A bisfullerene-bis(dipyrrinato)zinc complex: Electronic coupling and charge separation in an easy-to-assemble synthetic system

Article abstract of DOI:10.1002/chem.200802564

A study was conducted to demonstrate the synthesis and photophysical characterization of a bisfullerene-bis(dipyrrinato)zinc complex for facile tethering of two fullerene subunits. Significant information was gathered about the synthesis and photopysical

Full text of DOI:10.1002/chem.200802564

DOI: 10.1002/chem.200802564
A Bisfullerene–Bis(dipyrrinato)zinc Complex: Electronic Coupling and
Charge Separation in an Easy-to-Assemble Synthetic System
Yannick Rio,[a] David Sꢀnchez-Garcꢁa,[b] Wolfgang Seitz,[c] Tomꢀs Torres,*[a]
Jonathan L. Sessler,*[b] and Dirk M. Guldi*[c]
The use of a bis(dipyrrinato)zinc ((dpy)–Zn–(dpy)) linker
allows for the facile construction of a bis-C60 system that un-
dergoes effective charge separation upon laser photoexcita-
tion at l=480 nm.
connect fullerenes through platinum and copper complexes,
respectively. Moreover, our group and, separately, the team
of Martin, Rebek, and Guldi have taken advantage of
Watson–Crick base-pairing interactions to link C60 acceptors
to various donors through hydrogen bonds.[4] In a different
vein, Diederich et al. and, more recently, Khlobystov and
Schrçder et al. have used pyridine–metal coordination to
link two C60 derivatives.[2,5] Unfortunately, the first of these
approaches necessarily produces systems that are not sym-
metrical, whereas the latter gives rise to complexes without
apparent electronic interaction between the fullerene subu-
nits. Thus, we sought a new approach that would allow the
creation of a linked C60–C60 pseudodimer while providing
for effective coupling between the fullerene subunits.
Herein, we report the synthesis and preliminary photophysi-
cal characterization of system 1 (C60–(dpy)–Zn–(dpy)–C60),
which achieves these paired objectives; this new system is
Fullerenes, such as C60, have been extensively exploited as
electron-poor acceptors in the construction of model photo-
synthetic systems and in the development of rudimentary
photovoltaic devices. In the context of this general frame-
work, considerable effort has been devoted to the construc-
tion and study of dimeric systems in which two C60 subunits
are linked by various spacer units.[1] Such systems are of fun-
damental interest in terms of understanding the interaction
between ostensibly equivalent redox partners and the effect
that varying the coupling can have on these interactions.
However, a drawback of most such dimers is that they rely
on covalent synthesis to establish the requisite linkage.
Thus, there is an incentive to prepare systems that take ad-
vantage of easy-to-manipulate noncovalent interactions to
create the desired C60–C60 arrangements. Diederich et al.[2]
and Nierengarten et al.[3] have employed such a strategy to
based on the use of a bis
linkage.
ACHTUNGTRNE(NUNG dipyrrin)zinc ((dpy)–Zn–(dpy))
[a] Dr. Y. Rio, Prof. T. Torres
Departamento de Quimica Organica
Universidad Autonoma de Madrid
Cantoblanco, 28049-Madrid (Spain)
Fax : (+34)91-497-3966
E-mail: tomas.torres@uam.es
[b] Dr. D. Sꢀnchez-Garcꢁa, Prof. Dr. J. L. Sessler
Department of Chemistry & Biochemistry
The University of Texas at Austin
Austin, TX 78712-0165 (USA)
Fax : (+1)512-471-7550
E-mail: sessler@mail.utexas.edu
Dipyrrin systems and their applications in the construc-
tion of model electron- and energy-transfer systems have
been extensively studied. For instance, dipyrrins[6] have been
used as auxiliary dyes and linkers in multiporphyrin arrays
and in a wide variety of solar harvesting models.[7] Almost
without exception, these systems have been predicated on
the use of functionalized boron complexes of dipyrrins (e.g.,
borondifluorides). As such, they do not rely on dipyrrin as a
linking moiety. On the other hand, bis(dipyrrinato)metal
[c] W. Seitz, Prof. D. M. Guldi
Department of Chemistry and Pharmacy & Interdisciplinary Center
for Molecular Materials
University of Erlangen-Nurnberg
Egerlandstrrasse 3, 91058 Erlangen (Germany)
Fax : (+49)913-1852-8307
E-mail: dirk.guldi@chemie.uni-erlangen.de
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.200802564.
3956
ꢂ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 3956 – 3959

COMMUNICATION
complexes of divalent metal ions, such as zinc, have been
known since the days of Fischer,[8] and have been exploited
as a linking subunit by various researchers since the late
1960s.[6,9] Moreover, Bocian, Holten, Lindsey et al. have ex-
ploited the (dpy)–Zn–(dpy) linkage to create bisporphyrin-
based energy-transfer systems.[10] However, to the best of
our knowledge bridging bis(dipyrrinato)zinc motifs have not
been used as linking moieties to produce charge-separating
devices. Complex 1 was designed to test this potential utility.
The synthesis of complex 1 is summarized in Scheme 1. It
involves several key precursors. The first of these, fullero-
to)zinc linker provides effective electronic coupling between
the two fullerene subunits.
Deeper insight into the extent of electronic communica-
tion between the constituents in 1 came from fluorescence
measurements. In line with their different 0–*0 transitions
(l=485 nm for (dpy)–Zn–(dpy) and l=690 nm for C60),
(dpy)–Zn–(dpy) and C60 exhibit *0–0 fluorescence maxima
around l=520 (quantum yield 6.0ꢃ10À3) and 715 nm (quan-
tum yield 6.0ꢃ10À4), respectively.[14] It is fairly well estab-
lished that substituents, if they are redox- and photoinactive,
play an insignificant role in modulating the basic properties
of fulleropyrrolidines.[15] This appreciation led us to chose
N-methyl fulleropyrrolidine as a reference. In the case of 1,
excitation was effected at a number of different wave-
lengths, namely l=330, 400, 435, and 485 nm, which corre-
spond to a C60 maximum, a (dpy)–Zn–(dpy) absorption max-
imum, nearly equal C60/ACTHNUGRTENUNG(dpy)–Zn–(dpy) absorption intensi-
ties, and a maximal absorption for the (dpy)–Zn–(dpy) frag-
ment, respectively. Taken in concert, these studies thus
served to confirm the observation of little appreciable emis-
sion from the (dpy)–Zn–(dpy) fragment (>99% quenching
in the case of excitation at l=485 nm) and the presence of
appreciable C60-centered fluorescence at l=715 nm (see
Figure 1). Although the quantum yield of the latter emission
was found to vary somewhat based on the specific choice of
excitation wavelength and solvent, such findings are fully
consistent with effective energy transfer from the (dpy)–Zn–
(dpy) linker to the flanking C60 acceptor subunits. Notably,
they provide support for a facile but not necessarily quanti-
tative transduction of singlet excited states with efficiencies
of up to 50% (50% in toluene, 41% in THF, 33% in benzo-
nitrile). When excitation wavelengths of lex =330 and
400 nm are used, the quantum yields are close to 6.0ꢃ10À4
(i.e., identical to that of the C60 reference) in toluene and
THF. With lex =435 nm the quantum yield is somewhat
lower (i.e., 3.6ꢃ10À4), and with an excitation wavelength of
lex =485 nm, these quantum yields are around 3.0ꢃ10À4,
2.5ꢃ10À4, and 2.0ꢃ10À4 in toluene, THF, and benzonitrile,
respectively.
Scheme 1. Synthesis of complex 1. i) C60/N-octylglycine, toluene, reflux
then HCl, 37%; ii) pyrrole, trifluoroacetic acid, 62%; iii) p-chloranil;
iv) [ZnACHTUNGTRENNUNG(OAc)2]·H2O, CHCl3/MeOH, 74%.
pyrrolidine 4 (Scheme 1), was obtained in 37% yield from a
1,3-dipolar cycloaddition of an azomethine ylid, which was
generated in situ from 4-(diethoxymethyl)benzaldehyde and
N-octylglycine,[11] to C60 followed by acidic cleavage of the
resulting acetal.[12] The resulting aldehyde was then used to
prepare the corresponding 5-aryl-substituted dipyrrome-
thane 3 by treatment with an excess of pyrrole in the pres-
ence of acid and in the absence of air (Scheme 1).[13] Oxida-
tion with p-chloranil (to produce a putative dipyrromethene
2; structure not shown) followed by treatment with zinc ace-
tate gave complex 1 (for experimental details see the Sup-
porting Information).
Complex 1 gave NMR spectroscopic and mass spectro-
metric data consistent with its proposed structure (see the
Supporting Information). The spectra also confirmed that
the ground state UV/Vis absorption spectrum of 1 is more
than just the linear sum of the component spectra. Whereas
the long wavelength absorption (i.e., lmax for the 0–*0 transi-
tion) for (dpy)–Zn–(dpy) appears at l=485 nm and C60
shows a characteristic 0–*0 transition at l=690 nm (and
also maxima at l=435 and 330 nm), complex 1 gives rise to
a spectrum that is best described as a broad absorption with
scarcely visible peaks at l=485 and 435 nm (see the Sup-
porting Information). Such differences are consistent with
the presence of strong electronic interactions between the
individual constituents (i.e., C60 and (dpy)–Zn–(dpy)) in
complex 1 and leads us to conclude that the bis(dipyrrina-
Figure 1. Fluorescence spectra of (dpy)–Zn–(dpy) (black) and 1 (grey) in
THF at RT, with an identical optical absorbance intensity of 0.2 at an ex-
citation of l=480 nm; please note that the l=650–800 nm region has
been amplified by a factor of 10 to facilitate interpretation.
Chem. Eur. J. 2009, 15, 3956 – 3959
ꢂ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3957

T. Torres, J. L. Sessler, D. M. Guldi et al.
Further support for the proposed energy transfer mecha-
nism came from an analysis of the excitation spectrum;
here, the C60 fluorescence at l=715 nm was used. A spec-
trum was observed under these conditions that is in reasona-
ble agreement with the ground-state absorption, including
the l=435, 485, and 690 nm features that are ascribed to
both constituents.
with solvent polarity. For example, in toluene the lifetime is
as short as 80 ps, whereas in anisole, THF, and ortho-dichlor-
obenzene the lifetimes are 2300, 900 and 1700 ps, respective-
ly. Mechanistically, we assign this process to an intersystem
crossing.
In contrast with the above, photoexcitation of 1 at l=
480 nm reveals a new transient spectrum that is distinctly
different from the singlet or triplet features of the C60 and
(dpy)–Zn–(dpy) constituents (see Figure 2), in addition to
Additional support for the (dpy)–Zn–(dpy)-to-C60 energy
transduction hypothesis came from monitoring the fluores-
cence lifetime of 1 at l=715 nm following excitation at l=
480 nm. This linked complex failed to give any detectable
fluorescence signal at l=520 nm (corresponding to (dpy)–
Zn–(dpy) emission; vide supra) within our instrumental
time resolution of 0.1 ns, a lifetime of 1.6 ns was observed
for the emission at l=715 nm in all solvents. This lifetime
matches that seen for the C60 reference and is thus fully con-
sistent with the proposed energy-transfer process. Consider-
ing that the C60 fluorescence lifetime is not affected by the
presence of (dpy)–Zn–(dpy), we postulate that yet another
deactivation mechanism is operative in photoexcited form
of 1. It is also worth noting that quantum yield for the C60
emission following irradiation of the (dpy)–Zn–(dpy) linker
is reduced compared with that obtained upon exclusive C60
excitation. On this basis, we conclude that an electron-trans-
fer process could be competing with the proposed energy-
transfer deactivation.
Conclusive information about the photoproducts, that is,
electron versus energy transfer, came from transient absorp-
tion spectroscopy. As a predicate for studying linked system
1, a differential spectrum was recorded immediately after
subjecting C60 to excitation with a l=387 nm laser pulse;
this produced a spectrum characterized by a broad absorp-
tion maximum at l=880 nm, which is attributed to the C60
singlet excited state (Esinglet =1.76 eV), a species formed with
a rate constant of 1ꢃ1012 sÀ1. The features of the singlet ex-
cited state decay slowly (8.0ꢃ108 sÀ1) through intersystem
crossing to the energetically lower-lying triplet excited state
(Etriplet =1.50 eV), with a triplet quantum yield close to
unity. In contrast, femtosecond transient absorption meas-
urements of the control linker (dpy)–Zn–(dpy) (model com-
pound 5; see the Supporting Information) with excitation at
Figure 2. Differential absorption spectra (visible and near infrared) ob-
tained upon femtosecond flash photolysis (l=480 nm) of 1 (ca. 1.0ꢃ
10À6 m) in THF, 20 ps after the laser excitation, at RT. Inset: Time-absorp-
tion profile of the spectra shown above at l=1005 nm, reflecting the for-
mation and decay of a radical ion pair state.
the typical transient bleaching of the excited (dpy)–Zn–
(dpy) state seen upon irradiation of the linker alone. For in-
stance, transient maxima at l=510 and 1005 nm are seen
roughly 50 ps after the laser pulse. The l=1005 nm feature
is a fingerprint absorption characteristic of the one-electron-
reduced radical anion of C60.[14] On the basis of comparison
with the spectrum of the radiolytically generated one-elec-
tron-oxidized radical cation of (dpy)–Zn–(dpy), the l=
510 nm feature is assigned to an oxidized bis(dipyrrinato)-
zinc linker. Thus, the observation of maxima at l=510 and
1005 nm is taken as unambiguous evidence for a fast intra-
molecular charge-separation process involving formation of
a radical-ion-pair state.
l=480 nm revealed strong singlet–singlet features (Esinglet
=
2.47 eV) that start to grow in immediately following photo-
excitation.
The radical-ion-pair state produced following l=480 nm
photoexcitation is metastable and recombines on the time
scale of several hundred picoseconds. Applying unimolecu-
lar fitting functions to the transitions in the near-infrared
(i.e., one-electron-reduced radical anion of C60) allowed the
following radical-ion-pair-state lifetimes to be determined:
1130 ps in toluene (with l=387 nm photoexcitation),
1516 ps in anisole (with l=387 nm photoexcitation), 990 ps
in THF (with l=480 nm photoexcitation), and 1290 ps in
ortho-dichlorobenzene (with l=480 nm photoexcitation).
Similar lifetimes were also derived from the analogous mul-
tiwavelength analysis of the l=510 nm transition (i.e., one-
This fast process, with a half-life that varies from 9 to
22 ps depending on the solvent polarity, is followed by a
much slower process. This latter slower-decaying component
exhibits a lifetime that is typically in the order of thousands
of picoseconds; however, its kinetics of decay vary strongly
3958
www.chemeurj.org
ꢂ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 3956 – 3959

A Bisfullerene–Bis(dipyrrinato)zinc Complex
COMMUNICATION
electron-oxidized radial cation spectrum of (dpy)–Zn–
(dpy)); in this latter case, competing processes, such as
energy transfer and intersystem crossing, hampered an accu-
rate kinetic analysis.[16]
[1] J. L. Segura, N. Martin, Chem. Soc. Rev. 2000, 29, 13–25; J.-J. Yin,
Y.-G. Li, B. Li, W.-X. Li, L.-M. Jin, J.-M. Zhou, Q.-Y. Chen, Chem.
Commun. 2005, 3041–3043.
[2] T. Habicher, J.-F. Nierengarten, V. Gramlich, F. Diederich, Angew.
Chem. 1998, 110, 2019–2022; Angew. Chem. Int. Ed. 1998, 37, 1916–
1919.
[3] Y. Rio, G. Enderlin, C. Bourgogne, J.-F. Nierengarten, J.-P. Gissel-
brecht, M. Gross, G. Accorsi, N. Armaroli, Inorg. Chem. 2003, 42,
8783–8793.
Analyzing the transient changes that are still present at
the end of the charge-recombination process, namely, at
around 3000 ps following excitation, reveals a reasonable
agreement with the features that corresponds to the C60 trip-
let excited state, a species characterized by a maximum at
l=690 nm. The same l=690 nm feature is also observed in
complementary nanosecond experiments performed in tolu-
ene and THF. Thus, it appears that the immediate fate of
the charge-separated species is not regeneration of the
ground state of 1, but rather a longer-lived triplet. Such a
finding is, of course, fully consistent with the proposed initial
charge-separation process.
Electrochemical experiments with 1 in dichloromethane
provided thermodynamic insights into the competition be-
tween electron and energy transfer. By simply adding the
first oxidation potential of (dpy)–Zn–(dpy) and the first re-
duction potential of C60, the radical-ion-pair-state energy
was determined to be 1.9 eV in dichloromethane. The C60
singlet excited state, as the other deactivation product of 1,
is around 1.76 eV. It is implicit that changing the solvent po-
larity from, for example, toluene to benzonitrile impacts the
radical-ion-pair-state energy, whereas the singlet excited
state energies of (dpy)–Zn–(dpy) and C60 remain invariant.
In other words, electron transfer becomes increasingly favor-
able, in a thermodynamic sense, over energy transfer.
In summary, the use of a (dpy)–Zn–(dpy) linker allows
the facile tethering of two C60 subunits and gives rise to an
electronically coupled system that allows for effective
charge separation following photoexcitation. This linker acts
as a versatile redox relay and ensures the efficient mediation
of charges across this unit. The ease of this approach leads
us to propose that it can be used to produce a range of
charge-separating and photovoltaic devices. In an effort to
build on the present approach and create systems that might
be capable of stabilizing the formation of longer-lived
charged separated states, current work is focused on the syn-
thesis of multicomponent donor-acceptor systems linked
through (dpy)–Zn–(dpy) motifs.
[4] L. Sꢀnchez, N. Martin, D. M. Guldi, Angew. Chem. 2005, 117, 5508–
5516; Angew. Chem. Int. Ed. 2005, 44, 5374–5382; L. Sꢀnchez, M.
Sierra, N. Martin, A. J. Myles, T. J. Dale, J. Rebek, Jr., W. Seitz,
D. M. Guldi, Angew. Chem. 2006, 118, 4753–4757; Angew. Chem.
Int. Ed. 2006, 45, 4637–4641; J. L. Sessler, J. Jayawickramarajah, A.
Gouloumis, G. Dan Pantos, T. Torres, D. M. Guldi, Tetrahedron
2006, 62, 2123–2131; J. L. Sessler, C. M. Lawrence, J. Jayawickra-
marajah, Chem. Soc. Rev. 2007, 36, 314–325; T. Torres, A. Goulou-
mis, D. Sꢀnchez-Garciꢀ, J. Jayawickramarajah, W. Seitz, D. M.
Guldi, J. L. Sessler, Chem. Commun. 2007, 292–294; J. L. Sessler, J.
Jayawickramarajah, A. Gouloumis, T. Torres, D. M. Guldi, S. Maldo-
nado, K. J. Stevenson, Chem. Commun. 2005, 1892–1894.
[5] J. Fan, Y. Wang, A. J. Blake, C. Wilson, E. S. Davies, A. N. Khlobys-
tov, M. Shroder, Angew. Chem. 2007, 119, 8159–8162; Angew.
Chem. Int. Ed. 2007, 46, 8013–8016.
[6] E. Wood Tabitha, A. Thompson, Chem. Rev. 2007, 107, 1831–1861.
[7] R. Ziessel, G. Ulrich, A. Harriman, New J. Chem. 2007, 31, 496–
501.
[8] H. Fischer, M. Schubert, Ber. 1924, 57B, 610–617; H. Fischer, H.
Orth, Die Chemie des Pyrrols, Akademische Verlag, Leipzig, 1937.
[9] Y. Zhang, A. Thompson, S. J. Rettig, D. Dolphin, J. Am. Chem. Soc.
1998, 120, 13537–13538; A. A.-S. Ali, R. E. Benson, S. Blumentritt,
T. S. Cameron, A. Linden, D. Wolstenholme, A. Thompson, J. Org.
Chem. 2007, 72, 4947–4952.
[10] L. Yu, K. Muthukumaran, I. V. Sazanovich, C. Kirmaier, E. Hindin,
J. R. Diers, P. D. Boyle, D. F. Bocian, D. Holten, J. S. Lindsey, Inorg.
Chem. 2003, 42, 6629–6647.
[11] J. Li, H. Grennberg, Chem. Eur. J. 2006, 12, 3869–3875.
[12] M. Prato, M. Maggini, Acc. Chem. Res. 1998, 31, 519–526.
[13] C.-H. Lee, J. S. Lindsey, Tetrahedron 1994, 50, 11427–11440; D. T.
Gryko, C. Clausen, K. M. Roth, N. Dontha, D. F. Bocian, W. G.
Kuhr, J. S. Lindsey, J. Org. Chem. 2000, 65, 7345–7355; P. D. Rao, S.
Dhanalekshmi, B. J. Littler, J. S. Lindsey, J. Org. Chem. 2000, 65,
7323–7344.
[14] D. M. Guldi, M. Prato, Acc. Chem. Res. 2000, 33, 695–703.
[15] D. M. Guldi, M. Maggini, Gazz. Chim. Ital. 1997, 127, 779–785.
[16] When excitation wavelengths of l=330 and 400 nm are used, the
quantum yields are close to 6.0ꢃ10À4 (i.e., identical to that of the
C60 reference) in toluene and THF. With lex =435 nm, the quantum
yield is somewhat lower (3.6ꢃ10À4), and with an excitation wave-
length of l=485 nm, the quantum yields are around 3.0ꢃ10À4, 2.5ꢃ
10À4, and 2.0ꢃ10À4 in toluene, THF, and benzonitrile, respectively.
Further support for this proposed energy transfer mechanism comes
from the observation that using excitation at the C60 fluorescence
maximum (i.e., l=715 nm) gives rise to a spectrum that is in reason-
able agreement with the ground-state absorption spectrum, includ-
ing the l=435, 485, and 690 nm features that are characteristic of
both constituents. However, because the quantum yield for the C60
emission following irradiation of the (dpy)–Zn–(dpy) linker is re-
duced compared to that obtained upon direct C60 activation, we con-
clude that the intramolecular charge separation is not fully efficient.
We have no evidence for a dissociation of 1, either in the ground,
excited, or radical-ion-pair states.
Acknowledgements
This work has been supported by the Spanish MEC (CTQ-2005-08933-
BQU, Consolider-Ingenio 2010, Nanociencia Molecular-CSD2007-00010),
the Comunidad de Madrid (S-0505/PPQ/000225), the ESF-MEC
(SOHYDS, MAT-2006-28180-E), the Robert A. Welch Foundation (grant
F-1018), the National Science Foundation (CHE-0749571), the Deutsche
Forschungsgemeinshaft (SFB 583), FCI, the COST (D35), and the Office
of Basic Energy Science of the US Department of Energy.
Keywords:
electron
transfer
·
fullerenes
·
Received: December 6, 2008
supramolecular chemistry
Published online: March 12, 2009
Chem. Eur. J. 2009, 15, 3956 – 3959
ꢂ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3959