A ferrocene-C60-dinitrobenzene triad: Synthesis and computational, electrochemical, and photochemical studies

Article abstract of DOI:10.1021/jp0136415

Synthesis and physicochemical characterization of a molecular triad comprised of ferrocene, C₆₀, and dinitrobenzene entities is reported. Electrochemical studies revealed multiple redox processes involving all three redox active, ferrocene, C

Full text of DOI:10.1021/jp0136415

J. Phys. Chem. A 2002, 106, 649-656
649
A Ferrocene-C60-Dinitrobenzene Triad: Synthesis and Computational, Electrochemical,
and Photochemical Studies
Francis D’Souza,*,† Melvin E. Zandler,† Phillip M. Smith,† Gollapalli R. Deviprasad,†
Klykov Arkady,† Mamoru Fujitsuka,‡ and Osamu Ito*,‡
Department of Chemistry, Wichita State UniVersity, 1845 Fairmount, Wichita, Kansas 67260-0051, and
Institute of Multidisciplinary Research for AdVanced Materials, Chemical Reaction Science, Tohoku UniVersity,
Katahira, Sendai, 980-8577 Japan
ReceiVed: September 27, 2001; In Final Form: NoVember 12, 2001
Synthesis and physicochemical characterization of a molecular triad comprised of ferrocene, C60, and
dinitrobenzene entities is reported. Electrochemical studies revealed multiple redox processes involving all
three redox active, ferrocene, C60, and dinitrobenzene entities. A total of eight reversible redox couples within
the accessible potential window of o-dichlorobenzene containing 0.1 M (TBA)ClO4 are observed. A comparison
between the measured redox potentials with those of the starting compounds revealed absence of any significant
electronic interactions between the different redox entities. The geometric and electronic structure of the
triad is elucidated by using ab initio B3LYP/3-21G(*) methods. In the energy optimized structure, as predicted
by electrochemical studies, the first HOMO orbitals are found to be located on the ferrocene entity, whereas
the first LUMO orbitals are mainly on the C60 entity with small orbital coefficients on the dinitrobenzene
entity. The subsequent LUMO’s track the observed site of electrochemical reductions of the triad. The
photochemical events in the triad are probed by both steady-state and time-resolved techniques. The steady-
state emission spectra of the triad and the starting dyad, 2-(ferrocenyl)fulleropyrrolidine, are found to be
completely quenched compared to fulleropyrrolidine bearing no redox active substituents. The subpicosecond
and nanosecond transient absorption spectral studies reveal efficient charge separation in the triad. As suggested
by the coefficients of the LUMO orbitals, the transient absorption spectrum of the triad revealed bands
corresponding to the formation of the fulleropyrrolidine anion species. The estimated rise and decay rate
constants are found to be 2.5 × 1011 and 8.5 × 109 s-1 for the triad, and these values compare with a rise and
decay rate constants of 2.2 × 1011 and 4.5 × 109 s-1 for the ferrocene-C60 dyad in benzonitrile. The observed
faster rate constants are attributed to the close spacing of the redox entities of the triad to one another.
Introduction
ferrocene-C60 (donor-acceptor)32-36 and C60-dinitrobenzene
(acceptor-acceptor)37-38 type dyads. The presently developed
Fullerenes, C60 and C70, linked to multiple redox- and/or
photoactive molecular entities1-5 are useful candidates to build
molecular/supramolecular electronic devices6-11 and artificial
light energy harvesting systems.12-26 Toward this, electron
donors such as porphyrin, ferrocene, N,N-dimethylaminophenyl,
ruthenium(II) trisbipyridine, and tetrathiafulvarene have been
employed to form fullerene-electron donor type dyads, whereas
electron acceptors such as benzoquinone derivatives, spiro-
annelated methano groups, cyano groups, fluorine atoms, TCNQ
and DCNQI derivatives, ammonium cations, and nitroaromatic
substituents have been employed to form covalently linked
fullerene-electron acceptor type dyads.1-5,26-28 In addition, a
few triads and tetrads bearing one or more different electron
donor and/or acceptor entities have been elegantly designed,
synthesized, and studied.15,26-31 Much important and useful
information has been obtained from studies involving these
molecular and supramolecular systems.
molecular system combines these two dyads to form a hybrid
triad bearing three redox active entities. The geometric and
electronic structure of this newly developed triad has been
elucidated from optical absorption, electrochemical, and ab initio
B3LPY/3-21G(*) methods. Using steady-state and time-resolved
emission techniques, the excited-state photoinduced electron-
transfer reactions are also probed in the triad.
Experimental Section
Chemicals. Buckminsterfullerene, C60 (+99.95%), was from
BuckyUSA (Bellaire, TX). o-Dichlorobenzene in sure seal
bottle, ferrocenaldehyde, 3,5-dinitrobenzoyl chloride, glycine,
and tetra-n-butylammonium perchlorate, (TBA)ClO4, were from
Aldrich Chemicals (Milwaukee, WI). All chemicals were used
as received unless otherwise stated. o-Dichlorobenzene for
electrochemical studies was dried over CaH2 and distilled under
vacuum prior to the experiments. (TBA)ClO4 was recrystallized
from ethanol and dried in a vacuum oven at 35 °C for 10 days.
Synthesis. 2-(Ferrocenyl)fulleropyrrolidine. This was syn-
thesized according to a general procedure developed for the
preparation of substituted fulleropyrrolidines based on 1,3-
dipolar cycloaddition of azomethine ylides to C60.33,39 For this,
a mixture of C60 (100 mg), glycine (53 mg), and ferrocenalde-
In the present study, we report on the synthesis and physi-
cochemical characterization of a ferrocene-C60-dinitrobenzene,
that is, a donor-(acceptor 1)-(acceptor 2)-type triad (Figure
1). Earlier, several groups including us have reported on
* To whom correspondence should be addressed.
† Wichita State University.
‡ Tohoku University.
10.1021/jp0136415 CCC: $22.00 © 2002 American Chemical Society
Published on Web 01/04/2002

650 J. Phys. Chem. A, Vol. 106, No. 4, 2002
D’Souza et al.
Computational Calculations. The computational calculations
were performed by using either Gaussian 98,40 SPARTAN,41
or Cache MOPAC42 software packages. The graphics of the
HOMO and LUMO molecular orbital coefficients were gener-
ated with the help of GaussView software. The compounds were
fully optimized to a stationary point on the Born-Openheirmer
potential energy surface. The frontier HOMO and LUMO
molecular orbital coefficients were calculated on the fully
optimized structures.
Time-Resolved Absorption Measurements. The subpico-
second transient absorption spectra were recorded by the pump
and probe method. The samples were excited with a second
harmonic generation (SHG, 388 nm) of output from a femto-
second Ti:sapphire regenerative amplifier seeded by SHG of a
Er-dropped fiber (Clark-MXRCPA-2001 plus, 1 kHz, fwhm 150
fs). The excitation light was depolarized. The monitor white
light was generated by focusing the fundamental of laser light
on the flowing D2O/H2O cell. The transmitted monitor light was
detected with a dual MOS linear image sensor (Hamamatsu
Photonics, C6140) or InGaAs photodiode array (Hamamatsu
Photonics, C5890-128).
Nanosecond transient absorption spectra in the NIR region
were measured by means of laser-flash photolysis; 532 nm light
from a Nd:YAG laser was used as the exciting source, and a
Ge-avalanche-photodiode module was used for detecting the
monitoring light from a pulsed Xe lamp as described in our
previous report.43-45
Figure 1. Structure of the investigated (a) 2-(ferrocenyl)fulleropyr-
rolidine dyad and (b) N-(3′,5′-dinitrobenzoyl)-2-(ferrocenyl)fulleropyr-
rolidine triad.
hyde (75 mg) was refluxed in toluene for 6 h. The compound
was purified over silica gel column using toluene/hexane (90:
10 v/v) as eluent. Yield 23%. 1H NMR in CDCl3/CS2 (1:1 v/v),
δ ppm, 5.57 (s, 1H, pyrrolidine H), 5.01, 4.88 (d, d, 2H,
pyrrolidine H), 4.55, 4.46, 4.21 (s, s, s, 1H, 1H, 2H, ferrocene
H), 4.36 (s, 5H, ferrocene H). The 13C NMR spectrum in CDCl3/
CS2 (1:2 v/v) revealed all of the expected peaks.33 UV-visible
in n-hexane, λmax nm, 245 (sh), 255, 307 (br), 325, 429. ESI
mass in CH2Cl2: calcd, 947.65; found, 947.1.
Results and Discussion
The procedure developed for the synthesis of the triad bearing
three different redox active entities, namely, ferrocene, C60, and
dinitrobenzene, is found to be an elegant one. The method of
covalent linking dinitrobenzoyl entity to the fulleropyrrolidine
nitrogen is not much different from the earlier reported Sanger’s
reaction which involved a direct reaction between the fullero-
pyrrolidine secondary nitrogen and 1-fluoro-2,4-dinitroben-
zene.37 The ESI mass spectrum of the triad in the CH2Cl2 matrix
revealed the molecular ion peak in agreement with the calculated
N-(3′,5′-Dinitrobenzoyl)-2-(ferrocenyl)fulleropyrrolidine. This
was synthesized by reacting 2-(ferrocenyl)fulleropyrrolidine (50
mg) and an excess of 3,5-dinitrobenzoyl chloride in CH2Cl2
containing triethylamine (1.1 eq) for 3 h. The compound was
purified over silica gel column using toluene/ethyl acetate (90:
10 v/v) as eluent. Yield 38%. 1H NMR in CDCl3/CS2 (1:1 v/v),
δ ppm, 9.19 (s, 1H, dinitrobenzoyl H), 9.14 (s, 2H, dinitroben-
zoyl H), 5.57 (s, 1H, pyrrolidine H), 5.01, 4.88 (d, d, 2H,
pyrrolidine H), 4.50, 4.21 (m, 4H, ferrocene H), 4.36 (s, 5H,
ferrocene H). UV-visible in n-hexane, λmax nm, 221, 254 (sh),
309, 326 (br), 414 (br). ESI mass in CH2Cl2: calcd, 1141.67;
found, 1141.1.
1
one. The integrated peak intensities of the H NMR spectrum
clearly established the molecular integrity of the triad. The
resonance peak positions of the pyrrolidine and ferrocene entities
were found at positions almost identical to that of the starting
material, 2-ferrocenylfulleropyrrolidine. Only the dinitrobenzoyl
proton resonance peaks revealed a small shielding (0.1-0.2
ppm) as compared to its starting material, 3,5-dinitrobenzoyl
chloride. The optical absorption spectrum of the triad exhibited
absorption bands corresponding to the C60 and dinitrobenzoyl
entities. The absorption peaks corresponding to the ferrocene
entity were merged into the strong absorption bands of C60 and
dinitrobenzoyl entities in the 225-425 nm wavelength region.
Cyclic Voltammetric Studies. Electrochemical studies using
the cyclic voltammetric technique have been performed to
evaluate potentials of the different redox active entities and also
to visualize any electronic interactions between them. Figure 2
shows the cyclic voltammograms of 2-(ferrocenyl)fulleropyr-
rolidine and N-(3′,5′-dinitrobenzoyl)-2-(ferrocenyl)-fulleropyr-
rolidine in dry o-dichlorobenzene containing 0.1 M (TBA)ClO4.
For 2-(ferrocenyl)fulleropyrrolidine, the reversible redox po-
tential corresponding to the oxidation of the ferrocene entity is
located at 0.06 V vs Fc/Fc+, whereas the potentials for the first
three reversible reductions of the C60 entity are located at -1.19,
-1.57, and -2.11 V vs Fc/Fc+. The anodic to cathodic peak
separations were found to range between 100 and 110 mV and
plots of peak current vs square root of scan rates were linear
Instrumentation. The UV-visible spectral measurements
were carried out with a Shimadzu model 1600 UV-visible
spectrophotometer. The fluorescence was monitored by using
a Spex Fluorolog tau-3 spectrometer. A right angle detection
1
method was used. The H NMR studies were carried out on a
Varian 400 MHz spectrometer. Tetramethylsilane (TMS) was
used as an internal standard. Cyclic voltammograms were
obtained by using a conventional three electrode system on a
model AFCB1 bipotentiostat of Pine Instrument Co. (Grove
City, PA). A platinum disk electrode was used as the working
electrode. A platinum wire served as the counter electrode. An
Ag/AgCl electrode, separated from the test solution by a fritted
supporting electrolyte/solvent bridge, was used as the reference
electrode. The potentials were referenced to an internal fer-
rocene/ferrocenium redox couple. All of the solutions were
purged prior to spectral and electrochemical measurements using
argon gas.

A Ferrocene-C60-Dinitrobenzene Triad
J. Phys. Chem. A, Vol. 106, No. 4, 2002 651
Figure 3. Space filling model of the B3LPY/3-21G(*) optimized
structure of N-(3′,5′-dinitrobenzoyl)-2-(ferrocenyl)fulleropyrrolidine.
Importantly, the observed eight reversible redox processes
(one oxidation and seven reductions) within the accessible
potential window of the solvent under ordinary solution condi-
tions are noteworthy. It is now well-known that under special
solution conditions, that is, solvents with extended potential
window and reduced temperature, pristine fullerenes (C60 and
C70) exhibit six one-electron reversible reductions.47 Owing to
the presence of a large number of reversible reductions,
fullerenes are considered potential candidates to build molecular
electronic devices.4-11 The present triad which undergoes eight
reversible redox reactions under ordinary solutions conditions
is more appealing in this regard than pristine fullerenes because
a large number of reversible redox reactions can be ac-
complished under normal experimental conditions. Our present
results suggest that fullerenes bearing one or more redox active
groups undergoing reversible redox reactions are potentially
more useful than pristine C60 specially to build molecular
electronic devices.
Figure 2. Cyclic voltammograms of (a) 2-(ferrocenyl)fulleropyrrolidine
and (b) N-(3′,5′-dinitrobenzoyl)-2-(ferrocenyl)fulleropyrrolidine in 0.1
M (TBA)ClO4, o-dichlorobenzene. Scan rate ) 100 mV/s.
SCHEME 1
for these redox reactions under the present solution conditions.
The measured redox potentials agreed well with the potentials
reported earlier for ferrocene linked fullerene derivatives33 and
fulleropyrrolidine derivatives46 in general.
Computational Studies
Interestingly, the voltammogram of the triad exhibited revers-
ible redox processes corresponding to the reduction of the
appended dinitrobenzoyl entity, in addition to the ferrocene and
C60 entities. The electrode processes corresponding to the
reduction of the dinitrobenzoyl entity are located at -1.35 and
-1.76 V vs Fc/Fc+. The anodic to cathodic peak separation is
found to vary between 160 and 170 mV, and multiple scanning
of the voltammograms in the potential range did not change
the current or shape of the voltammograms. A comparison
between the peak currents for these processes with that of the
ferrocene and fullerene redox couples in Figure 2b suggests that
each of the redox processes of the dinitrobenzoyl group involves
two one-electron reactions, presumably involving simultaneous
reduction of the two nitro groups of the dinitrobenzoyl entity.
The reversible redox potential corresponding to the appended
ferrocene oxidation is located at 0.05 V, whereas the first three
reversible fulleropyrrolidine reductions are located at -1.19,
-1.58, and -2.13 V vs Fc/Fc+. An examination of these
potentials with that of the starting material, 2-(ferrocenyl)-
fulleropyrrolidine in Figure 2a suggests the absence of signifi-
cant electronic interactions between the different redox entities.
The sequence of the occurrence of different electrode reactions
in the triad is summarized in Scheme 1, where Fc represents
the ferrocenyl entity, C60 represents the fulleropyrrolidine entity
while, and DNB represents the dinitrobenzoyl entity of the triad.
Prediction of the accurate geometric and electronic structure
of supramolecular systems bearing one or more different types
of redox and/or photoactive groups poses a challenge for
researchers working in this area.48 The geometric parameters
are obtained easily by semiempirical PM3 and AM1 methods
because these methods possess built-in parameters. However,
semiempirical methods often yield errors in the calculated
electronic structure. Ab initio calculations based on either density
functional theory or Hartree-Fock methods at a moderate level
often yield correct geometry and electronic structures which
agree with the experimental results. Here, we have utilized both
semiempirical PM3 and ab initio methods at the 3-21G(*) level
to probe the geometry and electronic structure of the triad.
In the present study, the ab initio calculations at the B3LYP/
3-21G(*) level and the HF/3-21G(*) level have been performed
using Gaussian 98 software, whereas the semiempirical calcula-
tions have been performed using Cache MOPAC or SPARTAN
programs. As discussed below, however, only the B3LYP/3-
21G(*) methods resulted in the prediction of plausible geometry
and electronic structure. Figure 3 shows the space-filling model
of the energy minimized structure on the Born-Openheirmer
potential energy surface, whereas Figure 4 shows the first
HOMO and the first five LUMO frontier molecular orbitals. In
the optimized structure, the pyrrolidine ring is puckered in such

652 J. Phys. Chem. A, Vol. 106, No. 4, 2002
D’Souza et al.
Figure 4. Frontier HOMO, LUMO, LUMO+1, LUMO+2, LUMO+3, and LUMO+4 orbitals of the investigated N-(3′,5′-dinitrobenzoyl)-2-
(ferrocenyl)fulleropyrrolidine triad.
a way that the 2-ferrocenyl entity is anti to the nitrogen atom
and the N-dinitrobenzoyl group is trans49 to the 2-ferrocenyl
group. The two nitro groups of the dinitrobenzoyl entity are
coplanar with the phenyl ring, whereas the carbonyl group is
out of the plane of the dinitrobenzene ring with a dihedral angle
of about 25°. The carbonyl group lies along the N-C(ferrocene)
bond of the pyrrolidine ring with a dihedral angle of 2°. The
two cyclopentadienyl rings of the ferrocene entity form an
almost eclipse configuration. The carbon atoms of the 6,6-ring
junction from which the pyrrolidine ring is formed are nearly
0.35 Å away from the positions of the normal C60 spheroid
carbons. Earlier, an X-ray analysis of fulleropyrrolidine in a
self-assembled 2-(4′-pyridyl)fulleropyrollidine and tetraphen-
ylporphyrinatozinc(II)50 had revealed such structural distortions
in agreement with calculations at the B3LYP/STO-3G* level.
In the energy minimized structure, the center-to-center distance
between the C60 spheroid and ferrocene entities is found to be
8.28 Å, whereas this difference between the C60 spheroid and
dinitrobenzoyl entities is around 9.06 Å, and between the
ferrocene and dinitrobenzoyl entities, this distance is about 6.50
Å. The edge-to-edge distance between the C60 spheroid and
ferrocene entities is found to be 2.59 Å, whereas this distance
between the C60 spheroid and dinitrobenzoyl entities is about
4.32 Å, and between the ferrocene and dinitrobenzoyl entities,
this distance is about 4.55 Å.
In agreement with the electrochemical results, the frontier
HOMO is found to be located on the ferrocene entity (Figure
4a). Interestingly, as expected from the electrochemical data,
the majority of the frontier LUMO is located on the C60
spheroid, whereas a small portion of the orbital is also found
to exist on one of the nitro groups of the dinitrobenzoyl entity
(Figure 4b). The calculated (gas phase) HOMO-LUMO gap
is found to be 2.12 eV, and this compares with the electro-
chemically measured (difference between the first oxidation and

A Ferrocene-C60-Dinitrobenzene Triad
J. Phys. Chem. A, Vol. 106, No. 4, 2002 653
Figure 7. (a) Transient absorption spectra of N-(3′,5′-dinitrobenzoyl)-
2-(ferrocenyl)fulleropyrrolidine in benzonitrile at 4 and 760 ps after
the laser irradiation (388 nm, <150 fs). (b and c) Absorption time
profiles at 970 nm. Solid lines were fitted curves assuming rise and
decay of 2.5 × 1011 and 8.5 × 109 s-1 of rate constants, respectively.
Figure 5. Steady-state fluorescence emission spectrum of (i) 2-(phen-
yl)fulleropyrrolidine, (ii) 2-(ferrocenyl)fulleropyrolidine, and (iii) N-(3′,5′-
dinitrobenzoyl)-2-(ferrocenyl)fulleropyrrolidine in o-dichlorobenzene;
λex ) 390 nm. Concentration of fullerenes employed ) 10-4 M.
Figure 8. Transient absorption spectra of (a) fulleropyrrolidine, (b)
2-(ferrocenyl)fulleropyrrolidine, and (c) N-(3′,5′-dinitrobenzoyl)-2-
(ferrocenyl)fulleropyrrolidine in benzonitrile at 100 ns after the laser
irradiation (352 nm, 6 ns). Figure inset: absorption time profiles at
720 nm.
especially B3LYP, is recently being demonstrated by Schaefer
and co-workers on electron affinities of aromatic compounds.52
As mentioned earlier, our attempts using semiempirical PM3
and Hartree-Fock ab initio methods at a 3-21G(*) level resulted
either in incorrect geometry or electronic structure for the studied
triad. The PM3 calculations using Cache MOPAC revealed a
plausible geometry (later used as the starting geometry in the
B3LYP and HF calculations) but the HOMO orbitals were found
to be on the C60 entity. The PM3 calculations performed using
SPARTAN revealed at least three stable structures, the two low
energy structures had HOMO’s on ferrocene, but revealed
interactions between the carbonyl group or phenyl ring hydro-
gens of dinitrobenzoyl entity and the ferrocene iron resulting
into a distorted ferrocene (the coplanarity of the two cyclopen-
tadienyl rings was lost). The third SPARTAN structure was
similar to the Cache MOPAC structure with HOMO on the
fullerene entity. Interestingly, the structure calculated using the
Hartree-Fock 3-21G(*) method optimized to a geometry close
to the B3LYP structure, but like the Cache MOPAC PM3
results, the HOMO was found to be on the fullerene entity. In
the absence of any spectroscopic evidence for such distorted
geometry and electronic properties, we did not further use these
methods and adopted only the density functional method
(B3LYP) where the calculated geometry and electronic proper-
ties agreed well with the experimental observations.
Figure 6. (a) Transient absorption spectra of 2-(ferrocenyl)fullero-
pyrrolidine in benzonitrile at 4 and 760 ps after the laser irradiation
(388 nm, <150 fs). (b and c) Absorption time profiles at 970 nm. Solid
lines were fitted curves assuming rise and decay of 2.2 × 1011 and
4.5 × 109 s-1 of rate constants, respectively.
first reduction potential) value of 1.23 V. The orbital energies
of the first HOMO and the first five LUMO orbitals are found
to be -5.723, -3.597, -3.572, -3.520, -3.343, and -3.239
eV, respectively. The subsequent virtual orbitals, that is,
LUMO+1, LUMO+2, LUMO+3, and LUMO+4, are shown
in Figures 4c-d. These virtual orbitals are localized as: C60,
dinitrobenzene, C60, dinitrobenzene, and C60 entities. Interest-
ingly, this trend tracks the sequence of the site of electron
transfer of the triad as depicted in Scheme 1. To our knowledge,
this is the first example where, in a supramolecular system
comprised of three redox entities, the calculated HOMO and
virtual orbitals track the observed multistep electrochemical
redox reactions. It may be mentioned here that only recently
the validity of molecular orbitals generated by density functional
methods is being recognized.51 The accuracy of these methods,

654 J. Phys. Chem. A, Vol. 106, No. 4, 2002
D’Souza et al.
Figure 9. Energy level diagrams of (a) fulleropyrrolidine (C60), (b) 2-(ferrocenyl)fulleropyrrolidine (Fc-C60), and (c) N-(3′,5′-dinitrobenzoyl)-2-
(ferrocenyl)fulleropyrrolidine (Fc-C60-DNB).
Photochemical Studies. Figure 5 shows the room-tempera-
ture fluorescence emission spectra of the 2-(ferrocenyl)fullero-
pyrrolidine dyad, the ferrocene-C60-dinitrobenzene triad, and
a reference compound, 2-(phenyl)fulleropyrrolidine, bearing no
ferrocene or dinitrobenzene entities in o-dichlorobenzene. In
agreement with the earlier report,35 fulleropyrrolidine exhibited
a broad emission band at around 712 nm when excited at 390
nm. Changing the wavelength of excitation to other absorption
wavelengths of fulleropyrrolidine, revealed no change in the
emission maxima. Interestingly, the emission of both the dyad
and the triad is found almost completely quenched, indicating
the occurrence of excited-state reactions from the singlet excited
fulleropyrrolidine. Such efficient quenching is also observed in
benzonitrile and acetonitrile solvents. To verify the quenching
mechanism and characterize the reaction products, subpico-
second as well as nanosecond transient absorption studies on
both the dyad and triad have been performed.
Transient spectral results for the dyad, 2-(ferrocenyl)fullero-
pyrrolidine, observed after the 150 fs laser excitation at 388
nm in benzonitrile are shown in Figure 6a. Immediately after
the laser pulse, a broad transient absorption band appeared in
the wavelength region of 800-900 nm and this has been
attributed to the singlet-singlet absorption of the fullero-
pyrrolidine entity.44 In addition, a sharp band at 970 nm
corresponding to the radical anion of the fulleropyrrolidine
entity45 is also observed. A nearly complete decay of both the
transient absorption bands is observed after a delay time of 760
ps. As shown in Figure 6b, the rise of the 970 nm band takes
place as fast as 10 ps and this may correspond to the occurrence
of charge separation from the singlet excited state. From the
rise of this band, the evaluated rate of the charge separation
generating the ion-pair state is found to be 2.2 × 1011 s-1. The
970 nm band is found to decay with the rate constant of 4.5 ×
109 s-1, and this has been ascribed to the charge recombination
of the ion-pair state. The residue of the decay at 1000 ps in
Figure 6c may be attributed to the tail of the triplet excited state
of the fullerene entity, whose absorption maximum may be
shorter than 800 nm (spectrum at 760 ps in Figure 6a).
has been attributed to the radical anion of the C60 entity.44 The
broad background absorption observed in the whole spectral
region has been attributed to the singlet-singlet absorption. At
760 ps, most of the broad singlet-singlet absorption disappears
leaving traces of the absorption corresponding to the radical
anion of the C60 entity at 970 nm and the triplet-triplet
absorption at wavelengths shorter than 850 nm. The rise of the
760 nm band is shown in Figure 7b, from which the evaluated
charge-separation rate is found to be 2.5 × 1011 s-1. A
comparison between the rates obtained for the dyad and the
triad suggests that the dinitrobenzene entity slightly accelerates
the charge-separation process. The decay rate of the triad
estimated from Figure 7c is found to be 8.5 × 109 s-1, which
is twice as large as the corresponding rates obtained for
2-(ferrocenyl)fulleropyrrolidine. It is interesting to note that the
presence of the dinitrobenzene entity does not slow down the
charge-recombination process, although such effect is anticipated
from the observed small coefficients of the LUMO on the
dinitrobenzene entity (Figure 4b). It may be mentioned here
that the dinitrobenzene entity actually assists the charge-
separation and charge-recombination processes between ferro-
cene and fulleropyrrolidine entities upon photoexcitation.
Figure 8 shows the transient absorption spectra observed after
nanosecond laser excitation of fulleropyrrolidine and 2-(ferro-
cenyl)fulleropyrrolidine and N-(3′,5′-dinitrobenzoyl)-2-(ferro-
cenyl)fulleropyrrolidine in benzonitrile. The observed intense
transient band of fulleropyrrolidine at 720 nm with a shoulder
at 850 nm was attributed to the triplet-triplet absorption,
generated via the intersystem crossing from the singlet excited
state within 1.2 ns. The intensity of the absorption maximum
decays slowly with a rate constant of 1.6 × 105 s-1 as shown
in the inserted time-profile in Figure 8. In the case of
2-(ferrocenyl)fulleropyrrolidine, the shape of the absorption band
is similar to that of fulleropyrrolidine, indicating the formation
of the triplet state of the fullerene entity which is also supported
by a similar decay rate constant of 2.5 × 105 s-1 (inset of Figure
8). However, the absorption intensity is reduced by a factor of
about 1/8, suggesting that most of the excited singlet state
undergoes the charge-separation process rather than formation
of a C60 triplet state.
The transient absorption spectra and the time profiles observed
after laser excitation of the ferrocene-C60-dinitrobenzene triad
are shown in Figure 7a, in which the 970 nm band seen at 4 ps

A Ferrocene-C60-Dinitrobenzene Triad
J. Phys. Chem. A, Vol. 106, No. 4, 2002 655
The absorption intensity of the transient absorption band of
N-(3′,5′-dinitrobenzoyl)-2-(ferrocenyl)fulleropyrrolidine is found
to be considerably lower than that of fulleropyrrolidine (1/60).
Although the peak at 720 nm is broad, this weak absorption
has also been attributed to the triplet state because the lifetime
is almost the same as that obtained for fulleropyrrolidine (rate
constant ) 1.9 × 105 s-1). One of the reasons for the low
distribution of the triplet state may be the faster charge-
separation process of the triad than that of dyad.
Acknowledgment. The authors are thankful to the donors
of the Petroleum Research Fund, administered by the American
Chemical Society, National Institutes of Health, National
Science Foundation CCLI and A&I (to FD), and the Dreyfus
Foundation (to MES) for support of this work. The authors are
also thankful to the High Performance Computing Center of
the Wichita State University for lending SGI ORIGIN 2000
computer time.
Supporting Information Available: The ESI-mass spectra
in CH2Cl2 matrix, UV-visible spectra in n-hexane, and 1H NMR
spectra in CDCl3/CS2 (1:1 v/v) of 2-(ferrocenyl)fulleropyrroli-
dine and N-(3′5′-dinitrobenzoyl)-2-(ferrocenyl)-fulleropyrroli-
dine. This material is available free of charge via the Internet
at http://pubs.acs.org.
Figure 9 compares the energy level diagrams for the different
photochemical events of fulleropyrrolidine, 2-(ferrocenyl)-
fulleropyrrolidine and N-(3′,5′-dinitrobenzoyl)-2-(ferrocenyl)-
fulleropyrrolidine. The high triplet yield in fulleropyrrolidine
is attributed to the high quantum yield of intersystem crossing
of the singlet excited state (Figure 9a). The observed fast charge-
separation rates in the dyad and triad are in good agreement
with lower ion-pair states compared with the triplet states. That
is, the energy gaps between the singlet excited states and the
ion pair states are about 0.5 eV (energy diagrams b and c in
Figure 9), which may be normal or top region of the Marcus
parabola for fullerene derivatives.53 This suggests the unlikely
formation of a direct triplet state of the fullerene entity via
intersystem crossing of the singlet excited state, which is in
good agreement with the observed decrease in the formation of
the triplet excited state in Figure 8. The faster charge-
recombination rate of the ion-pair of the triad compared to that
of the dyad could be attributed to spatial back electron transfer
from the dinitrobenzene entity to the ferrocene entity. This
provides indirect evidence for the migration of the electron
toward the dinitrobenzene entity in the ion pair state of the triad,
although this process is energetically unfavorable. This unique
photochemical and photophysical behavior is further supported
by the ab initio calculated distribution of LUMO as shown in
Figure 4b.
References and Notes
(1) Martin, N.; Sachez, B.; Illescas, B.; Perez, I. Chem. ReV. 1998, 98,
2427.
(2) Prato, M.; Maggini, M. Acc. Chem. Res. 1998, 31, 519.
(3) Carano, M.; Ceroni, P.; Paolucci, F.; Roffia, S.; Da Ros, T.; Prato,
M.; Sluch, M. I.; Pearson, C.; Petty, M. C.; Bryce, M. R. J. Mater. Chem.
2000, 10, 269.
(4) Fullerene and Related Structures; Hirsch, A., Ed.; Springer: Berlin,
1999; Vol. 199.
(5) Long, N. J. Angew. Chem., Int. Ed. Engl. 1995, 34, 21.
(6) Molecular Electronic-Science and Technologie; Aviram, A., Ed.;
Engineering Foundation: New York, 1989.
(7) Aviram, A.; Ratner, M. Chem. Phys. Lett. 1974, 29, 277.
(8) Launay, J. P. In Molecular Electronics in Granular Nanoelectronics;
Ferry, D. K., Ed.; New York, 1995.
(9) Introduction to Molecular Electronics; Petty, M. C., Bryce, M. R.,
Bloor, D., Eds.: Oxford University Press: New York, 1995.
(10) Prasad, P. N.; Williams, D. J. Introduction to Nonlinear Optical
Effects in Molecules and Polymers; Wiley: New York, 1991.
(11) Nalwa, M. S. AdV. Mater. 1993, 5, 341.
(12) Photoinduced Electron Transfer; Fox, M. A., Channon, M., Eds.;
Elsevier: Amsterdam, 1988; Parts A-D.
(13) Kurreck, H.; Huber, M. Angew. Chem., Int. Ed. Engl. 1995, 34,
849.
(14) Staab, H. A.; Feurer, A.; Hauck, R. Angew. Chem., Int. Ed. Engl.
1994, 33, 2428.
Summary
(15) Roest, M. R.; Verhoeven, J. W.; Schiddeboon, W.; Warman, J. M.;
Lawson, J. M.; Paddon-Row, M. N. J. Am. Chem. Soc. 1996, 118, 1762.
(16) Maruyama, K.; Osuka, A. Pure Appl. Chem. 1990, 62, 1511.
(17) Gust, D.; Moore, T. A. Science 1989, 244, 35.
(18) (a) Gust, D.; Moore, T. A. Top. Curr. Chem. 1991, 159, 103. (b)
Gust, D.; Moore, T. A.; Moore, A. L. Pure Appl. Chem. 1998, 70, 2189.
(19) Gust, D.; Moore, T. A.; Moore, A. L. Acc. Chem. Res. 1993, 26,
198.
(20) Wasielewski, M. R. Chem. ReV. 1992, 92, 435.
(21) Paddon-Row, M. N. Acc. Chem. Res. 1994, 27, 18.
(22) Sutin, N. Acc. Chem. Res. 1983, 15, 275.
(23) Bard, A. J.; Fox, M. A. Acc. Chem. Res. 1995, 28, 141.
(24) Meyer, T. J. Acc. Chem. Res. 1989, 22, 163.
Synthesis of a supramolecular triad comprised of three redox
active entities, namely, ferrocene, C60, and dinitrobenzene, has
been accomplished. Electrochemical studies revealed multiple
redox processes involving all three redox active entities. The
observed eight reversible redox couples within the accessible
potential window of o-dichlorobenzene suggest that fullerene
bearing redox active groups which undergo reversible redox
reactions may be more suitable for building molecular charge
storage devices than pristine fullerenes. Computational calcula-
tions performed using the ab initio B3LYP/3-21G(*) method
predicted a plausible geometry and electronic structure of the
investigated triad, whereas such calculations performed by
semiempirical PM3 or Hartree-Fock 3-21G(*) revealed either
an incorrect geometry or incorrect electronic structure which
did not agree with the experimental observations. The calculated
HOMO orbital by the B3LYP/3-21G(*) method was found to
lie primarily on the ferrocene entity, whereas the LUMO orbitals
are mainly located on the C60 entity with small orbital
coefficients on the dinitrobenzene entity. The orbital coefficients
of the subsequent LUMO orbitals track the observed site of
electrochemical reductions of the triad. Both steady-state
fluorescence and time-resolved picosecond transient absorption
studies revealed efficient charge separation and charge recom-
bination in the studied triad, and this has been attributed to the
close spacing of the entities of the triad to one another. The
estimated rise and decay rate constants are found to be slightly
higher for the triad as compared to the corresponding rates of
the dyad.
(25) Piotrowiak, P. Chem. Soc. ReV. 1999, 28, 143.
(26) Imahori, H.; Sakata, Y. AdV. Mater. 1997, 9, 537.
(27) Diederich, F.; Gomez-Lopez, M. Chem. Soc. ReV. 1999, 28, 263.
(28) Guldi, D. M. Chem. Commun. 2000, 321.
(29) (a) Liddell, P. A.; Kuciauskas, D.; Sumida, J. P.; Nash, B.; Nguyen,
D.; Moore, A. L.; Moore, T. A.; Gust, D. J. Am. Chem. Soc. 1997, 119,
1400. (b) Kuciauskas, D.; Liddell, P. A.; Lin, S.; Johnson, T. E.; Wighorn,
S. L.; Lindsey, J. S.; Moore, A. L.; Moore, T. A.; Gust, D. J. Am. Chem.
Soc. 1999, 121, 8604. (c) Carbonera, D.; Di Valentin, M.; Corvaja, C.;
Agostini, G.; Giacometti, G.; Liddell, P. A.; Kuciauskas, D.; Moore, A. L.;
Moore, T. A.; Gust, D. J. Am. Chem. Soc. 1998, 120, 4398.
(30) Luo, C.; Guldi, D. M.; Imahori, H.; Tamaki, K.; Sakata, Y. J. Am.
Chem. Soc. 2000, 122, 6535.
(31) D’Souza, F.; Deviprasad, G. R.; El-Khouly, M. E.; Fujitsuka, M.;
Ito, O. J. Am. Chem. Soc. 2001, 123, 5277-5284 and references therein.
(32) Maggini, M.; Karlsson, A.; Scorrano, G.; Sandona, G.; Farnia, G.;
Prato, M. J. Chem. Soc., Chem. Commun. 1994, 589.
(33) Prato, M.; Maggini, M.; Giacometti, C.; Scorrano, G.; Sandona,
G.; Farnia, G. Tetrahedron 1996, 52, 5221.
(34) Noworyta, K.; Krinichnaya, E. P.; Kutner, W.; Smith, P. M.;
Deviprasad, G. R.; D’Souza, F. Fullerenes 2000: Electrochemistry and
Photochemistry. Proc. Electrochem. Soc. 2000, 8, 54; Fukuzumi, S.,
D’Souza, F., Guldi, D. M., Eds.

656 J. Phys. Chem. A, Vol. 106, No. 4, 2002
D’Souza et al.
(35) Guldi, D. M.; Maggini, M.; Scorrano, G.; Prato, M. J. Am. Chem.
Soc. 1997, 119, 974.
(36) Poplawska, M.; Byszewski, P.; Kowalska, E.; Diduszko, R.;
Radomska, J. Synth. Met. 2000, 109, 239.
(37) Deviprasad, G. R.; Rahman, M. S.; D’Souza, F. Chem. Commun.
1999, 849.
(38) de la Cruz, P.; de la Hoz, A.; Langa, F.; Martin, N.; Perez, M. C.;
Sanchez, L. Eur. J. Org. Chem. 1999, 3433.
(43) (a) Matsumoto, K.; Fujitsuka, M.; Sato, T.; Onodera, S.; Ito, O. J.
Phys. Chem. B 2000, 104, 11632. (b) Komamine, S.; Fujitsuka, M.; Ito,
O.; Morikawa, K.; Miyata, T.; Ohno, T. J. Phys. Chem. A 2000, 104, 11497.
(44) Nojiri, T.; Watanabe, A.; Ito, O. J. Phys. Chem. A 1998, 102, 5215.
(45) Fujitsuka, M.; Ito, O.; Yamashiro, T.; Aso, Y.; Otsubo, T. J. Phys.
Chem. A 2000, 104, 4876.
(46) Kutner, W.; Noworyta, K., Deviprasad, G. R.; D’Souza, F. J.
Electrochem. Soc. 2000, 147, 2647.
(47) Xie, Q.; Perez-Cordero, E.; Echegoyen, L. J. Am. Chem. Soc. 1992,
114, 3978.
(48) (a) Hehre, W. J. Practical Strategies for Electronic Structure
Calculation; Wavefunction, Inc.: Irvine, CA, 1996. (b) Hehre, W. J. Critical
Assessment of Modern Electronic Structure Methods; Wavefunction, Inc.:
Irvine, CA, 1995.
(49) D’Souza, F.; Zandler, M. E.; Deviprasad, G. R.; Kutner, W. J. Phys.
Chem. A 2000, 104, 6887.
(39) Maggini, M.; Scorrano, G.; Prato, M. J. Am. Chem. Soc. 1993, 115,
9798.
(40) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb,
M. A.; Cheeseman, J. R.; Zakrzewski, V. G.; Montgomery, J. A., Jr.;
Stratmann, R. E.; Burant, J. C.; Dapprich, S.; Millam, J. M.; Daniels, A.
D.; Kudin, K. N.; Strain, M. C.; Farkas, O.; Tomasi, J.; Barone, V.; Cossi,
M.; Cammi, R.; Mennucci, B.; Pomelli, C.; Adamo, C.; Clifford, S.;
Ochterski, J.; Petersson, G. A.; Ayala, P. Y.; Cui, Q.; Morokuma, K.; Malick,
D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.; Cioslowski, J.;
Ortiz, J. V.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi,
I.; Gomperts, R.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.;
Peng, C. Y.; Nanayakkara, A.; Gonzalez, C.; Challacombe, M.; Gill, P. M.
W.; Johnson, B. G.; Chen, W.; Wong, M. W.; Andres, J. L.; Head-Gordon,
M.; Replogle, E. S.; Pople, J. A. Gaussian 98, revision A.7; Gaussian,
Inc.: Pittsburgh, PA, 1998.
(50) D’Souza, F.; Rath, N. P.; Deviprasad, G. R.; Zandler, M. E. Chem.
Commun. 2001, 267.
(51) Stowasser, R.; Hoffmann, R. J. Am. Chem. Soc. 1999, 121, 3414.
(52) (a) Brown, S. T.; Rienstra-Kiracofe, J. C.; Schaefer, H. F. J. Phys.
Chem. A 1999, 103, 4065. (b) Rienstra-Kiracofe, J. C.; Barden, C. J.; Brown,
S. T.; Schaefer, H. F. J. Phys. Chem. A 2001, 105, 524.
(41) SPARTAN PRO; Wavefunction, Inc.: Irvine, CA 92715.
(42) Cache MOPAC-2000, version 4.4; Fujitsu Systems, Inc.:
http://www.cachesoftware.com.
(53) Imahori, H.; Tamaki, K.; Gulidi, D. M.; Luo, C.; Fujitsuka, M.;
Ito, O.; Sakata, Y.; Fukuzumi, S. J. Am. Chem. Soc. 2001, 123, 2607.