Excited state intramolecular proton transfer and metal ion complexation of 2-(2′-hydroxyphenyl)benzazoles in aqueous solution

Article abstract of DOI:10.1021/jp014634j

The excited-state intramolecular proton transfer (ESIPT) of a series of water-soluble 2-(2′-hydroxyphenyl)-benzazole derivatives has been studied under physiological conditions using absorbance and steady-state emission spectroscopy. At neutral pH in the presence of 0.1 M ionic background, the fluorescence properties of these derivatives differ substantially compared to previously reported data in nonaqueous solvents. The ESIPT process is disrupted, presumably due to intermolecular hydrogen bonding with surrounding water molecules combined with increased stabilization of the trans-rotamer, which cannot undergo the ESIPT process. The emission spectrum of the benzimidazole derivative depends significantly on the solvent polarity, as revealed by titrations with Zn(II) in methanol, ethanol, and under physiological conditions. Inhibition of ESIPT via metal coordination shows a significant wavelength shift together with a substantial ratio increase by a factor of 13.7. Titration of the benzoxazole derivative with Zn(II) yielded a 50-fold increased emission intensity. The fluorescence increase is specific for Zn(II), and with a logK of 3.93 (K_d = 117 μM) the ligand would be suitable as a fluorescence probe in a biological environment to gauge Zn(II) concentrations in the range from 10 μM to 1 mM.

Full text of DOI:10.1021/jp014634j

5210
J. Phys. Chem. A 2002, 106, 5210-5220
Excited State Intramolecular Proton Transfer and Metal Ion Complexation of
2-(2′-Hydroxyphenyl)benzazoles in Aqueous Solution
Maged M. Henary and Christoph J. Fahrni*
School of Chemistry and Biochemistry, Georgia Institute of Technology, 770 State Street,
Atlanta, Georgia 30332
ReceiVed: December 20, 2001; In Final Form: March 4, 2002
The excited-state intramolecular proton transfer (ESIPT) of a series of water-soluble 2-(2′-hydroxyphenyl)-
benzazole derivatives has been studied under physiological conditions using absorbance and steady-state
emission spectroscopy. At neutral pH in the presence of 0.1 M ionic background, the fluorescence properties
of these derivatives differ substantially compared to previously reported data in nonaqueous solvents. The
ESIPT process is disrupted, presumably due to intermolecular hydrogen bonding with surrounding water
molecules combined with increased stabilization of the trans-rotamer, which cannot undergo the ESIPT process.
The emission spectrum of the benzimidazole derivative depends significantly on the solvent polarity, as revealed
by titrations with Zn(II) in methanol, ethanol, and under physiological conditions. Inhibition of ESIPT via
metal coordination shows a significant wavelength shift together with a substantial ratio increase by a factor
of 13.7. Titration of the benzoxazole derivative with Zn(II) yielded a 50-fold increased emission intensity.
The fluorescence increase is specific for Zn(II), and with a logK of 3.93 (Kd ) 117 µM) the ligand would be
suitable as a fluorescence probe in a biological environment to gauge Zn(II) concentrations in the range from
10 µM to 1 mM.
Introduction
has widespread implications in the action of many laser dyes,4,5
as high-energy radiation detectors,6-8 UV photostabilizers,9,10
and fluorescent probes.11-13 In the case of the adiabatic,
barrierless excited-state intramolecular proton transfer (ESIPT),
a covalently attached proton, typically of a hydroxyl or amino
group, in the electronically excited-state migrates to a neighbor-
ing hydrogen-bonded atom less than 2 Å away. The formed
phototautomer emits light and thermally equilibrates back to
the ground state with the proton bound to its original atom. This
process can proceed extremely fast and several studies have
found subpicosecond reaction rates for a variety of molecules,
even in rigid media or at temperatures as low as 4 K.14-21 Recent
femtosecond transient spectroscopy measurements show that the
proton takes 60 fs to arrive at its tautomer equilibrium position.
The electron distribution changes on the same time scale, but
the nuclear framework cannot adjust to the new structure as
quickly and starts to strongly vibrate in low-frequency skeletal
modes.22-24 Since the formed phototautomer is only more stable
in the excited state but not the ground state, the observed Stokes
shift of the fluorescence emission is unusually large with values
ranging from 100 to 500 nm.25 Furthermore, ESIPT molecules
often exhibit dual emission: besides the red-shifted tautomer
emission, a normal emission at higher energy can be observed.
The overall fluorescence emission spectrum depends on the
nature of both the ground and excited state of the molecule and
is often strongly influenced by the pH as well as the hydrogen-
bonding ability and polarity of the solvent.26,27 These properties
render ESIPT molecules interesting targets for the development
of ratiometric fluorescent probes.
Fluorescent sensor molecules for the detection of metal ions
have attracted considerable research interest and led to the
development of highly specific probes with a broad range of
applications in environmental chemistry, biochemistry, and cell
biology.1 The majority of these chemosensors are based on a
“switching on/off” mechanism, such that the fluorescent output
is quenched in the absence, and restored upon binding of the
analyte. Due to the linear relationship between fluorescence
intensity and analyte concentration, such probes can principally
be used for quantitative measurements. Since the emission
intensity depends both on the analyte and the probe concentra-
tion, it is necessary to know the latter. This is in particular a
limiting requirement for many biological applications since in
a cellular environment, the dye concentration depends on many
factors such as membrane permeability, incubation time and
temperature, or variations in the cell size. Measured fluorescence
intensities may vary significantly without actually reflecting
differences in analyte concentrations. This problem can be
solved using a fluorophore which displays a shift in the peak
excitation or emission wavelength upon binding of the analyte.
The ratio of the luminescence intensities at two excitation
wavelengths is sufficient to determine the analyte concentration,
independent of dye concentration or any instrument-related
parameters.2 Few such ratiometric probes have been developed
for intracellular applications at present.3 Due to the lack of
suitable probes, the cell biology of many biologically important
metal cations such as zinc, copper, manganese, and iron is still
unexplored.
In this study, we explore the influence of metal cation binding
on the proton-transfer process. The chelation of metal cations
often competes with protonation of the ligand donor atoms and
therefore the emission of ESIPT molecules is expected to be
responsive toward metal coordination. If the proton transfer
Molecules with intramolecular hydrogen bonds often exhibit
photoinduced proton-transfer processes, a phenomenon which
* To whom correspondence should be addressed. Phone: 404-385-1164/
fax: 404-894-2295. Email address: fahrni@chemistry.gatech.edu.
10.1021/jp014634j CCC: $22.00 © 2002 American Chemical Society
Published on Web 05/08/2002

Ion Complexation of 2-(2′-Hydroxyphenyl)benzazoles
J. Phys. Chem. A, Vol. 106, No. 21, 2002 5211
SCHEME 1
process is efficiently disrupted by the coordinated metal cation,
the red-shifted tautomer emission is expected to disappear in
favor of the normal emission of the metal-complex at higher
energy. Fluorescent probes which are based on inhibition of
ESIPT therefore have the potential to undergo large shifts of
the peak emission energy. A shift of the peak excitation energy
is generally preferred when monitoring the probe response in
biological samples with a standard fluorescence microscope
setup.28 Commercially available ratiometric imaging systems
offer up to millisecond time resolution using a high-speed
excitation monochromator. Nevertheless, with the development
of two-photon microscopy imaging, most of the popular
ratiometric probes used in cell biology are not applicable.29-31
These dyes often show only little shift in the peak emission
and can only be used for ratiometric measurements at two
excitation energies. Two-photon microscopy imaging requires
an expensive pumped-laser system, which does not allow fast
switching between two different excitation wavelengths. In
contrast, the sample emission can be monitored at two different
wavelengths with a rotating filter wheel without requiring
expensive modifications of the instrument. Consequently, to
benefit from the considerable advantages of two-photon mi-
croscopy, a new generation of ratiometric probes is required
which exhibits large shifts of the emission energy upon binding
of the analyte.
displacement of the hydroxyl proton, the ESIPT process is no
longer possible and a substantial blue shift in the peak emission
would be expected (Scheme 1).
Since our primary goal is to develop fluorescence probes for
biological applications, we conducted our studies under physi-
ological conditions in aqueous solution at neutral pH and limited
the selection of cations to biologically important metals such
as Ca(II), Mg(II), Zn(II), Fe(II), and Cu(II). Only heavy and
transition metal ions, such as Zn(II), Cd(II), Cu(II), or Fe(II),
bind with significant affinity to the 2-(2′-hydroxyphenyl)-
benzazole ligands under physiological conditions. Among the
biologically important metals, only Zn(II) ions form a fluores-
cent complex, whereas the complexes with paramagnetic
transition metal ions are nonfluorescent. For this reason, the
present work focuses primarily on the investigation of the Zn(II)
complexes by means of absorption and steady-state emission
spectroscopy. Due to the poor solubility of the benzazole ligands
and their transition metal complexes in aqueous solution at
neutral pH, the ligand framework was slightly modified by
introduction of a carboxylic acid functionality (Scheme 2). Since
carboxylic acids are fully dissociated at neutral pH, the solubility
of the ligands in water is greatly enhanced. The acid functional-
ity is spatially separated from the metal binding site and is not
expected to interfere with the ESIPT process or metal coordina-
tion.
In recent years, the general family of 2-(2′-hydroxyphenyl)-
benzazoles has been studied extensively to elucidate the
mechanism of the excited-state intramolecular proton-transfer
process.21,32-44 The wealth of information available about the
photophysics of benzazole derivatives prompted us to choose
this class of molecules as an attractive starting point to explore
the influence of cation binding on the proton-transfer process.
The presence of an intramolecular hydrogen bond is vital to
the ESIPT process. At the same time, the special properties of
water with its high polarity and capacity to form hydrogen bonds
might interfere and ESIPT processes that occur in nonpolar
solvents cannot necessarily be extrapolated to water. Surpris-
ingly, the enol form of 2-(2′-hydroxyphenyl)benzimidazole
(HBI) undergoes ESIPT with almost 100% efficiency in
nonpolar as well as protic solvents such as ethanol or neat
water.34 If coordination of a metal cation is accompanied by
Experimental Section
Materials and Reagents. 2-Aminophenol (Aldrich, 99%),
2-amino-thiophenol (Aldrich, tech., 90%), o-phenylenediamine
(Aldrich, 99%), barium manganate (Aldrich, tech., 90%), 1,4-
benzoquinone (Aldrich, 98%), ferric chloride hexahydrate
(Aldrich, 98%), (4-formyl-3-hydroxy-phenoxy)-acetic acid ethyl-
ester (1) was synthesized following a literature procedure.45
NMR: δ in ppm vs SiMe4 (0 ppm, 1H, 300 MHz). MS: selected
peaks; m/z. Melting points are uncorrected. Flash chromatog-
raphy (FC): Merck silica gel (70-230 mesh). TLC: 0.25 mm,
Merck silica gel 60 F254, visualizing at 254 nm or with 2%
KMnO4 solution. Analytical experiments: Buffer solutions were
made with ultrapure deionized water (18 Ω resistance). PIPES
(Sigma-Aldrich, 99%), Zn(OTf)2 (Aldrich), Cu(NO3)2‚5H2O

5212 J. Phys. Chem. A, Vol. 106, No. 21, 2002
Henary and Fahrni
SCHEME 2
(Aldrich), Fe(BF4)2‚6H2O (Strem), Ca(NO3)2‚4H2O (Aldrich),
Mg(NO3)2‚6H2O, KNO3 (Aldrich, 99%), KCl (Aldrich, 99%).
7.54 (d, J ) 7.7 Hz, 1H), 7.62 (d, J ) 8.8 Hz, 1H), 8.94 (s,
1H), 13.07 (s, broad, 1H).
Synthesis. (a) {3-Hydroxy-4-[(2-hydroxy-phenylimino)-
methyl]-phenoxy}-acetic acid ethyl ester (2). A solution of
aldehyde 1 (250 mg, 1.12 mmol) and 2-amino-phenol (128 mg,
1.17 mmol) in benzene (15 mL) was heated under reflux using
a Dean-Stark condenser for continuous removal of water. After
12 h the reaction mixture was cooled to room temperature and
concentrated in vacuo. The yellow solid residue was recrystal-
lized from ether providing 251 mg (0.79 mmol, 71%) of Schiff
base 2 as yellow solid: mp 206-208 °C; Rf 0.25 (1:1 hexane:
(c) (4-Benzoxazol-2-yl-3-hydroxy-phenoxy)-acetic acid eth-
yl ester (4a). A mixture of imine 2 (215 mg, 0.68 mmol) and
BaMnO4 (700 mg, 2.73 mmol) in benzene (10 mL) was heated
under reflux for 3 h. The mixture was cooled to room
temperature and the residue filtered through Celite and con-
centrated in vacuo. The black residue was further purified on
silica gel (FC, hexane-CH2Cl2, 3:1 f 1:1, to give 33 mg (0.11
mmol, 15%) of oxazole 4a as colorless solid: mp 123-124
°C; Rf 0.38 (3:1 hexanes-EtOAc); 1H NMR (CDCl3, 300 MHz)
δ 1.32 (t, J ) 7.1 Hz, 3H), 4.30 (q, J ) 7.1, 2H), 4.68 (s, 2H),
6.59-6.65 (m, 2H), 7.32-7.39 (m, 2H), 7.55-7.60 (m, 1H),
7.66-7.72 (m, 1H), 7.95 (d, J ) 8.8 Hz, 1H), 11.65 (s, 1H);
MS (70 eV) 313 (M+, 100), 210 (32), 182 (22), 36 (18). EI-
HRMS m/e calculated for C17H15NO5 313.09502, found
313.09747.
1
EtOAc); H NMR (DMSO-d6, 300 MHz) δ 1.22 (t, J ) 7.1
Hz, 3H), 4.18 (q, J ) 7.1 Hz, 2H), 4.83 (s, 2H), 6.36 (s, 1H),
6.48 (d, J ) 8.8 Hz, 1H), 6.86 (t, J ) 7.4 Hz, 1H), 6.94 (d, J
) 8.2 Hz, 1H), 7.09 (t, J ) 7.7 Hz, 1H), 7.35 (d, J ) 7.7 Hz,
1H), 7.47 (d, J ) 8.8 Hz, 1H), 8.86 (s, 1H), 9.80 (s, 1H), 14.38
(s, broad, 1H); MS (70 eV) 315 (M+, 100), 314 (48), 120 (24),
43 (35). EI-HRMS m/e calculated for C17H17NO5 315.11067,
found 315.11088.
(d) (4-Benzoxazol-2-yl-3-hydroxy-phenoxy)-acetic acid (4b).
To a solution of LiOH‚H2O (97 mg) in a mixture of H2O (1.5
mL), THF (2 mL), and MeOH (1.5 mL) was added 4a (55 mg,
0.18 mmol). The reaction mixture was heated under reflux for
2.5 h, cooled to room temperature, and the organic solvents
were removed in vacuo. The mixture was diluted with H2O (2
mL) and the product precipitated by dropwise addition of aq.
HCl (1 M). The product was filtered off, washed with little H2O,
and dried in vacuo affording 37 mg (0.13 mmol, 74%) of acid
(b) {3-Hydroxy-4-[(2-mercapto-phenylimino)-methyl]-phen-
oxy}-acetic acid ethyl ester (3). A solution of aldehyde 1 (208
mg, 0.93 mmol) and 2-aminothiophenol (116 mg, 0.93 mmol)
in EtOH (8 mL) was heated under reflux for 12 h. The mixture
was cooled to room temperature, concentrated in vacuo and the
residue crystallized from ether yielding 193 mg (0.58 mmol,
63%) of Schiff base 3 as yellow solid: mp 129-130 °C; Rf
1
1
0.23 (1:1 hexane:EtOAc); H NMR (DMSO-d6, 300 MHz) δ
4b as colorless solid. mp 232-234 °C; H NMR (DMSO-d6,
1.22 (t, J ) 7.1 Hz, 3H), 4.19 (q, J ) 7.1 Hz, 2H), 4.88 (s,
2H), 6.52 (s, 1H), 6.62 (d, J ) 8.2 Hz, 1H), 7.26 (t, J ) 7.1
Hz, 1H), 7.35 (t, J ) 7.1 Hz, 1H), 7.48 (d, J ) 7.7 Hz, 1H),
300 MHz) δ 4.77 (s, 2H), 6.62-6.68 (m, 2H), 7.38-7.45 (m,
2H), 7.75-7.82 (m, 2H), 7.92 (d, J ) 8.8 Hz, 1H), 11.35 (s,
1H), 13.10 (s, broad, 1H); MS (70 eV) 285 (M+, 100), 227

Ion Complexation of 2-(2′-Hydroxyphenyl)benzazoles
J. Phys. Chem. A, Vol. 106, No. 21, 2002 5213
(18), 198 (38). EI-HRMS m/e calculated for C15H11NO5
dust particles or fibers. UV-vis absorption spectra were
recorded at 25 °C using a Varian Cary Bio50 UV-vis
spectrometer with constant-temperature accessory. Steady-state
emission and excitation spectra were recorded with a PTI
fluorimeter and FELIX software. Throughout the titration the
sample solution was stirred with a magnetic stirring device. For
all titrations the path length was 1 cm with a cell volume of
3.0 mL. All fluorescence spectra have been corrected for the
spectral response of the detection system (emission correction
file provided by instrument manufacturer) and for the spectral
irradiance of the excitation channel (via calibrated photodiode).
Quantum yields were determined using quinine sulfate dihydrate
in 1.0 N H2SO4 (Φf ) 0.54 ( 0.05) as fluorescence standard.46
Electrode Calibration in Aqueous Solution. Measurements
were performed with a Mettler Inlab combination glass electrode
(Model No. 423). For the determination of the pK values in
aqueous solution, the electrode was calibrated for -log[H3O+]
by titration of a standardized HCl solution (Aldrich, 0.1 N
volumetric standard) with KOH (Aldrich, 0.1 N volumetric
standard) at 25 °C and 0.1 M ionic strength (KCl). The end
point, E°, and slope were determined using Gran’s method47 as
implemented in the software GLEE.48 The calibration procedure
was repeated three times prior to each pK value determination.
The electrode potential was measured with the Corning pH/Ion
Analyzer 355 and the emf measurements were reproducible with
(0.1 mV accuracy. For the determination of the apparent
stability constants, the electrode was calibrated using commercial
pH reference solutions.
Determination of pK Values. All protonation constants
reported in this study were determined from absorption meas-
urements. The UV-vis absorption spectra of the ligands were
monitored for a series of solution in which -log[H3O+] was
varied between 5 and 11. The emf of each solution was directly
measured in the UV quartz cell (electrode diameter 3 mm) and
converted to -log[H3O+] using E° and slope as obtained from
the electrode calibration procedure described above. The raw
spectral and emf data were processed via nonlinear least-squares
fit analysis using the SPECFIT software package.49
Complex Stability Constants. All measured cation stability
constants K, reported herein, were measured at pH 7.2 (PIPES
10 mM, 0.1 M KNO3 ionic strength) and are therefore apparent
stability constants. To 3.0 mL of a 30 µM solution of the
corresponding ligand in PIPES buffer (10 mM, pH 7.20, 0.1 M
KNO3) were added 5 µL aliquots of Zn(OTf)2 stock solution
(concentration 5 mM, 50 mM or 0.5 M). Upon each addition
the solution was stirred for 15 min to reach equilibrium and
the UV-vis spectrum was subsequently monitored. The re-
corded UV-vis traces were analyzed by a nonlinear least-
squares fit algorithm using an equilibrium model with 1:1
complex stoichiometry (SPECFIT software).49
285.06372, found 285.05910.
(e) [4-(1H-Benzimidazol-2-yl)-3-hydroxy-phenoxy)]-acetic
acid ethyl ester (5a). A solution of aldehyde 1 (200 mg, 0.89
mmol) and o-phenylenediamine (96 mg, 0.89 mmol) in EtOH
(10 mL) was heated under reflux for 2 h. After addition of 1,4-
benzoquinone (192 mg, 1.78 mmol) the mixture was further
refluxed for 0.5 h. The reaction mixture was cooled to room
temperature, and quenched by dropwise addition of saturated
aq. NaHCO3 (20 mL). The mixture was extracted with EtOAc
(2 × 20 mL) and the combined organic layers were washed
with H2O (2 × 10 mL), dried (MgSO4), and concentrated in
vacuo. The residue was further purified on silica gel (FC,
hexanes-EtOAc, 4:1 f 2:1, yielding 245 mg (0.78 mmol, 87%)
of imidazole 5a as colorless solid. mp 194-196 °C; Rf 0.36
1
(2:1 hexane:EtOAc); H NMR (CDCl3, 300 MHz) δ 1.23 (t, J
) 7.1 Hz, 3H), 4.20 (q, J ) 7.1 Hz, 2H), 4.85 (s, 2H), 6.57 (d,
J ) 2.5 Hz, 1H), 6.65 (dd, J ) 8.8, 2.5 Hz, 1H), 7.22-7.28
(m, 2H), 7.62 (s, broad, 2H), 7.95 (d, J ) 8.8 Hz, 1H), 13.0 (s,
broad, 1H), 13.4 (b, broad, 1H); MS (70 eV) 312 (M+, 100),
197 (30). EI-HRMS m/e calculated for C17H16N2O4 312.11101,
found 312.11186.
(f) [4-(1H-Benzimidazol-2-yl)-3-hydroxy-phenoxy)]-acetic
acid (5b). Ester 5a (200 mg, 0.64 mmol) was hydrolyzed
following the procedure as described for ester 4a yielding 135
mg (0.47 mmol, 74%) of acid 5b as tan solid. mp > 300 °C;
1H NMR (DMSO-d6, 300 MHz) δ 4.75 (s, 2H), 6.55 (d, J )
2.2 Hz, 1H), 6.64 (dd, J ) 8.8, 2.2 Hz, 1H), 7.27 (dd, J ) 6.0,
3.0 Hz, 2H), 7.63 (dd, J ) 6.0, 3.1 Hz, 2H), 7.99 (d, J ) 8.8
Hz, 1H), 13.18 (s, broad, 3H); MS (70 eV) 284 (M+, 100), 197
(61), 36 (23). EI-HRMS m/e calculated for C15H12N2O4
284.07971, found 284.07969.
(g) (4-Benzothiazol-2-yl-3-hydroxy-phenoxy)-acetic acid
ethyl ester (6a). To a hot solution of FeCl3‚6 H2O (462 mg,
1.71 mmol) in EtOH (5 mL) a solution of imine 3 (142 mg,
0.43 mmol) in EtOH (2 mL) was added, and the resulting
mixture was heated under reflux for 0.5 h. The dark mixture
was cooled to room temperature, diluted with H2O (3 mL) and
extracted with EtOAc (2 × 10 mL). The combined organic
extracts were dried (MgSO4) and concentrated in vacuo afford-
ing a brown residue, which was further purified on silica gel
(FC, hexanes-EtOAc, 5:1 f 3:1, providing 76 mg (0.23 mmol,
54%) of benzothiazole 6a as pale greenish solid mp 137-139
1
°C; Rf 0.38 (3:1 hexane:EtOAc); H NMR (CDCl3, 300 MHz)
δ 1.32 (t, J ) 7.1 Hz, 3H), 4.30 (q, J ) 7.1 Hz, 2H), 4.67 (s,
2H), 6.55-6.62 (m, 2H), 7.38 (t, J ) 7.1 Hz, 1H), 7.49 (t, J )
8.2 Hz, 1H), 7.60 (d, J ) 8.8 Hz, 1H), 7.87 (d, J ) 8.8 Hz,
1H), 7.94 (d, J ) 8.2 Hz, 1H), 12.71 (s, broad, 1H); MS (70
eV) 329 (M+, 100), 256 (20), 226 (38), 214 (43), 198 (31).
EI-HRMS m/e calculated for C17H15NO4S 329.07218, found
329.07211.
Results and Discussion
(h) (4-Benzothiazol-2-yl-3-hydroxy-phenoxy)-acetic acid
(6b). Ester 6a (44 mg, 0.13 mmol) was hydrolyzed following
the procedure as described for ester 4a affording 31 mg (0.103
mmol, 77%) of acid 6b as pale tan solid: mp 226-228 °C; 1H
NMR (DMSO-d6, 300 MHz) δ 4.75 (s, 2H), 6.57 (d, J ) 2.8
Hz, 1H), 6.62 (dd, J ) 11.1, 2.6 Hz, 1H), 7.40 (td, J ) 8.7 Hz,
1.6 Hz, 1H), 7.52 (td, J ) 8.7 Hz, 1.6 Hz, 1H), 7.98-8.10 (m,
3H), 11.81 (s, broad, 1H), 13.12 (s, broad, 1H); MS (70 eV)
301 (M+, 100), 214 (45), 186 (18). EI-HRMS m/e calculated
for C15H11NO4S 301.04088, found 301.04140.
Synthesis. The preparation of all three benzazole derivatives
4-6 is based on the salicylic aldehyde precursor 1 which was
synthesized according to a published procedure (Scheme 2).45
The benzoxazole derivative 4a was obtained by condensation
of 1 with 2-aminophenol to give the Schiff base 2. Initial
attempts to cyclize and oxidize 2 with Fe(III) chloride resulted
in hydrolysis of the imino group, and aldehyde 1 was the only
observed product. In contrast, treatment with barium man-
ganate gave the desired oxazole 4a, but only with low yield.
The low yield is presumably due to decomposition, since no
other products could be isolated. Benzimidazole 5a was
synthesized via condensation of 1 with o-phenylenediamine
followed by oxidation of the dihydroimidazole intermediate with
Steady-State Absorption and Fluorescence Spectroscopy.
All sample stock solutions and buffer solutions were filtered
through 0.2-µm nylon membrane filters to remove interfering

5214 J. Phys. Chem. A, Vol. 106, No. 21, 2002
Henary and Fahrni
TABLE 1: Protonation Constants and Photophysical Data of Benzazole 4b, 5b, and 6b in Aqueous Solutiona
species
data
oxazole (4b)
imidazole (5b)
thiazole (6b)
pK1
pK2
L2-
[LH-]/[L2-][H+]
9.28 ( 0.01
8.63 ( 0.01
8.25 ( 0.01
n.a.
369, 16300
370 (448)e
448 (350)f
0.79
[LH2]/[LH-][H+]
n.a.
5.90 ( 0.01
340, 22200
341 (396)e
396 (350)f
0.63
Absorptionbλmax (nm), ꢀ(cm-1M-1
Excitationc λmax (nm)
Emissionc λmax (nm)
Quantum Yielde
)
347, 18500
347 (420)e
420 (350)f
0.29
Stokes Shift (cm-1
)
5008
4159
4778
LH-
Absorptionb λmax (nm), ꢀ(cm-1M-1
Excitationd λmax (nm)
)
)
317, 29400
329 (417)e, 326 (364)e
364 (320)f, 417 (350)f
4073, 7564
0.07
316, 22400
332 (355)e
355 (332)f, 424 (372)f
3476, 8060
0.25
323, 19900
340 (436)e
436 (340)f
8023
0.05
n.a.
Emissiond λmax (nm)
Stokes Shift (cm-1
Quantum Yield
)
LH2
Absorptionb λmax (nm), ꢀ(cm-1M-1
n.a.
325, 28100
a 0.1 M KNO3, 25 °C. b Deconvoluted spectra from pH titration. c pH 11.20, 0.1M KNO3. d pH 7.20, 10 mM PIPES. e Emission wavelength in
parentheses. f Excitation wavelength in parentheses.
1,4-benzoquinone to afford 5a in 87% yield. Hydroquinone
which was formed as a byproduct from the reduction of 1,4-
benzoquinone could be easily separated by extraction with
aqueous bicarbonate solution. Altogether, the one-pot procedure
provides access to gram quantities of the benzimidazole
derivative under mild conditions and with high yield. Finally,
condensation of 1 with 2-aminothiophenol gave imine 3 with
63% yield. In contrast to the benzoxazole derivative 4, cycliza-
tion and oxidation by Fe(III) chloride were successful and
provided the corresponding benzothiazole derivative 6a with
54% yield. All analytical studies were carried out with the acids
4b, 5b, and 6b, which were obtained by hydrolysis of the
corresponding esters with lithium hydroxide in a methanol-THF-
water solvent mixture.
Most previously published procedures for the synthesis of
benzazoles are based on carboxylic acids, acid chlorides, or
esters as the condensation partner and require much harsher
reaction conditions (e.g. polyphosphoric acid at 250 °C or
nitrobenzene at 150 °C),50 which are not compatible with the
ester functionality. The outlined synthesis gives access to
benzazoles under mild conditions and should be suitable for
the synthesis of a broad range of functionalized benzazoles.
Protonation Equilibria
Absorption Spectra. The pK values for the three benzazole
ligands 4b, 5b, and 6b were measured by spectrophotometric
titrations. The UV absorption spectra were monitored for
solutions in which -log[H3O+] was varied between 5 and 11
while keeping the ionic strength constant with 0.1 M KNO3.
At higher H3O+ concentrations (-log[H3O+] < 5), the ligand
Figure 1. Deconvoluted UV-vis absorbance spectra for a) oxazole
4b, b) thiazole 6b and c) imidazole 5b. The spectra for the fully
deprotonated (L2-) and monoprotonated (LH-) ligand (0.1 M KNO3,
25 °C) were obtained from pH titrations (In case of ligand 5b a double-
protonated species (LH2) was detected). The plotted UV-vis traces
for the 1:1 zinc(II)-ligand complex (ZnL) were obtained from the
corresponding spectrophotometric Zn(II) titrations at constant pH 7.20
via deconvolution (10 mM PIPES, 0.1 M KNO3, 25 °C).
started to precipitate and therefore the pK values of the
carboxylic acid moiety could not be determined by this method.
The UV-vis traces were analyzed over the entire wavelength
range using a nonlinear least-squares fit algorithm as imple-
mented in the SPECFIT software package.49,51 The fitting
procedure not only yielded the protonation constants, but also
allowed spectral deconvolution to determine the UV spectra for
all involved species (Figure 1). A compilation of the curve fitting
results including photophysical properties of the individual
species is given in Table 1. Both the oxazole 4b and thiazole
6b gave UV traces with a single set of isosbestic points within
the studied -log[H3O+] range, suggesting the presence of a single
protonation equilibrium. The pK value can be assigned to
protonation of the phenolic oxygen which is consistent with
the observed red-shifted maximum in the spectrum of the fully
deprotonated ligand (L2-). The proton affinity of the thiazole
ligand was found to be substantially lower (8.25) as compared
to the oxazole derivative (9.28), which is consistent with the
differences in pi-donor and sigma-acceptor properties of sulfur
and oxygen. A similar trend was reported for the unsubstituted
2-(2′-hydroxyphenyl)-benzothiazole and -benzoxazole deriva-
tives in mixed aqueous-organic solvent. Under these conditions
the pK values also differ by approximately one order of
magnitude.52 In contrast, the UV-titration of the benzimidazole
derivative 5b shows two distinct regions, each containing a
characteristic set of isosbestic points (Figure 2a), which suggests

Ion Complexation of 2-(2′-Hydroxyphenyl)benzazoles
J. Phys. Chem. A, Vol. 106, No. 21, 2002 5215
Figure 2. a) UV-vis absorbance of ligand 5b as a function of pH
(-log[H3O+] ranging between 5 and 11, µ ) 0.1 M KNO3, 25 °C). b)
Calculated species distribution diagram as a function of -log[H3O+].
Figure 3. Normalized fluorescence emission and excitation spectra
for a) oxazole 4b, b) thiazole 6b and c) imidazole 5b in aqueous
solution. All spectra were monitored with 0.1 M KNO3 ionic back-
ground at 25 °C. The traces for the monoprotonated ligand (LH-) were
measured at pH 7.20 (10 mM PIPES), the spectra of the fully
deprotonated ligand (L2-) were acquired at pH 11.2 (unbuffered). The
spectra of the Zn(II) complex were recorded at 5 mM total Zn(OTf)2
for (a) and (c) and 500 mM Zn(OTf)2 for graph (b).
SCHEME 3
derivatives 4b and 6b is considerably less basic, and only a
single protonation equilibrium was observed for -log[H3O+]
between 5 and 11. Conclusively, under neutral conditions the
monoprotonated ligand is the predominant species for both
derivatives.
Fluorescence Spectra. The excitation and emission spectra
for the three benzazole ligands were recorded at pH 7.2 and
11.2 in buffered aqueous solution (Figure 3). Several previous
studies of unsubstituted 2-(2′-hydroxyphenyl)-benzazole were
performed in ethanol.32-34,36 To allow a direct comparison with
these literature data, we also measured the photophysical
properties of the ester derivatives 4a, 5a, and 6a in neat ethanol
and in basic ethanolic solution (Figure 4). A compilation of all
photophysical properties is given in Tables 1 and 2.
In aqueous solution at neutral pH, the emission spectrum of
the oxazole ligand 4b exhibits two maxima at 364 and 417 nm,
respectively, with a low quantum yield of 3% (Figure 3a).
Possible mechanisms accounting for the nonradiative loss of
energy include solute-solvent interactions, torsional motions
of the two rings, or triplet population.37,54,55 In ethanol, the higher
energy emission band was observed at approximately the same
wavelength (360 nm), whereas the second, lower energy
emission is considerably red-shifted to 464 nm (Figure 4a).
Previous studies have shown that the intensity of the normal
emission at shorter wavelength is strongly solvent dependent
and generally very weak in a nonpolar environment.33 The
emission has been assigned to the trans-rotamer Et, which cannot
undergo excited-state intramolecular proton transfer (Scheme
3). The excitation spectra corresponding to the two emission
bands differ substantially, both in ethanol and aqueous solution,
the presence of an additional protonation equilibrium in the
analyzed pH range. The first pK value was found to be 8.63
and lies between the pK of the thiazole and oxazole ligands,
and the absorption maximum of the fully deprotonated phenolate
anion (L2-) is also red-shifted compared to the mono- (LH-)
and diprotonated (LH2) species (Figure 1c). The second pK value
of 5.90 is too high for protonation of the carboxylic acid, but
within the expected range for protonation of the benzimidazole
nitrogen (pK of benzimidazole ) 5.68).53 Furthermore, the
deconvoluted UV spectrum of the monoprotonated ligand (LH-)
shows an additional absorption band at lower energy (360 nm),
which was not observed for the oxazole nor the thiazole ligand.
This absorption band has been previously assigned to the keto
tautomer K (Scheme 3), which is in equilibrium with the enol
form at neutral pH.34 The absence of the lower energy absorption
band in the spectra of the benzoxazole and thiazole ligands
suggests that the keto tautomer does not contribute to the
ground-state equilibrium of these derivatives. Based on the
measured pK values, a species distribution plot was calculated
for benzimidazole ligand 5b (Figure 2b). At neutral pH the
monoprotonated ligand shows an abundance of more than 90%
and is therefore the major species present in the equilibrium.
The benzazole nitrogen atom of the oxazole and thiazol

5216 J. Phys. Chem. A, Vol. 106, No. 21, 2002
Henary and Fahrni
TABLE 2: Photophysical Data of Benzazole 4a, 5a, and 6a in Ethanol at 25 °C
species
L-
data
oxazole (4a)
imidazole (5a)
thiazole (6a)
Absorptiona λmax (nm), ꢀ(cm-1M-1
Excitationa λmax (nm)
)
341, 17900
349 (420)c
420 (349)d
5515
318, 23600
300 (360)c
360, 464 (300)d
3668, 9895
356, 17400
339 (393)c
393 (339)d
2644
319, 26400
314 (435)c
435 (314)d
8359
380, 22800
366 (444)c
444 (366)d
3793
335, 26600
320 (375)c
375 (320)d
3184
Emissiona λmax (nm)
Stokes Shift (cm-1
)
LH
Absorptionb λmax (nm), ꢀ(cm-1M-1
Excitationb λmax (nm)
)
Emissionb λmax (nm)
Stokes Shift (cm-1
)
a Addition of excess NaOEt. b Neat EtOH. c Emission wavelength in parentheses. d Excitation wavelength in parentheses.
ability of water to form hydrogen bonds disrupts the intra-
molecular H-bond required for the ESIPT process and de-
stabilizes the planar conformation. The experimental results
suggest that in buffered neutral aqueous solution the intra-
molecular proton-transfer process is disrupted in favor of the
normal emission of the trans rotamer Et and the phenolate anion.
The emission and excitation spectra of the benzothiazole
ligand 6b resemble those of the oxazole derivative, but with a
bathochromic shift of all maxima by 20 to 30 nm (Figure 3b).
At neutral pH in aqueous solution the fluorescence spectrum
of the thiazole ligand also shows dual emission with a maximum
at 436 nm and a weaker shoulder at higher energy (375 nm).
The relative intensity of the two bands is just reversed compared
to the oxazole emission (4b) under the same conditions. The
more intense lower-energy band also matches the peak emission
of the deprotonated phenolate anion. The excitation spectrum
of this band shows a shoulder at 370 nm, which coincides with
the excitation maximum of the phenolate anion. As already
observed for the oxazole derivative, the lower-energy band can
be assigned to emission of the phenolate anion. The pK value
of the thiazole derivative 6b is one order of magnitude lower
compared to the oxazole ligand (8.25 vs 9.28), which implies
that at neutral pH the phenolate anion is present with ap-
proximately 10% abundance. This is consistent with the fact
that the fluorescence spectrum of the thiazole derivative 6b is
dominated by the lower-energy emission of the phenolate anion,
whereas the normal emission at higher energy is the major
component for the oxazole ligand 4b. Interestingly, the emission
spectrum of ester 6a in ethanol exhibits only a single band at
375 nm, which presumably corresponds to the normal emission
of the trans-rotamer Et (Figure 4b). No significant lower-energy
band was observed which could be assigned to emission of the
keto-tautomer or the phenolate anion. As already found for the
oxazole derivative 4b, the fluorescence spectrum of the thiazole
6b in neutral aqueous solution is determined by normal emission
of the phenolate anion and the trans-rotamer. The ESIPT process
is inhibited due to the formation of strong intermolecular
hydrogen bonds with solvent molecules, and emission of the
keto-tautomer (K*) was not observed, neither in aqueous
solution nor in ethanol.
Figure 4. Normalized fluorescence emission and excitation spectra
for a) oxazole 4a, b) thiazole 6a and c) imidazole 5a in ethanol at 25
°C. The traces for the protonated ligands (neutral net charge) were
measured in neat ethanol, the spectra of the deprotonated ligand
(phenolate anion) were recorded after addition of excess NaOEt. The
spectra of the Zn(II) complex were obtained at 5 mM total Zn(OTf)2
for (a) and (c) and 500 mM Zn(OTf)2 for graph (b).
which is in agreement with the presence of two distinctly
different species. In ethanol, the largely Stokes-shifted band at
464 nm can be attributed to the ESIPT emission of the keto
tautomer (1K*).33 In aqueous solution, the lower energy band
at 417 nm is considerably less Stokes shifted (7564 vs 9895
cm-1), and is presumably not due to the emission of the keto
tautomer. The band appears at the same energy as the peak
emission of the fully deprotonated ligand under basic conditions
(Figure 3a). The excitation spectrum recorded at 417 nm shows
a shoulder at 350 nm, which also coincides with the observed
excitation maximum of the deprotonated phenolate anion.
Hence, these data suggest that the lower-energy emission in
aqueous solution can be assigned to fluorescence of the
deprotonated phenolate anion. Given the pK value of 9.23 for
the phenol oxygen, at neutral pH the ground-state equilibrium
is essentially dominated by the monoprotonated species LH-
(>99%). Therefore, the phenolate anion must be formed in the
excited state via intermolecular proton transfer to the solvent.
Similar observations were reported for 2-(2′-hydroxyphenyl)-
benzimidazole (HBI) in water or binary solvents.34,35 The strong
The fluorescence spectrum of the benzimidazole ligand 5b
is composed of three different emission bands (Figure 3c). The
quantum yield was measured to be 25% and is comparable to
the reported quantum yield of the unsubstituted 2-(2′-hydroxy-
phenyl)-benzimidazole (HBI) ligand.33 In aqueous solution at
neutral pH, the spectrum is dominated by the highest energy
band at 355 nm, which is presumably due to normal emission
of the trans-rotamer Et. This assignment is consistent with
previous studies of HBI in ethanol and water, where a maximum
of 350 nm was reported for the emission of the trans-rotamer
Et.32,34,36 Ligand 6b not only exhibits the short-wavelength
emission, but also two weaker, red-shifted emission bands which
appear only as shoulders at 396 and 430 nm. The presence of

Ion Complexation of 2-(2′-Hydroxyphenyl)benzazoles
J. Phys. Chem. A, Vol. 106, No. 21, 2002 5217
TABLE 3: Thermodynamic and Photophysical Data for the Complexation of Benzazole 4b, 5b, and 6b with Zn(II) in Aqueous
Solutiona
data
oxazole (4b)
imidazole (5b)
thiazole (6b)
logKb
3.93 ( 0.01
347, 25100
350 (411)e
411 (350)f
4487
4.32 ( 0.01
332, 23200
340 (393)e
393 (340)f
4675
1.97 ( 0.01
366, 24500
364 (442)e
438 (370)f
4491
Absorptionc λmax (nm), ꢀ(cm-1M-1
Excitationd λmax (nm)
)
Emissiond λmax (nm)
Stokes Shift (cm-1
Quantum Yield
)
0.26
0.23
0.07
a 0.1 M KNO3, 25 °C, PIPES 10 mM pH 7.20. b Apparent binding constant at pH 7.2. c Deconvoluted spectra from UV-vis titration. d PIPES
10 mM, pH 7.20, 0.1M KNO3, 32 µM ligand, 5 mM Zn(OTf)2. e Emission wavelength in parentheses. f Excitation wavelength in parentheses.
three distinctly different species in the excited-state equilibrium
is further supported by the fact that the relative intensity of the
three bands depends strongly on the excitation wavelength. With
decreasing excitation energy, the intensity of the two lower-
energy bands increases. In contrast, the fluorescence spectrum
of ester 5a in ethanol shows only two bands of which the lower-
energy emission at 435 nm is considerably stronger. Based on
published data, the large Stokes shift of 8395 cm-1 of this band
is consistent with emission of the keto-tautomer (1K*), whereas
the higher energy band at 353 nm can be assigned to the normal
emission of the trans-rotamer (1Et*).32,34,36 The observed lower-
energy shoulder at 430 nm in aqueous solution is presumably
due to emission of the keto-form, whereas the second shoulder
at 396 nm can be assigned to the emission of the deprotonated
phenolate anion. The emission spectrum under basic conditions
in buffered aqueous solution or ethanol (pH > 11) shows a
maximum at 396 and 393 nm respectively, which confirms this
assignment. Consistent with the emission properties of the
thiazole and oxazole ligands, the phenolate emission band is
observed in aqueous solution but not in ethanol.
As already described for the thiazole and oxazole derivatives,
the ability of water to form strong intermolecular hydrogen
bonds with the ligand results in stabilization of the two isomers
Et and Eo, which do not undergo ESIPT (Scheme 3). The
fluorescence properties of 5b at neutral pH are therefore mostly
determined by the emission of these two species. This observa-
tion contrasts with previously reported results for 2-(2′-
hydroxyphenyl)-benzimidazole (HBI) in unbuffered water.34
Under these conditions, the keto-isomer (1K*) is responsible
for the major emission channel, whereas fluorescence of the
trans-rotamer (1Et*) is substantially weaker. The difference must
be ascribed to the presence of PIPES buffer (10 mM) and the
0.1 molar ionic background. Both components would be
expected to further stabilize intermolecular hydrogen bonds.
Complexation Studies. The purpose of this study is exploring
the fluorescence properties of benzazole ligands with biologi-
cally important metals. Under physiological conditions, com-
plexation with transition metal ions such as Zn(II), Fe(II), Cu(II),
or Ni(II) leads to a change in both the absorption and emission
spectra of the benzazole ligands. No effect was observed for
Ca(II), Mg(II), Na(I), or K(I). Since the complexes with the
paramagnetic transition metal ions Cu(II), Fe(II), and Ni(II) are
nonfluorescent, the binding studies focused exclusively on the
investigation of the Zn(II) complexes by means of absorption
and steady-state spectroscopy in aqueous solution at neutral pH.
Absorption Spectra. The binding affinities with Zn(II) for
all three ligands 4b, 5b, and 6b were determined by UV-vis
spectrophotometric titrations in PIPES buffered solution at pH
7.2 and constant ionic strength (0.1 M KNO3). The ions of the
ionic background (potassium nitrate), as well as the triflate anion
introduced as counterion of the Zn(II) salt, have a very low
affinity to the ligand and Zn(II) respectively, and should
therefore not disturb the equilibrium measurements. The UV
absorption spectra were recorded for a set of solutions in which
the metal concentration was increased stepwise from 0 to 1 mM
(oxazole 4b and imidazole 6b) or 50 mM in the case of the
thiazole 5b. The spectra were analyzed over the entire wave-
length range using a nonlinear least-squares fit algorithm as
implemented in the SPECFIT software package.49,51 For all three
ligands a single set of isosbestic points was observed, which
suggests a simple complexation equilibrium. Fitting of the
titration curves was found to yield acceptable results only for a
1:1 complexation model. Attempts to use 2:1 or 3:1 stoichio-
metries yielded deconvoluted spectra with negative absorption
values, which further supports the presence of a single equi-
librium system with 1:1 complex stoichiometry. A compilation
of the curve fitting results as well as the photophysical properties
of the three Zn(II) complexes is given in Table 3, and the
deconvoluted UV-vis spectra are included in Figure 1. All
tabulated binding affinities were measured in buffered solution
at constant pH and ionic strength, and therefore correspond to
apparent stability constants (logK′) under these specific condi-
tions. For all ligands, the lowest energy absorption of the Zn(II)
complex shows the typical features of organic chromophores
with a broad, structureless band and high extinction coefficient
(> 20′000 M-1cm-1), which indicates a ligand-centered ππ*
electronic transition. Compared to the monoprotonated, uncom-
plexed ligands, the absorption maxima are considerably red-
shifted, and essentially overlap with the lowest energy band of
the corresponding phenolate anion (Figure 1). This similarity
suggests that coordination of the metal ion is accompanied by
deprotonation of the phenol oxygen, which also acts as donor
atom in the Zn(II) complex. The imidazole ligand 5b has the
highest affinity for Zn(II) with logK ) 4.32. As would be
expected from the smaller π-donor strength of oxygen compared
to nitrogen, the oxazole derivative 4b exhibits a slightly lower
affinity with logK ) 3.93. Following the same trend, the thiazole
ligand, 6b, binds Zn(II) considerably less tightly with logK )
1.97. The differences in affinity are best visualized in a species
distribution plot showing the relative abundance of the complex
as a function of total Zn(II) present (Figure 5). The graph
illustrates that the oxazole and imidazole derivatives both reach
Zn(II) saturation at concentrations around 5 mM, whereas the
thiazole ligand 5b requires a metal concentration of 0.5 M.
Fluorescence Spectra. Using identical conditions as de-
scribed for the UV-vis binding studies, fluorescence titrations
with Zn(II) were performed in aqueous solution for each of the
ligands 4b, 5b, and 6b. The 1:1 zinc-complex of oxazole 4b
revealed a more than 50-fold increase in fluorescence intensity
compared to the unbound ligand (Figure 6a). A quantum yield
of 27% was obtained at Zn(II) saturation, which is nine times
greater than the quantum yield of the free ligand under identical
conditions. Coordination of Zn(II) appears to inhibit major
nonradiative decay pathways leading to an electronically dif-
ferent lowest excited singlet state. Not only are the relative
energies of neighboring singlet nπ* or triplet states expected

5218 J. Phys. Chem. A, Vol. 106, No. 21, 2002
Henary and Fahrni
Figure 5. Calculated species distribution diagram as a function of the
total Zn(II) concentration for oxazole 4b, imidazole 5b, and thiazole
6b in aqueous solution at pH 7.20 and 25 °C (30 µM total ligand
concentration).
Figure 7. UV-vis absorbance (left) and fluorescence emission spectra
(right) of imidazole ligand 5 as a function of added Zn(II) (excitation
at the isosbestic point of the UV titration). a) Titration of 5a in ethanol
b) Titration of 5a in methanol c) Titration of 5b in aqueous solution
(pH 7.20, 10 mM PIPES, 0.1 M KNO3).
Figure 6. a) Left: UV-vis absorbance of oxazole 4b as a function of
added Zn(II) in 10 mM PIPES at 25 °C. Right: fluorescence emission
spectra under identical conditions, excitation at 350 nm. b) Left: UV-
vis absorbance of thiazole 6b as a function of added Zn(II) in 10 mM
PIPES at 25 °C. Right: fluorescence emission spectra under identical
conditions, excitation at 364 nm.
Figure 8. Relative fluorescence intensity (at 411 nm) of ligand 4b in
aqueous solution as a function of various metal cations (pH 7.20, 10
mM PIPES, 0.1 M KNO3).
to be altered, but Zn(II) coordination rigidifies the ligand
backbone and prevents nonradiative deactivation through tor-
sional motions of the two aryl moieties. The emission maximum
was observed at 411 nm, which is only slightly blue-shifted
compared to the emission of the deprotonated phenolate anion.
As already discussed in the previous paragraph, the photophysi-
cal properties of the Zn(II)-bound ligand are very similar
compared to the deprotonated phenolate anion. The excitation
spectrum recorded at 411 nm matches the spectrum of the
phenolate anion very closely, with almost identical maxima at
350 and 347 nm, respectively. The dramatic increase in
fluorescence intensity upon Zn(II) binding prompted us to
further explore the suitability of oxazole ligand 4b as a Zn(II)
specific fluorescence probe. Indeed, titration with Fe(II), Cu(II),
Ca(II), and Mg(II) did not show any increase in fluorescence
intensity, even at high metal concentrations (Figure 8).56 The
titration of thiazole derivative 5b with Zn(II) showed qualita-
tively the same result as observed for oxazole 4b, but the
fluorescence intensity increases only about four times compared
to the unbound ligand (Figure 6b). The peak emission appears
red-shifted at 438 nm, and is again close to the maximum of
446 nm measured for the phenolate anion.
The fluorescence quantum yields of the oxazole and thiazole
derivatives vary significantly between the unbound and Zn(II)-
saturated form. In contrast, the imidazole dervative 5b is strongly
fluorescent in the presence and absence of Zn(II). Conclusively,
this ligand is the only derivative which would be suitable for
the development of ratiometric probes as dicussed in the
introductory paragraph. The formation of the keto-tautomer by
ESIPT is largely inhibited under physiological conditions,
presumably due to the ionic background and buffer used. To
explore the influence of metal coordination on the ESIPT
process in more detail, we performed Zn(II) titrations not only
in aqueous solution, but also in methanol and ethanol using the
ester derivative 5a (Figure 7). As previously shown, in ethanol
the ESIPT process yielded strong emission of the keto-tautomer

Ion Complexation of 2-(2′-Hydroxyphenyl)benzazoles
J. Phys. Chem. A, Vol. 106, No. 21, 2002 5219
(1K*). Addition of increasing amounts of Zn(II) revealed a
substantial blue-shift in the peak emission from 435 to 393 nm.
The keto-tautomer emission as well as the normal emission band
of the trans-rotamer (1Et*) at higher energy disappeared
completely upon saturation with excess Zn(II). Clearly, the
intramolecular hydrogen bond has been replaced by coordination
to the metal cation, resulting in inhibition of the ESIPT process.
Interestingly, the UV-vis spectra measured for an equivalent
set of solutions suggest formation of the Zn(II) complex with
2:1 stoichiometry. After addition of 0.5 molar equivalents of
Zn(II), the spectral changes became much smaller and a large
excess of Zn(II) was required to further affect the absorption
spectrum and shift the equilibrium toward the 1:1 complex.
Since the titration in ethanol was performed in unbuffered
solution, coordination of Zn(II) lowers the pH and therefore
complicates the equilibrium system throughout the titration. It
is therefore not surprising that nonlinear least-squares analysis
of the UV-vis traces using 1:1 and 2:1 complex models was
not successful. An identical fluorescence titration in methanol
yielded qualitatively the same result. Most importantly, the
maxima of the observed emission bands did not shift and
therefore allow a direct comparison with the titration carried
out in ethanol. The higher polarity and stronger hydrogen-
bonding ability of methanol further destabilizes the intra-
molecular hydrogen bond required for ESIPT. As a result the
normal emission of the trans-rotamer, 1Et* is significantly
increased relative to the keto-emission. Addition of Zn(II) again
leads to displacement of the intramolecular hydrogen bond by
the metal cation and yields an emission band virtually identical
to the one observed in ethanol. A qualitative analysis of the
corresponding UV-vis titration in methanol reveals a single
set of isosbestic points with no indication for the formation of
a 2:1 complex. The greater polarity of methanol apparently
stabilizes the positively charged 1:1 complex, whereas in ethanol
the equilibrium is more in favor of the neutral 2:1 species.
Finally, an analogous titration was performed with acid deriva-
tive 5b in aqueous solution at neutral pH. As already pointed
out previously, the fluorescence spectrum of 5b is dominated
by the normal emission of the trans-rotamer, whereas the lower
energy keto-tautomer emission band appears only as a shoulder
(Figure 7c). Addition of Zn(II) again generated a single emission
band with a maximum at 393 nm, which overlaps with the
maxima observed in ethanol and methanol. Nonlinear least-
squares fit analysis of the UV-vis traces for an identical titration
are in agreement with a 1:1 complex stoichiometry, which is
additionally supported by the presence of a single set of
isosbestic points. In the presence of excess Zn(II), both emission
and excitation spectra were independent of the excitation or
emission wavelength at which the spectra were recorded, an
observation which is consistent with the formation of a single
species (1:1 complex).
carboxylate group. Under physiological conditions, the fluo-
rescence properties of all three derivatives 4, 5, and 6, differ
substantially with respect to previously published results in
nonaqueous solvents. Most important, in all cases the ESIPT
process is disrupted, presumably due to intermolecular hydrogen
bonding with surrounding water molecules and increased
stabilization of the trans-rotamer Et, which cannot undergo the
ESIPT process. Titration of the oxazole derivative 4b with Zn(II)
at neutral pH revealed a strong increase in the fluorescence.
The response is specific for Zn(II) and was not observed for
other biologically important metals such as Fe(II), Cu(II),
Mg(II), or Ca(II). With a logK of 3.93 (corresponding to the
dissociation constant Kd ) 117 µM), oxazole 4b would be
suitable as a fluorescence probe in a biological environment to
gauge Zn(II) concentrations in the range from 10 µM to 1 mM.
The thiazole derivative 6b shows qualitatively a similar
behavior, but with substantially lower binding affinity and lower
quantum yield. In contrast, the imidazole derivative 5b revealed
a slightly higher affinity for Zn(II) with a logK of 4.32 (Kd )
47 µM). Furthermore, the titration in ethanol, methanol, and
buffered aqueous solution showed that inhibition of ESIPT via
metal coordination is a promising concept for the development
of fluorescence probes with ratiometric emission properties. The
changes of the fluorescence intensity ratio at two selected
wavelengths going from the unbound ligand to its metal-
saturated form, determine the dynamic resolution of the probe
and is therefore of great importance.2 Typically, with increasing
shift of the peak emission the overall change of emission ratios
is also increasing, which generally improves the dynamic
resolution. Inhibition of ESIPT via metal coordination as
demonstrated by titration of 5a with Zn(II) in ethanol shows a
significant wavelength shift, together with a substantial ratio
increase by a factor of 13.7. For biological applications, the
solvent polarity dependence of the emission spectrum is a
limiting factor, and might lead to artifacts due to partitioning
of the probe into vesicular membranes. Future research will
concentrate on the identification of ESIPT molecules which do
not allow rotational isomerism in the groundstate, and therefore
eliminate the polarity dependent normal emission observed for
2-(2′-hydroxyphenyl)benzazole derivatives.
Acknowledgment. Financial support via startup funds from
the Georgia Institute of Technology is gratefully acknowledged.
We also thank S. Shealy and D. Bostwick for mass spectral
data, C. Burda (Case Western Reserve University, Cleveland)
for valuable discussions, and L. Lawson and B. Ayres for a
critical review of the manuscript.
References and Notes
(1) DeSilva, A. P.; Gunaratne, H. Q. N.; Gunnlaugsson, T.; Huxley,
A. J. M.; McCoy, C. P.; Rademacher, J. T.; Rice, T. E. Chem. ReV. 1997,
97, 1515.
(2) Grynkiewicz, G.; Poenie, M.; Tsien, R. Y. J. Biol. Chem. 1985,
260, 3440.
Conclusions
(3) Haugland, R. P. Handbook of Fluorescent Probes and Research
Chemicals; Molecular Probes, Inc.: Eugene, OR, 1996.
(4) Chou, P. T.; Martinez, M. L.; Clements, J. H. Chem. Phys. Lett.
1993, 204, 395.
(5) Parthenopoulos, D. A.; McMorrow, D.; Kasha, M. J. Phys. Chem.
1991, 95, 2668.
(6) Chou, P. T.; Martinez, M. L. Radiat. Phys. Chem. 1993, 41, 373.
(7) Sytnik, A.; Kasha, M. Rad. Phys. Chem. 1993, 41, 331.
(8) Renschler, C. L.; Harrah, L. A. Nucl. Instrum. Methods Phys. Res.,
Sect. A 1985, A235, 41.
(9) O’Connor, D. B.; Scott, G. W.; Coulter, D. R.; Yavrouian, A. J.
Phys. Chem. 1991, 95, 10252.
Numerous studies focused on the characterization of intra-
molecular excited-state proton transfer of 2-(2′-hydroxyphenyl)-
benzazoles in nonpolar and polar solvents, but only few
investigations were carried out in aqueous solution, and none
of them under physiological conditions in the presence of 0.1
M ionic background. The low solubility of 2-(2′-hydroxy-
phenyl)benzazoles in water, particularly in the presence of
additional ionic background, hampers photophysical studies
under these conditions. To overcome these limitations, we
synthesized a series of derivatives, which are well soluble at
neutral pH due to the presence of a negatively charged
(10) Keck, J.; Kramer, H. E. A.; Port, H.; Hirsch, T.; Fischer, P.; Rytz,
G. J. Phys. Chem. 1996, 100, 14468.
(11) Sytnik, A.; Delvalle, J. C. J. Phys. Chem. 1995, 99, 13028.

5220 J. Phys. Chem. A, Vol. 106, No. 21, 2002
Henary and Fahrni
(12) Sytnik, A.; Kasha, M. Proc. Natl. Acad. Sci. U.S.A. 1994, 91, 8627.
(13) Sytnik, A.; Gormin, D.; Kasha, M. Proc. Natl. Acad. Sci. U.S.A.
1994, 91, 11968.
(14) Barbara, P. F.; Walsh, P. K.; Brus, L. E. J. Phys. Chem. 1989, 93,
29.
(35) Roberts, E. L.; Dey, J.; Warner, I. M. J. Phys. Chem. A 1997, 101,
5296.
(36) Sinha, H. K.; Dogra, S. K. Chem. Phys. 1986, 102, 337.
(37) Grellmann, K. H.; Mordzinski, A.; Heinrich, A. Chem. Phys. 1989,
136, 201.
(15) Laermer, F.; Elsaesser, T.; Kaiser, W. Chem. Phys. Lett. 1988, 148,
119.
(16) Arthen-Engeland, T.; Bultmann, T.; Ernsting, N. P.; Rodriguez, M.
A.; Thiel, W. Chem. Phys. 1992, 163, 43.
(17) Douhal, A.; Fiebig, T.; Chachisvilis, M.; Zewail, A. H. J. Phys.
Chem. A 1998, 102, 1657.
(38) Itoh, M.; Fujiwara, Y. J. Am. Chem. Soc. 1985, 107, 1561.
(39) Nagaoka, S.; Itoh, A.; Mukai, K.; Hoshimoto, E.; Hirota, N. Chem.
Phys. Lett. 1992, 192, 532.
(40) Stephan, J. S.; Grellmann, K. H. J. Phys. Chem. 1995, 99, 10066.
(41) Tanaka, K.; Deguchi, M.; Yamaguchi, S.; Yamada, K.; Iwata, S.
J. Heterocycl. Chem. 2001, 38, 131.
(18) Marks, D.; Prosposito, P.; Zhang, H.; Glasbeek, M. Chem. Phys.
Lett. 1998, 289, 535.
(19) Nagaoka, S.; Nagashima, U.; Ohta, N.; Fujita, M.; Takemura, T.
J. Phys. Chem. 1988, 92, 166.
(42) Elsaesser, T.; Schmetzer, B.; Lipp, M.; Baeuerle, R. J. Chem. Phys.
Lett. 1988, 148, 112.
(43) Potter, C. A. S.; Brown, R. G.; Vollmer, F.; Rettig, W. J. Chem.
Soc., Faraday Trans. 1994, 90, 59.
(20) Frey, W.; Elsaesser, T. Chem. Phys. Lett. 1992, 189, 565.
(21) Frey, W.; Laermer, F.; Elsaesser, T. J. Phys. Chem. 1991, 95, 10391.
(22) Lochbrunner, S.; Schultz, T.; Schmitt, M.; Shaffer, J. P.; Zgierski,
M. Z.; Stolow, A. J. Chem. Phys. 2001, 114, 2519.
(23) Lochbrunner, S.; Wurzer, A. J.; Riedle, E. J. Chem. Phys. 2000,
112, 10699.
(24) Ernsting, N. P.; Kovalenko, S. A.; Senyushkina, T.; Saam, J.;
Farztdinov, V. J. Phys. Chem. A 2001, 105, 3443.
(25) Nagaoka, S.; Kusunoki, J.; Fujibuchi, T.; Hatakenaka, S.; Mukai,
K.; Nagashima, U. J. Photochem. Photobiol. A 1999, 122, 151.
(26) Elsaesser, T.; Schmetzer, B. Chem. Phys. Lett. 1987, 140, 293.
(27) Krishnamurthy, M.; Dogra, S. K. J. Photochem. 1986, 32, 235.
(28) Wang, X. F.; Herman, B. In Fluorescence Imaging Spectroscopy
and Microscopy; Winefordner, J. D., Ed.; John Wiley & Sons: New York,
1996; Vol. 137.
(29) Ko¨nig, K.; Simon, U.; Halbhuber, K. J. Cell. Mol. Biol. 1996, 42,
1181.
(30) So, P. T. C.; Dong, C. Y.; Masters, B. R.; Berland, K. M. Annu.
ReV. Biomed. Eng. 2000, 2, 399.
(31) Ko¨nig, K. J. Microscopy 2000, 200, 83.
(32) Das, K.; Sarkar, N.; Majumdar, D.; Bhattacharyya, K. Chem. Phys.
Lett. 1992, 198, 443.
(33) Das, K.; Sarkar, N.; Ghosh, A. K.; Majumdar, D.; Nath, D. N.;
Bhattacharyya, K. J. Phys. Chem. 1994, 98, 9126.
(34) Mosquera, M.; Penedo, J. C.; Rios Rodr´ıguez, M. C.; Rodr´ıguez-
Prieto, F. J. Phys. Chem. 1996, 100, 5398.
(44) Brewer, W. E.; Martinez, M. L.; Chou, P. T. J. Phys. Chem. 1990,
94, 1915.
(45) Shamim, M. T.; Ukena, D.; Padgett, W. L.; Hong, O.; Daly, J. W.
J. Med. Chem. 1988, 31, 613.
(46) Demas, J. N.; Crosby, G. A. J. Phys. Chem. 1971, 75, 991.
(47) Gran, G. Analyst (London) 1951, 77, 661.
(48) Gans, P.; O’Sullivan, B. Talanta 2000, 51, 33.
(49) Binstead, R. A.; Zuberbu¨hler, A. D. SPECFIT Global Analysis
System; 3.0.27 ed.; Spectrum Software Associates, Marlborough MA 01752,
2001.
(50) Mickelson, J. W.; Jacobsen, E. J.; Carter, D. B.; Im, H. K.; Im, W.
B.; Schreur, P.; Sethy, V. H.; Tang, A. H.; McGee, J. E.; Petke, J. D. J.
Med. Chem. 1996, 39, 4654.
(51) Maeder, M.; Zuberbu¨hler, A. D. Anal. Chem. 1990, 62, 2220.
(52) Durmis, J.; Karvas, M.; Manasek, Z. Collect. Czech. Chem.
Commun. 1973, 38, 215.
(53) Catalan, J.; Claramunt, R. M.; Elguero, J.; Laynez, J.; Menendez,
M.; Anvia, F.; Quian, J. H.; Taagepera, M.; Taft, R. W. J. Am. Chem. Soc.
1988, 110, 4105.
(54) Yang, G.; Morlet-Savary, F.; Peng, Z.; Wu, S.; Fouassier, J.-P.
Chem. Phys. Lett. 1996, 256, 536.
(55) Mordzinski, A.; Grellmann, K. H. J. Phys. Chem. 1986, 90, 5503.
(56) Similar results were reported very recently for 2-(3,5,6-trifluoro-
2-hydroxy-4-methoxyphenyl)benzoxazole in organic and mixed aqueous
solvents: Tanaka, K.; Kumagai, T.; Aoki, H.; Deguchi, M.; Iwata, S. J.
Org. Chem. 2001, 66, 7328.