Ultrafast charge transfer in amino-substituted boron dipyrromethene dyes and its inhibition by cation complexation: A new design concept for highly sensitive fluorescent probes

Article abstract of DOI:10.1021/jp982701c

The photophysical behavior of a newly synthesized aza crown-substituted boron-dipyrromethene (BDP) dye and its dimethylamino analogue were investigated with steady-state and time-resolved fluorometry and compared to a reference compound. In solvents more polar than hexane, excitation of the dyes leads to a fast charge transfer from the locally excited (LE) state to a weakly emissive charge-transfer (CT) state. The donor-substituted compounds show dual emission from the LE and CT state, both fluorescence quantum yields being low. The rate constant of excited-state charge separation, calculated from the global analysis of time-resolved emission data, was determined to 1.6 × 10¹¹ s^-1 in 1,4-dioxane. The crowned compound forms 1:1-complexes with various alkali and alkaline-earth metal ions, which exist as two conformers in solution. In these complexes, coordination of the cation to the nitrogen donor atom of the crown inhibits the charge-transfer process, leading to a cation-dependent enhancement of the LE emission and the fluorescence lifetimes by factors > 10³. This efficient "switching on" of the fluorescence renders the crowned BDP dye an extremely sensitive fluorescent probe for metal ions.

Full text of DOI:10.1021/jp982701c

J. Phys. Chem. A 1998, 102, 10211-10220
10211
Ultrafast Charge Transfer in Amino-Substituted Boron Dipyrromethene Dyes and Its
Inhibition by Cation Complexation: A New Design Concept for Highly Sensitive
Fluorescent Probes
†
,‡
§
,†
Matthias Kollmannsberger, Knut Rurack,* Ute Resch-Genger, and J o1 rg Daub*
Institut f u¨ r Organische Chemie der UniVersit a¨ t Regensburg, D-93040 Regensburg, Germany, Institut f u¨ r
Physikalische und Theoretische Chemie, Humboldt UniVersit a¨ t zu Berlin, Bunsenstrasse 1,
D-10117 Berlin, Germany, and Bundesanstalt f u¨ r Materialforschung und -pr u¨ fung (BAM),
Rudower Chaussee 5, D-12489 Berlin, Germany
ReceiVed: June 22, 1998; In Final Form: September 23, 1998
The photophysical behavior of a newly synthesized aza crown-substituted boron-dipyrromethene (BDP) dye
and its dimethylamino analogue were investigated with steady-state and time-resolved fluorometry and
compared to a reference compound. In solvents more polar than hexane, excitation of the dyes leads to a fast
charge transfer from the locally excited (LE) state to a weakly emissive charge-transfer (CT) state. The
donor-substituted compounds show dual emission from the LE and CT state, both fluorescence quantum
yields being low. The rate constant of excited-state charge separation, calculated from the global analysis of
1
1
-1
time-resolved emission data, was determined to 1.6 × 10
s
in 1,4-dioxane. The crowned compound
forms 1:1-complexes with various alkali and alkaline-earth metal ions, which exist as two conformers in
solution. In these complexes, coordination of the cation to the nitrogen donor atom of the crown inhibits the
charge-transfer process, leading to a cation-dependent enhancement of the LE emission and the fluorescence
3
lifetimes by factors >10 . This efficient “switching on” of the fluorescence renders the crowned BDP dye
an extremely sensitive fluorescent probe for metal ions.
Introduction
cence intensity and lifetime but provides additional information
via spectral shifts in both absorption and emission. Whereas
The use of fluorescent sensor molecules for the detection of
metal ions, organic, and biological analytes1,2,3 has focused much
the latter effect is advantageous in terms of spectral discrimina-
tion of analytically relevant species, the fluorescence of these
intrinsic probes is either quenched by cation complexation or
the fluorescence enhancement factors (FEF) are comparatively
small (factors of ca. 2-5). Systems showing large complex-
ation-induced FEF, which are highly desirable in terms of low
detection limits and high signal-to-noise ratios, are rare, and
FEF between 30 and ca. 500 have been so far only reported for
research interest on the synthesis as well as the understanding
of the photophysical mechanisms governing the spectroscopic
behavior of such systems.4
-12
These fluorescent probes are
either constructed as fluorophore-spacer-receptor or as in-
trinsic fluorescent probes.13 In the former case fluorophore and
receptor are separated by a short alkyl spacer, whereas in the
latter case the receptor is part of the chromophore π-electron
system. Here, the receptor usually acts as an electron-donating
and the basic fluorophore as an electron-accepting moiety. For
both types of fluorescent probes, nitrogen-containing crown
ethers are often used as receptors, especially for metal ion
recognition.
The optical properties of the intrinsically donor-acceptor-
substituted fluoroionophores (ICT probes) are determined by a
strong solvent-dependent behavior, and the fluorescence ob-
served under many analytic conditions, i.e., in polar liquid
media, is largely Stokes-shifted. In fluorophore-spacer-
receptor systems, electronic interaction is only possible via long-
range forces such as photoinduced electron transfer (PET).
Binding of the analyte to the receptor decreases its electron-
donating ability and thus leads to the inhibition of the quenching
process(es) resulting in the “switching on/off” of the fluores-
cence; however, no spectral changes occur. In contrast,
coordination to the nitrogen donor atom of the receptor of an
intrinsic fluoroionophore not only leads to changes in fluores-
6
,8,13
some fluorophore-spacer-receptor PET systems.
During the last years, molecules consisting of aromatic amine
donor groups directly and covalently bound to fluorophores such
1
4-20
20,21
22
as anthracene,
pyrene,
phenanthrene, or acridine/
5,23
ium have been studied extensively to elucidate the mechanism
of the charge-transfer (CT) process that takes place upon
excitation of the fluorophore. In polar solvents, emission occurs
no longer from the locally excited (LE) state of the fluorophore
but from a highly polar CT state that is either directly accessible
via absorption or is formed in the picosecond time range after
excitation. However, the exact mechanism of this process is
still not fully understood, and different intramolecular reaction
pathways during the lifetime of the excited state (e.g., the twisted
24,25
intramolecular charge-transfer “TICT” model ) are discussed
for these molecules. In some systems, charge transfer can be
sufficiently described with a simple three-state model (e.g.,
22
p-phenanthryl dimethylaniline ). However, a nonexponential
fluorescence decay behavior is observed for others (e.g.,
1
5-17,20
p-anthracenyl-dimethylaniline, ADMA
) suggesting a
†
Institut f u¨ r Organische Chemie.
Institut f u¨ r Physikalische und Theoretische Chemie.
Bundesanstalt f u¨ r Materialforschung und -pr u¨ fung.
‡
more complex situation involving different transient species.
In contrast, in H-acidic solvents, protonation of the amino group
§
1
0.1021/jp982701c CCC: $15.00 © 1998 American Chemical Society
Published on Web 11/21/1998

10212 J. Phys. Chem. A, Vol. 102, No. 50, 1998
Kollmannsberger et al.
blocks the charge-transfer process and reestablishes emission
from the locally excited state as the major deactivation
route.
Recently, we have shown that the use of difluorobordiaza-
s-indacenes (boron-dipyrromethene dyes, BDPs) as acceptor
moieties and fluorophores in such systems yields highly sensitive
Steady-State Spectroscopy. UV/vis spectra were obtained
on a Perkin-Elmer Lambda9 UV/vis/NIR spectrophotometer and
a Specord M400/M500 absorption spectrometer from Carl Zeiss.
Steady-state emission spectra were recorded on a Hitachi F-4500
and a Spectronics Instruments 8100 spectrofluorometer. For
the fluorescence experiments, only dilute solutions with an
optical density (o.d.) below 0.01 at the excitation wavelength
(o.d. < 0.04 at the absorption maximum) were used. For the
determination of the relative fluorescence quantum yields (φf),
the optical densities of the solutions at the excitation wave-
lengths were adjusted to an optical density of 0.1 ( 0.001 in a
100 mm absorption cell. These solutions were then transferred
to a 10 mm quartz cell, and the fluorescence measurements were
performed with a 90° standard geometry and an emission
polarizer set at 54.7°. Fluorescein 27 in 0.1 N NaOH (φf 0.90
( 0.03) and DCM in methanol (φf 0.43 ( 0.08) were used as
1
3,26,27
2
8-30
fluorescent probes.
BDP dyes combine the advantages of
high molar extinction coefficients (ꢀ > 70 000 M- cm ) and
high fluorescence quantum yields (φf ca. 0.5-0.8) and can be
excited at relatively long wavelengths (ca. 500 nm). They are
currently in use as laser dyes and fluorescent labels for
1
-1
3
1
biomolecules and have been incorporated in an electron-
32
transfer probe of radical ion pair generated electric fields. The
spectroscopically advantageous properties and the high electron
affinity of the BDP fluorophore which promises fast and efficient
charge transfer, led us to synthesize and investigate the crown
ether analogue 1 of the dimethylanilino-substituted BDP dye 2
under the premise of developing new and highly sensitive
fluorescent probes for metal ions. Here, for a better understand-
ing of the operative fluorescence mechanisms, the photophysical
properties of 1, 2, and the reference compound 3 are investigated
by steady-state absorption and emission spectroscopy as well
as time-resolved fluorometry in different solvents. Furthermore,
the complexation of 1 with various alkali and alkaline-earth
metal ions is spectroscopically studied and discussed.
34
fluorescence standards. All the fluorescence spectra presented
here are corrected for the spectral response of the detection
system (calibrated quartz halogen lamp placed inside an
integrating sphere; Gigahertz-Optik) and for the spectral irra-
diance of the excitation channel (calibrated silicon diode
mounted at a sphere port; Gigahertz-Optik). The fluorescence
quantum yields were calculated from multiple measurements
(N ) 6), and the uncertainties of the measurement were
determined to (5% (for 1.0 > φ > 0.2), (10% (for 0.2 > φ
f
f
-
3
>
0.02), (20% (for 0.02 > φf > 5 × 10 ), and (30% (for 5
-3
×
10 > φf), respectively.
Time-Resolved Spectroscopy. Fluorescence lifetime (τf)
measurements were performed with a unique laser impulse
fluorometer with picosecond time resolution described else-
3
5
where. The sample was excited with the second harmonic
output (LBO crystal) of a regenerative mode-locked argon ion
laser-pumped Ti:sapphire laser at a repetition rate of either 82
or 4 MHz (reduction by synchronized pulse selection). The
emitted fluorescence was collected at right angles (polarizer set
at 54.7°; monochromator with spectral bandwidths of 4, 8, and
1
6 nm), and the fluorescence lifetime profiles were recorded
with a time-correlated single photon counting setup and a time
-
1
-1
division of 5.2 ps chn (82 MHz version) and 52.6 ps chn
4 MHz version) yielding the corresponding experimental
(
accuracies of (3 ps (82 MHz version) and (0.04 ns (4 MHz
version), respectively. The temporal response of the system
was 35 ps, and the calibration of the experimental setup was
36
checked with pinacyanol in ethanol (τf ) 13 ( 1 ps), rose
3
7
bengal in methanol (τf ) 0.50 ( 0.02 ns), and fluorescein 27
3
8
in 0.1 N NaOH (τf ) 4.50 ( 0.03 ns). For the estimation of
the color shift (wavelength-dependent temporal response of the
detection system) between the instrumental response function
and the actual fluorescence decay, the fluorescence of fluores-
cein 27 (in 0.1 N NaOH) and DCM (in ethanol) were quenched
by saturation with potassium iodide, and their decay profiles
were recorded at the corresponding emission wavelengths.
Analysis of the fluorescence decay data was performed with a
PC using the software packages IBH Decay Analysis Software
V4.2 (IBH Consultants Ltd.) and Globals Unlimited V2.2 Time
Experimental Section
Materials. Metal perchlorates purchased from Merck, Acros,
and Aldrich were of highest purity available and dried in a
vacuum oven before use.33 All the solvents employed were of
spectrometric grade. For the determination of the complex
stability constants, acetonitrile was distilled from CaH2 prior
to use.
Domain (Globals Unlimited). The goodness of the fit of the
2
R
single decays, as judged by reduced chi-squared (ø ), the
autocorrelation function C(j) of the residuals, and the Durbin
Watson parameter (DW), was always acceptable yielding values
2
R
of ø < 1.2 and DW > 1.8. In the case of the global analysis
Apparatus. Melting points (uncorrected) were measured on
of the decays recorded at different emission wavelengths, the
decay times of the species are linked while the program varies
the preexponential factors and time constants until the changes
a Kofler hot plate. IR spectra were obtained on a BioRad FTS
1
1
55 spectrometer. H NMR spectra were recorded on a Bruker
2
AC250 spectrometer.
in the error surface (ø surface) are minimal, i.e., convergence

Charge Transfer in Amino-Substituted BDP Dyes
J. Phys. Chem. A, Vol. 102, No. 50, 1998 10213
is reached. The fitting results are judged for every single decay
3: Yield: 280 mg (0.86 mmol, 43%). Green crystals, mp
172-175 °C. IR (KBr): 1544, 1510 (CdC, CdN), 1190 (B-
2
2
(
local ø ) and for all the decays (global ø ). The errors for all
R
R
2
1
F); H NMR (CDCl3): 1.37 (s, 6H, CH3), 2.56 (s, 6H, CH3),
the measurements presented here were below global ø ) 1.2.
R
5
.98 (s, 2H,), 7.26-7.30 (m, 2H, phenyl), 7.47-7.50 (m, 3H,
The rate constants of radiative kf and nonradiative knr
deactivation were generally calculated from the measured
fluorescence quantum yields and fluorescence lifetimes accord-
ing to eqs 1 and 2; other equations used to calculate rate
constants of reactions in the excited state are given in the
corresponding paragraphs.
•
+
phenyl). MS (EI, 70 eV): 324 (100%, M ), 309 (65%, M -
•
+
•+
CH3 ), 304 (67%, M - HF ). Calculated for C19H19N2BF2:
C 70.41, H 5.87, N 8.65. Found: C 70.33, H 5.93, N 8.64.
Results and Discussion
Absorption and Emission of 1, 2, and 3 in Different
φf
kf ) τf
(1)
(2)
Solvents. 1, 2, and 3 show identical absorption spectra with
41
an absorption pattern typical for BDP chromophores.
In
hexane, a strong S0-S1 transition with a maximum at 500 nm
(
1 - φf)
-1
-1
(ꢀ ) 88 000 M
cm ) and a shoulder at the shorter wavelength
k )
nr
τf
side as well as a broader, much weaker transition at 380 nm (ꢀ
-
1
-1
)
6000 M cm ) are observed. No long wavelength CT
Complex Stability Constants. The complex stability con-
absorption is present in 1 and 2. The similar absorption spectra
of 1, 2, and 3 indicate the absence of donor-acceptor interac-
tions in the ground state, which is obviously due to the sterically
demanding methyl groups that force the molecule into a twisted
conformation. An X-ray structure of a similar compound
showed the twist angle between the phenyl ring and the BDP
chromophore to be 75°. Furthermore, the absorption spectra
are only barely affected by solvent polarity, the maximum being
slightly shifted hypsochromically (ca. 5 nm) with increasing
polarity for 1-3. (Table 1).
-
6
stants K were measured by titrating 10 M solutions of 1 in
acetonitrile with stock solutions of metal perchlorates and
monitoring the changes in fluorescence. From a least-squares
analysis of a plot according to eq 3, K can be determined from
the slope; the reported values are mean values of at least two
measurements with correlation coefficients >0.99.
2
8
If - I
0
)
K[M]
(3)
Imax - Ir
Unlike their absorption properties, the emission behavior of
1 and 2 is strongly solvent-dependent (Figure 1). In hexane, 1
and 2 show emission from the locally excited state of the BDP
fluorophore at 508 nm shaped like a mirror image of the
absorption spectrum. Similar effects occur for the reference
compound 3. In more polar solvents, this emission is strongly
quenched; in diethyl ether by factors of 65 (1) and 25 (2) and
in acetonitrile by factors of 1.3 × 10 (1) and 1 × 10 (2),
respectively. Along with the decrease in LE emission, the
formation of a second broad emission band is observed at longer
wavelengths showing rather low fluorescence quantum yields
and a strongly solvatochromic behavior (Table 1 and Figure
1). The fluorescence quantum yields of the two emission bands
of 1 and 2 in diethyl ether, 1,4-dioxane, and tetrahydrofuran
were calculated after spectral separation. The LE band was
modeled by fitting the emission spectrum of 3 in the corre-
sponding solvent to two log-normal functions describing the
sub and the main band centered at 530 and 508 nm. In a fit of
the spectra showing dual emission, the position, half-width, and
skewness of these two components were kept fixed (within a
certain interval) and a third log-normal function was included
to describe the remaining long wavelength band. An example
of a fit is given in Figure 2, and the calculated fluorescence
quantum yields are included in Table 1.
The intensity of the long wavelength emission band decreases
with increasing solvent polarity, and the center of the band shifts
to the red; in acetonitrile, it could not be detected at all.
Fluorescence excitation spectra of 1 and 2 recorded with the
detection wavelength set in both regions, i.e., the LE as well as
the long wavelength emission band, have identical shapes and
match the UV/vis absorption spectra. This indicates that the
long wavelength emission originates from a species formed after
excitation of the BDP chromophore and both the solvato-
chromism and the positive solvatokinetic effect suggest a highly
dipolar nature. We attribute this red emission to a species
having pronounced charge-transfer (CT) character similar to that
in p-anthracenyl dimethylaniline and related compounds, which
is formed by a fast reaction in the excited state.
The values of K, measured at two different excitation (480 and
95 nm) and detection (505 and 510 nm) wavelengths, are in
good agreement.
4
Syntheses. Compounds 1-3 were prepared from 2,4-
3
9
40
dimethylpyrrole and p-formylphenylaza-15-crown-5 (1),
p-dimethylaminobenzaldehyde (2), and benzaldehyde (3) in a
one-pot reaction. General procedure: 4 mmol of 2,4-dimeth-
ylpyrrole and 2 mmol of the aldehyde were dissolved in 250
mL of absolute methylene chloride under nitrogen atmosphere.
One drop of trifluoroacetic acid was added and the solution
stirred at room temperature until TLC-control (silica; 1, ethyl
acetate; 2, 3, CH2Cl2) showed complete consumption of the
aldehyde. At this point, a solution of 2 mmol dichlorodicy-
anobenzoquinone (DDQ) in CH2Cl2 was added, and stirring was
continued for 15 min followed by addition of 4 mL of N-ethyl-
N,N-diisopropylamine and 4 mL of BF3‚Et2O. After stirring
for another 30 min the reaction mixture was washed with water
and dried, and the solvent was evaporated. The residue was
chromatographed twice on a silica column (1, ethyl acetate/
methanol 10:1, second chromatography CH2Cl2/MeOH 20:1;
3
3
2
, 3, CH2Cl2). Recrystallization from ethyl acetate/hexane (1)
or CHCl3/hexane (2, 3) yielded analytically pure samples.
: Yield: 758 mg (1.40 mmol, 70%). Red crystals, mp 178-
79 °C. IR (KBr): 1545, 1509 (CdC, CdN), 1198 (B-F).
1
1
1
H NMR (CDCl3): 1.50 (s, 6H, CH3), 2.54 (s, 6H, CH3), 3.61-
3
6
.71 (m, 16H, crown), 3.77-3.82 (m, 4H, crown), 5.97 (s, 2H,),
.7-7.1 (m, 4H, phenyl). MS (EI, 70 eV): m/z (%) 541.4
•+
(
100%, M ). Calculated for C29H38N3O4BF2: C 64.34, H 7.07,
N 7.76. Found: C 64.37, H 7.07, N 7.62.
: Yield: 270 mg (0.74 mmol, 37%). Orange needles, mp
27 °C. IR (KBr): 1542, 1509 (CdC, CdN), 1196 (B-F).
H NMR (CDCl3): 1.49 (s, 6H, CH3), 2.55 (s, 6H, CH3), 3.02
s, 6H, N(CH3)2), 5.96 (s, 2H), 6.77 (m, 2H, phenyl), 7.08 (m,
2
2
1
(
•
+
2
H, phenyl). MS (EI, 70 eV): 367 (100%, M ), 352 (24%,
•+
•+
M - CH3 ), 347 (34%), (M - HF ). Calculated for C21H24N3-
BF2: C 68.68, H 6.59, N 11.45. Found: C 68.53, H 6.66, N
11.43.

10214 J. Phys. Chem. A, Vol. 102, No. 50, 1998
Kollmannsberger et al.
TABLE 1: Absorption and Emission Properties of 1, 2, and 3 in Different Solvents61
9
-1
9
-1
λ
abs/nm
λ
em (LE)/nm
λ
em (CT)/nm
Φ
f
(LE)
Φ
f
(CT)
Φ
f
(CT)/Φ
f
(LE)
τ
1
/ps
τ
3
/ps
τ
2
/ns
k
f
/10 s
k
nr/10 s
1
hexane
Et
dioxane
THF
500
498
500
500
496
508
510
511
511
505
0.26
1590
15
6
<3
5
0.16
0.19
0.46
0.43
2
O
613
637
684
a
0.004
0.001
0.0003
0.0002
0.050
0.033
0.014
a
12
33
47
405
150
76
3.65
3.54
2.45
2.56
MeCN
2
hexane
500
498
500
500
496
508
509
510
512
505
0.31
1610
10
6
Et
2
O
615
645
706
a
0.012
0.004
0.001
0.0003
0.050
0.029
0.009
a
4.2
7.2
9.0
3.54
3.50
2.40
2.79
dioxane
THF
MeCN
4
3
3
hexane
Et O
2
dioxane
THF
MeCN
500
498
500
500
497
510
510
510
510
505
0.50
0.59
0.58
0.56
0.60
2720
2980
3350
3060
3170
0.18
0.20
0.17
0.18
0.19
0.18
0.14
0.12
0.14
0.13
a
Too low to be measured
Figure 2. Fit of the LE and CT band of the emission spectrum of 2
in diethyl ether. The LE band is modeled from the main band and the
subband of 3 in diethyl ether (full lines, measured spectrum and fit;
dotted lines, single components; excitation at 475 nm).
the attached dimethylamino and crown substituents or to the
interaction with an energetically low-lying local triplet state
4
2
(
increase of kisc). In solvents more polar than hexane, the
fluorescence decays of both 1 and 2 could be only fitted to the
sum of at least two exponentials leading to lifetimes in the 3-10
ps (fast component) and the 2-3 ns (slow component) time
range. Measurements of the fluorescence decay profiles as a
function of emission wavelength in the region of 495-690 nm
revealed a decrease of the relative amplitude of the fast
component with increasing detection wavelength and the vice
versa behavior of the relative amplitude of the slow component.
This is consistent with a mechanism involving the excitation
of a single ground-state species to a LE state followed by a fast
reaction to a CT state (Scheme 1). Thus, the lifetime of the
LE state is substantially shortened and its fluorescence decays
with a high rate constant, whereas the slow component can be
attributed to the CT state.
Figure 1. Absorption and emission spectra of 2 in hexane (A), diethyl
ether (B), 1,4-dioxane (C), tetrahydrofuran (D), and acetonitrile (E).
The spectra are normalized for better comparison (excitation at 475
nm).
The fluorescence decay profiles of the reference compound
3
could be described by a single-exponential fit (fluorescence
lifetimes τf of 2.7-3.3 ns) in all the solvents investigated; in
Having this picture in mind, the occurrence of a rise time
should be expected in the spectral region of the CT emission.
This rise time should be similar to the main decay time of the
LE band, i.e., the lifetime of the fast decay component, because
owing to sterical restriction and fast charge transfer, a substantial
contribution from other transient species seems to be rather
unlikely. However, it is difficult to detect the rise times owing
9
-1
fact, the radiative rate constants of 0.18 × 10 s are identical
within experimental errors (Table 1). In hexane, the emission
of 1 and 2 centered at 508 nm decays monoexponentially as
well, yielding similar radiative rate constants. The nonradiative
rate constants are slightly higher than for 3, which may be
attributed either to the higher conformational flexibility due to

Charge Transfer in Amino-Substituted BDP Dyes
J. Phys. Chem. A, Vol. 102, No. 50, 1998 10215
SCHEME 1: Generalized Kinetic Scheme for the
Photophysical Processes Governing the Excited-State
Behavior of 1 and 2 Involving Two or Three Excited
Species (LE, CT, (TS)) in Possible Parallel or
Consecutive Reaction Mechanismsa
Figure 3. Normalized fluorescence decay curves of 2 in diethyl ether
at 500 nm (curve A) and 670 nm (curve B); instrumental response
a
For 2, only the LE and CT state are involved and thus kLT ) kTL
) kTC ) kCT ) 0; (TS) ) transitional or third species.
function (IRF) shown for comparison. Global analysis with τ
ns and τ ) 11ps yields A11 ) 1.36 and A12 ) 24.2 for curve A and A21
5.1 and A22 ) -4.9 for curve B, respectively (excitation at 475 nm).
1
) 3.42
2
)
to the large differences in the two lifetimes and the limitations
of our experimental setup. We were at least able to detect these
rise times for 1 and 2 in diethyl ether and for 2 in 1,4-dioxane,
and, indeed, the rise times are very similar to the decay times
of the fast component. In acetonitrile, the emission from the
CT state is extremely weak, and therefore, the relative amplitude
of the long-lived decay component is very small (<0.5%)
making the determination of the rise time impossible.
TABLE 2: Calculated Photophysical Parameters of the
Dimethylamino Compound 2
9
-1
9
-1
CT)-1/10 s
9
-1
k
f f
CT/k
LE
solvent
k
LC/10 s
k
CL/10 s
(τ
0
diethyl ethera
1,4-dioxaneb
85.8
163
4.8
2.9
0.29
0.32
0.234
0.132
a
See Figure 3 for values of Aij. b Global analysis with τ
) 6 ps yields A11 ) 1.32 and A12 ) 73.8 for a fluorescence
decay monitored at 500 nm.
1
) 3.09 ns
and τ
2
Furthermore, for 1, a third decay component with a relative
amplitude of e0.05 and time constants of τ3 ) 405 ps (diethyl
ether), 150 ps (1,4-dioxane), and 76 ps (tetrahydrofuran) was
observed in these three solvents. However, such a third decay
component could not be detected for 2. Obviously, a third,
weakly fluorescing species with pronounced CT character is
involved in the fluorescence behavior of 1. This third decay
component (TS) could be due to a double minimum in the
potential surface of the excited CT state as a function of the
twist angle between the donor and the acceptor, a mechanism
recently observed in highly pretwisted biphenyls.43 However,
it was not possible to detect a second rise time, and the limiting
experimental possibilities, especially in the NIR region, do not
allow for a global analysis of the wavelength-dependent
fluorescence decay data with respect to the verification of such
an excited-state reaction mechanism.44 Therefore, the real nature
of this species and the deactivation process of the initially
excited 1 (Scheme 1) remain unclear at present.
reasonable estimate for τ0LE, an analytical solution of eqs 6-9
yields the photophysical parameters given in Table 2.4
5-47
-
1
A12 X - τ1
)
(6)
-1
A
11
τ2 - X
LE -1
X ) kLC + (τ0
)
(7)
(8)
CT -1
Y ) kCL + (τ0
)
-1
1
2
2
1/2
τ1,2 ) {(X + Y) ( [(X - Y) + 4k k ] }
(9)
LC CL
The resulting rate constants for the formation of the CT state
10 -1
11 -1
of 8.6 × 10
s
in diethyl ether and 1.6 × 10
s
in 1,4-
dioxane document the ultrafast charge transfer that occurs after
excitation of the BDP fluorophore. The high efficiency of this
charge-transfer process in such solvents of medium polarity is
To determine the rate constants governing the excited-state
reaction and relaxation processes in 2, we performed a global
analysis of the time- and wavelength-resolved emission data in
diethyl ether and 1,4-dioxane. The temporal evolution of a
fluorescence decay corresponding to a two-state model, i.e.,
emission from the LE and CT state, can be described by a sum
of two exponentials:
also exemplified by the high ratio of k /k of 18 in diethyl
LC CL
ether and 56 in 1,4-dioxane.
Having Scheme 1, the kinetic parameters given in Table 2,
and the efficient excited-state CT reaction in mind, construction
of the decay as well as the species associated spectra (DASi(λ)
and SASi(λ)) allows for a verification of the kinetic model
t(τ1)-
1
-t(τ2)-1
-
employed. The DASi(λ) and SASi(λ) are given by eqs 10-
I (t) ) A e
+ A e
(4)
(5)
LE
11
12
48
1
4.
1
-t(τ2)-1
-
t(τ1)-
-t(τi)-
1
I (t) ) A e
+ A e
CT
21
22
Itot(λ, t) ) Ai(λ) e
(10)
(11)
∑i
The two decay rates τi- directly observed in the experiment
1
tot
X
Ai(λ) F
(λ)
SS
are linked to the species-related fluorescence lifetimes τ0 (X
DAS (λ) )
)
LE, CT) through eqs 6-9 involving the rate constants of the
i
45,46
excited-state reactions, kLC and kCL.
Using the decay times
Ai(λ)τi
∑i
τi and amplitudes Aij given in the caption of Figure 3 and a

1
0216 J. Phys. Chem. A, Vol. 102, No. 50, 1998
Kollmannsberger et al.
TABLE 3: Protonation-Induced Changes in the
Ii
∞
tot
SS
SAS (λ) ) F (λ) 0
∫
(λ, t) dt, with i ) LE or CT
Photophysical Properties of 1 and 2 in MeCN
i
Itot
9
-1
9
-1
λ
abs/nm
λ
em/nm
Φ
f
τ (LE)/ps
k
f
/10 s
k
nr/10 s
(
12)
1
2
1
2
3
496
496
500
500
497
505
505
510
510
505
0.0002
0.0003
0.68
0.49
0.60
5
2.6
3800
2900
3170
SAS (λ) ) [DAS (λ) + DAS (λ)]τ τ Y
(13)
LE
1
2
1 2
+
+
-H
-H
0.179
0.169
0.19
0.084
0.176
0.13
-1
2
-1
SAS (λ) ) [(τ - Y)DAS (λ) - (Y - τ1 )DAS (λ)]τ τ
1 2
CT
1
2
(14)
yield is reestablished. Time-resolved fluorescence measure-
ments yield monoexponential decay kinetics, the lifetime of the
LE state being increased by a factor of ca. 1000, and calculated
radiative rate constants similar to that of the reference compound
where Itot(λ, t), Ai(λ), and τi are the overall fluorescence intensity,
the amplitudes, and corresponding fluorescence decay times of
tot
the species i from the time-resolved measurements and FSS -
(λ) is the integral steady-state emission spectrum. Furthermore,
3
.
assuming that in the high-energy region of the spectrum only
tot
the LE species emits, SASLE(λ) can be substituted by FSS (λ)
Complexation Studies
in eq 13. In this (short or “blue” wavelength) spectral region
where ICT f 0, convergence of a plot of Y(λ) vs λ yields the
true value of Y (for 2 in diethyl ether Y(λ f λblue) ) 4.99 ×
Having the strong and analytically valuable proton-induced
effects on the absorption and emission behavior of the amino-
substituted BDP derivatives in mind, we examined the photo-
physical changes of 1 upon complexation to the alkali and
9
-1 49
1
0 s ). When no backward reaction occurs, this value must
irr
-1
equal the value for Y in an irreversible process, Y ) τ1
With τ1 ) 3.42 ns, as indicated in the caption of Figure 3, a
value of Y ) 2.93 × 10 s is obtained, clearly smaller than
that found in the experiment. This corroborates the fact that
although the forward CT process is fast and efficient, the
backward reaction is not negligible.
.
+
+
2+
2+
2+
alkaline-earth metal cations Li , Na , Mg , Ca , Sr , and
2
+
irr
8
-1
Ba in acetonitrile in order to test its possible application as a
fluorescent probe. The complexation-induced effects on the
steady-state absorption and emission spectra are similar to those
found in the protonation studies, i.e., only a small change in
absorption (Table 4) but a strong increase in the BDP LE
emission. Moreover, the fluorescence excitation spectra mea-
sured at different emission wavelengths always match and
resemble the corresponding absorption spectra. No cation-
dependent shifts of the emission band were detectable, which
leads to the conclusion that binding of a cation to the crown
ether moiety only suppresses the charge-transfer process owing
to electrostatic interaction between the positively charged cation
and the nitrogen electron donor atom. As can be deduced from
the data in Table 4, complexation of the employed alkali and
alkaline-earth metal ions leads to FEF whose size depend on
both the charge density and the extent of coordination to the
crown nitrogen (Figure 4). Analysis of the fluorescence titration
curves according to eq 3 (see Experimental Section) shows that
in all the cases studied complexes of a well-defined 1:1-
stoichiometry were formed (Figure 5). The effect of complex-
ation on the fluorescence quantum yields is very remarkable,
The rate constants for fluorescence deactivation of both
emitting states can be compared when plotting the concentration
ratio cCT/cLE as a function of time (see eq 15) and calculating
CT LE
kf /kf from the ratio of both the fluorescence intensities and
concentrations of the species (LE and CT) at equilibrium (t f
) according to eq 16.50 In both solvents analyzed, this ratio
∞
is less than 1 (Table 2) pointing to a strongly forbidden transition
in the CT state of 2.
t(τ1)-
1
-t(τ2)-1
-
c (t)
k (e
- e
)
CT
LC
)
(15)
(16)
-t(τ1)-1
-1
-1
+ (X - τ1-1)e-t(τ2)
c (t)
LE
(τ2 - X)e
CT
CT
ΦCT
ΦLE
kf c (t)
CT
kf kLC
)
)
LE
LE
CT -1
kf c (t) kf [kLC + (τ0 ) ]
LE
+
The main driving force for the very fast formation of the
highly polar, weakly emissive CT state is the high electron
affinity of the BDP fluorophore documented by its reduction
i.e., the chelation-induced FEF vary from 90 for Li to 2250
2
+
for Mg . To the best of our knowledge, the FEF observed for
the alkaline-earth metal ions are among the highest fluorescence
enhancement values reported in the literature until today.
In contrast to the steady-state fluorescence measurements,
which suggest a quite simple complexation behavior of 1,
fluorescence decay measurements reveal a more complex
situation. The lifetime of the LE emission increases drastically
upon complexation (Table 1/Table 4), but the fluorescence decay
curves of all the metal ion complexes measured can be described
only by a biexponential fit, with the exception of the Mg2
complex. In all the cases studied, the relative amplitudes of
the two distinct decay components are independent of excitation
(for wavelengths <495 nm) and observation wavelength, and
no rise times were found.
28
potential (-1.26 V vs SCE for 3 compared to, e.g., -2.22 V
vs SCE51 for phenylanthracene). Similar fast charge-transfer
27
reactions in the excited state have been observed in rhodamine
systems where the fluorophore is also very electron deficient
5
2
(
-0.8 to -0.9V vs SCE for several rhodamines ).
Because of the nearly perpendicular conformation of 1 and
2
in the CT state, the transition moment is very small and the
+
radiative deactivation strongly forbidden, which results in small
CT
rate constants for fluorescence deactivation and a ratio of kf
/
LE
kf < 1. A similar tendency has been reported for ADMA
and its preorientated derivative 3,5-dimethyl-ADMA, which has
a twist angle of ca. 65°.1
4,16,17
The charge-transfer process can be efficiently blocked by
protonation. In acetonitrile, the effect of protonation on the
absorption behavior of 1 and 2 is small and only a slight
bathochromic shift can be observed. In contrast, the emission
properties are drastically altered (Table 3). Independent of
solvent, protonation turns 1 and 2 into highly fluorescent
molecules: the CT emission disappears and the emission from
the LE state with its characteristically high fluorescence quantum
Recalling the circumstances for the appearance of two distinct
fluorescence lifetimes, they can be caused either by a photo-
induced process in the excited state (an intra- or an intermo-
lecular photophysical reaction) or by simultaneous excitation
of two different ground-state species. The absence of any shifts
in the steady-state emission spectra and the lack of a rise time
at the red edge of the emission band do not suggest the
involvement of an excited-state reaction such as a charge-transfer

Charge Transfer in Amino-Substituted BDP Dyes
J. Phys. Chem. A, Vol. 102, No. 50, 1998 10217
TABLE 4: Complex Formation and Photophysical Data of the Complexes of 1 with Alkali and Alkaline-Earth Metal Ions in
MeCN (Cavity Size of the Aza Crown: 1.7-1.8 Å)a
b
2
-1
F
SS1/FSS2
λabs/nm
λ
em/nm cation size /Å charge density/Z Å
log Ktot
Φ
f
FEF
τ
1
/ps
A
1
/(A
1
+ A
2
)
τ
2
/ns
1
1
1
1
1
1
1
-Li+
498
498
498
498
498
498
500
508
508
508
508
508
508
510
1.52 (6)
2.04 (6)
1.44 (6)
2.12 (7)
2.52 (8)
2.94 (9)
1.32
0.98
5.55
3.77
3.17
2.72
2.55
2.20
2.90
4.40
3.34
3.10
0.018
0.097
0.45
0.34
0.25
0.12
0.68
90
485
2250
1700 1050
1250
600
45
64
0.38
0.45
0.54
1.75
2.95
3.30
3.18
2.75
2.90
0.05
0.02
+
-Na
2
+
-Mg
2
+
-Ca
-Sr
-Ba
0.56
0.71
0.86
0.41
0.64
0.92
2
+
820
430
2
+
+
-H
3400
a
Value taken from ref 62. b Taken from ref 58; in brackets: coordination numbers of the cation, taken from ref 5c.
TABLE 5: Fluorescence Decay Times of the Ba2 Complex
+
as a Function of Ion-to-ligand Ratio
ratio Ba2+/1
0:1
00:1
250:1
The value of τ
τ
/ps
A
/(A
+A
)
τ
2
/nsa
1
1
1
2
4
488
460
424
0.85
0.86
0.85
2.59
2.54
2.52
4
2
a
2
given in Table 5 is shorter compared to that given
in Table 4. This is due to a different repetition rate (here: 82 MHz
with a time window of 12 ns) used in these time-resolved experiments
and the software-dependent neglect of the predecessor decay in the
fits given in Table 5. The value obtained in the 4 MHz experiments
(Table 4) is correct.
and only differ in their fluorescence quantum yields, it is very
difficult to discriminate between them using steady-state
measurements.
Figure 4. Absorption spectrum of 1 (left) and emission spectra of the
2
+
2+
2+
2+
+
complexes with (from top to bottom) Mg , Ca , Sr , Ba , Na ,
+
For a better understanding of this behavior, we performed
time-resolved measurements with solutions containing 1 and
Li , and the free ligand in MeCN (excitation at 480 nm).
2
+
Ba in the ratios of 2250:1 (full complexation), 400:1, and
0:1 (Table 5). Again, these experiments revealed no depen-
4
dence of the fluorescence lifetimes on the detection wavelength
as well as no dependence of their relative amplitudes on the
5
3
metal-to-ligand-ratios used. This excludes the consecutive
formation of two species with a 1:1 and a 1:2 stoichiometry
and different complexation constants in a titration. The two
species must be therefore formed simultaneously. The two
decay times of the free ligand in the partly complexed solutions
could not be detected here because of the very low emission of
the ligand with respect to the complexes.
We tentatively attribute this behavior to the existence of two
distinct conformers of the complexes in the ground state with
very similar absorption and LE emission spectra but different
fluorescence quantum yields and lifetimes.
This implies that that the fluorescence quantum yields
obtained for the cation complexes are average values of the two
conformers weighted by their relative concentrations cr, except
for the Mg complex, which exists only in one stable ground-
state conformation:
Figure 5. Changes in fluorescence intensity upon addition of metal
ions (excitation at 480 nm, detection at 515 nm).
2
+
process or the photoinduced decoordination of the cation. The
product of a charge-transfer reaction would have a more polar
CT character, its emission should appear red-shifted compared
to the precursor’s, and a rise time would be expected. Especially
in the case of the alkaline-earth metal cations showing two
fluorescence decay times in the nanosecond time range with
comparable relative amplitudes, such a rise time would have
been detected by our instrumental setup. In case of photode-
coordination of the cation,4d,9c the fluorescence lifetimes of the
decoordinated complexes are similar for different cations, in
sharp contrast to our measurements. These excited-state de-
coordination reactions are very fast (picosecond time scale), and
emission is only observed from the decoordinated species; i.e.,
monoexponential decay kinetics are observed.9 On the other
hand, the lack of any dependence of the fluorescence excitation
spectra on the emission wavelength does not favor the existence
of two different ground-state species at first sight. Yet, if two
species have nearly the same absorption and emission spectra
Φ ) cr1Φ
+ cr2Φ
(17)
1
2
With eq 17 it is not possible to determine the ratio of the
conformers, the individual Φi, and the corresponding radiative
and nonradiative rate constants, but the contributions of both
species to the overall fluorescence quantum yield can be derived
from a global analysis of the wavelength-resolved fluorescence
decay data.
Assuming that the equilibrium of conformers in the ground
and in the excited state is the same and that any reaction in the
excited state is slower than the involved relaxation processes,
it is possible to construct the decay-associated spectra from the
time-resolved fluorescence data according to eqs 10 and 11.48
In the absence of any excited-state reaction, these DASi(λ)
c
i
values are identical with the SAS, i.e., the contribution FSS of

10218 J. Phys. Chem. A, Vol. 102, No. 50, 1998
Kollmannsberger et al.
It is important to notice that, in contrast to the aforementioned,
the determination of the complex stability constants is inde-
pendent of the existence of differently emitting conformers as
long as their absorption spectra do not differ and their
concentration ratio remains constant during the titration. Since
conformers are equilibrated, the latter assumption is always true;
the former follows from the steady-state absorption and emission
data. In this case the intensity of the emitted light is proportional
to the total concentration and the determined complexation
constants refer to the total complexation yielding the overall
complex stability constant Ktot.
Although the exact molecular structure of the conformers
remains obscure, the main conformational differences should
reside at the nitrogen atom of the crown. The two conformers
could differ in their twist angles between the phenyl ring and
the BDP chromophore, which will vary their amount of CT
interaction. As a second possibility, having the endo/exo
isomerization of N-substituted aza crowns in mind, an equilib-
rium of a loose and a strong complex could affect both the
pyramidalization and conformational flexibility at the nitrogen
atom of the crown. Scheme 2 illustrates the different equilibria
involved in complex formation of substituted aza crown ethers
and metal ions in solution. Within this multistep Eigen-
Winkler scheme, for a given aza crown ether the ratio of k2/k-2
varies with the metal ion, counterion, and solvent, mainly
depending on the thermodynamic parameters ∆S and ∆H for
de/solvation of cation and crown, ion (pair) association, fit of
i
Figure 6. DAS (λ) (upper part) and spectral contributions FSS (lower
2
i
+
part) of the two decaying species of the Ba complex at M:L ) 400:
1
4
. The indices 1 and 2 correspond to the species decaying with τ
60 ps (1) and τ
f
)
f
) 2.54 ns (2). In the lower part, the full line represents
cation into the cavity, and coordination of the cation to the
the steady-state emission spectrum of the mixture, and the same
spectrum (dotted line) is included in the upper part for better comparison
54,55
crown’s heteroatoms.
The influence of both reaction
coordinates, pyramidalization, and conformational flexibility
e.g., twisting), on the LE and CT emission is well-documented
(excitation at 485 nm, explanation see text).
(
for related donor-acceptor-substituted compounds and was
the single species to the overall fluorescence intensity (see Table
):
intensively studied for alkylaminobenzonitriles. Nonetheless,
4
9
d,46
the real mechanism is still not clear yet.
We assume that
both coordinates cannot be seen separately and depend on each
other for the complexes studied here. As a direct consequence,
electronic and conformational differences of the two conformers
lead to differences in orbital overlap between the lone electron
pair of the nitrogen atom and the aromatic π-system resulting
in different CT quenching constants. Hence, different fluores-
cence decay times are observed for the LE emission.56
DAS (λ)τi- FSS (λ)
1
tot
i
i
FSS (λ) )
(18)
-1
DAS (λ)τi
∑i
i
For a better illustration, the DASi(λ) and corresponding
spectral contours FSS of the two excited species of the Ba
complex are presented in Figure 6.
Although a calculation of the individual fluorescence quantum
yields is impossible, it is reasonable to assume that the rate
constants of fluorescence deactivation of the alkaline-earth metal
ion complexes lie in the same range as those of the Mg
complex, 3, 1-H , and 2-H , which are very similar (Table 3).
The fluorescence decay times of the long-lived decay compo-
i
2+
The stability constants measured for the cation complexes
2
+
2+
2+
2+
+
of 1 are in the order of Ca > Sr > Ba > Mg > Li >
+
Na , which can be explained by the cation size and the diameter
of the crown moiety (Table 4) as well as the higher charge
density of the divalent cations. Hence, divalent cations form
more stable complexes compared to the monovalent ones. The
value of the complexation constant K correlates well with the
2
+
+
+
5
7
2 n+
charge densities n /r as a measure for the electrostatic
2+
2+
nents of the Ca and Sr complexes are longer than that of
attraction in the complexes (ionic radii taken from ref 58). The
the Mg2 complex, pointing to higher fluorescence quantum
+
exceptional behavior of the Mg complex showing the highest
2+
2+
yields of these species compared to that of the Mg complex.
Thus, the higher overall fluorescence quantum yield of the Mg
FEF value, but only a moderate stability constant can be ascribed
2+
2+
to the fact that the Mg ion, which is too small to fit well into
complex suggests the existence of a second weaker emitting
the cavity, has a strong affinity to the nitrogen donor atom of
the crown. In contrast, the comparatively small Li ion is
conformer in the Ca2+ and Sr complexes.
2+
+
SCHEME 2: Generalized Kinetic Scheme for the Processes Participating in the Complex Formation of 1 and Metal Ion
+
Perchlorates in Solution: Complex Formation of M with a Substituted Aza Crown Ether Involving the Formation of a
Contact Pair (Loose Complex, Step 1) and a (Strong) Complex (Step 2)

Charge Transfer in Amino-Substituted BDP Dyes
J. Phys. Chem. A, Vol. 102, No. 50, 1998 10219
supposed to be coordinated mainly by the four crown oxygen
atoms and one or two solvent molecules. This accounts for
the much lower fluorescence enhancement because the inhibition
of the nonradiative deactivation pathway can be only achieved
by effective coordination of the metal ion to the crown nitrogen.
fluorescent probes for metal ions showing intense absorption
and large fluorescence changes in an analytically useful
wavelength region.
Acknowledgment. Financial support by the Deutsche For-
schungsgemeinschaft (DFG), the Fonds der Chemischen Indus-
trie, and the Bayerische Staatsregierung (Sonderprogramm f u¨ r
Infrastrukturmassnahmen) is gratefully acknowledged.
2
n+
The good correlation of FEF and n /r for all the cations studied
+
except for Li confirms this complexation behavior. Similar
effects have been found by us for intramolecular CT fluorescent
probes as well.5
9
References and Notes
A closer examination of the fluorescence lifetimes of the
(1) Lakowicz, J. R. Topics in Fluorescence Spectroscopy; Plenum
2
+
2+
2+
+
+
complexes with Ca , Sr , Ba , Li , and Na reveals the
following tendencies: (i) both fluorescence lifetimes increase
with increasing values for FEF, and (ii) the increase in the
relative amplitude of the fast component is in good correlation
with the cation radius. The first correlation suggests that the
increase in the fluorescence lifetimes of both components seems
to be the direct consequence of the complex formation in the
ground state and the emission of both excited complex conform-
Press: New York, 1994; Vols. 1-4.
(2) Czarnik, A. W. Fluorescent Chemosensors for Ion and Molecule
Recognition, ACS Symposium Series 538, American Chemical Society:
Washington, D.C., 1993.
(
3) Lockhart, J. C. Chemical sensors. In ComprehensiVe Supramo-
lecular Chemistry; Lehn, J.-M., Ed.; Pergamon: New York, 1996; Vol. 1.
(4) (a) Fery-Forgues, S.; Le Bris, M. T.; Guett e´ , J. P.; Valeur, B. J.
Phys. Chem. 1988, 92, 6233. (b) Bourson, J.; Valeur, B. J. Phys. Chem.
1
989, 93, 3871. (c) Bourson, J.; Pouget, J.; Valeur, B. J. Phys. Chem. 1993,
97, 4552. (d) Martin, M. M.; Plaza, P.; Meyer, Y. H.; Badaoui, F.; Bourson,
2
+
ers. Ca forms the most stable complexes in the ground state,
and they show the longest fluorescence lifetimes and highest
fluorescence quantum yields. Assuming comparable rate con-
J.; Lefevre, J.-P.; Valeur, B. J. Phys. Chem. 1996, 100, 6879.
5) (a) Jonker, S. A.; Ariese, F.; Verhoeven, J. W. Recl. TraV. Chim.
(
Pays-Bas 1989, 108, 109. (b) Jonker, S. A.; van Dijk, S. I.; Goubitz, K.;
Reiss, C. A.; Schuddeboom, W.; Verhoeven, J. W. Mol. Cryst. Liq. Cryst.
1990, 183, 273. (c) Jonker, S. A.; Verhoeven, J. W.; Reiss, C. A.; Goubitz,
K.; Heijdenrijk, D. Recl. TraV. Chim. Pays-Bas 1990, 109, 154.
stants of fluorescence deactivation for all the complexes
studied,60 the increase in τf and Φf suggest a decrease in knr
(
6) Huston, M. E.; Haider, K. W.; Czarnik, A. W. J. Am. Chem. Soc.
with increasing complex stability.
1
988, 110, 4460. Huston, M. E.; Engleman, C.; Czarnik, A. W. J. Am.
On the other hand, the second correlation observed favors a
predominantly sterically controlled mechanism being responsible
for the ratio of the conformers formed. And indeed, the increase
in the relative amplitude of the fast component with increasing
cation radius suggests that the equilibrium of conformers in the
ground state is mainly dependent on the size of the cation. The
larger the cation complexed, the higher the relative amplitude
of the weakly fluorescing conformer, the lowest relative
Chem. Soc. 1990, 112, 7054. Czarnik, A. W. Acc. Chem. Res. 1994, 27,
302.
(
7) Alfimov, M. V.; Gromov, S. P.; Lednev, I. K. Chem. Phys. Lett.
1
991, 185, 455. Lednev, I. K.; Ye, T.-Q.; Hester, R. E.; Moore, J. N. J.
Phys. Chem. A 1997, 101, 4966.
(8) Bissell, R. A.; de Silva, A. P.; Gunaratne, H. Q. N.; Lynch, P. L.
M.; Maguire, G. E. M.; Sandanayake, K. R. A. S. Chem. Soc. ReV. 1992,
1
87. de Silva, A. P.; Gunaratne, H. Q. N.; Kane, A. T. M.; Maguire, G. E.
M. Chem. Lett. 1995, 125.
(9) (a) L e´ tard, J.-F.; Lapouyade, R.; Rettig, W. Pure Appl. Chem. 1993,
65, 1705. (b) L e´ tard, J.-F.; Delmond, S.; Lapouyade, R.; Rettig, W. Recl.
TraV. Chim. Pays-Bas 1995, 114, 517. (c) Mathevet, R.; Jonusauskas, G.;
Rulli e` re, C.; L e´ tard, J.-F.; Lapouyade, R. J. Phys. Chem. 1995, 99, 15709.
(d) L e´ tard, J.-F.; Delmond, S.; Lapouyade, R.; Braun, D.; Rettig, W.;
Kreissler, M. Recl. TraV. Chim. Pays-Bas 1995, 114, 517. (e) Cornelissen-
Gude, C.; Rettig, W.; Lapouyade, R. J. Phys. Chem. A 1997, 101, 9673.
+
amplitude being found in the case of Li .
Regarding the emission behavior of the complexes and the
lack of a detectable CT emission, the excited complexes seem
to be stable. No photoejection or decoordination of the cation
could be observed. The complexes of 1 behave much more
like PET probe complexes than typical ICT probe complexes,
which leads to these analytically very valuable fluorescence
enhancement factors.
(10) Van den Bergh, V.; Meuwis, K.; Boens, N.; De Schryver, F. C.;
Vincent, M.; Gallay, J.; Ameloot, M. Proc. SPIE-Int. Soc. Opt. Eng. 1994,
2137, 782. Meuwies, K.; Boens, N.; de Schryver, F. C.; Ameloot, M.;
Gallay, J.; Vincent, M. J. Phys. Chem. B 1998, 102, 641.
(
11) Fabbrizzi, L.; Poggi, A. Chem. Soc. ReV. 1995, 197. De Santis, G.;
Fabbrizzi, L.; Licchelli, M.; Sardone, N.; Velders, A. H. Chem. Eur. J.
1996, 2, 1243.
Concluding Remarks
(12) Rurack, K.; Bricks, J. L.; Kachkovski, A. D.; Resch, U. J. Fluoresc.
A comparison of the solvent-dependent absorption and
emission properties of amino-substituted BDP dyes revealed that
in a nonpolar solvent such as hexane emission occurs only from
a LE state, whereas in more polar solvents an ultrafast excited-
state charge-transfer reaction from the amino donor to the basic
fluorophore takes place. This results in strong quenching of
the LE emission and the appearance of a bathochromically
shifted emission from a lower lying CT state, both fluorescence
quantum yields being low. The efficient nonradiative deactiva-
tion process can be utilized to construct a very efficient
molecular switch. Protonation as well as complexation blocks
the CT process and switches on the LE emission again; i.e., a
strong increase in LE fluorescence quantum yield and lifetime
1997, 7, 63S. Rurack, K.; Bricks, J. L.; Slominski, J. L.; Resch, U. Dyes
Pigm. 1998, 36, 121.
(13) de Silva, A. P.; Gunaratne, H. Q. N.; Gunnlaugsson, T.; Huxley,
A. J. M.; McCoy, C. P.; Rademacher, J. T.; Rice, T. E. Chem. ReV. 1997,
9
7, 1515 and references therein.
(14) Siemiarczuk, A.; Grabowski, Z. R.; Krowczinski, A.; Asher, M.;
Ottolenghi, M. Chem. Phys. Lett. 1977, 51, 315.
(
(
15) Siemiarczuk, A.; Ware, W. R. J. Phys. Chem. 1987, 91, 3677.
16) Siemiarczuk, A.; Koput, J.; Pohorille, A. Z. Naturforsch. 1982, 37a,
598.
(17) Okada, T.; Mataga, N.; Baumann, W.; Siemiarczuk, A. J. Phys.
Chem. 1987, 91, 4490.
(
18) Mataga, N.; Nishikawa, S.; Asahi, T.; Okada, T. J. Phys. Chem.
1
1
1
988, 92, 6233.
(19) Lee, S.; Arita, K.; Kajimoto, O.; Tamao, K. J. Phys. Chem. A 1997,
01, 5228.
(20) Wiessner, A.; H u¨ ttmann, G.; K u¨ hnle, W.; Staerk, H.J. Phys. Chem.
3
with very large fluorescence enhancement factors >10 results.
995, 99, 14923.
Time-resolved measurements suggest that, with the exception
(21) Herbich, J.; Kapturkiewicz, A. Chem. Phys. 1993, 170, 221.
(22) Onkelinx, A.; De Schryver, F. C.; Viaene, L.; van der Auweraer,
M.; Iwai, K.; Yamamoto, M.; Ichikawa, M.; Masuhara, M.; Maus, M.;
2
+
of Mg , all the complexes studied exist as two conformers in
solution, which most likely differ in their conformation at the
crown nitrogen.
Rettig, W. J. Am Chem. Soc. 1996, 118, 2892.
(23) Herbich, J.; Kapturkiewicz, A. J. Am Chem. Soc. 1998, 120, 1014.
The crowned BDP dye presented here is an example for a
very efficient CT system that can be transferred to a LE system
by stable cation complexation in polar acetonitrile. Thus, with
dye 1, we present a new design concept for extremely sensitive
(24) Grabowski, Z. R.; Rotkiewicz, K.; Siemiarczuk, A.; Cowley, D.
J.; Baumann, W. NouV. J. Chim. 1979, 3, 443.
(
25) Rettig, W. Angew. Chem., Int. Ed. Engl. 1986, 25, 971.
(26) Shiguza, H.; Ogiwara, T.; Kimura, E. J. Phys. Chem. 1985, 89,
4302.

10220 J. Phys. Chem. A, Vol. 102, No. 50, 1998
Kollmannsberger et al.
(
27) Plaza, P.; Dai Hung, N.; Martin, M. M.; Meyer, Y. H.; Vogel, M.;
Rettig, W. Chem. Phys. 1992, 168, 365.
28) Kollmannsberger, M.; Gareis, T.; Heinl, S.; Breu, J.; Daub, J.
Angew. Chem., Int. Ed. Engl. 1997, 36, 1333.
29) Gareis, T.; Huber, C.; Wolfbeis, O. S.; Daub, J. Chem Commun.
997, 1717.
30) Werner, T.; Huber, C.; Heinl, S.; Kollmannsberger, M.; Daub, J.;
Wolfbeis, O. S. Fresenius J. Anal. Chem. 1997, 359, 150.
energetically unfavorable in the polar solvents diethyl ether and 1,4-dioxane.
Thus, knr should decrease resulting in more similar values of knr for 1, 2,
LE
(
and 3 in the medium polar solvents, and τf of 3 is taken as τ0 . However,
CT
the calculated values of kLC, kCL, and τ0 show only minor changes (<5%)
LE
(
when choosing τf of 2 in hexane instead as value for τ0
(48) L o¨ froth, J.-E. J. Phys. Chem. 1986, 90, 1160.
.
1
(
(49) Note that the value obtained by graphically analyzing this short
wavelength part of the spectrum is in good agreement with the value of Y
(
31) Molecular Probes Inc., Eugene, OR; URL: http://www.probes.com.
) 5.11 × 109 s-1 obtained from a single set of amplitudes measured at the
(32) Debreczeny, M. P.; Svec, W. A.; Wasielewski, M. R. Science 1996,
blue edge of the spectrum (cf. eq 6).
2
74, 584
33) Gmelins Handbuch der Anorganischen Chemie, Chlor, 8th ed.;
VCH: Berlin, 1927; p 402
34) Olmsted, J., III. J. Phys. Chem. 1979, 83, 2581. Drake, J. M.;
Lesiecki, M. L.; Camaioni, D. M. Chem. Phys. Lett. 1985, 113, 530.
(50) The ratio cCT/cLE in eq 16 is a value obtained for photostationary
(
conditions; i.e., for cCT/cLE (tstat) g 0.99cCT/cLE (t∞). This equilibrium is
reached after ca. 85ps in diethyl ether and 60ps in 1,4-dioxane, respectively.
(
(
(
(
51) Svaan, M.; Parker, V. D. Acta Chem. Scand. B 1982, 35, 559.
52) Piechowski, A. P. J. Electroanal. Chem. 1983, 145, 67.
53) The small decrease in τf when going from 40:1 to 2250:1 is
(35) Resch, U.; Rurack, K. Proc. SPIE-Int. Soc. Opt. Eng. 1997, 3105,
9
6.
(
attributed to fluorescence quenching due to the heavy atom effect induced
by the large excess of Ba ions in the solution.
36) Boens, N.; Tamai, N.; Yamazaki, I.; Yamazaki, T. Photochem.
2+
Photobiol. 1990, 52, 911.
(54) Gokel, G. W.; Echegoyen, L.; Kim, M. S.; Eyring, E. M.; Petrucci,
(
(
(
(
(
37) Eaton, D. F. J. Photochem. Photobiol., B 1988, 2, 523.
38) Velapoldi, R. A.; Epstein, M. S. ACS Symp. Ser. 1989, 383, 98.
39) Treibs, A.; Schulze L. Liebigs Ann. Chem. 1970, 739, 225.
40) Dix, J. P.; Voegtle, F. Chem. Ber. 1980, 113, 457.
S. Biophys. Chem. 1987, 26, 225. Echegoyen, L.; Gokel, G. W.; Kim, M.
S.; Eyring, E. M.; Petrucci, S. J. Phys. Chem. 1987, 91, 3854.
(
55) Salomon, M.; Hefter, G. T. Pure Appl. Chem. 1993, 65, 1533.
D’Aprano, A.; Salomon, M.; Mauro, V. J. Solution Chem. 1995, 24, 685.
56) Note that the detection of any CT emission was not possible in
acetonitrile.
57) Here, the charge density is expressed as n /rn+, the so-called “class
41) de Wal, E. V.; Pardoen, J. A.; van Knoeveringe, J. A.; Lugtenburg,
(
J. Recl. TraV. Chim. Pays-Bas 1977, 96, 306. Karolin, J.; Johansson, L.
B.-A.; Strandberg, L.; Ny, T. J. Am. Chem. Soc. 1994, 116, 7801.
2
(
(
42) Sch u¨ tz, M.; Schmidt, R. J. Phys. Chem. 1996, 100, 2012.
A” or ionic index (cf. Nieboer, E.; Richardson D. H. S. EnViron. Pollut.
Ser. B 1980, 1, 3.).
(43) Rettig, W.; Maus, M.; Lapouyade, R. Ber. Bunsen-Ges. Phys. Chem.
1
2
996, 100, 2091. Maus, M.; Rettig, W.; Lapouyade, R. J. Inf. Rec. 1996,
2, 451.
(58) Shannon, R. D. Acta Crystallogr. 1976, A32, 751.
(59) Rurack, K.; Bricks, J. L.; Resch-Genger, U. To be published.
(44) Strehmel, B.; Seifert, H.; Rettig, W. J. Phys. Chem. B 1997, 101,
2
1
232.
(60) The values of kf given in Tables 1 and 3 support the assumption
that the rate constants of fluorescence deactivation of the complexes are
comparable to those observed for 1, 2, and 3 under conditions when only
the LE emission is present. The measured fluorescence lifetime and quantum
yield, τf and Φf, and the radiative and nonradiative rate constants are linked
(45) Birks, J. B. Photophysics of Aromatic Molecules; Wiley: London,
970.
(46) Schuddeboom, W.; Jonker, S. A.; Warman, J. M.; Leinhos, U.;
K u¨ hnle, W.; Zachariasse, K. A. J. Phys. Chem. 1992, 96, 10809. Leinhos,
U.; K u¨ hnle, W.; Zachariasse, K. A. J. Phys. Chem. 1991, 95, 2013.
by the following equations: τf ) 1/(kf + knr) and Φf ) kf/(kf + knr). With
LE
CT
Φ ) k τ , not only the values obtained for F i from the time-resolved
(47) τ0 and τ0 are the fluorescence lifetimes of the LE and CT state,
f
f
f
SS
respectively. The calculation of the photophysical parameters is only possible
if one of these fluorescence lifetimes is known. However, τ0 can only be
estimated from the fluorescence lifetime of a model compound that does
not show the excited-state reaction investigated. Because of the independence
on solvent polarity of the photophysical parameters kf and knr of 3 and the
similarity of these parameters for 1 and 2 in hexane, the value of τf of 3 in
the corresponding solvent was chosen for τ0 . Although in hexane, knr of
1
to a low-lying nonpolar triplet state, the population of this state should be
measurements (see eqs 10, 11, 18), but also the actual fluorescence quantum
LE
yields Φ of the complexes should increase with increasing complex stability.
i
(61) Measurements in chlorinated solvents were hampered by traces of
HCl formed during the experiments. The much stronger fluorescence of
the protonated species made the determination of spectral parameters of
the free base impossible.
LE
(62) Frensdorff H. K. J. Am. Chem. Soc. 1971, 93, 600. Dalley, N. K.
In Synthetic Multidentate Macrocyclic Compounds; Izatt, R. M., Christensen,
J. J., Eds.; Academic Press: New York, 1978; p 207.
and 2 are somewhat larger than knr of 3, which is most likely attributed