, taking into account both the relative emission intensity and

amplitudes of the long-lived components, the amounts could be esti-

mated to be 0.003 to 0.015%, implying 99.997 and 99.985% purity

(!) of the dyads. A second hypothesis considers that this long-lived

component could be also due to a non-active conformation of the

dyad.

[73] R. Bonneau, I. Carmichael, G. L. Hug, Pure Appl. Chem. 1991 , 63 ,

28 9 –299.

[62] R. Gvishi, R. Reisfeld, Z. Burshtein, Chem. Phys. Lett. 1987, 213,

338–344.

[74] J. Li, N. Sun, Z.-X. Guo, C. Li, Y. Li, L. Dai, D. Zhu, J. Phys. Chem.

B 2002 , 106 , 1150 9 –11514.

[63] a) M. E. El-Khouly, P. Padmawar, Y. Araki, S. Verma, L. Y. Chiang,

O. Ito, J. Phys. Chem. A 2006 , 110 , 88 4 –891; b) T. Nakamura, H.

Kanato, Y. Araki, O. Ito, K. Takimiya, T. Otsubo, Y. Aso, J. Phys.

Chem. A 2006 , 110 , 347 1 –3479.

[75] a) C. Hippius, F. Schlosser, M. O. Vysotsky, V. Bçhmer, F. Würthner,

J. Am. Chem. Soc. 2006 , 128 , 387 0 –3871; b) C. Hippius, I. H. M. van

Stokkum, M. Gsänger, M. M. Groeneveld, R. M. Williams, F. Würth-

ner, J. Phys. Chem. A 2008, 112, 2476–2486.

[64] A. Sautter, B. K. Kaletas, D. G. Schmid, R. Dobrawa, M. Zimine, G.

Jung, I. H. M. van Stokkum, L. De Cola, R. M. Williams, F. Würth-

ner, J. Am. Chem. Soc. 2005 , 127 , 671 9 –6729.

[76] J. Baffreau, S. Leroy-Lhez, H. Derbal, A. R. Inigo, J.-M. Nunzi,

M. M. Groeneveld, R. M. Williams, P. Hudhomme, Eur. Phys. J.

Appl. Phys. 2006 , 36 , 30 1 –305.

[65] J. Baffreau, S. Leroy-Lhez, P. Hudhomme, M. M. Groeneveld,

I. H. M. van Stokkum, R. M. Williams, J. Phys. Chem. A 2006 , 110 ,

1312 3 –13125.

Received: January 25, 2008

Published online: April 16, 2008

4992

www.chemeurj.org

ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Chem. Eur. J. 2008, 14, 4974 –4992

Article Doi

Fullerene C60-perylene-3,4:9,10-bis(dicarboximide) light-harvesting dyads: spacer-length and bay-substituent effects on intramolecular singlet and triplet energy transfer

DOI: 10.1002/chem.200800156

Source and publish data:

Authors:

Article abstract of DOI:10.1002/chem.200800156

Full text of DOI:10.1002/chem.200800156

Products guided by the article

R&D Labs maybe for 1046791-02-9

Relevant to this article

Hot Product