Fluorescent perylene diimide rotaxanes: Spectroscopic signatures of wheel-chromophore interactions

Article abstract of DOI:10.1002/chem.200601014

[2]- and [3]-rotaxanes with a tetraphenoxy perylene diimide core were synthesized. Hydrogen bonding between the wheel and the imide changes the optical properties of the perylene chromophore: the absorption and fluorescence spectra are red-shifted. The decay times of the rotaxanes are shorter in comparison with that of the axle. Single molecule fluorescence measurements reveal relatively narrow distributions of emission maxima and decay times. The averages are in agreement with ensemble measurements. The observed red shifts make the perylene diimide a suitable chromophore for sensing the position of the wheel on the axle.

Full text of DOI:10.1002/chem.200601014

FULL PAPER
DOI: 10.1002/chem.200601014
Fluorescent Perylene Diimide Rotaxanes: Spectroscopic Signatures of
Wheel–Chromophore Interactions
Jacob Baggerman,[a] Dhiredj C. Jagesar,[a] Renaud A. L. VallØe,[b] Johan Hofkens,[b]
Frans C. De Schryver,[b] Frauke Schelhase,[c] Fritz Vçgtle,[c] and Albert M. Brouwer*[a]
Abstract: [2]- and [3]-rotaxanes with a tetraphenoxy perylene diimide core were
synthesized. Hydrogen bonding between the wheel and the imide changes the opti-
cal properties of the perylene chromophore: the absorption and fluorescence spec-
tra are red-shifted. The decay times of the rotaxanes are shorter in comparison
Keywords: fluorescence spectrosco-
py
· rotaxanes · single-molecule
with that of the axle. Single molecule fluorescence measurements reveal relatively
narrow distributions of emission maxima and decay times. The averages are in
agreement with ensemble measurements. The observed red shifts make the pery-
lene diimide a suitable chromophore for sensing the position of the wheel on the
axle.
studies · time-resolved spectrosco-
py · UV/Vis spectroscopy
Introduction
tecting the shuttling process in rotaxanes is less
common.[7–12] It is well known in the field of molecular sen-
sors that fluorescence is one of the most sensitive means of
detection.[13] Moreover, fluorescence spectroscopy is very
suitable for kinetic measurements with high time resolution
and high dynamic range. For fluorescence to be useful as a
probe for the position of the wheel it is important that inter-
action between the wheel and the chromophore changes the
properties of the latter.
In this paper we describe a [2]rotaxane and a [3]rotaxane
with a strongly fluorescent perylene diimide group. In the
rotaxanes, the presence of the wheel(s) leads to changes in
both the absorption and fluorescence properties of the chro-
mophore. The axle is used as reference compound. Further-
more the (spectro)electrochemical properties of the axle
and the rotaxanes are reported. The measurements show
that the wheel(s) stabilize(s) the anionic species. Also the
fluorescence of individual molecules is studied. These are
the first studies of a mechanically interlocked system at the
single molecule level.
Rotaxanes are supramolecular nanoscale architectures
which have recently attracted much attention because of
their possible applications as molecular switches and
motors.[1–3] Rotaxanes consist of a large ring, the wheel, and
an axle, which is threaded through the wheel and kept in
place by bulky stoppers at the end of the axle. Controlling
the relative position of the wheel with respect to the axle is
one of the main challenges for the desired applications.
Movement of the wheel can be induced using redox,[4] acid/
base[5] or photochemical stimulation.[6] The displacement of
the wheel is often monitored using NMR or changes in the
optical absorbance spectra. The use of fluorescence for de-
[a] Dr. J. Baggerman, D. C. Jagesar, Dr. A. M. Brouwer
Van ’t Hoff Institute for Molecular Sciences
University of Amsterdam, Nieuwe Achtergracht 129
1018 WS Amsterdam (The Netherlands)
Fax : (+31)20-525-5670
E-mail: A.M.Brouwer@uva.nl
[b] Dr. R. A. L. VallØe, Prof. Dr. J. Hofkens, Prof. Dr. F. C. De Schryver
Department of Chemistry, KU Leuven, Celestijnenlaan 200F
3001 Heverlee (Belgium)
Results and Discussion
[c] Dr. F. Schelhase, Prof. Dr. F. Vçgtle
KekulØ-Institut für Organische Chemie und Biochemie
der Universität Bonn
Synthesis: The rotaxanes were prepared using the so-called
trapping method (Scheme 1).[14] The 4-tritylphenolate ion
forms a complex with the wheel. This complex acts as a
supramolecular nucleophile and reacts with the benzylic
bromide to form the rotaxane. Because of the presence of
Gerhard-Domagk-Strasse 1, 53121 Bonn (Germany)
Supporting information for this article is available on the WWW
under http://www.chemeurj.org/ or from the author.
Chem. Eur. J. 2007, 13, 1291 – 1299
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1291

two bromides it is possible to get [2]- and [3]rotaxanes 2 and
3. If one equivalent of the wheel is used the main product is
the [2]rotaxane 2. The third product is a small fraction of
the axle 1. This is formed by reaction between the uncom-
plexed phenolate and the benzylic bromide. The three com-
pounds were separated using column chromatography.
are much smaller in DMSO. The maxima of the axle 1,
[2]rotaxane 2 and [3]rotaxane 3 are 577, 580, and 587 nm,
respectively, for the absorption and 615, 620, and 624 nm,
respectively, for the fluorescence spectra. Also in trifluoro-
ethanol the differences between the spectra of the axle and
the rotaxanes are relatively small but now the absorption
maxima are around 600 nm and the fluorescence maxima
are around 650 nm.
The results show that the excitation energies of the pery-
lene imide chromophore are lower in the rotaxanes 2 and 3
than in the axle 1. We attribute this to hydrogen bonding be-
tween the carbonyl groups of the perylene imide and the
amide hydrogens of the wheel. The fact that shifts are seen
in the absorption spectra, not only in the emission spectra,
demonstrates that the hydrogen bonds exist already in the
ground state, in contrast to systems studied previously.[6,7]
The shifts of the fluorescence maxima are larger compared
with those of the absorption. This indicates stronger hydro-
gen bonding in the excited state, which is the reason for the
red shift of both the absorption and the emission bands.
The Stokes shifts for the axle 1 are small in nonpolar sol-
vents (820–890 cmꢀ1), slightly larger in polar solvents with-
Electronic absorption and fluorescence spectra: To study
the influence of the wheel on the perylene imide chromo-
phore, absorption and fluorescence spectra were measured
in several solvents. The absorption spectra in methylcyclo-
hexane and dimethylsulfoxide (DMSO) are depicted in
Figure 1. The fluorescence spectra in methylcyclohexane
and DMSO are shown in Figure 2. The data for all solvents
used are compiled in Tables 1 and 2. Because of the small
amount available of [3]rotaxane 3 this was studied in a limit-
ed number of solvents. In methylcyclohexane the axle 1 has
its absorption maximum at 570 nm. The maxima of the
[2]rotaxane 2 and the [3]rotaxane 3 are red shifted to 593
and 608 nm, respectively. In the fluorescence spectra a simi-
lar red shift is observed from 598 nm for the axle 1 to 631
and 648 nm for 2 and 3, respectively. The observed red shifts
Scheme 1. Synthesis of the rotaxanes studied.
1292
www.chemeurj.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2007, 13, 1291 – 1299

Perylene Diimide Rotaxanes
FULL PAPER
Table 1. Absorption maxima [nm] of axle 1, [2]rotaxane 2 and [3]ro-
taxane 3 in different solvents, and shifts induced by the presence of the
macrocyclic ring(s) [cmꢀ1].
Solvent
1
2
3
Shift 1[a]
Shift 2[b]
methylcyclohexane
toluene
chloroform
tetrahydrofuran
ethanol
trifluoroethanol
acetonitrile
DMSO
570
576
587
568
578
593
570
577
593
598
602
590
595
609
586
580
608
680
639
424
656
494
443
479
90
1096
610
642
612
587
523
295
[a] Energy difference between the first absorption maximum of the axle
and that of the [2]rotaxane. [b] Energy difference between the first ab-
sorption maximum of the axle and that of the [3]rotaxane.
Table 2. Fluorescence maxima [nm] of axle 1, [2]rotaxane 2 and [3]ro-
taxane 3 in different solvents, and shifts induced by the presence of the
macrocyclic ring(s) [cmꢀ1].
Solvent
1
2
3
Shift 1[a]
Shift 2[b]
methylcyclohexane
toluene
chloroform
tetrahydrofuran
ethanol
trifluoroethanol
acetonitrile
DMSO
598
607
622
603
622
645
610
615
631
640
642
638
643
654
639
620
648
646
649
875
849
501
910
525
213
744
131
1290
995
669
651
624
143
235
Figure 1. Absorption spectra of the axle 1 (c), [2]rotaxane 2 (a) and
[3]rotaxane 3 (d) in methylcyclohexane (top) and DMSO (bottom).
All spectra are scaled to the same maximum absorbance.
[a] Energy difference between the first fluorescence maximum of the
axle and that of the [2]rotaxane. [b] Energy difference between the first
fluorescence maximum of the axle and that of the [3]rotaxane.
out H-bond donor capabilities (chloroform, tetrahydrofuran,
acetonitrile, and DMSO; 960–1150 cmꢀ1) and still larger in
hydrogen bonding ethanol (1224 cmꢀ1) and especially tri-
fluoroethanol (1360 cmꢀ1). The solvatochromic effect in the
non-specifically interacting solvents is small compared with
the effect of hydrogen bond donating solvents. For the [2]ro-
taxane 2 all Stokes shifts are in the range 1000–1400 cmꢀ1,
that is, generally larger than in the axle 1 because in all
cases hydrogen bonds are stronger in the excited state. Hy-
drogen bond donating solvents in this case do not cause es-
pecially large shifts, as they do in 1. Similar spectral shifts
were observed in studies of the binding between melamine
and perylene imides by Würthner et al.[15,16] They attributed
this effect to the polarization of the imide carbonyl groups
by hydrogen bonding. In their case the shifts are smaller
than in the rotaxanes discussed here.
The effect of the different types of solvent/solute interac-
tion on the absorption and fluorescence maxima can be ana-
lyzed using linear solvation energy relationships
(LSER).[17,18] One of the most successful proposals for the
description of LSER is the Kamlet-Taft expression [Eq. (1)].
*
~
n ¼ n0 þ aa þ bb þ sp
ð1Þ
Figure 2. Fluorescence spectra of the axle 1 (c), [2]rotaxane 2 (a)
The coefficients a, b and s describe the sensitivity of, in this
and [3]rotaxane
3 (d) in methylcyclohexane (top) and DMSO
(bottom). All spectra are scaled to the same maximum intensity.
case, the absorption or fluorescence maxima for the differ-
Chem. Eur. J. 2007, 13, 1291 – 1299
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
1293

A. M. Brouwer et al.
ent solvent properties. The solvent properties are described
by the parameters a, b and p*. The parameter a indicates
the hydrogen bond donation ability, the b parameter ac-
counts for the hydrogen bond acceptance ability and p* is a
polarity/polarizability parameter. The observed absorption
and fluorescence maxima (in wavenumbers) together with
the parameters for the solvents used are shown in Table 3.
Fitting the data and the parameters with Equation (1) gives
the coefficients in Table 4.
the fluorescence maxima. The effect may be due to stabiliza-
tion of the ground state by electron donating solvents. The
effects related to the hydrogen bonding donating and ac-
cepting ability are similar for the axle 1 and the [2]rotaxane
2. This is remarkable because it was expected that hydrogen
bond accepting solvents would have a stronger effect in case
of the [2]rotaxane 2 due to their tendency to break the inter-
action between the chromophore and the wheel.
The coefficient s is negative for the axle 1 and slightly
positive for the [2]rotaxane 2.
The negative value for the ab-
Table 3. Observed absorption and fluorescence maxima in wavenumbers [cmꢀ1] of the axle 1 and [2]rotaxane
2 in different solvents and the solvent parameters[17] for the Kamlet–Taft expression.
sorption and fluorescence indi-
cate a stabilization of the excit-
ed state by polar solvents. In
case of the [2]rotaxane 2, how-
ever, this stabilization is absent.
This can be explained by a
weakening of the hydrogen
bonds between the wheel and
the perylene imide carbonyl
groups in more polar solvents.
The weakening of the hydrogen
bonds reduces the stabilization
of the excited state. The desta-
Solvent
Axle 1
[2]Rotaxane 2
Parameters
nabs
nem
nabs
nem
a
b
p*
methylcyclohexane
toluene
chloroform
tetrahydrofuran
ethanol
trifluoroethanol
acetonitrile
DMSO
17544
17361
17036
17606
17301
16863
17544
17331
16722
16474
16077
16584
16078
15504
16393
16260
16863
16722
16611
16949
16807
16420
17065
17241
15845
15625
15576
15674
15552
15291
15649
16129
0.00
0.00
0.44
0.00
0.83
1.51
0.19
0.00
0.00
0.11
0.00
0.55
0.77
0.00
0.31
0.76
0.00
0.54
0.58
0.58
0.54
0.73
0.75
1.00
Table 4. Coefficients and standard deviations for the Kamlet–Taft ex-
pression obtained by fitting the data in Table 3 to Equation (1).
bilizing effect on the hydrogen bonds is apparently larger
than the stabilizing effect of the polarity, which gives a net
destabilizing effect.
Coefficient
Axle 1
[2]Rotaxane 2
Fluorescence
Absorption Fluorescence Absorption
n0
a
b
17530ꢁ160 16710ꢁ120
16730ꢁ130 15680ꢁ160
Electrochemistry: Perylene diimides have beside their
strong absorption and high fluorescence quantum yield also
useful electrochemical properties. They have high reduction
potentials, that is, they are strong electron acceptors. The
electrochemical properties of the axle 1 and the rotaxanes 2
ꢀ270ꢁ130
400ꢁ250
ꢀ420ꢁ290
0.728
ꢀ540ꢁ100
270ꢁ190
ꢀ600ꢁ220
0.927
ꢀ270ꢁ110
460ꢁ220
73ꢁ250
0.799
ꢀ300ꢁ140
25ꢁ260
26ꢁ300
0.662
s
R2
In Figure 3 the observed absorption and fluorescence
maxima are plotted versus the maxima predicted by Equa-
tion (1) using the coefficients obtained by fitting. These
plots show the correlation between the experimental data
and the model. The correlation is clearly present although
the regression coefficients (R2 in Table 4) are not very large.
Only for the fluorescence maxima of the axle 1 a good cor-
relation is found. The obtained coefficients are therefore
not very reliable but they can still be used to extract some
qualitative information about the influence of the different
solvent properties on the absorption and fluorescence spec-
tra.
The coefficient a is negative, indicating that hydrogen
bond donating solvents stabilize the excited state. This
effect is stronger for fluorescence than for absorption be-
cause solvent relaxation is required. This is stronger for the
axle 1 than for the [2]rotaxane 2 because in the latter the in-
teraction of the chromophore with the wheel competes with
the interaction of the chromophore with the solvent. The
positive coefficient b shows that the hydrogen bond accept-
ing solvents have the opposite effect on the spectra: they in-
crease the energy gap between the ground state and the ex-
cited state. This effect is larger for the absorption than for
Figure 3. Plots of the maxima obtained from fitting the absorption (a and
c) and fluorescence (b and d) maxima of the axle 1 (a and b) and [2]ro-
taxane 2 (c and d) to the Kamlet-Taft equation versus the observed
maxima.
1294
www.chemeurj.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2007, 13, 1291 – 1299

Perylene Diimide Rotaxanes
FULL PAPER
and 3 were determined by means of cyclic voltammetry
(Figure 4). The measured redox potentials are summarized
in Table 5.
60 mV. This indicates that chemical processes are taking
place at the timescale of the electrochemical measurement.
Probably, these involve conformational changes in which the
overall interaction between the wheel and imide increases.
The reduction potentials of the [3]rotaxane 3 are affected
even more strongly. Now both the first and second reduction
are shifted to higher potentials. Also the oxidation of the re-
duced species is affected more. These effects indicate that
the wheels influence the stability of the anions. The fact that
the re-oxidation is affected more than the reduction shows
that interaction of the wheel with the anions is stronger than
with the neutral species. This is also visible in the absorption
spectra of the anions (Figure 5).
Figure 4. Cyclic voltammograms of the axle 1 and the rotaxanes 2 and 3
in dichloromethane, sweep rate 0.1 Vsꢀ1, concentration ca. 1 mm, sup-
porting electrolyte n-Bu4PF6.
Table 5. Peak potentials of the reduction and back oxidation of the axle
1 and the rotaxanes 2 and 3 in dichloromethane, and halfwave potentials
for the reversible reduction of the axle. All relative to Fc/Fc+.
Compound
First reduction
Second reduction
Ered [V] Eox [V] E1/2 [V] Ered [V] Eox [V] E1/2 [V]
axle 1
ꢀ1.20
ꢀ1.11
ꢀ0.86
ꢀ0.67
ꢀ1.15
ꢀ1.34
ꢀ1.34
ꢀ1.18
ꢀ1.26
ꢀ1.14
ꢀ0.94
ꢀ1.30
[2]rotaxane 2 ꢀ1.10
[3]rotaxane 3 ꢀ0.88
Figure 5. Absorption spectra of the radical anion (top) and the dianion
(bottom) of the axle 1 (a) and [2]rotaxane 2 (c) in dichlorome-
thane.
The halfwave reduction potentials found for the axle 1,
ꢀ1.15 and ꢀ1.30 V, are similar to the values reported for
similar compounds.[15,19,20] The separation between the re-
duction and oxidation waves is about 80 mV, which is only
slightly larger than the theoretically expected 60 mV for a
reversible process. The presence of the wheel changes the
redox behavior of the perylene core. In the case of the [2]ro-
taxane 2 the first reduction is shifted to a higher potential,
but the second reduction appears almost unchanged. The
position of the oxidation wave of the reduced species is,
however, changed for both the radical anion and the dia-
nion. The peaks are shifted +0.25 and +0.12 V, respectively.
The distance between the corresponding reduction and oxi-
dation waves is much larger than the theoretically expected
Only the spectra for the axle 1 and the [2]rotaxane 2 were
measured because the amount of [3]rotaxane 3 was too
small to obtain good results. The spectrophotometer used
could only measure wavelengths between 190 and 1100 nm.
This is the reason why the lowest energy absorption band of
the radical anion is not completely visible. The spectra of
the radical anion and the dianion of the axle 1 and the [2]ro-
taxane 2 are similar to the spectra observed for similar pery-
lene diimide derivatives.[15,19,20] The absorption maxima and
estimated molar absorption coefficients are shown in
Table 6. The molar absorption coefficients were estimated
using the coefficients measured for the axle 1 and [2]ro-
taxane 2 in chloroform and assuming full conversion of the
Chem. Eur. J. 2007, 13, 1291 – 1299
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
1295

A. M. Brouwer et al.
Table 6. Absorption maxima [nm] and estimated molar absorption coeffi-
cients [mꢀ1 cmꢀ1] in parentheses of the neutral species and anions of the
axle 1 and the [2]rotaxane 2 in dichloromethane.
chloroform are found to be comparable to those of similar
perylene bisimide derivatives.[21,22] In DMSO the decay is
faster than in toluene and chloroform, and cannot be de-
scribed with a single exponential. This behavior is independ-
ent of concentration, and is therefore not associated with ag-
gregation.[19] Intramolecular electron transfer between the
perylene diimide and the 4-tritylphenoxy stopper group is
unlikely to occur on thermodynamic grounds. The reduction
Compound
Neutral
Radical anion
Dianion
lmax (e)
lmax1 (e)
lmax2 (e)
lmax3 (e)
lmax (e)
Axle 1
580
1082
976
790
682
G
U
(13”103)
G
ACHTREUNG
[2]rotaxane 2
1075
974
G
G
U
N
N
potential of the excited chromophore is Ered + eE00
=
(ꢀ1.1 + 2.1) V = +1 V versus Fc/Fc+, which is too low to
allow oxidation of a mono-alkoxybenzene unit.[23] It is
known, however, that the solvent DMSO can be oxidized by
photoexcited C60, which in its T1 state has a reduction poten-
tial of about +0.8 V (vs Fc/Fc+).[24,25] DMSO acts as a sacri-
ficial electron donor,[26] so that a substantial concentration
of the C60 radical anion can be generated. Irradiation with
visible light of axle 1 in argon-saturated DMSO indeed pro-
duced the characteristic absorption band of the radical
anion at 790 nm. Thus, electron transfer between the photo-
excited perylene imide chromophore and the solvent is a
quenching pathway for the former. Upon prolonged irradia-
tion, however, other bands were observed as well, which in-
dicates that follow-up or side reactions occur.
The need to invoke a second exponential term to describe
the fluorescence decay can be attributed to conformational
heterogeneity. Different orientations of the aryloxy substitu-
ents at the bay positions give rise to different conformation-
al species which are likely to have slightly different excita-
tion energies and reduction potentials, which affect the elec-
tron transfer rates.
For the [2]rotaxane 2 the decay time is slightly shorter
than for the axle 1. In DMSO, the difference is bigger than
in toluene and chloroform, and the short component has
gained importance. This can be understood as a result of the
increased reduction potential of the perylene chromophore
in 2, which leads to an increased rate and efficiency of elec-
tron transfer quenching. It should be realized that this effect
is limited by the additional structural reorganization re-
quired to form the optimally hydrogen bonded radical
anion, as shown above.
The second wheel of the [3]rotaxane 3 has a larger effect
on the decay time. In this case the decay times in toluene
and chloroform are both substantially shortened. The same
is observed for the long lived component in DMSO. In the
case of the [3]rotaxane 3 a small contribution from an addi-
tional short component can be distinguished at 640 nm. The
time constant of this component is 0.9 ns in toluene and
0.2 ns in DMSO, respectively, and may be due to some con-
formational relaxation in the excited state which leads to
red shift of the spectrum in time.[7] A corresponding rise, ex-
pected at the long wavelength side of the emission band,
could not be resolved.
neutral species to the anions. The absorption spectra of the
radical anions of the [2]rotaxane 2 and the axle 1 are almost
the same. The maxima that occur at 976 and 1082 nm in the
case of axle 1 are slightly blue-shifted in case of the [2]ro-
taxane 2. The maximum at 790 nm shows a small red shift.
Bigger differences are observed in the range 400–600 nm,
but these are partly due to the presence of small amounts of
neutral molecules. The blue shift indicates a larger stabiliza-
tion of the ground state in comparison to the excited state.
The extra stabilization of the ground state relative to that of
the excited state in the radical anion of the [2]rotaxane 2 is
apparently quite small. In the case of the dianion this extra
stabilization is larger. The spectrum of the dianion of [2]ro-
taxane 2 is narrowed and has a higher molar absorption co-
efficient at the maximum. This might be due a change in the
conformation of the perylene core. The stronger binding of
the wheel can give rise to some steric repulsion between
wheel and the bulky tert-butyl phenoxy side chains of the
perylene core. This results in a change of the structure and
an increase of the oscillator strength. The narrowing of the
spectrum might also be caused by the reduction of the
number of accessible conformations.
Fluorescence dynamics: Besides the spectral properties, also
the excited state decay times can be influenced by hydrogen
bonding between the wheel and the (excited) chromophore.
To study this, the fluorescence decay times (Table 7) of the
axle 1 and the rotaxanes 2 and 3 in toluene, chloroform and
DMSO were obtained by single photon counting experi-
ments. The solutions were excited with 614 nm laser pulses
and the fluorescence was detected at three different wave-
lengths: 640, 680, and 720 nm. Identical decay traces were
obtained at the three wavelengths, with one exception, dis-
cussed below. The decay times for the axle 1 in toluene and
Table 7. Fluorescence decay times [ns] and amplitudes (in parentheses)
of the axle 1 and the rotaxanes 2 and 3 in some solvents obtained by
single photon counting experiments with excitation at 614 nm.
Solvent
Axle 1
[2]Rotaxane 2
[3]Rotaxane 3
t1 [ns] t2 [ns] t3 [ns]
t1 [ns] t2 [ns] t1 [ns] t2 [ns]
toluene
–
–
6.3
6.4
3.6
–
–
6.0
6.1
3.1
0.9
(0.09)[a] (0.91)
0.9 4.6
(0.16)[a] (0.84)
0.2 1.3
4.4
–
The fluorescence quantum yield in chloroform was mea-
sured using two different references (Table 8). Using the
decay times in chloroform (Table 7) and the quantum yields
(Table 8) the natural rate constant for fluorescence can be
calculated using kf0 = Ff/tf. The natural rate constant is
chloroform
DMSO
1.2
1.3
2.4
(0.09) (0.91) (0.58) (0.42) (0.16)[a] (0.48) (0.36)
[a] Only observed at 640 nm.
1296
www.chemeurj.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2007, 13, 1291 – 1299

Perylene Diimide Rotaxanes
FULL PAPER
1.4”108 sꢀ1 for all three compounds. This shows that the
lowering of the quantum yield and the shortening of the life-
time are due to an increased rate of nonradiative decay.
Table 8. Fluorescence quantum yields of the axle 1, and the rotaxanes 2
and 3 in chloroform.
Reference
Axle 1
[2]Rotaxane 2
[3]Rotaxane 3
perylene red (F=0.96)[27]
cresyl violet (F=0.54)[28]
average
0.94
0.88
0.91
0.91
0.86
0.88
0.63
0.64
0.63
A likely reason for the shorter decay times of the rotax-
anes 2 and 3 in toluene and chloroform compared with
those of the axle 1 is the enhancement of the internal con-
version due to the hydrogen bonding between the chromo-
phore and the wheel. If the shortening of the decay time of
the rotaxanes 2 and 3 is due to the hydrogen bonding then
there could be a correlation between the emission maximum
and the decay time: when the rotaxane is in a conformation
in which the wheel is not close to the perylene imide the
decay time and emission spectrum should be similar to
those of the axle 1. On the other hand when the wheel is hy-
drogen bonded to the imide the decay time will be shorter
and the spectrum red-shifted. Thus, it can be expected that
spectral and temporal characteristics will be correlated be-
cause they are related to the conformational state of the
molecules. Such a correlation between two properties of
sub-ensembles cannot be easily determined using bulk ex-
periments, but it can be readily revealed by single molecule
fluorescence spectroscopy.[21]
Figure 6. Distributions of fluorescence maxima (a and c) and decay times
(b and d) of the axle 1 (a and b) and the [2]rotaxane 2 (c and d) in a
PMMA film.
able. The bulky wheel may have a direct influence on the
relative energies of the conformers, which differ in the ori-
entation of the phenoxy groups,[21,22,29] but it is also possible
that it prevents the trapping of energetically less favorable
conformations in the spin-coated polymer film by creating
some local free volume.[30] This leads to the disappearance
of the short decay times. For the axle 1, no correlation be-
tween the decay times and the emission maxima was found
(see Figure 7). In the case of [2]rotaxane 2 a weak correla-
tion appears to be present, longer decay times being associ-
ated with red-shifted emission maxima. This is opposite to
what was observed in solution, where red shifts due to hy-
drogen bonding between wheel and chromophore cause en-
hanced nonradiative decay. It appears that the restricted dy-
namics in the polymer matrix suppress this effect.
Single molecule spectroscopy: The fluorescence spectra and
decay times of individual molecules of the axle 1 and [2]ro-
taxane 2 in poly(methyl methacrylate) (PMMA) films were
measured. The films were spin coated from chloroform.
Figure 6 shows the distributions of fluorescence maxima and
decay times for the axle 1 and the [2]rotaxane 2. Here a sim-
ilar red shift of the fluorescence is observed as in the ensem-
ble measurements. In the case of the axle 1 the occurrence
maxima are centered around 580 and 600 nm. For the [2]ro-
taxane 2 the occurrence maxima are around 600 and
620 nm. The main maximum of the axle 1 at 580 nm shifts to
620 nm in case of the rotaxane 2. This shift corresponds with
observed red shifts in solution. The occurrence maximum at
600 nm is there in both cases. This maximum might be due
to hydrogen bonding of some molecules with the polymer
(PMMA). This leads to competition between the polymer
and the wheel in the case of the [2]rotaxane 2.
The decay times for both compounds have occurrence
maxima around 5.9 ns. This decay time corresponds with the
decay time in solution. In case of the axle 1 there is an addi-
tional occurrence maximum around 2.9 ns. Such a bimodal
distribution of decay times has been found for related pery-
lene imides.[21,22] This was attributed to the presence of dif-
ferent conformers. Apparently, the presence of the wheel
makes the conformation with the short decay time less prob-
Conclusion
The absorption and fluorescence spectra of the perylene di-
imide chromophore are red-shifted due to hydrogen bond-
ing between the amide groups in the wheel and the carbonyl
groups of the imide. The solvents modulate this effect of the
wheel on the spectral properties of the chromophore. The
largest spectral shift in going from axle to rotaxane is ob-
served in nonpolar solvents such as methylcyclohexane and
the smallest in solvents such as DMSO and trifluoroethanol.
In the case of nonpolar solvents hydrogen bonding between
the chromophore and the wheel is very favorable, but in
DMSO, a strong hydrogen bond acceptor, and trifluoroetha-
nol, a strong hydrogen-bond donor, the red shift is small.
This is due to weakening of the hydrogen bonding between
the imide and the wheel by competing interactions with the
solvent.
Chem. Eur. J. 2007, 13, 1291 – 1299
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
1297

A. M. Brouwer et al.
monochromators in the excitation and emission channels. The excitation
light source was a 450 W Xe lamp and the detector a Peltier cooled
R636-10 (Hamamatsu) photomultiplier tube. The fluorescence spectra
were corrected for the wavelength response of the detection system. The
fluorescence of fluorophores in solution was detected in a right angle ge-
ometry using solutions with low absorbances (below 0.10). Fluorescence
quantum yields were obtained using the comparative method and calcu-
lated with Equation (2):[31]
ðAreference Isample n 2Þ
Fsample ¼ Freference
ð2Þ
ðAsample Ireference n02Þ
The subscripts indicate whether the parameter refers to the sample or
the reference and the symbols have the following meaning: F is the
quantum yield, A is the absorption, I is the integrated emission intensity
across the band, and n and n0 are the refractive indices of the solvent of
the sample and the reference, respectively.
Time-resolved fluorescence: Time-resolved fluorescence was measured
with a time-correlated single-photon counting (TCSPC) setup.[32] This
setup consists of a cavity dumped DCM dye laser (Coherent model 700)
pumped by a mode-locked Ar+ laser (Coherent 486 AS Mode Locker,
Coherent Innova 200 laser). The fluorescence was collected through a po-
larizer at the magic angle (54.78) with respect to the (vertical) polarisa-
tion of the laser beam to exclude polarization effects. A microchannel
plate (Hamamatsu R3809) was used as the detector. The overall response
function (IRF) was measured from the Rayleigh scattering of colloidal
silicon dioxide (LUDOX from DuPont).
The fluorescence decay times were obtained from the fluorescence
decays by fitting the data with the convolution of the appropriate
number of exponentials and the IRF. The fitting was done by numerical
iterative reconvolution with a home written program. The program was
implemented in Igor Pro 5 and uses a Levenberg–Marquardt algorithm
to minimize c2.
Figure 7. Plots of the emission maxima versus the decay times of the axle
1 (top) and the [2]rotaxane 2 (bottom).
The spectral changes make the perylene diimide a suitable
chromophore for reporting the position of the wheel on the
axle. The presence of the wheel also gives rise to an increase
of the reduction potentials indicating a stabilization of the
anionic species by the wheel. This is also visible in the ab-
sorption spectra of the anions which shift to the blue. The
fluorescence decay times become shorter in the series axle 1
> [2]rotaxane 2 > [3]rotaxane 3. A similar correlation be-
tween the decay times and emission maxima was, however,
not found with single molecule spectroscopy of 1 and 2 in
PMMA films.
In DMSO, partial quenching of the fluorescence occurs
due to electron transfer between the imide and the solvent,
which acts as a sacrificial electron donor. The efficiency of
this process increases with increasing reduction potential,
that it, it is more effective in the series axle 1 < [2]rotaxane
2 < [3]rotaxane 3.
Single-molecule fluorescence spectroscopy: For this work a confocal fluo-
rescence microscope setup was used at the Katholieke Universiteit
Leuven, which is described elsewhere.[33]
Cyclic voltammetry: Cyclic voltammograms were obtained of ꢂ1 mm sol-
utions of the compound of interest with tetrabutyl ammonium hexafluor-
ophosphate as supporting electrolyte. The measurements were done in
gas-tight, single-compartment, three electrode cells equipped with plati-
num working, coiled platinum wire auxiliary and silver wire pseudorefer-
ence electrodes. The cell was connected to a computer controlled PAR
model 283 potentiostat. The redox potentials were determined against
the ferrocene/ferrocenium redox couple.
Spectroelectrochemistry: UV/Vis spectroelectrochemical measurements
were done in optically transparent thin-layer electrochemical (OTTLE)
cells with a Pt minigrid working electrode and CaF2/quartz windows.[34,35]
An HP 8453 diode array spectrophotometer was used to measure the
spectra.
Acknowledgements
ˇ
We have greatly appreciated the help of Frantisek Hartl and Taasje Ma-
habiersing with the electrochemical experiments. This research was sup-
ported in part by The Netherlands Organization for the Advancement of
Research (NWO). We are grateful for support of the European Transna-
tional Research Project CERC3 and of the Deutsche Forschungsgemein-
schaft. We also thank our cooperation partners Professor A. von Zelew-
sky (Fribourg), and Professor J.-P. Sauvage (Strasbourg) in our joint pro-
ject “Topological chiral and cycloenantiomeric catenanes, rotaxanes and
pretzelanes”.
Experimental Section
Steady-state absorption: Electronic absorption spectra were recorded on
two spectrophotometers. The first one is a single beam HP 8453 diode
array spectrophotometer with a spectral range from 190 to 1100 nm and
a spectral resolution of about 2 nm. The second one is a double beam
Varian Cary 3E spectrophotometer, spectral range 190 to 900 nm with
spectral bandwidths down to 0.2 nm. The spectra were recorded in rec-
tangular 1 cm quartz cuvettes.
Steady-state fluorescence: Steady-state fluorescence spectra were record-
[1] A. R. Pease, J. O. Jeppesen, J. F. Stoddart, Y. Luo, C. P. Collier, J. R.
ed on a Spex Fluorolog 3 spectrometer, equipped with double grating
Heath, Acc. Chem. Res. 2001, 34, 433.
1298
www.chemeurj.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2007, 13, 1291 – 1299

Perylene Diimide Rotaxanes
FULL PAPER
[2] R. Ballardini, V. Balzani, A. Credi, M. T. Gandolfi, M. Venturi, Acc.
Chem. Res. 2001, 34, 445.
[20] C.-C. You, F. Würthner, J. Am. Chem. Soc. 2003, 125, 9716.
[21] J. Hofkens, T. Vosch, M. Maus, F. Kçhn, M. Cotlet, T. Weil, A. Herr-
mann, K. Müllen, F. C. De Schryver, Chem. Phys. Lett. 2001, 333,
255.
[22] R. A. L. VallØe, M. Cotlet, M. van der Auweraer, J. Hofkens, K.
Müllen, F. C. De Schryver, J. Am. Chem. Soc. 2004, 126, 2296.
[23] A. Zweig, W. G. Hodgson, W. H. Jura, J. Am. Chem. Soc. 1964, 86,
4124.
[3] J.-P. Collin, C. Dietrich-Buchecker, P. Gavina, M. C. Jimenez-
Molero, J.-P. Sauvage, Acc. Chem. Res. 2001, 34, 477.
[4] A. Altieri, F. G. Gatti, E. R. Kay, D. A. Leigh, D. Martel, F. Paoluc-
ci, A. M. Z. Slawin, J. K. Y. Wong, J. Am. Chem. Soc. 2003, 125,
8644.
[5] J. D. Badjic, V. Balzani, A. Credi, S. Silvi, J. F. Stoddart, Science
2004, 303, 1845.
ˇ
[24] V. Brezovꢁ, A. Gügel, P. Rapta, A. Stasko, J. Phys. Chem. 1996, 100,
[6] A. M. Brouwer, C. Frochot, F. G. Gatti, D. A. Leigh, L. Mottier, F.
Paolucci, S. Roffia, G. W. H. Wurpel, Science 2001, 291, 2124.
[7] G. W. H. Wurpel, A. M. Brouwer, I. H. M. van Stokkum, A. Farran,
D. A. Leigh, J. Am. Chem. Soc. 2001, 123, 11327.
[8] P. N. Taylor, A. J. Hagan, H. L. Anderson, Org. Biomol. Chem. 2003,
1, 3851.
[9] D. H. Qu, Q. C. Wang, J. Ren, H. Tian, Org. Lett. 2004, 6, 2085.
[10] E. M. Perez, D. T. F. Dryden, D. A. Leigh, G. Teobaldi, F. Zerbetto,
J. Am. Chem. Soc. 2004, 126, 12210.
16232.
[25] R. R. Hung, J. J. Grabowski, J. Phys. Chem. 1991, 95, 6073.
[26] M. Koncka, Z. Stojek, Electroanalysis 1995, 7, 1010.
[27] R. Gvishi, R. Reisfeld, Z. Burshtein, Chem. Phys. Lett. 1993, 213,
338.
[28] D. Magde, J. H. Brannon, T. L. Cremers, J. Olmsted, III, J. Phys.
Chem. 1979, 83, 696.
[29] P. Osswald, D. Leusser, D. Stalke, F. Würthner, Angew. Chem. 2005,
117, 254; Angew. Chem. Int. Ed. 2005, 44, 250.
[11] D. A. Leigh, M. A. F. Morales, E. M. Perez, J. K. Y. Wong, C. G.
Saiz, A. M. Z. Slawin, A. J. Carmichael, D. M. Haddleton, A. M.
Brouwer, W. J. Buma, G. W. H. Wurpel, S. Leon, F. Zerbetto,
Angew. Chem. 2005, 117, 3122; Angew. Chem. Int. Ed. 2005, 44,
3062.
[12] H. Onagi, J. Rebek, Jr., Chem. Commun. 2005, 4604.
[13] A. P. de Silva, H. Q. N. Gunaratne, T. Gunnlaugsson, A. J. M.
Huxley, C. P. McCoy, J. T. Rademacher, T. E. Rice, Chem. Rev. 1997,
97, 1515.
[14] C. Seel, F. Vçgtle, Chem. Eur. J. 2000, 6, 21.
[15] F. Würthner, A. Sautter, Chem. Commun. 2000, 445.
[16] F. Würthner, C. Thalacker, A. Sautter, Adv. Mater. 1999, 11, 754.
[17] M. J. Kamlet, J. L. M. Abboud, M. H. Abraham, R. W. Taft, J. Org.
Chem. 1983, 48, 2877.
[30] P. D. Zoon, A. M. Brouwer, R. A. L. VallØe, J. Hofkens, F. C. De
Schryver, unpublished results, 2006.
[31] D. F. Eaton, in Handbook of Organic Photochemistry, Vol. 1 (Ed.:
J. C. Scaiano), CRC Press, Boca Raton, Florida, 1989.
[32] S. I. van Dijk, P. G. Wiering, C. P. Groen, A. M. Brouwer, J. W. Ver-
hoeven, W. Schuddeboom, J. M. Warman, J. Chem. Soc. Faraday
Trans. 1995, 91, 2107.
[33] J. Hofkens, M. Maus, T. Gensch, T. Vosch, M. Cotlet, F. Kçhn, A.
Herrmann, K. Müllen, F. De Schryver, J. Am. Chem. Soc. 2000, 122,
9278.
ˇ
ˇ
[34] M. Krejeik, M. Danik, F. Hartl, J. Electroanal. Chem. Interfacial
Electrochem. 1991, 317, 179.
[35] F. Hartl, H. Luyten, H. A. Nieuwenhuis, G. C. Schoemaker, Appl.
Spectrosc. 1994, 48, 1528.
[18] Y. Marcus, Chem. Soc. Rev. 1993, 22, 409.
[19] F. Würthner, A. Sautter, D. Schmid, P. J. A. Weber, Chem. Eur. J.
2001, 7, 894.
Received: July 15, 2006
Published online: October 26, 2006
Chem. Eur. J. 2007, 13, 1291 – 1299
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
1299