Solvent-dependent steady-state fluorescence spectroscopy for searching ESPT-dyes: Solvatochromism of HPTS revisited

Article abstract of DOI:10.1039/b816695a

We reinvestigated the solvatochromism of 8-hydroxypyrene-1,3,6-trisulfonate (pyranine) in conjunction with that of 8-methoxypyrene-1,3,6-trisulfonate and of 1-hydroxypyrene (pyrenol) by use of 25 different solvents. Conclusions for the prediction of ESPT

Full text of DOI:10.1039/b816695a

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/pccp | Physical Chemistry Chemical Physics
Solvent-dependent steady-state fluorescence spectroscopy for searching
ESPT-dyes: solvatochromism of HPTS revisited
Gregor Jung,* Stephan Gerharz and Alexander Schmitt
Received 23rd September 2008, Accepted 18th December 2008
First published as an Advance Article on the web 3rd February 2009
DOI: 10.1039/b816695a
We reinvestigated the solvatochromism of 8-hydroxypyrene-1,3,6-trisulfonate (pyranine) in
conjunction with that of 8-methoxypyrene-1,3,6-trisulfonate and of 1-hydroxypyrene (pyrenol) by
use of 25 different solvents. Conclusions for the prediction of ESPT behaviour of synthetic dyes
were drawn by comparison with the solvatochromism of p-hydroxystyryl Bodipy dyes. Solvents
were chosen according to their Kamlet–Taft parameters a and b for elucidating the acidicity of
the dyes and the basicity of their conjugated bases in the ground and excited state. Comparison of
the spectra of pyranine and pyrenol in solvents with varying b-values revealed that the acidity of
both dyes is similar therein. The well-known ESPT behaviour of pyranine in water is assigned to
a change of the electronic state at a-values B0.7 to 0.8. The high acidity of this excited state also
appears in the vanishing solvatochromism of the photoproduct fluorescence. However, prediction
of an ESPT tendency of synthetic dyes might fail when only fluorescence emission data are
considered. We propose to refer instead to the energetic difference of the 0–0 transition in
absorption together with the solvatochromism of the acidic form in aprotic solvents of similar
polarity.
ESPT in ultrasensitive spectroscopy challenging. Even in
1. Introduction
cases, where ESPT could be initiated by near UV light,
Excited state proton transfer (ESPT) of 8-hydroxypyrene-
photodestruction and photoconversion counteract these
experiments.21,22 This deficiency of known molecules provokes
1,3,6-trisulfonate (HPTS; pyranine) to water was first
described by Forster1,2 more than 50 years ago and is still
¨
the focus of scientific research.3 This photochemical reaction
our search for dye molecules which exhibit ESPT and strong
visible fluorescence in both states. Though, straightforward
exemplifies the more general case of various aromatic alcohols
and systematic synthetic work relies on a tool with which the
which undergo an increase of their acidity upon excitation (for
synthetic direction can be confined and, finally, the ability of
reviews, see ref. 4–6). Naphthol and pyrene derivatives were
ESPT can be predicted. Steady-state fluorescence and UV-Vis
especially thoroughly investigated in the past, but ESPT
spectroscopic signatures are most convenient in this approach
reactions were also detected in stilbene derivatives, fluorescent
as they can be applied to screening techniques.
dyes and proteins.7–11 ESPT is not limited to water as final
The increase of acidity of a dye upon excitation can roughly
acceptor but protonation of other solvents can be observed in
be estimated by thermodynamics applied to optical spectroscopy.
super-photoacids.5,12,13 One reason for the ongoing interest is
A red-shift in the lowest electronic transition of the conjugate
that proton transfer reactions are the most basic chemical
base compared to the acidic form results in a change of the
reactions with an overwhelming importance in many areas of
acidity constant pKA according to
chemistry. Especially helpful are ESPT systems where both
ðhnA ꢁ hnBÞ
substrate and product are fluorescent. Thus, triggering of these
DpKA
ꢀ
ð1Þ
kT ln 10
reactions with short excitation pulses allows for studies of the
underlying photophysical mechanisms which precede the
protolysis with time-resolved fluorescence spectroscopy.3,10–18
Additionally, the fluorescent product states are long-living
(Bns), and the kinetics of diffusion-controlled recombination
processes also can be monitored by the same techniques.6,18–20
In most of the cited systems, excitation has to be performed
in the UV or near-UV range. Ubiquitously excited background
fluorescence, enhanced Raman scattering and the energy
content of the absorbed photons make an observation of
where nA and nB are the frequencies of 0–0 optical transitions
of the acid and the base, respectively.5,6 Time constants for
proton transfer to solvent ought to be faster than the radiation
emission, otherwise ESPT remains hidden. Therefore, kinetic
parameters have to be converted to spectroscopic data to be
useful in screening applications.
Such a more-kinetic approach is based on the analysis of
spectroscopic data of aromatic alcohols, recorded in different
solvents with varying hydrogen-bonding capability.23–26 Most
often, the Kamlet–Taft parameters a and b are used to
describe the ability of a solvent to donate (a) or accept (b) a
hydrogen-bond (HB).24–29 Solvatochromism of a solute, i.e.
the difference of the transition frequency ni to a reference
transition frequency ni0 in dependence of the solvent, then
Biophysical Chemistry, Saarland University, Campus, Building B2 2,
D-66123 Saarbruecken, Germany.
E-mail: g.jung@mx.uni-saarland.de; Fax: +49 (0)681 302 64846;
Tel: +49 (0)681 302 64848
ꢂc
This journal is the Owner Societies 2009
1416 | Phys. Chem. Chem. Phys., 2009, 11, 1416–1426

View Article Online
provides a tool to quantify the stabilization of the excited state
with respect to the ground state. The index i describes whether
properties of the ground state (i = exc) or the excited state
(i = em) of the probe molecule are investigated.6 This
classification results from the different time scales of the
instantaneous absorption/emission process and the subsequent
rearrangement of nuclei. Eqn (2) comprehends solvatochromism
in a mathematical way with p* as a measure of the dipolarity
of the solvent.
since the acidity of the photoacid is increased in the
excited state.
The contribution of HB-donating, protic solvents to
the stabilization of the excited state before ESPT is
ambiguous.24,26 In HPTS, a strong dependence of the red-shift
in fluorescence on parameter a is observed, whereas in
naphthol and its derivatives, only the spectral position of the
excitation maximum is sensitive to a.6 A recently accepted
explanation resolves this discrepancy by ascribing the impact
of a to charge stabilization at the sulfonate groups of HPTS in
protic solvents.24 Recent experiments with derivatives of
HPTS show that strongly withdrawing substituents are
necessary to enable charge transfer in the excited state.13,14
ni = ni0 + pip* + bib + aia
(2)
ai, bi and pi are parameters of the molecule under investigation.
The latter value is related to the change of charge distribution
within the molecule whereas the first two parameters
characterize therein the acidity (ai) and basicity (bi) in the
considered electronic state. Applied to HPTS, specific
interactions like the HBs depicted in Scheme 1 can be probed
in the respective electronic state if reference molecules like
1-hydroxypyrene (pyrenol, PyOH) and the methoxy-derivative
of HPTS, 8-methoxypyrene-1,3,6-trisulfonate (MPTS), are at
disposal.
As
a consequence, HBs directly from solvent to the
hydroxyl-moiety (a1 in Scheme 1) therefore appear less
important or even detrimental for the initial stages of ESPT.
In conclusion, solvatochromism of fluorescent acids in sol-
vents of varying b-values as well as solvatochromism of the
conjugated base in solvents of varying a-values is promising in
search of new ESPT dyes.
In our contribution, we reinvestigate the solvent dependence
of spectra of paradigmatic HPTS and compare it to
solvatochromism of its sulfonato-deficient derivative PyOH
and its methoxy-derivative MPTS. To date, solvent variation
of MPTS and HPTS was mainly performed with polar and
protic solvents, of which the H-bond donating capability
exerted a stronger effect on spectroscopic band position and
shape of the solute than their HB accepting power.16,23,30,31
Multiparameter fitting according to eqn (2) was used to extract
the values of ai and bi.24 However, when the assumed linearity
in the energetics is no longer maintained, such analyses are
misleading. Our strategy separates both contributions: in the
first part, we focus on the dependence of fluorescence
excitation and emission spectra on the Kamlet–Taft parameter
b of the solvent which aims at finding reliable values for
bi. Next, we analyse the influence of solvents on the spectra
of MPTS. The reduced basicity of the photoproduct 8-oxypyrene-
1,3,6-trisulfonate (OPTS) is characterized by the impact of
HB-donating solvents on its spectra. Solubility problems are
overcome by the usage of crown ether. The presented results
are discussed with respect to our ongoing synthetic work of
borondipyrromethene (Bodipy, BDP)-dyes.22,32
Nonspecific dipolar solvation, which contributes to the
Stokes-shift in systems with a considerable change of the static
dipole moment, can be eliminated by comparison of the spectra
of hydroxyl-arenes and their methoxy-derivatives (‘‘differential
solvatochromism’’).6,26,27 Only a few ESPT systems were
investigated so far, but the known examples share a depen-
dence of solvatochromic shifts on the solvents’ b-values. These
shifts are more pronounced in fluorescence than in absorption
and are explained by a stronger HB from the solute to the
solvent in the excited state. This is not unexpected because
proton release occurs along this coordinate.
ESPT can also be investigated from the product side, i.e. by
inspecting solvatochromism of the conjugated base. All
studied examples of naphtholate bases show that the excitation
spectra are more affected by the solvent acidity than the
emission spectra.6 Again, this is reasonable as the basicity of
the ESPT products is reduced compared to the ground state
2. Experimental
Materials and synthesis
HPTS (Aldrich, 97%), MPTS (Fluka, 98%), Pyrenol (Aldrich,
98%) and 18-Crown-6 (Aldrich, 99%) were used without
further purification. The solvents used in this study were of
highest available purity (Spectranal, Chromasolv, Uvasol or
HPLC-grade) and—if not—tested for fluorescent impurities.
All solvents were used as received.
The p-hydroxystyryl- and p-methoxystyryl–Bodipy dyes
(HO-BDP, MeO-BDP; Scheme 2) were synthesized by
Knoevenagel condensation in boiling toluene of methylated,
green fluorescent Bodipy dyes22,32–34 and p-hydroxybenzaldehyde
(Aldrich, 98%) or p-methoxybenzaldehyde (Aldrich, 98%),
respectively.35 The raw product was purified twice by column
Scheme 1 Structure of HPTS, its anionic form OPTS and the
reference dyes MPTS and PyOH. HB between protic solvents (HS)
and basic solvents (S) are probed by varying a- and b-values of the
solvent. The effect of HB at the donor and the sulfonate-groups differs
as indicated by the subscript.
ꢂc
This journal is the Owner Societies 2009
Phys. Chem. Chem. Phys., 2009, 11, 1416–1426 | 1417

View Article Online
illumination showed us whether enough dye was dissolved
for recording reliable fluorescence and excitation spectra. In
the cases where no fluorescence could be detected, 10–25 mg of
crown ether was added in order to improve the solubility of the
multiply charged dye anions by complexation of sodium
counter ions. The final concentration of the crown ether
(o20 mM) resulted in an ionic strength far below those which
were reported to influence the spectra of HPTS.36 Please note
also, that the actual concentration of sodium ions is still
considerably smaller than millimolarity as the only source of
these counter ions was the dye. However, we cross-checked
our procedure with solvents where fluorescence was sufficient
without addition of the helper molecule (Table 1).
Scheme 2 Structures of the synthesized Bodipy dyes. HO-BDP
(R1 HO–) and MeO-BDP (R1 HO–) were carrying
=
=
a
phenyl-group in the meso-position (R2 = phenyl–). Only for the
measurements of the spectra of O-BDP (Fig. 5a) we used a better
soluble and more fluorescent HO-BDP (R1 = HO–. R2 = H–).
Only in a few cases (chloroform (CHCl3), methylene
chloride (CH2Cl2) and bromobenzene (BrB)), saturated
solutions of the crown ether were able to dissolve some
amount of the charged dyes. We were not able to record
fluorescence excitation or emission spectra of HPTS and
MPTS in cyclohexane, toluene, acetophenone, ethyltrichloro-
acetate or of HPTS in BrB.
chromatography to finally yield the desired product. Details
will be described elsewhere. The solubilities of HO-BDP and
MeO-BDP were high enough to record spectra in all solvents;
however, due to the vanishingly small fluorescence quantum
yield of deprotonated HO-BDP (O-BDP), we synthesized for
these measurement HO-BDP lacking the phenyl moiety in the
meso-position.33,34
Sample preparation—spectra of 8-oxypyrene-1,3,6-trisulfonate
(OPTS)
Sample preparation—spectra of acidic forms in apolar solvents
An aliquot of HPTS was taken as described above, and the
organic solvent was added after complete drying. Except from
water, the solutions showed typical blue fluorescence of HPTS.
Therefore, 10–25 mg crown ether and additional, solid NaOH
(B5–10 mg) was added. Greenish fluorescence indicated
successful deprotonation. Spectra of OPTS could be detected
only in a few, highly polar solvents, in which HPTS itself was
already analysed. The solubility in protic solvents was higher;
PyOH, MPTS and HPTS were dissolved in methanol (MeOH)
or ethanol (EtOH), and an aliquot was taken for each sample,
which ensured that the optical density in the final experiments
was low enough in each case to avoid spectral distortion due to
innerfilter effects and emission reabsorption. After complete
evaporation of EtOH, the final solvent was added. Visual
inspection of the fluorescence by means of blacklight
Table 1 Used solvents in this study and their physical (refractive index n0, dielectric constant erel) and solvatochromic parameters (a, b, p*)24,28,43
#
Solvent
a
b
p*
erel
20.56
35.94
25.20
5.40
40.96
4.81
2.02
n0
PyOH
HPTS
MPTS
OPTS
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
Acetone
Acetonitrile
Benzonitrile
Bromobenzene
Butyrolactone
Chloroform
Cyclohexane
Dichloromethane
Dimethyl formamide
Dimethyl sulfoxide
Dioxane
0.08
0.19
0
0
0
0.44
0
0.30
0
0
0
0.83
0
0.90
0.71
1.96
0
0.48
0.31
0.41
0.07
0.49
0
0.71
0.75
0.90
0.71
0.87
0.58
0
0.82
0.88
1.00
0.55
0.54
0.55
0.92
0.97
0.65
0.87
1.359
1.344
1.528
1.557
1.437
1.446
1.426
1.424
1.431
1.479
1.422
1.361
1.372
1.432
1.448
1.275
1.459
—
—
—
+
+
—
+
—
+
+
+
—
+
—
—
—
+
—
—
+
—
+
—
—
—
+
+
+
—
0
+
+
+
—
+
0
—
0
—
—
—
—
—
—
0
+
—
+
—
+
+
0
0
0
—
0
8.93
0.69
0.76
0.37
0.77
0.45
0.52
0.48
0
36.71
46.45
2.21
24.55
6.02
37.7
111.0
16.70
29.6
+
+
+
+
0
+
+
+
+
Ethanol
Ethyl acetate
Ethylene glycol
Formamide
Hexafluoro-2-propanol
Hexamethyl-phosphoric
triamide
—
—
—
+
1.05
0
18
19
20
21
22
23
24
25
Methanol
2-Propanol
Propylene carbonate
Tetrahydrofuran
Tetramethylurea
Toluene
0.93
0.76
0
0
0
0
1.51
1.17
0.62
0.95
0.40
0.55
0.80
0.11
0
0.60
0.48
0.83
0.58
0.83
0.54
0.73
1.09
32.66
19.92
64.92
7.58
23.60
2.38
1.328
1.377
1.422
1.407
1.449
1.497
1.30
—
—
+
+
+
+
—
—
—
—
+
+
+
—
—
—
+
+
+
+
+
—
+
+
+
+
+
—
0
—
+
+
Trifluoroethanol
Water
26.53
78.30
0.47
1.333
Definition of symbols: +: good spectra; 0: noisy spectra or spectra under extreme conditions recorded; —: no spectra obtained; ꢁ: no spectra
recorded.
ꢂc
This journal is the Owner Societies 2009
1418 | Phys. Chem. Chem. Phys., 2009, 11, 1416–1426

View Article Online
only in HFiP, no reliable excitation spectra and a very weak
fluorescence spectrum of OPTS was detectable.
Fluorimetry
Depicted fluorescence excitation and emission spectra of the
blue-green emitting dyes were recorded with a fluorescence
spectrophotometer (Fluoromax-3, Jobin-Yvon and FP 6500,
Jasco) with 1 nm resolution. Afterwards, spectra were
converted to the wavenumber range with the necessary
corrections.37 As far as comparison to reported values is
possible, the evaluated maxima of our study lie within
o100 cmꢁ1 of ref. 24. Slight deviations (o200 cmꢁ1) from
the fluorescence values of ref. 30 are noticed.
3. Results and discussion
3.1 Correlation of spectroscopic maxima and b-values
of the solvents
All solvents used in this section share the property that they
are aprotic, i.e. a = 0. Excitation and emission spectra of
PyOH and HPTS in several solvents with similar polarity, i.e.
similar values of Kamlet–Taft parameter p*, are displayed in
Fig. 1 and 2, respectively. Table 2 summarizes the found
spectroscopic maxima. Both dyes exhibit red-shifts of the
maxima upon increasing b values of the solvents. These
dependencies are displayed in Fig. 1c and 2c. The regression
lines represent single parameter fits according to eqn (2), of
which the slope is bi and the intercept at the ordinate is an
extrapolation to the transition energy ni0 of a non-interacting
solvent. bi is a property of the solute’s electronic state and
describes its tendency to release the proton. The results of
these one-parameter fits are summarized in Table 4.
Solvents with similar p* values are evenly distributed
around the regression lines. PyOH was reported to exhibit
only a small change of its static dipole moment.24 MPTS,
which is taken as analogue to HPTS, does not show a clear
dependence in a Lippert–Ooshika–Mataga plot (data not
shown). This is compatible with a minor change of the static
dipole moment upon excitation, which is predicted by
theoretical calculations.30 It is therefore justified to neglect
the dependency of the energy gap on p* while applying eqn (2)
to solvatochromism of HPTS and PyOH.
Fig. 1 Solvatochromism of PyOH in solvents with varying b-values
(a = 0). Depicted numbers correspond to the solvent numbers in
Table 1. (a) Fluorescence excitation spectra (ldet = 380–420 nm).
(b) Fluorescence emission spectra (lexc = 365–375 nm). (c) Plot of all
spectroscopic maxima versus the solvent’s b-value. The grey lines
represent linear fits to the excitation maxima (open circles) and
emission maxima (filled triangles), respectively. For PyOH,
bexc = ꢁ390 (ꢃ40) cmꢁ1 and bem = ꢁ530 (ꢃ50) cmꢁ1. The data of
bromobenzene (4) were not considered for the fit.
The obtained bi values are close for HPTS and PyOH
(ꢁ320 cmꢁ1 vs. ꢁ390 cmꢁ1 for bexc
;
ꢁ560 cmꢁ1 vs.
ꢁ530 cmꢁ1 for bem). The hydroxyl-moieties are more strongly
interacting with the solvent in the excited than in the ground
state as indicated by |bem| 4 |bexc|. This is in agreement with a
higher acidity in the excited state and resembles the behaviour
of naphthol derivatives.6,26
spectroscopic differences can be traced back to HB differences
among different solvents. It is assumed that other solvent–
solute interactions are cancelled. Taking the differences
For comparison, we also investigated the solvatochromism
of MPTS. The bi-values for MPTS are close to 0 within their
error limits (Fig. 2d). This is expected due to lack of an acidic
proton. The data also verify that other acidic hydrogen atoms
in the core structure are missing.
between the excitation and emission maxima yields b0exc
=
ꢁ360 (ꢃ90) cmꢁ1 and b0em = ꢁ630 (ꢃ170) cmꢁ1, and thus,
similar values as above are obtained within the error limits
(Table 4). Bathochromic shifts in HPTS with increasing
HB-accepting power of the solvent, therefore, reflect solely
the increased HB-strength in the excited state of the solute
compared to its ground state.
Specific and reliable assignment of spectroscopic changes to
an altered local environment can be found out by differential
solvatochromism.6,26 Here, the frequency values ni of HPTS
are subtracted from those of MPTS and, subsequently, the
ꢂc
This journal is the Owner Societies 2009
Phys. Chem. Chem. Phys., 2009, 11, 1416–1426 | 1419

View Article Online
Fig. 2 Solvatochromism of HPTS in solvents with varying b-values (a = 0). Depicted numbers correspond to the solvent numbers in Table 1.
(a) Fluorescence excitation spectra (ldet = 400–450 nm). (b) Fluorescence emission spectra (lexc = 380–390 nm). (c) Plot of all spectroscopic
maxima versus the solvent’s b-value. The grey lines represent linear fits to the excitation maxima (open circles) and emission maxima (filled
triangles), respectively. For HPTS, bexc = ꢁ320 (ꢃ80) cmꢁ1 and bem = ꢁ560 (ꢃ140) cmꢁ1. The data of benzonitrile (3) were not considered for the
fit. Differential solvatochromism verifies these values. (d) For comparison: Plot of the excitation and emission maxima of MPTS vs. the solvent’s
b-value. The slopes are close to 0 within the error margins.
Table 2 Fluorescence excitation and emission maxima of PyOH, HPTS and MPTS in b-solvents (a = 0) (ꢃ30 cmꢁ1
)
Excitation maximum/cmꢁ1
Emission maximum/cmꢁ1
PyOH
HPTS
MPTS
PyOH
HPTS
MPTS
#
Solvent
3
4
5
7
9
10
11
13
17
Benzonitrile
—
24 913
—
25 120
—
25 111
24 951
25 099
25 093
24 890
25 106
—
25 222
—
25 274
25 155
25 218
25 285
25 299
—
24 308
—
24 675
—
24 671
24 373
24 665
24 581
24 271
24 613
—
24 827
—
24 926
24 694
24 848
24 902
24 943
Bromobenzene
Butyrolactone
Cyclohexane
Dimethylformamide
Dimethyl sulfoxide
Dioxane
Ethyl acetate
Hexamethylphosphoric
triamide
26 010
25 963
26 167
25 881
25 793
26 001
26 040
25 751
25 820
25 775
26 071
25 638
25 569
25 834
25 884
25 507
20
21
22
23
Propylene carbonate
Tetrahydrofuran
Tetramethylurea
Toluene
26 016
25 957
25 863
26 060
25 150
25 075
25 066
—
25 182
25 271
25 283
—
25 824
25 783
25 632
25 916
24 722
24 615
24 570
—
24 706
24 888
24 921
—
We conclude that the acidity of HPTS upon excitation is
increasing a-values of the solvent. Previously, HPTS was
examined in a variety of solvents, and multiparameter fitting
of the spectroscopic maxima was performed.24 While bexc
reported in this reference is close to our value, bem deviates
strongly. As redoing the same measurements presumably
would not lead to different results, we performed fluorescence
spectroscopy of MPTS in different solvents (Fig. 3 and
Table 3). MPTS can substitute HPTS due its spectroscopic,
chemical and electronic similarity.16,30 The extrapolated values
rather similar to the increased acidity of PyOH. However, even
a small discrepancy might not explain the differences in the
ESPT behaviour: HPTS easily releases a proton to water
whereas PyOH requires ‘‘catalysis’’ by hydrogenphosphate.31
3.2 Influence of HB-donating solvents on the spectra of MPTS
The missing differences of spectroscopic behaviour between
PyOH and HPTS encouraged us to investigate the influence of
ꢂc
This journal is the Owner Societies 2009
1420 | Phys. Chem. Chem. Phys., 2009, 11, 1416–1426

View Article Online
Table 3 Fluorescence excitation and emission maxima of MPTS and
OPTS in a-solvents (ꢃ30 cmꢁ1
)
Excitation
maximum/cmꢁ1
Emission
maximum/cmꢁ1
MPTS
OPTS
MPTS
OPTS
#
Solvent
1
Acetone
25 370
25 297
25 056
25 076
25 076
24 862
24 788
—
25 035
24 886
24 563
24 592
24 450
23 753
23 474
23 310
24 213
24 691
23 419
22 936
—
2
6
8
Acetonitrile
Chloroform
Dichloromethane
20 619
—
—
21 459
21 645
21 142
—
21 838
21 097
23 095
21 692
19 493
—
—
12 Ethanol
14 Ethylene glycol
15 Formamide
19 608
19 531
19 268
19 763
19 646
19 569
19 685
19 569
16 Hexafluoro-2-propanol 24 629
18 Methanol
19 2-Propanol
24 Trifluoroethanol
25 Water
25 049
25 136
24 779
24 797
emission. This behaviour is different to what has been
described for naphthol-derivatives and PyOH.24,26 A blue-shift
of the electronic spectra, which corresponds to positive ai
values, is explained by a HB between the oxygen atoms of
the aromatic alcohols and the solvent (e.g. a1 in Scheme 1)
which is weakened upon excitation. In terms of the molecular
electronics of the probe molecule, electron density is
transferred from the donor to the aromatic core upon excitation.
A plot of the spectroscopic maxima against a can return aexc
and aem of MPTS after fitting, which is a measure of the
HB-accepting character of the respective electronic state
(Fig. 3c). The extracted value for aexc, ꢁ350 (ꢃ70) cmꢁ1, is
of the same order as the value given by ref. 24 for HPTS. A
negative value (in HPTS) was traced back to the increasing
electron withdrawing power of sulfonate groups in conjunction
with the donor ability of a hydroxyl-group. The sulfonamide
derivative of HPTS, in which the electron-withdrawing
strength of the substituents is further enhanced, also exhibits
red-shifted spectra compared to HPTS.13,15 The same
argumentation also holds for MPTS. It is worth mentioning
that the observed red-shift is reduced most likely as a result of
the blue-shift due to HB directly at the methoxy-position.
The influence of a on the fluorescence maxima is more
complex. While at low HB-donating ability, the fluorescence
is (more or less linearly) bathochromically shifted by a higher
solvent acidity. The observed Stokes-shifts (nexc ꢁ nem) are in
the range of the extrapolated values of HPTS and MPTS in the
b-dependence plot (Fig. 2c and d; nexc,0 ꢁ nem,0). In contrast,
the fluorescence emission maxima are nearly constant at high
a values. A change of behaviour appears around an a-value of
Fig. 3 Solvatochromism of MPTS in solvents with varying a-values.
Depicted numbers correspond to the solvent numbers in Table 1.
(a) Fluorescence excitation spectra (ldet = 450 nm). (b) Fluorescence
emission spectra (lexc = 380–390 nm). (c) Plot of all fitted spectro-
scopic maxima versus the solvent’s a-value. The grey line represents a
linear fit to the excitation maxima (open circles) and yields for MPTS,
a
exc = ꢁ350 (ꢃ70) cmꢁ1. The value of aem was not determined by a fit
to the emission maxima (filled triangles), see text for further details.
of ni0 in Fig. 2c and d coincide within the error limits (Table 4)
and, thus, verify this similarity. MPTS, however, is convenient
as it does not exhibit a spectroscopic dependence on the
HB-accepting power of the solvent. As shown in section 3.1,
pi and bi of MPTS are approximately 0. Hence, solvatochromism
can be analysed on the basis of varying a-values independent
of the solvent’s b-value and multiparameter fits can be
avoided. This is helpful in screening applications where
elaborate computations should be avoided. Fig. 3a shows
the excitation and Fig. 3b the emission spectra of MPTS in
several solvents with differing a-values. Increasing a-values are
accompanied by bathochromic shifts both in excitation and
0.7–0.8. Both sections are separated by a step of B1000 cmꢁ1
.
Also the line shape changes.38 In HPTS under highly acidic
conditions, this change of behaviour was attributed to charge
redistribution, but was explicitly excluded for MPTS.14 We,
however, argue by means of a rough estimate that, indeed,
MPTS and HPTS are identical in the occurrence of the charge
transfer state: nem of MPTS in water is B22 900 cmꢁ1 which is
larger by B300 cmꢁ1 than the value of nem of HPTS.24 If we
add the bathochromic shift due to HB between the base H2O
and the photoacid HPTS, then we further reduced the
transition energy by bem*b (the b-value of H2O is B0.5) which
is in the range of the missing 300 cmꢁ1. The mentioned change
ꢂc
This journal is the Owner Societies 2009
Phys. Chem. Chem. Phys., 2009, 11, 1416–1426 | 1421

View Article Online
Table 4 Fitting results of solvatochromism experiments. All data are obtained using linear fits according to eqn (2). R denotes the correlation
coefficient
Spectra: solvent dependence
Fig.
Ordinate intercept/cmꢁ1
Slope ai, bi/cmꢁ1
R
Exc.
PyOH: b
HPTS: b
MPTS: b
Diff. solvato-chrom.
HPTS
HO-BDP
MeO-BDP
Diff. solvato-chrom.
HO-BDP
1a,c
2a,c
2d
n
n
n
exc0: 26 150 (ꢃ20)
exc0: 25 260 (ꢃ50)
exc0: 25 190 (ꢃ50)
bexc: ꢁ390 (ꢃ40)
bexc: ꢁ320 (ꢃ80)
bexc: +90 (ꢃ80)
b0exc: ꢁ360 (ꢃ90)
0.96
0.84
0.38
0.82
—
30 (ꢃ60)
5b,d
5e
nexc0: 17 635 (ꢃ40)
exc0: 17 620 (ꢃ50)
bexc: ꢁ350 (ꢃ70)
bexc: ꢁ160 (ꢃ80)
b0exc: ꢁ190 (ꢃ30)
0.87
0.58
0.91
n
10 (ꢃ20)
MPTS: a
OPTS: a
O-BDP: a
PyOH: b
HPTS: b
MPTS: b
Diff. solvato-chrom.
HPTS
HO-BDP
MeO-BDP
Diff. solvato-chrom.
HO-BDP
3a
4a
5a
1b,c
2b,c
2d
n
n
n
n
n
n
exc0: 25 280 (ꢃ60)
exc0: 20 830 (ꢃ160)
exc0: 17 190 (ꢃ80)
em0: 26 040 (ꢃ30)
em0: 24 920 (ꢃ90)
em0: 24 760 (ꢃ100)
aexc: ꢁ350 (ꢃ70)
aexc: +780 (ꢃ210)
aexc: +190 (ꢃ80)
bem: ꢁ530 (ꢃ50)
bem: ꢁ560 (ꢃ140)
bem: +150 (ꢃ150)
b0em: ꢁ630 (ꢃ170)
0.86
0.75
0.74
0.96
0.83
0.36
0.80
Em.
—
100 (ꢃ110)
5c,d
5f
—
nem0: 17 340 (ꢃ50)
em0: 17 320 (ꢃ50)
bem: ꢁ525 (ꢃ90)
bem: ꢁ310 (ꢃ80)
b0em: ꢁ220 (ꢃ40)
0.91
0.81
0.87
n
10 (ꢃ30)
MPTS: a
OPTS: a
O-BDP: a
3b
5
5a
Not determined
Not determined
aem: +240 (ꢃ60)
Not determined
—
0.75
—
nem0: 19330 (ꢃ50)
Not determined
of spectroscopic behaviour, although not further quantified,
has impact on previous solvatochromic investigations: it
indicates that multiparameter fitting procedures are misleading
when a linear dependence is not given. This might especially be
important in the investigation of chemical reactions.
position, and might be the reason for the low correlation value
of aem and aexc (R = 0.75) in our one-parameter correlation.
However, a correlation with polarity (p* or the dielectric
constant erel) was not found. In contrast to the excited state,
the ground state is more sensitive to HB as expected for a base
(aexc = +780 (ꢃ210) cmꢁ1). The acidity of HFiP (pKA = 9.3),
which shifts the equilibrium to HPTS, might therefore be the
reason, why only rather poor spectra of OPTS could be
obtained in this solvent.
3.3 Influence of HB-donating solvents on the spectra of OPTS
The deprotonated OPTS is the photoproduct after ESPT of
HPTS. The idea behind the investigation of its solvatochromism
in a-solvents is that we now explore ESPT dyes from the
photoproduct side. We would like to find out whether this
approach is appropriate to find target structures. First of all,
the response of the OPTS-spectra to stronger HB from the
solvent allows verification of whether the basicity of the
excited state is reduced compared to the ground state. For
very strong photoacids, |aem| o |aexc| was found in agreement
with a higher excited state acidity and opposite to the effect of
HB-donating solvents at the reactant side. The values are
positive indicating spectral blueshifts with increasing solvent
acidity (Fig. 4a). For the search of ESPT dyes, the aem values,
which are specific for the properties of the excited state, are
more important.
It is interesting to compare the effect of solvents with high
a-values on the emission spectra of MPTS, which is similar to
HPTS shortly before ESPT, and OPTS, which is HPTS after
the ESPT. Although the HB at the electron-donor moiety
(methoxy group vs. oxy-group) is different in both species, the
HB at the sulfonate-groups is nearly identical. aem of MPTS,
although not further determined (Fig. 3c), and aem of OPTS
are in a similar range. This comparison shows that, once
charge redistribution occurred at a a-value of B0.8, excited
HPTS becomes more acidic than before and only then the
proton can be transferred to the solvent without the need of
catalysts. This is also reflected by the equilibrium constant of
HPTS + H2O/OPTS + H3O+ which is shifted to the neutral
chromophore state on a time scale before charge redistribution
occurs.18 It is not possible to discuss whether 1La/1Lb state
reversal or not occurs along the ESPT-coordinate on the basis
of our experiments.16,24,38 Nevertheless, we adhere to the
mentioned acidity change in the excited state of HPTS as the
difference to PyOH.
Fluorescence emission spectra of OPTS in different solvents
are depicted in Fig. 4b, and the spectral maxima positions
versus the solvents’ a-value are shown in Fig. 4c. The spectra in
highly protic solvents are broader than in aprotic solvents;
however the centre is hardly affected by the a-value: aem
,
obtained as +240 (ꢃ60) cmꢁ1 from the linear fit in Fig. 4c,
indeed is comparably small with respect to other deprotonated
ESPT dyes.6 The peak positions of fluorescence in purely
aprotic solvents (a = 0) cover a spectral range which is as
large as the whole energetic variation due to HB. Thus, these
non-HB interactions are equally important for the spectral
3.4 Discussion of solvent effects with respect to HO-BDP
According to the revisited solvatochromism of HPTS, the
wealth of available dyes can be narrowed down during the
search of ESPT-dyes by the investigation of solvatochromism.
ꢂc
This journal is the Owner Societies 2009
1422 | Phys. Chem. Chem. Phys., 2009, 11, 1416–1426

View Article Online
fluorescein dyes in terms of photostability and are therefore
appropriate for our long-term goal of ESPT-dyes in the visible
range.22,32–34 The dye in the focus of this section has in its
neutral form lexc = 569 nm, lem = 581 nm and in its anionic
form lexc = 573 nm, lem B 710 nm (Fig. 5a). According to
eqn (1), one calculates DpKa = 0.25 on the basis of the
excitation data and DpKa = 6.55 on the basis of the emission
data. Even if we take the average, as suggested by ref. 5, then a
distinct increase of the acidity by more than 3 orders of
magnitude could be expected.
Other dyes with styryl-moieties were shown to exhibit
ESPT.7 However, we were led by the spectroscopic similarity
of ammonium-groups and hydroxyl-groups and their depro-
tonated counterparts, i.e. amino-groups and oxy-substituents.
The comparability is e.g. manifested in substituted xanthene
dyes, i.e. rhodamine vs. fluoresceine dyes. With respect to
ESPT, the similarity was already noticed by Forster1,2 and,
¨
quite recently, by time-resolved spectroscopy.14 Differences
+
between isoelectronic NH3 and OH-groups concern the
pKa values of the considered dyes.
The synthesis and spectroscopic properties of p-dimethyl-
aminostyryl Bodipy dyes (NMe2-BDP) were published in the
last years,35,39 and recently, we performed theoretical
calculations to elucidate its ESPT-behaviour.40 Whereas the
protonated dye exhibits only small Stokes-shifts upon
excitation, the neutral chromophore undergoes charge-
transfer with a large change of the static dipole moment.
The reported spectroscopic values of excitation and emission
maxima for NMe2-BDP and its conjugated acid
NMe2H+-BDP are close to our values of the pair
HO-BDP/O-BDP. We therefore assume that all conclusions
drawn for NMe2-BDP/NMe2H+-BDP pair also hold for
HO-BDP/O-BDP although we did not perform such elaborate
solvatochromic studies as in ref. 39. We summarize that if
ESPT occurs then charge-transfer as the thermodynamic
driving force would act only after ESPT.
We studied the solvatochromism of HO-BDP in b-solvents
(Fig. 5b–d). We renounced the aromatic solvents benzonitrile
(3), bromobenzene (4) and toluene (23) which showed systematic
deviations in the spectra of PyOH and HPTS. We assign this to
putative p–p interactions between the solvent and the aromatic
solute. The bi-values, bexc and bem, are close to the values of
PyOH and HPTS (Table 4). However, the uncritical
interpretation as an increased acidity in the excited state like
in the case of PyOH and HPTS fails: the b-dependence of the
spectra of the methoxy-derivative MeO-BDP as well as
differential solvatochromism provide evidence that the acidity
of the OH-group is not substantially raised and that some other
slightly acidic proton exists within the molecule (Table 4). Very
acidic solvents like hexafluoro-2-propanol shift both excitation
and emission spectra of MeO-BDP in equal measure to the blue
(Fig. 5e and f). The blue-shift is in agreement with a reduction of
the donor properties of the methoxy-moiety; the likewise
blue-shift in the ground state as well as in the excited state
imply, however, that aexc B aem and that no additional electron
density is shifted from the electron donor toward the aromatic
system in the excited state.
Fig. 4 Solvatochromism of OPTS in solvents with varying a-values.
Depicted numbers correspond to the solvent numbers in Table 1.
(a) Fluorescence excitation spectra (ldet = 530 nm). (b) Fluorescence
emission spectra (lexc = 420–460 nm). (c) Plot of all fitted spectro-
scopic maxima versus the solvent’s a-value. The grey lines represent
linear fits to the excitation maxima (open circles) and emission maxima
(filled triangles), respectively. For OPTS, aexc = +780 (ꢃ210) cmꢁ1
and aem = +240 (ꢃ60) cmꢁ1. For the fit of the excitation maxima, not
the maximum in trifluoroethanol (24), but the shoulder at l = 452 nm
(see a) was used. An excitation maximum in hexafluoro-2-propanol
(16) was not found.
In this last section, we would like to discuss the example where
an acidity increase of an aromatic alcohol is not observed
although fluorescence emission data misled us to assume its
ESPT-capability. This is the case of HO-BDP (Scheme 2).
We applied the solvatochromic method to HO-BDP in
order to find out whether this compound might be a target
structure for an ESPT-dye in the visible range. Bodipy dyes
exhibit high fluorescence quantum yields, are superior to
We also performed excitation and emission spectra of
O-BDP, the conjugated base of HO-BDP, in dependence of
ꢂc
This journal is the Owner Societies 2009
Phys. Chem. Chem. Phys., 2009, 11, 1416–1426 | 1423

View Article Online
Fig. 5 Solvatochromism of Bodipy dyes. Depicted numbers correspond to the solvent numbers in Table 1. (a) Fluorescence excitation (full lines)
of HO-BDP (grey, ldet = 590 nm) and its conjugated base O-BDP (black, ldet = 710 nm) and fluorescence emission spectrum of O-BDP
(black-dotted, lexc = 585 nm) in water. The large Stokes-shift of the anionic form was considered as an indication of ESPT. Note also the very
uncommon line shape of the excitation spectrum of O-BDP. (b) Fluorescence excitation spectra (ldet = 600–635 nm) of HO-BDP in solvent with
varying b-values. (c) Fluorescence emission spectra (lexc = 520–550 nm) of HO-BDP in solvent with varying b-values. (d) Plot of all spectroscopic
maxima of HO-BDP versus the solvent’s b-value. The grey lines represent linear fits to the excitation maxima (open circles) and emission maxima
(filled triangles), respectively. For HO-BDP, bexc = ꢁ350 (ꢃ70) cmꢁ1 and bem = ꢁ525 (ꢃ90) cmꢁ1. However, differential solvatochromism yields
b0exc = ꢁ190 (ꢃ30) cmꢁ1 and b0em = ꢁ220 (ꢃ40) cmꢁ1. (e) Fluorescence excitation spectra (ldet = 600–615 nm) of MeO-BDP in different solvents.
Note the spectral shifts of MeO-BDP in highly basic solvent (17). (f) Fluorescence emission spectra (lexc = 525–550 nm) of MeO-BDP in solvent
with varying b-values. Note the spectral shifts of MeO-BDP in highly basic solvent (17).
the a-value of some protic solvents (solvents 12, 14, 15, 18, 19,
24, 25; Fig. 5a). They share as common property a similar
Lippert solvation parameters Df B 0.3
contribution of unspecific dipolar solvation is cancelled. A
slight blue-shift of the excitation spectra was found for
O-BDP, aexc = +190 cmꢁ1, which is unexpectedly small for
a base on the basis of naphtholate or OPTS data.6 One has to
consider the noticed solvatochromism of MeO-BDP in the
sense that here also the b-values of the solvents have an impact
on the spectral position. For aem, however, we could not find a
clear dependence. All fluorescence maxima lie within a range
erel ꢁ 1
n2 ꢁ 1
Df ¼
ꢁ
ð3Þ
2erel þ 1 2n2 þ 1
where erel is the dimensionless dielectric constant and n the
refractive index of the solvent. This should ensure that the
ꢂc
This journal is the Owner Societies 2009
1424 | Phys. Chem. Chem. Phys., 2009, 11, 1416–1426

View Article Online
of 705 to 710 nm. The fact that all recorded emission data are
broad in combination with very weak fluorescence aggravates
here the analysis and a proper assignment of the emission
maxima is hardly achievable. Moreover, it is questionable
whether one-parameter fitting can fully describe the
solvatochromism of a charge-transfer dye, even if solvents
with similar Df-values but differing a-values are used.
Nevertheless, our data are in good agreement with the
behaviour of NMe2-BDP where vanishingly small aexc and
aem were found by multiparameter fitting.39
(butyrolactone, dimethylformamide, hexamethylphosphoric
triamide, propylene carbonate and tetramethylurea) covers a
large range of b-values and should already provide an insight
into the ESPT behaviour. Finally, we do not recommend
judging the ESPT tendency on the basis of aem since dipolar
relaxation might mask subtle effects of the solvent acidity.
Acknowledgements
The authors thank M. Wild for assistance in the synthetic work
and I. Bernhardt for generous support. The work was
supported by the German Science Foundation (DFG JU650/1).
In summary, the situations for HPTS (at early times), PyOH
and HO-BDP are not so different: whenever a suitable proton
acceptor is so close that proton transfer can occur, this will
happen in the excited state.31,41 For example, ultrafast ESPT
to acetate was observed for HPTS on a time scale before
charge redistributation occurs.36 However, differences among
the investigated dyes concern the time scales of proton transfer.
The rate constants for proton transfer are related to the free
energy release DG during the reaction.42 This is also exemplified
in the protonation kinetics of pyrene-1-carboxylate by various
acids.20 While the acidity increase of HPTS (at early times)
and PyOH, which is probed by the variation of the solvent
b-value, consequently enables proton transfer during the
excited state lifetime, ESPT is rather unlikely to proceed in
HO-BDP. If, however, ESPT accidentally occurs in HO-BDP,
then the energy release due to solvation in the anionic state
would lock the thermodynamics.
References
1 Th. Forster, Naturwissenschaften, 1949, 36, 186.
¨
2 Th. Forster, Z. Elektrochem., 1950, 54, 42.
¨
3 O. Mohammed, D. Pines, J. Dreyer, E. Pines and E. Nibbering,
Science, 2005, 310, 83.
4 P. Wan and D. Shukla, Chem. Rev., 1993, 93, 571.
5 L. Tolbert and K. Solntsev, Acc. Chem. Res., 2003, 35, 19.
6 N. Agmon, J. Phys. Chem. A, 2005, 109, 13.
7 F. Lewis, L. Sinks, W. Weigel, M. Sajimon and E. Crompton,
J. Phys. Chem. A, 2005, 109, 2443.
8 B. Cohen and D. Huppert, J. Phys. Chem. A, 2001, 105, 7157.
9 A. Orte, L. Crovetto, E. Talavera, N. Boens and J. Alvarez-Pez,
J. Phys. Chem. A, 2005, 109, 734.
10 M. Chattoraj, B. King, G. Bublitz and S. Boxer, Proc. Natl. Acad.
Sci. U. S. A., 1996, 93, 8362.
11 H. Lossau, A. Kummer, R. Heinecke, F. Pollinger-Dammer,
¨
We therefore conclude that the spectroscopic signature
which points to the ESPT propensity is likely the difference
of the excitation maxima of acid and conjugated base, i.e. the
transition energy difference Dnexc0 before solvent relaxation
occurs. Dnexc0 be regarded as a measure for the gradient of
ESPT although subtleties of the potential energy surface are
ignored. In the case of HO-BDP, Dnexc0 is 100–150 cmꢁ1 which
is in contrast to a roughly ten-times-larger value of PyOH.31
C. Kompa, G. Bieser, T. Jonsson, C. Silva, M. Yang,
D. Youvan and M. Michel-Beyerle, Chem. Phys., 1996, 213, 1.
12 B. Cohen, J. Segal and D. Huppert, J. Phys. Chem. A, 2002, 106,
7462.
13 D. Spry and M. Fayer, J. Chem. Phys., 2007, 127, 204501.
14 D. Spry and M. Fayer, J. Chem. Phys., 2008, 128, 084508.
15 E. Pines, D. Pines, Y.-Z. Ma and G. Fleming, ChemPhysChem,
2004, 5, 1315.
16 O. Mohammed, J. Dreyer, B. Magnes, E. Pines and E. Nibbering,
ChemPhysChem, 2005, 6, 625.
17 T. Tran-Thi, T. Gustavsson, C. Prayer, S. Pommeret and J. Hynes,
Chem. Phys. Lett., 2000, 329, 421.
18 P. Leiderman, L. Genosar and D. Huppert, J. Phys. Chem. A,
2005, 109, 5965.
4. Conclusion
In our contribution, we investigated the solvatochromism of
several dyes with the aim to predict the ESPT behaviour of
freshly synthesized dyes. The dyes HPTS, PyOH and HO-BDP
are all aromatic alcohols but decreasingly tend to undergo
ESPT. The strong photoacidity of HPTS presumably results
from charge redistribution in succession of solvent rearrangement,
i.e. the formation of strong electron withdrawing substituents
by HB from solvent to HPTS.14,24 PyOH, HO-BDP and many
naphtholes have in common that acidic solvents do not
accelerate ESPT. We therefore propose the following criteria
to distinguish potential ESPT dyes like PyOH from slower
reacting dyes like HO-BDP. Firstly, differences in Dnexc0 are
related to the thermodynamics and should hint at the overall
ESPT driving force. The best procedure would be to compare
the spectra in a solvent of low polarity; solubility problems can
be overcome by use of crown ether. Secondly, spectra of the
acid in solvents with varying b-values should explore their
kinetic tendency of ESPT. Differential solvatochromism
ensures to assign observed effects to a certain proton. In the
case where methoxy-derivatives are not available, solvents
of similar polarity are preferential. A set of 5 solvents
19 K. Solntsev, D. Huppert and N. Agmon, Phys. Rev. Lett., 2001, 86,
3427.
20 B. Zelent, J. Vanderkooi, R. Coleman, I. Gryczynski and
Z. Gryczynski, Biophys. J., 2006, 91, 3864.
21 A. Kotlyar, N. Borovok, S. Raviv, L. Zimanyi and M. Gutman,
Photochem. Photobiol., 1996, 63, 448.
22 B. Hinkeldey, A. Schmitt and G. Jung, ChemPhysChem, 2008, 9,
2019.
23 N. Barrash-Shiftan, B. Brauer and E. Pines, J. Phys. Org. Chem.,
1998, 11, 743.
´
24 T.-H. Tran-Thi, C. Prayer, Ph. Millie, P. Uznanski and J. Hynes,
J. Phys. Chem. A, 2002, 106, 2244.
25 K. Solnstev, D. Hupert, L. Tolbert and N. Agmon, J. Am. Chem.
Soc., 1998, 120, 7981.
26 K. Solntsev, D. Huppert and N. Agmon, J. Phys. Chem. A, 1998,
102, 9599.
27 K. Solntsev, D. Huppert and N. Agmon, J. Phys. Chem. A, 1999,
103, 6984.
28 M. Kamlet, J. Abboud, M. Abraham and R. Taft, J. Org. Chem.,
1983, 48, 2877.
29 C. Reichardt, Solvent and Solvent Effects in Organic Chemistry,
VCh, Weinheim, 2nd edn, 1988.
30 R. Jimenez, D. Case and F. Romesberg, J. Phys. Chem. B, 2002,
106, 1090.
31 B. Milosavljevic and J. Thomas, Photochem. Photobiol. Sci., 2002,
1, 100.
ꢂc
This journal is the Owner Societies 2009
Phys. Chem. Chem. Phys., 2009, 11, 1416–1426 | 1425

View Article Online
32 A. Schmitt, B. Hinkeldey, M. Wild and G. Jung, J. Fluoresc., DOI:
10.1007/s10895-008-0446-7.
38 D. Spry, A. Goun, C. Bell and M. Fayer, J. Chem. Phys., 2006,
125, 144514.
39 M. Baruah, W. Qin, C. Flors, J. Hofkens, R. Vallee, D. Beljonne,
´
M. Van der Auweraer, W. Borggraeve and N. Boens, J. Phys.
Chem. A, 2006, 110, 5998.
40 O. Clemens, M. Basters, M. Wild, S. Wilbrand, C. Reichert,
M. Bauer, M. Springborg and G. Jung, J. Mol. Struct.
(THEOCHEM), 2008, 866, 15.
33 G. Ulrich, R. Ziessel and A. Harriman, Angew. Chem., 2008, 120,
1202 (Angew. Chem., Int. Ed., 2008, 47, 1184).
34 A. Loudet and K. Burgess, Chem. Rev., 2007, 107, 4891.
35 K. Rurack, M. Kollmannsberger and J. Daub, Angew. Chem.,
2001, 113, 396.
36 L. Genosar, D. Cohen and D. Huppert, J. Phys. Chem. A, 2000,
104, 6689.
37 J. Lakowicz, Principles of Fluorescence Spectroscopy, 2nd edn,
2000.
41 T. Thun, J. Fluoresc., 2003, 13, 323.
42 E. Pines and G. Fleming, J. Phys. Chem., 1991, 95, 10448.
43 I. Renge, J. Phys. Chem. A, 2000, 104, 7452.
ꢂc
This journal is the Owner Societies 2009
1426 | Phys. Chem. Chem. Phys., 2009, 11, 1416–1426