Solvatochromism of 3-[2-(4-diphenylaminophenyl)benzoxazol-5-yl]alanine methyl ester: A new fluorescence probe

Article abstract of DOI:10.1016/j.saa.2004.06.035

The photophysical properties of 3-[2-(4-diphenylaminophenyl)benzoxazol-5-yl]alanine methyl ester (1b) and its Boc derivative (1a) were studied in a series of solvents. Its UV-Vis absorption spectra are less sensitive to the solvent polarity than the corre

Full text of DOI:10.1016/j.saa.2004.06.035

Spectrochimica Acta Part A 61 (2005) 1133–1140
Solvatochromism of 3-[2-(4-diphenylaminophenyl)
benzoxazol-5-yl]alanine methyl ester
A new fluorescence probe
Katarzyna Guzow, Magda Milewska, Wiesław Wiczk∗
Faculty of Chemistry, University of Gdan´sk, Sobieskiego 18, 80-952 Gdan´sk, Poland
Received 5 May 2004; accepted 21 June 2004
Abstract
The photophysical properties of 3-[2-(4-diphenylaminophenyl)benzoxazol-5-yl]alanine methyl ester (1b) and its Boc derivative (1a) were
studied in a series of solvents. Its UV–Vis absorption spectra are less sensitive to the solvent polarity than the corresponding fluorescence
spectra which show pronounced solvatochromic effect leading to large Stokes shifts. Using an efficient solvatochromic method, based on the
molecular-microscopic empirical solvent polarity parameter ETN , a large change of the dipole moment on excitation has been found. From
an analysis of the solvatochromic behaviour of the UV–Vis absorption and fluorescence spectra in terms of bulk solvent polarity functions,
f(εr,n) and g(n), a large excited-state dipole moment (µe = 11 D), almost perpendicular to the smaller ground-state dipole moment, was
observed. This demonstrates the formation of an intramolecular charge-transfer excited state. Large changes of the fluorescence quantum
yields as well as the fluorescence lifetimes with an increase of a solvent polarity cause that the new non-proteinogenic amino acid, 3-[2-
(4-diphenylaminophenyl)benzoxazol-5-yl]-alanine methyl ester, is a new useful fluorescence probe for biophysical studies of peptides and
proteins.
© 2004 Elsevier B.V. All rights reserved.
Keywords: Solvatochromism; Excited-state dipole moment; Fluorescence lifetime; Internal charge transfer; Fluorescence probe
1. Introduction
to quantify the relative contributions from these two effects.
In order to obtain an insight to the various models of solva-
tion, several empirical solvent polarity parameters have been
proposedtocharacterizeandrankempiricallythepolarityofa
solvent [2–6]. One of the most popular solvent polarity scale,
ET (30), based on Vis spectroscopic band shifts of a negatively
solvatochromic standard probe [3,4] is particularly useful for
the study of a solvent dependence on UV–Vis absorption and
emission energies. The multiple linear regression, approach
of Kamlet and co-workers [2,5,6], has also been used to corre-
late UV–Vis absorption and emission energies with an index
of the solvent dipolarity/polarizability which is a measure
of the solvent’s ability to stabilize a charge or dipole through
nonspecific dielectric interactions (π∗), and indices of the sol-
vent’s hydrogen-bond donor strength (α) and hydrogen-bond
acceptor strength (β), [7–10] according to Eq. (1):
In case of different electron densities in the electronic
ground and excited state of a light-absorbing molecule, its
dipole moment is different in these two states. Thus, a change
of the solvent affects the ground and excited state differ-
ently. It is commonly acknowledged that solvent-dependent
UV–Vis spectral band shifts can arise from either general
and/or specific solvent effects [1]. The first effect results from
isotropic interactions of the chromophore dipole moment
with the reaction field induced in the surrounding solvent.
Specific effects result from the short-range anisotropic inter-
actions between the chromophore with one or more solvent
molecules in its first solvation shell, an important example of
which is the formation of hydrogen bonds. It is thus of interest
∗
Corresponding author. Tel.: +48 58 34 50353; fax: +48 58 34 50472.
E-mail address: ww@chem.univ.gda.pl (W. Wiczk).
v˜A,F = v˜0 + Aπ∗ + Bα + Cβ
(1)
1386-1425/$ – see front matter © 2004 Elsevier B.V. All rights reserved.
doi:10.1016/j.saa.2004.06.035

1134
K. Guzow et al. / Spectrochimica Acta Part A 61 (2005) 1133–1140
The intramolecular charge-transfer excited state can also be
investigated through the difference in the dipole moments
between the first excited singlet state (µe) and the ground
state (µg), e.g. ꢀµ = µe − µg. For spherical molecules, in
the case of isotropic polarizability of a molecule based on the
theory of Bilot and Kawski [11,12] and Bakshiev [13], Eqs.
(2) and (3) hold:
of µe, and µg as well as µe can be calculated directly from
the spectroscopic measurements [12,14].
As explained by Ravi et al. [15] and Kumar et al. [16] the
problem associated with the Onsager radius estimation can
be minimized since a ratio of two Onsager radii is involved
according to Eq. (13):
ꢄ
ꢇ
ꢀ
ꢁ
2
ꢅ
ꢆ
3
∆µ
aB
a
ν˜a − ν˜f = 11307.6
ETN + constant (13)
ν˜A − ν˜F = m1f(εr, n) + constant
ν˜A + ν˜F = −m2[f(εr, n) + 2g(n)] + constant
(2)
(3)
∆µB
where ꢀµB and aB are the dipole moment change and
Onsager cavity radius, respectively, for a pyridinium N-
phenolate betaine dye used to determine the ET (30) and re-
where ν˜A and ν˜F are the absorption and emission band shifts
(in wavenumbers) measured in solvent of different relative
permittivities εr and different refractive indices n,
spective ETN values (ꢀµB =9 D and aB =6.2 A), whereas ꢀµ
˚
2(µꢀe − µꢀg)2
2(µe + µg − 2µeµgcos ψ)
2
2
and a are the corresponding quantities for the molecule un-
der study. ETN is a dimensionless solvent polarity parameter
introduced by Reichardt [4] as a normalized ET (30) solvent
polarity parameter according to Eq. (14).
m1 =
m2 =
=
(4)
(5)
hca3
hca3
2(µ2e − µ2g)
hca3
ET (solvent) − ET (TMS)
ETN
=
(14)
µe and µg are the dipole moments in the excited and ground
state, respectively, h is Planck’s constant, c is the velocity
of light in vacuum, a is Onsager’s interaction radius of the
solute, and f(εr, n) are the solvent polarity functions given by
Eqs. (6) and (7).
ET (water) − ET (TMS)
where TMS denotes tetramethylsilane (ET (30) = 30.7, [4]).
From the proteinogenic ␣-amino acids only phenylala-
nine, tyrosine, and tryptophan possesses a fluorescence
in the near-UV region and are, therefore, frequently used
in the biophysical studies of peptides and proteins [17].
Among these three ␣-amino acids, only tryptophan shows
an evident solvatochromic effect [18,19] and is used for
the estimation of the solvent accessibility to the indole
chromophore [20]. To by-pass the limitation connected with
tryptophan, we synthesized non-proteinogenic aromatic
␣-amino acids based on the (benzoxazol-5-yl)alanine
skeleton which have remarkable photophysical properties
[21–24]. In this paper, we present a new ␣-amino acid
Boc-3-[2-(4-diphenylaminophenyl)benzoxazol-5-yl]alanine
methyl ester (1a, Fig. 1) exhibiting a large Stokes shift.
ꢀ
ꢁ
2n2 − 1 εr − 1 n2 − 1
f(εr, n) =
−
(6)
(7)
n2 + 2 εr + 2 n2 + 2
3 n4 − 1
g(n) =
2
2
2
(n + 2)
Generally, the dipole moments µe and µg are not parallel to
each other but make an angle ψ.
The use of Eqs. (4) and (5) leads than to Eqs. (8) and (9)
[12,14]:
ꢀ
ꢁ
1/2
1
2
2
µe = µg
+
m2hca3
(8)
(9)
ꢂ
ꢃ
1
m1
m2
cos ψ =
(µ2g + µ2e) −
(µ2e − µ2g)
2µeµg
If the ground and excited state dipole moments are parallel,
based on Eqs. (4) and (5), Eqs. (10)–(12) can be derived:
ꢀ
ꢁ1/2
ꢁ1/2
m2 − m1 hca3
µg =
(10)
(11)
2
2m1
ꢀ
m2 + m1 hca3
µe =
or
2
2m1
m1 + m2
m2 − m1
µe =
µg for (m2 > m1)
(12)
From Eqs. (10) and (11) arise that for parallel orientated
dipole moments in the ground and excited state there is no
need to estimate Onsager’s cavity radius for the calculation
Fig. 1. Molecular structure of 1a and 1b.

K. Guzow et al. / Spectrochimica Acta Part A 61 (2005) 1133–1140
1135
2. Experimental
M.p. 133–134 ◦C; IR (KBr): ν˜max (cm−1) = 3398, 3062,
2973, 1720, 1703, 1594, 1495, 1393, 1323, 1286, 1269, 1208,
1164, 1056, 1025, 984, 840, 762, 699; 1H NMR (400 MHz,
CDCl3): δH (ppm) = 1.42 (s, 9H, (CH3)3), 3.16–3.27 (m, 2H,
2.1. Synthesis
N-Boc-3-[ 2-( 4-diphenylaminophenyl )benzoxazol-5-yl ]
alanine methyl ester (1a) was synthesized from N-
Boc-3-nitro-l-tyrosine methyl ester [25,26] and 4-
(diphenylamino)benzaldehyde (Lancaster) which gives
the intermediate Schiff base. This base was oxidized to the
corresponding heterocyclic compound with lead tetraacetate
in DMSO [23]. The Boc group was removed to give 1b
using a mixture of trifluoroacetic acid and CH2Cl2 (50:50,
v/v).
C␤H2), 3.74 (s, 3H, OCH3), 4.64 (d, 1H, C␣H, J = 8.01 Hz),
ꢁ
5.04 (d, 1H, NH, J = 8.81 Hz), 7.06–7.19 (m, 9H, C6H, C3 H,
ꢁ
C5 H, C2aH, C2bH, C4aH, C4bH, C6aH, C6bH), 7.31–7.35
(m, 4H, C3aH, C3bH, C5aH, C5bH), 7.46 (d, 2H, C4H, C7H,
ꢁ
ꢁ
J = 8.01 Hz), 8.04 (dt, 2H, C2 H, C6 H, J = 2.40 Hz, J =
7.01 Hz); 13C NMR (100 MHz, CDCl3): δC (ppm) = 28.51
(CH3)3, 38.56 C␤, 52.53 OCH3, 54.76 C␣, 68.81 Ct-Bu
,
110.44 C7, 120.31 C4, 121.28 C4a, C4b, 124.54 C2a, C2b, C6a,
ꢁ
ꢁ
C6b, 125.92 C3 , C5 , 125.97 C6, 128.94 C3a, C3b, C5a, C5b,
ꢁ
ꢁ
ꢁ
A mixture of N-Boc-3-nitro-l-tyrosine methyl ester
(0.68 g, 2.0 mmol) and 10% palladium on active carbon in
MeOH (20 ml) was stirred under a hydrogen atmosphere at
room temperature for about 90 min. (TLC monitoring, Merck
silica-gel plates (Kieselgel 60 F254), (CH2Cl2/MeOH/AcOH
100:10:1), Rf = 0.9 (N-Boc-3-nitro-l-tyrosine methyl ester),
Rf = 0.72 (N-Boc-3-amino-l-tyrosine methyl ester)). The
catalyst was filtered off and the solvent evaporated in vacuo
to give a brownish oily product which was dissolved in
anhydrous EtOH and mixed with the solution of the aldehyde
(0.57 g, 2.1 mmol) in anhydrous EtOH. The mixture was
stirred at RT overnight (TLC monitoring (AcOEt/petroleum
ether 1:2), Rf = 0.85). After this time, the solvent was
removed by evaporation and the Schiff base obtained was
dissolved in DMSO and lead tetraacetate (1.33 g, 3.0 mmol)
was added. The mixture was stirred at RT for about 2 h (TLC
monitoring (AcOEt/petroleum ether 1:5), Rf = 0.16) and
then dissolved in AcOEt and washed in turns with a saturated
aqueous solution of NaCl (1×), a 5% solution of NaHCO3
(2×), a saturated aqueous solution of NaCl (3×), and dried
with anhydrous MgSO4. The solvent was evaporated in
vacuo and the product was isolated by means of column
chromatography (Merck, Silica gel 60, 0.040–0.063 mm),
using a mixture of AcOEt/petroleum ether 1:3 as an eluent.
The crude product was recrystallized from a mixture of
MeCN/water, giving a yellow solid (0.64 g, 1.1 mmol)
with 57% yield. The purity of the obtained compound was
checked by means of TLC (Rf = 0.49, AcOEt/petroleum ether
1:3) and analytical RP-HPLC (Kromasil column, C-8, 5 ␮m,
250 mm long, i.d. = 4.5 mm) with detection at λ = 223 nm.
The mobile phase was a gradient running from 0 to 100% of
B (A = water with addition of 0.01% trifluoroacetic acid, B
= 80% of an aqueous solution of acetonitrile with addition
of 0.08% trifluoroacetic acid) over 60 min plus 100% of B
over 10 min (tR = 67.8 min). The melting point (m.p.) was
determined in a capillary tube using a Gallenkamp Griffin
MPA-350.MB2.5 apparatus and is uncorrected. The identi-
129.79 C2 , C6 , 132.50 C5, 135.01 C1 , 142.51 C9, 146.08
ꢁ
C4 , 146.96 C8, 148.11 NHCO, 151.43 C1a, C1b, 164.27 C2,
171.90 CO; MS m/z (MALDI): 563.2 (93%, M+); anal. calcd.
for C34H33N3O5 (%): C, 72.45; H, 5.90; N, 7.46; found: C,
72.04; H, 5.66; N, 7.41.
2.2. Spectroscopic measurements
The absorption spectra of 1a and 1b in all solvents studied,
which were of the highest available quality (spectroscopic or
HPLC grade), were measured with a Perkin-Elmer Lambda
40P spectrophotometer, whereas emission spectra were mea-
sured using a Perkin-Elmer LS 50B spectrofluorimeter. Fluo-
rescence quantum yields (QY) were calculated with quinine
sulphate in 0.5 M H2SO4 (QY= 0.53 0.02) as reference
and were corrected for different refractive indexes of solvents
[17]. In all fluorimetric measurements, the optical density of
solution does not exceed 0.1.
The fluorescence lifetimes were measured with a time-
correlated single-photon counting apparatus Edinburgh CD-
900. The excitation source was a NanoLed 03 (UV led
370 nm) from IBH. The half-width of the response function of
apparatus was about 1.0 ns using a Ludox solution as a scatter.
The excitation and emission wavelengths were isolated using
a monochromator (about 12 nm spectral band-width). Fluo-
rescence decay data were fitted by the iterative convolution
to the sum of exponents according to Eq. (15),
ꢀ
ꢁ
t
I(t) =
αi exp −
(15)
τi
i
where αi is the pre-exponential factor obtained from the
fluorescence intensity decay analysis and τi the decay time
of the i-th component, using a software supported by the
manufacturer. The adequacy of the exponential decay fitting
was judged by visual inspection of the plots of weighted
residuals as well as by the statistical parameter X2R and
shape of the autocorrelation function of the weighted
residuals and serial variance ratio (SVR). Because there is
no difference in the experimental data, obtained for both
compounds (1a and 1b), within the range of experimental
error, only one set of them is presented and discussed in this
paper.
fication of the product was based on the H and 13C NMR
1
(Varian, Mercury-400 BB spectrometer (400 and 100 MHz,
respectively)) in CDCl3, FTIR (Bruker IFS-66 instrument),
and mass spectra (Bruker Biflex III (MALDI-TOF)) as well
as by elemental analysis (Carlo Erba CNSO Eager 200
instrument).

1136
K. Guzow et al. / Spectrochimica Acta Part A 61 (2005) 1133–1140
Fig. 2. The UV–Vis absorption and emission spectra of 1b measured in n-hexane (16), methanol (1), DMF (21), ethyl acetate (19) and propylene carbonate
(23).
3. Results and discussion
tion of the emission band depends more on the nonspecific
dielectric interactions than on the solvent’s hydrogen-bond
donor and acceptor strengths which both have similar in-
fluence. Contrary to the absorption, these parameters have
greater influence on the energy of the emission transition.
The energy of absorption and emission transitions are also
correlated with the empirical solvent polarity parameter ETN.
As in the case of multiple regression analysis, the correlation
of the energy of absorption transition with the solvent polar-
ity parameter ETN is poor (r = 0.605, Np = 22) because of the
only small changes of the absorption maxima position with
the solvent polarity. Much better correlations with ETN is ob-
tained for the energy of emission transition (r = 0.906, Np =
22). This correlation is better when the solvents were divided
into non-hydrogen-bond donor (non-HBD) and hydrogen-
bond donor (HBD) solvents. For the non-HBD solvents, the
following correlation equation was obtained:
Representative UV–Vis absorption and fluorescence
spectra of 1b in n-hexane, methanol, DMF, and propylene
carbonate are presented in Fig. 2, whereas the spectral
and photophysical parameters (maxima of absorption and
emission spectra in wavenumbers, fluorescence quantum
yields (QY), fluorescence lifetimes (␯), and quality of fit
parameter (X2R)) in all solvents studied are collected in
Table 1. The absorption spectra do not show a vibrational
structure, even not in nonpolar solvents such as n-hexane,
whereas the emission spectra measured in nonpolar solvents
show well-resolved vibrational structures (Fig. 2). An
increase of solvent polarity causes a bathochromic shift of
the long-wavelength absorption and fluorescence bands,
however, the red shift of the emission band is substantially
longer (Fig. 2, Table 1). A multiple regression analysis of
the absorption transition energy using Eq. (1) with the data
collected in Tables 1 and 2 gives Eq. (16)
ν˜F(cm−1) = 25090( 230) − 9140( 770)ETN,
r = 0.963, Np = 13 (18)
ν˜A(cm−1) = ν˜A0 − 110( 160)π∗ − 370( 110)α
+ 100( 140)β,
while for the transition energy of emission the following
parameters were obtained (Eq. (17)) with a much better
correlation coefficient.
whereas for HBD solvents,
r = 0.736
(16)
ν˜F(cm−1) = 24650( 470) − 5100( 640)ETN,
r = 0.949, Np = 9. (19)
ν˜F(cm−1) = ν˜F0 − 3510( 470)π∗ − 1120( 300)α
The change of solvent polarity influences not only the
energy of absorption and emission transition but also the
fluorescence quantum yield as well as the fluorescence
lifetimes (Table 1). The highest fluorescence quantum yield
was observed in nonpolar solvents (0.99 in iso-octane, 0.98
in n-hexane) and the lowest in water (0.003). The dependence
of the fluorescence lifetime on solvent polarity is opposite
to the fluorescence quantum yield. The shortest fluorescence
− 1110( 400)β,
r = 0.940
(17)
The solvent’s hydrogen-bond donor strength influences
the energy of absorption transition more than the nonspecific
dielectric interactions, whereas the solvent’s hydrogen-bond
acceptor strength, exhibits a small influence only. The posi-

K. Guzow et al. / Spectrochimica Acta Part A 61 (2005) 1133–1140
1137
Table 1
Photophysical properties of 1b measured in 25 solvents of different polarity at room temperaturea
No.
Solvent
ν˜A (cm−1
)
ν˜F (cm−1
)
Quantum yield
Lifetime and pre-exponential factor
Quality of fit (X2R)
b
τ (ns)
ꢂτꢃ (ns)
α
1
2
3
4
5
6
MeOH
EtOH
1-BuOH
i-PrOH
1-PeOH
HCONH2
27393.5
27300.0
27240.5
27363.5
27192.4
27214.6
20576.1
21019.4
22002.2
21941.8
21990.1
20345.9
0.42
0.61
0.87
0.77
0.76
0.35
4.04
4.49
4.01
3.94
3.96
3.59
–
–
–
–
–
–
1.000
1.000
1.000
1.000
1.000
1.000
1.11
1.07
1.06
1.12
1.09
1.05
7
1,2-Pr(OH)2
27148.1
20714.7
0.63
4.53
4.68
0.51
1.000
0.512
0.488
3.04
1.05
4.43
8
9
10
11
TFE
27314.9
27247.9
27743.1
27739.3
20151.1
19881.0
21921.8
21470.7
0.05
0.003
0.68
0.62
–
–
3.96
4.60
–
–
–
–
–
–
–
–
1.06
1.03
Water
Acetone
MeCN
1.000
1.000
12
C6H5CN
27151.8
21680.2
0.26
3.52
2.87
5.85
1.000
0.871
0.129
4.49
1.11
3.56
13
14
15
16
17
18
19
20
21
22
23
CH2Cl2
CHCl3
Cx
27309.3
27081.9
27647.2
27793.2
27674.0
27658.7
27739.2
27559.6
27548.2
27472.5
27666.3
22160.7
22471.9
25300.4
25461.5
25316.4
25380.7
22284.1
22369.4
21656.7
21074.8
21097.0
0.87
0.88
0.87
0.98
0.94
0.99
0.80
0.77
0.72
0.69
0.67
3.57
3.60
1.29
1.18
1.26
1.23
3.00
2.95
4.57
4.93
5.19
–
–
–
–
–
–
–
–
–
–
–
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.04
1.07
1.09
1.14
1.03
1.15
1.04
1.17
1.15
1.03
1.11
Hex
MeCx
iso-Oct
AcOEt
THF
DMF
DMSO
PC
24
di-MeOEt
27582.4
22038.6
0.44
2.52
2.62
0.87
1.000
0.700
0.300
1.57
2.40
2.80
1.02
2.70
25
1,4-Dioxane
27540.6
22896.4
0.47
2.86
1.16
3.07
1.000
0.333
0.667
1.02
a
Solvent abbreviations: MeOH, methanol; EtOH, ethanol; 1-BuOH, 1-butanol; i-PrOH, 2-propanol; 1-PeOH, 1-pentanol; 1,2-Pr(OH)2, propane-1,2-diol;
TFE, 2,2,2-trifluoroethanol; MeCN, acetonitrile; Cx, cyclohexane; Hex, n-hexane; MeCx, methylcyclohexane; iso-Oct, iso-octane; AcOEt, ethyl acetate; THF,
tetrahydrofuran; DMF, N,N-dimethylformamide; DMSO, dimethylsulphoxide; PC, propylene carbonate; di-MeOEt, 1,2-dimethoxyethane.
2
ꢈ
α τ
i
b
i
ꢂτꢃ =
.
i
i
α τ
i
lifetimes are observed in nonpolar solvents (1.18 ns in
n-hexane), whereas the longest was found for propylene
carbonate (5.2 ns) (because of the very low fluorescence in-
tensity, the fluorescence lifetime in water and TFE could not
be measured using our apparatus). The fluorescence intensity
decay of 1b in all solvents used, except 1,4-dioxane, TFE,
benzonitrile and propane-1,2-diol, are mono-exponential.
Heterogenous fluorescence decay indicates some specific
interactions with the solute, therefore, the values measured in
these solvents were not included in the correlation analysis.
The fluorescence quantum yield as well as the lifetime were
correlated with the solvent parameters using Eq. (1). For the
fluorescence quantum yield, we obtained:
and for the fluorescence lifetimes:
τ(ns) = τ0 + 3.25( 0.60)π∗ + 0.55( 0.42)α
+0.98( 0.48)β,
r = 0.897.
(21)
In both cases, the nonspecific dielectric interactions
(measured by π∗) have a stronger influence on the measured
quantities than the two remaining parameters. However, the
hydrogen-bond acceptor strength influences the fluorescence
lifetime more than the hydrogen-bond donor strength of the
solvents. The opposite is true for the fluorescence quantum
yield (because of the very small C(β) value its influence can
be neglected). The fluorescence quantum yields were also
correlated, separately for the HBD and non-HBD solvents,
with the ETN parameter and good correlations were obtained
φ = φ0 − 0.40( 0.18)π∗ − 0.30( 0.11)α + 0.08( 0.15)β,
r = 0.731
(20)

1138
K. Guzow et al. / Spectrochimica Acta Part A 61 (2005) 1133–1140
Table 2
Solvent parameters applied in the calculations and values of solvent polarity function.
No.
Solvent
ETN a
π∗b
α2
β2
f(␧r,n)
g(n)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
a
MeOH
EtOH
0.76
0.6
0.98
0.85
0.79
0.77
–
0.62
0.77
0.88
0.95
–
–
–
–
0.18
0.48
0.31
0.41
0
0
0
–
–
0.8538
0.8135
0.7512
0.7786
0.7032
0.8948
0.8216
0.8369
0.9127
0.7904
0.8625
0.7607
0.5929
0.3650
−0.0033
−0.0015
−0.0111
−0.0228
0.4941
0.5500
0.8362
0.8405
0.8796
0.2804
0.0470
0.2252
0.2447
0.2698
0.2569
0.2785
0.3032
0.2930
0.2045
0.2269
0.2443
0.2356
0.3558
0.2877
0.3002
0.2893
0.2535
0.2888
0.2661
0.2535
0.2747
0.2924
0.3238
0.2857
0.2498
0.2864
0.65
0.58
0.55
0.56
0.77
0.75
0.89
1.00
0.35
0.47
0.33
0.30
0.25
−0.003
0.00
−0.003
0.02
0.22
0.19
0.38
0.44
0.47
0.22
0.16
0.54
0.46
0.47
–
0.85
–
0.56
1.09
0.68
0.75
0.9
0.802
0.76
0
1-BuOH
i-PrOH
1-PeOH
HCONH2
1,2-Pr(OH)2
TFE
0.77
–
1.66
1.1
0.1
0.19
0
0.3
0.44
0
–
–
–
0
0
0
0
0
0
0
Water
Acetone
MeCN
C6H5CN
CH2Cl2
CHCl3
Cx
Hex
MeCx
iso-Oct
AcOEt
THF
DMF
DMSO
PC
di-MeOEt
1,4-Dioxane
−0.08
–
–
–
0.55
0.58
0.88
1
0.81
0.53
0.55
0.45
0.55
0.69
0.76
0.4
0.41
0.37
Data from [4].
Data from [29].
b
(r = 0.891, Np = 13, for the non-HBD solvents, and r =
0.954, Np = 9 for the HBD solvents) (Fig. 3). Also, a very
good correlation of the fluorescence lifetimes with ETN
for non-HBD solvents was obtained (r = 0.992, Np = 13).
For HBD solvents a cluster of points was obtained and a
correlation with the solvent polarity parameter ETN was
not found (Fig. 4). A good linear correlation between the
logarithm of the fluorescence lifetime of a porphyrin deriva-
tive with the ET (30) parameter was also obtained by Drain
et al. [27].
According to Eq. (13), the linear correlation of the Stokes
shift with the solvent polarity parameter ETN allows to cal-
culate the excited-state dipole moment when the Onsager’s
cavity radius and the ground-state dipole moment are known.
The following correlation of (ν˜A − ν˜F) with ETN for non-HBD
Fig. 3. Linear correlation of the fluorescence quantum yield of 1b with the
solvent polarity parameter ETN . The numbers refer to the solvents of Table 1.
Fig. 4. Linear correlation of the fluorescence lifetimes of 1b with the solvent
polarity parameter ETN . The numbers refer to the solvents of Table 1.

K. Guzow et al. / Spectrochimica Acta Part A 61 (2005) 1133–1140
1139
4. Conclusion
3-[2-(4-Diphenylaminophenyl)benzoxazol-5-yl]alanine
methyl ester (1b) and its derivative 1a, because of large
solvent-dependent changes of the dipole moment on ex-
citation as well as the fluorescence quantum yields and
fluorescence lifetime, seems to be a useful fluorescence
probe. Moreover, the presence of the ␣-amino acid moiety
allows to incorporate it into a peptide/protein chain, thus it
could be used in studies of the microenvironment polarity of
biologically active compounds.
Acknowledgement
This work was financially supported by a grant
1005/T09/2003/24 from the Ministry of Science and Infor-
matics (Poland).
Fig. 5. Linear correlation of the Stokes shift (ν˜A − ν˜F) of 1b with the solvent
polarity parameter ETN . The numbers refer to the solvents of Table 1.
solvents was obtained:
References
(ν˜A − ν˜F) = 8940( 720)ETN + 2540( 210),
r = 0.966, Np = 13, (22)
[1] P. Suppan, N. Ghoneim, Solvatochromism, The Royal Society of
Chemistry, Cambridge, U.K., 1997.
[2] M.J. Kamlet, Prog. Phys. Org. Chem. 13 (1982) 485.
[3] C. Reichardt, Angew. Chem. 91 (1979) 119;
C. Reichardt, Angew. Chem. Int. Ed. 18 (1979) 98.
[4] C. Reichardt, Chem. Rev. 94 (1994) 2319.
[5] M.J. Kamlet, J.-L.M. Abboud, R.W. Taft, J. Am. Chem. Soc. 99
(1977) 6027.
[6] M.J. Kamlet, J.-L.M. Abboud, M.H. Abraham, R.W. Taft, J. Org.
Chem. 48 (1983) 2877.
[7] V. Nagarajan, A.M. Brearley, I.-J. Kang, P.F. Barbara, J. Phys. Chem.
86 (1987) 3183.
whereas for HBD solvents:
(ν˜A − ν˜F) = 5090( 690)ETN + 2630( 510),
r = 0.941, Np = 9
(23)
was found (Fig. 5). Based on the calculated values of On-
sager’s cavity radius for 1b (a = 6.4 A) and a ground-state
dipole moment of µg = 2.45 D, using a semiempirical PM3
method (MOPAC 2002 [28]), and the appropriate data for
˚
[8] K.M. Solntsev, D. Huppert, N. Agman, J. Phys. Chem. A 102 (1998)
9599.
˚
the betaine dye (ꢀµ = 9 D, a = 6.2 A), the following changes
[9] B. Ren, F. Gao, Z. Tong, Y. Yan, Chem. Phys. Lett. 307 (1999) 55.
[10] T.A. Fayed, J. Photochem. Photobiol. A: Chem. 121 (1999) 17.
[11] L. Bilot, A. Kawski, Z. Naturforsch. 17a (1962) 621.
[12] A. Kawski, in: J.F. Rabek (Ed.), Progress in Photochemistry and
Photophysics, 5, CRC Press, Boca Raton, USA, 1992, pp. 1–47
(Chapter 1).
[13] N.G. Bakshiev, Opt. Spektrosk. 16 (1964) 821.
[14] A. Kawski, B. Kuklin´ski, P. Bojarski, H. Diehl, Z. Naturforsch. 55a
(2000) 817.
[15] M. Ravi, A. Samanta, T.P. Radhakrishnan, J. Phys. Chem. 98 (1994)
9133.
[16] S. Kumar, V.C. Rao, R.G. Rastogi, Spectrochim. Acta, Part A 57
(2001) 41.
[17] J.R. Lakowicz, Principles of Fluorescence Spectroscopy, second ed.,
Kluwer Academic/Plenum Publishers, New York, 1999.
[18] M. Sun, P.-S. Song, Photochem. Photobiol. 25 (1977) 3.
[19] J.R. Lombardi, J. Phys. Chem. A 103 (1999) 6335.
[20] M.R. Eftink, C.A. Ghiron, Biochemistry 15 (1976) 672.
[21] K. Guzow, M. Szabelski, J. Malicka, W. Wiczk, Helv. Chim. Acta
84 (2001) 1086.
[22] A. Rzeska, J. Malicka, K. Guzow, M. Szabelski, W. Wiczk, J. Pho-
tochem,. Photobiol. A: Chem. 146 (2001) 9.
[23] K. Guzow, M. Szabelski, J. Malicka, J. Karolczak, W. Wiczk, Tetra-
hedron 58 (2002) 2201.
of dipole moment were obtained: ꢀµ = 8.4 D for non-HBD
solvents and ꢀµ =6.3 D for HBD solvents. Thus, a relatively
large charge transfer takes place during the excitation of 1b.
The excited-state dipole moment can be also calculated using
the correlation of the Stokes shifts with the solvent polarity
function of Eqs. (2) and (3). The slope obtained from the lin-
ear regression of (ν˜A − ν˜F) as a function of f(εr,n) (Eq. (2))
is equal to m1 =4540( 240), r = 0.963, Np = 13, and from
the linear regression of (ν˜A + ν˜F) as a function of f(εr,n) +
2g(n) (Eq. (3)) m2 = 4660( 400), (r = 0.963, Np = 13). Us-
ing Eq. (8), the theoretically calculated ground-state dipole
moment, Onsager’s cavity radius, Planck’s constant, and the
velocity of the light, the excited-state dipole moment of 1b
was determined to µe =11.3 D, which is in a good agreement
with the value obtained from Eq. (13). Using Eq. (9) and the
values of m1 and m2 as well as the ground- and excited-state
dipole moment values, the angle between the ground- and
excited-state dipole moment can be calculated. For the com-
pound studied (1b), its value is equal to ψ ≈ 74◦ (cos ψ =
0.273), thus, the dipole moments of the ground and excited
state are almost mutually perpendicular.
[24] K. Guzow, K. Mazurkiewicz, M. Szabelski, R. Ganzynkowicz, J.
Karolczak, W. Wiczk, Chem. Phys. 295 (2003) 119.

1140
K. Guzow et al. / Spectrochimica Acta Part A 61 (2005) 1133–1140
[25] E. Wu¨nsch, in: E. Mu¨ller (Ed.), Synthese von Peptiden, in Houben-
Weyl, Methoden der organischen Chemie, 15, Georg Thieme Verlag,
Stuttgart, 1974, p. 317.
[27] C.M. Drain, S. Gentemann, J.A. Roberts, N.Y. Nelson, C.J. Med-
forth, S. Jia, M.C. Simpson, K.M. Smith, J. Fajer, J.A. Shelnutt, D.
Holten, J. Am. Chem. Soc. 120 (1998) 3781.
[26] J.M. Stewart, J.D. Young, Solid Phase Peptide Synthesis,
second ed., Pierce Chemical Company, Rockford, IL, 1984,
p. 63.
[28] J.J.P.Steward, MOPAC 2002, Fujitsu Limited, 2001.
[29] A. de Juan, G. Fonrodona, E. Casassas, Trends Anal. Chem. 16
(1997) 52.