Merocyanine dyes containing an isoxazolone nucleus: Synthesis, X-ray crystal structures, spectroscopic properties and DFT studies

Article abstract of DOI:10.1016/j.dyepig.2011.10.003

Two merocyanine dyes, each with an isoxazolone nucleus, were synthesized and characterized by NMR, IR, UV-Vis spectroscopy, elemental analyses and single X-ray diffraction. Crystallographic data revealed that one dye crystallized in an orthorhombic system

Full text of DOI:10.1016/j.dyepig.2011.10.003

Dyes and Pigments 93 (2012) 1408e1415
Contents lists available at SciVerse ScienceDirect
Dyes and Pigments
journal homepage: www.elsevier.com/locate/dyepig
Merocyanine dyes containing an isoxazolone nucleus: Synthesis, X-ray crystal
structures, spectroscopic properties and DFT studies
,
a
a
a
b,
Xiang-Han Zhang a , Yong-Hua Zhan , Dan Chen , Fu Wang , Lan-Ying Wang
*
**
a Life Sciences Research Center, School of Life Sciences and Technology, Xidian University, Xi’an, Shaanxi 710071, PR China
b Key Laboratory of Synthetic and Natural Functional Molecule Chemistry (Ministry of Education), College of Chemistry and Materials Science, Northwest University,
Xi’an, Shaanxi 710069, PR China
a r t i c l e i n f o
a b s t r a c t
Article history:
Two merocyanine dyes, each with an isoxazolone nucleus, were synthesized and characterized by NMR,
IR, UVeVis spectroscopy, elemental analyses and single X-ray diffraction. Crystallographic data revealed
that one dye crystallized in an orthorhombic system, Cmca space group, while the other dye crystallized
in a monoclinic system, P21/n space group. In the case of both dyes, the molecules adopted the most
Received 3 August 2011
Received in revised form
26 September 2011
Accepted 5 October 2011
Available online 15 October 2011
stable ketonic tautomeric form, and the crystal packing was stabilized by hydrogen bonds and
pep
stacking interactions. The absorption spectra of the dyes in various solvents of different polarities were
studied at room temperature. The equilibrium geometries, vibrational data and electronic spectra of the
compounds were also investigated by the B3LYP method. Importantly, the predicted results fit to the
experimental data, showing that this quantum chemical technology is an efficient approach to predict
the properties of merocyanine dyes.
Keywords:
Merocyanine dyes
X-ray diffraction
H-bond
Ó 2011 Elsevier Ltd. All rights reserved.
Solvent effect
Vibrational spectroscopy
DFT calculations
1. Introduction
interactions. A number of details of the X-ray crystallography
leading to the molecular conformation or intermolecular interac-
Isoxazolone functional groups have attracted considerable
attention for use in the design of biologically active molecules and
advanced organic materials. Due to their excellent biological
activities and unique pharmacophores, isoxazolone derivatives
have been widely used as anti-androgens and inhibitors in the
therapy of diverse diseases [1e5], as well as in herbicides for
agriculture [6,7]. Furthermore, an isoxazolone nucleus is a good
proaromatic acceptor, when linked to aromatic donors, for conju-
tions of cyanine and merocyanine dyes remain useful and are
a view to their practical application [11]. Interestingly, isoxazolone
derivatives can exist in three tautomeric forms: the OH form, the
C]O form and the NH form. A mixture of these is always obtained,
but the ketonic forms are generally predominant [12]. Based on the
X-ray crystal results in this study, theoretical chemistry, like density
functional theory (DFT), is valuable [13e15] in validating the
experimental data for the tautomerism of merocyanine with an
isoxazolone nucleus.
In order to obtain greater insight into the multicolor of the dyes,
we mainly focused on the effect of the solvent on the electronic
absorption of the dyes by means of experimental and theoretical
methods. The TDDFT level with vertical SCRF approach reproduced
the solvatochromism of the dyes, which depended on the
surrounding solvent molecules and a polarizable continuum. It has
been shown that theoretical calculations on the chromophoric
properties of organic dyes are indispensable tools for practical
applications and should help in predicting and explaining the
structureeproperty relationships of the dyes [16e19]. Therefore,
resonance frequency calculations were carried out for preliminary
studies of the IR spectra of the dyes. For experimental vibrational
spectroscopy that lacks a direct relation between the spectra and
gated donor-acceptor (De
their good molar extinction coefficients, tunable absorption
spectra, and large first molecular hyperpolarizabilities ),
peA) merocyanine dyes [8,9]. Because of
(b
merocyanine dyes with an isoxazolone nucleus have been used for
optical recording and nonlinear optical research [10].
In this paper, two merocyanine dyes, each with an isoxazolone
nucleus, were synthesized. The X-ray crystal structures of the two
dyes correlated with the IR spectra, specifically demonstrating the
molecular conformations and intra- and intermolecular
* Corresponding author. Tel.: þ86 29 81891070; fax: þ86 29 81891060.
** Corresponding author.
E-mail addresses: zhangxianghan@life.xidian.edu.cn (X.-H. Zhang), wanglany@
nwu.edu.cn (L.-Y. Wang).
0143-7208/$ e see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.dyepig.2011.10.003

X.-H. Zhang et al. / Dyes and Pigments 93 (2012) 1408e1415
1409
CH3
the structure, theoretical predictions of vibrational spectra are of
OHC
+
R
R
practical value for the identification of known and unknown
compounds [20e22]. Thus, theoretical predictions, preferably those
used with DFT methods, are advantageous. In this work, we fol-
lowed this approach, and the detailed results and discussion are
elaborated in the following sections.
O
O
N
O
Pyridine, EtOH
Reflux
CH3CCH2COC2H5
NH2OH
+
N
H
N
H
O
1: R = H
1: R = H
2: R = OCH2Ph
2: R = OCH2Ph
Scheme 1. Synthesis of merocyanine dyes (1) and (2).
2. Experimental
sulphoxide, methanol, ethanol and ethyl acetate as solvents, each at
a concentration of 10ꢁ5 M, in 1 cm quartz cells.
2.1. Synthesis
Crystals with approximate dimensions of 0.38
ꢂ
0.25
ꢂ
The general route for the synthesis of dyes (1) and (2) is shown
in Scheme 1. To a stirred mixture of hydroxylamine hydrochloride
(4 mmol), pyridine (4 mmol) in EtOH (10 mL), ethyl acetoacetate
(4 mmol) was added via a dropping funnel over a 30 min time span.
Then, either 1H-indole-3-carbaldehyde (3.5 mmol) or 5-benzyloxy-
1H-indole-3-carbaldehyde (3.5 mmol) was added, and the mixture
was stirred under reflux for 2 h [23]. After cooling, the precipitated
crude dyes were filtered and then washed thoroughly with ether
and ethanol. The solid was purified by recrystallization using
EtOHeMeOH, resulting in pure dyes (1) and (2) (yields of 67% and
60%, respectively). Single crystals of dyes (1) and (2) were grown
from an EtOHeMeOH mixture (1:2) or EtOHeacetone mixture
(1:1), respectively, for X-ray diffraction experiments.
0.13 mm3 for dye (1) and 0.31 ꢂ 0.26 ꢂ 0.13 mm3 for dye (2) were
selected for data collection. The X-ray diffraction data were
collected using a Bruker SMART APEX II CCD X-ray crystallography
system that was equipped with graphite monochromated Mo K
a
radiation ( e2 scan technique at room
l
¼ 0.71073 Å) by using
u
q
temperature. The structures were solved by direct methods and
refined using the full-matrix least-squares method on F2 with
anisotropic thermal parameters for all non-hydrogen atoms.
Hydrogen atoms were generated geometrically. Crystallographic
computing was performed with the SHELXTL [24] and PLATON [25]
programs. The crystal data, including details concerning data
collection and structure refinement for dyes (1) and (2), are
summarized in Table 1. Parameters in CIF format are available as
Electronic Supplementary Information from Cambridge Crystallo-
graphic Data Center (CCDC 724598, 729689).
Dye (1) (4-[(1H-indole-3-yl)-methylene]-3-methyl-isoxazole-5-
one): Orange needle crystal, yield 67%, m.p.: 239e241 ꢀC. 1H NMR
(DMSO-d6, 400 MHz)
d (ppm): 2.35 (s, 3H, CH3), 7.32e7.36 (m, 2H,
ArH), 7.60 (s, 1H, eCH]), 8.16e8.21 (m, 2H, ArH,), 9.52 (s, 1H,
2.3. Computational details
J ¼ 2.4 Hz, pyrrole-H), 12.81 (s, 1H, NH). 13C NMR (DMSO-d6,
400 MHz) d (ppm): 10.7, 108.3, 112.2, 112.7, 118.4, 122.1, 123.5, 127.5,
The single molecular structures of dyes (1) and (2) were opti-
mized at the DFT level using 6-311G** and 6-31G* basis sets. DFT
calculations were carried out using a hybrid exchange-correlation
functional termed B3LYP, which refers to Becke’s three-parameter
exchange functional (B3) along with the nonlocal correlation
functional of Lee, Yang, and Parr (LYP). The PCM model [26] at the
B3LYP/6-311G** level was used to study the solvent polarity effect
135.9, 138.0, 140.0, 161.2, 169.9. IR (KBr)
3125, 3063 (m, ]CeH), 2939 (m, yCeH), 1701 (s, yC]O), 1599 (s, yC]
C), 1582, 1452 (vs, yC]C yC]N), 1377, 1348, 1298, 1219 (yC]C), 1109,
995 (dCH), 878 (yNeO), 812, 756 (m, ]CH). Anal. Calcd. for
y: 3458 (m. b, yNeH), 3157,
y
d
C13H10N2O2 ¼ 226.23: C, 69.02; H, 4.01; N,12.37; Found: C, 69.42; H,
4.26; N, 12.38. UVeVis (MeOH) lmax: 426.4 nm.
Dye
methyl-isoxazole-5-one): Red columnar crystal, yield 60%, m.p.:
240e242 ꢀC. 1H NMR (DMSO-d6, 400 MHz)
(ppm): 2.36 (s, 3H,
(2)
(4-[(5-Benzyloxy-1H-indole-3-yl)-methylene]-3-
Table 1
d
Crystal data of dyes (1) and (2).
CH3), 5.19 (s, 2H, CH2), 7.02 (dd, 1H, J1 ¼ 2.0 Hz, J2 ¼ 8.4 Hz, ArH),
7.36 (d, 1H, J ¼ 7.6 Hz, ArH), 7.42 (t, 2H, J1 ¼7.6 Hz, J2 ¼ 8.4 Hz, ArH),
7.52 (t, 3H, J1 ¼ J2 ¼ 8.4 Hz, ArH), 7.89 (d,1H, J ¼ 2.0 Hz, ArH), 8.20 (s,
1H, eCH]), 9.47 (s, 1H, Pyrrole-H), 12.71 (s, 1H, NH). 13C NMR
Identification code
(1)
(2)
Empirical formula
Formula weight
Temperature (K)
Wavelength (Å)
Crystal system
C13H10N2O2
226.23
298 (2) K
0.71073
Orthorhombic
Cmca
C20H16N2O3
332.35
298 (2) K
0.71073
Monoclinic
P21/n
(DMSO-d6, 400 MHz)
113.5, 127.4, 128.0, 128.7,130.8, 136.7, 138.4, 140.0, 154.9, 161.2, 170.1.
IR (KBr) : 3454 (m. b, yNeH), 3119, 3094, 3063 (m, ]CeH), 2922 (m,
d (ppm): 10.9, 69.6, 102.5, 107.6, 112.4, 113.3,
y
y
Space group
Unit cell dimensions (Å)
a ¼ 6.6348 (15)
b ¼ 19.189 (4)
c ¼ 18.074 (4)
a ¼ 7.3056 (19)
b ¼ 19.681 (5)
c ¼ 11.445 (3)
yCeH), 1703 (s, yC]O), 1595 (s, yC]C), 1580, 1439 (s, yC]C yC]N), 1375,
1337, 1288, 1200, 1132(yC]C), 1107, 990 (dCH), 874 (yNeO), 789, 727
(m,
d
]CH). Anal. Calcd. for C20H16N2O3 ¼ 332.35: C, 72.28; H, 4.85;
b
¼ 90ꢀ
b
¼ 105.063 (4)ꢀ
Volume (Å3)
2301.1 (9)
1589.0 (7)
N, 8.43; Found: C, 72.42; H, 4.62; N, 8.39. UVeVis (MeOH) lmax
:
Z
8
4
433.6 nm.
Calculated density (g/cm3)
Absorption coefficient (mmꢁ1
F (0 0 0)
1.323
0.091
968
1.389
0.095
696
)
2.2. Measurements
Crystal size (mm3)
q
0.38 ꢂ 0.25 ꢂ 0.13
0.31 ꢂ 0.26 ꢂ 0.13
Range for data collection
2.12e25.10
2.07e25.10
The starting materials were commercially available and used as
received without additional purification. The solvents used were of
analytical grade. Melting points were taken using an XT-4 micro-
melting apparatus and were uncorrected. Elemental analyses were
performed using a Vario EL-III instrument, while IR spectra, in
cmꢁ1, were recorded on a Bruker Equiox-55 spectrometer. 1H NMR
and 13C NMR spectra were recorded at 400 MHz using a Varian
Inova-400 spectrometer, and chemical shifts were reported relative
to the internal Me4Si. The absorption spectra were recorded on
a Shimadzu UV-2450 UVeVis spectrometer using water, dimethyl
Limiting indices h, k, l
ꢁ6/7, ꢁ22/22, ꢁ21/16 ꢁ8/5, e23/23, e13/7
Reflections collected/unique/Rint 5619/1121/0.0270
6425/2692/0.0338
95.4%
Completeness to
q
¼ 25.10ꢀ
100%
Data/restraints/parameters
1121/0/114
1.072
2692/0/227
1.083
Goodness-of-fit on F2
Final R indices [I > 2
s
(I)]
R1 ¼ 0.0388,
wR2 ¼ 0.1115
R1 ¼ 0.0486,
wR2 ¼ 0.1187
0.142 and ꢁ0.158
R1 ¼ 0.0618,
wR2 ¼ 0.1635
R1 ¼ 0.0835,
wR2 ¼ 0.1830
0.437 and ꢁ0.230
R indices (all data)
Largest diff. peak and
hole (e Åꢁ3
)

1410
X.-H. Zhang et al. / Dyes and Pigments 93 (2012) 1408e1415
CH
3
CH
3
CH
3
R
R
R
N
NH
N
O
O
O
N
N
O
NH form
N
O
HO
OH form
H
C=O form
Scheme 2. The equilibrium of three tautomeric forms.
on the absorption spectra. The analytic frequency calculations were
performed, and the absence of imaginary frequencies confirmed
that the optimized structures were at an energy minimum. The IR
spectra were calculated at the B3LYP/6-31G* level, where a modi-
fication of the calculated data using the empirical scaling factor
0.9613 [27] was used to achieve better correspondence between
the experimental and theoretical values. All calculations reported
in this work were performed using the GAUSSIAN 03 program [28].
Fig. 2. ORTEP diagram of (2) with ellipsoids drawn at the 50% probability level.
3. Results and discussion
of all the potential tautomers, and these results were also consis-
tent with the crystal structure analysis.
3.1. Synthesis and characterization
3.2. Crystal structure
Using a three-component, one-pot method, dyes (1) and (2)
were successfully synthesized and high yields were obtained (60%
and 67%, respectively). The products were purified by recrystalli-
zation using ethanol or methanol. The choice of the solvent had
a crucial effect on single crystal formation. A single solvent (e.g.,
ethanol, methanol or acetone) could not give suitable single crys-
tals, but a mixture of solvents (EtOHeMeOH or EtOHeacetone)
gave suitable crystals. The structures of dyes (1) and (2) were
confirmed by elemental analyses, UVeVis absorption spectra, IR, 1H
NMR and 13C NMR data. In all cases, the IR spectra showed typical
The molecular structures and atom numbering of (1) and (2) are
shown in Figs. 1 and 2, and the key geometric parameters of dyes
(1) and (2), based on X-ray diffraction and DFT calculations, are
presented in Fig. 3.
For dyes (1) and (2), it was found that the carbon-carbon bond
lengths of the molecular skeletons were intermediate between
typical CeC single (1.54 Å) and C]C double (1.34 Å) bonds and that
the carbondnitrogen bond lengths were also intermediate
between typical CeN single (1.47 Å) and C]N double (1.27 Å)
bonds. All bond angles were close to 120ꢀ in the benzene rings and
aromatic
absorption
(
y]CeH
cmꢁ1
yC]C , and d]CH
1599e1595 cmꢁ1
:
3157e3063
,
yCeH
:
2939e2922 cmꢁ1
,
:
:
812e727 cmꢁ1), resonance conjugated unsaturated stretching
modes (yC]C,
yC]N, 1582e1132 cmꢁ1), the absorption of an indole
ring (yNeH 3458e3454 cmꢁ1), and classical carbonyl stretching
vibration (yC]O, 1703e1701 cmꢁ1). A merocyanine dye with an
isoxazolone nucleus could have three tautomers: an OH form, a C]
O form or an NH form, as Scheme 2 shows. The IR spectra indicated
that the dyes were predominantly in the ketonic form. Moreover,
calculations were performed at the B3LYP/6-311G** level for all
possible structures of dyes (1) and (2) in Scheme 2. The single point
energies
of
the
C]O,
OH
and
NH
forms
were ꢁ761.512758, ꢁ761.4747822 and ꢁ761.104039 hartrees for
dye (1), and ꢁ1107.170345, ꢁ1107.133618 and ꢁ1107.1364861 har-
trees for dye (2). Obviously, the C]O form was the most stable form
Fig. 3. Selected bond lengths (Å) and torsion angles (ꢀ) of dye (1) and (2) (X-ray
diffraction data are in bold type and calculated results by B3LYP/6-311G** are in light
italics).
Fig. 1. ORTEP diagram of (1) with ellipsoids drawn at the 50% probability level.

X.-H. Zhang et al. / Dyes and Pigments 93 (2012) 1408e1415
1411
Table 2
The geometry of DeH/A hydrogen bonds of dye (1) and (2).
Dyes DeH/A
d(DeH)/Å D(H/A)/Å d(D/H)/Å :ðDHAÞ=ꢀ
(1)
N(1)eH(1)/N(2) 0.86
C(4)eH(4)/O(1) 0.93
N(1)eH(1)/O(2) 0.86
2.04
2.46
1.99
2.894(2)a
3.299(3)b
2.842c
171
150
168
(2)
a
Symmetry code: ꢁx, 1/2 ꢁ y, z þ 1/2.
Symmetry code: ꢁx, ꢁ1/2 þ y, 1/2 ꢁ z.
b
c
Symmetry code: 1/2 þ x, ꢁ1/2 ꢁ y, 1/2 þ z.
close to 108ꢀ in the five-membered rings. The
p electrons in the
whole dye molecule were delocalized, and methyl groups in the
isoxazolone rings were involved in the molecular conjugate
systems. For dye (1), all of the torsion angles were 180ꢀ or 0ꢀ, which
indicates that the molecular framework of (1) is planar. For dye (2),
the C(8)eO(1) (1.377 Å), C(7)eO(1) (1.422 Å) and C(6)eC(7)
(1.507 Å) bonds were shorter than conventional CeO single bonds
(1.43 Å) or CeC single bonds (1.54 Å), which indicates that the
benzyloxy group and the indole ring has pe
p conjugation. The
benzyloxy group was distorted both from planarity and coplanarity
within the indole ring, as indicated by the corresponding torsion
angles of C(7)eO(1)eC(8)eC(9) and C(6)eC(7)eO(1)eC(8)
(ꢁ3.9(4)ꢀ and ꢁ173.2ꢀ, respectively). The torsion angles of C(8)e
C(7)eC(9)eC(10) (0ꢀ) for (1) and C(15)eC(16)eC(17)eC(18) (ꢁ1.9ꢀ)
for (2) showed that the molecular framework of (1) was planar,
while the indole and isoxazolone groups were slightly distorted
from planarity in (2), and this was induced by the benzyloxy group.
It could also be seen that the computed bond lengths, bond angles
and torsion angles were in good agreement with the experimental
data, as shown in Fig. 3. With respect to the experimental values,
Fig. 4. Crystal packing of (1) along the a axis (a) and b axis (b).
Fig. 6. Molecular electrostatic potential map of (1) and (2) calculated at the B3LYP/6-
Fig. 5. Crystal packing of (2) along the c axis (a) and b axis (b).
311G** level.

1412
X.-H. Zhang et al. / Dyes and Pigments 93 (2012) 1408e1415
Table 3
The physical constants and the spectral data of dyes (1) and (2) in different solvents.
Solvents
Refractive
Dielectric
(1)
(2)
index/n
constant/ε
lmax/nm
ε ꢂ 10ꢁ4/L
lꢁ1 ꢂ 103
lmax/nm
ε ꢂ 10ꢁ4/L
lꢁ1ꢂ103
molꢁ1 cmꢁ1
(nmꢁ1
)
molꢁ1 cmꢁ1
(nmꢁ1
)
Ethyl acetate
Ethanol
Methanol
DMSO
1.372
1.361
1.329
1.477
1.333
6.08
25.30
33.10
47.24
80.18
412.2
426.2
426.4
432.0
434.0
2.1
3.1
2.7
2.7
2.7
2.43
2.35
2.35
2.31
2.31
418.4
433.2
433.6
436.4
438.8
2.8
2.8
2.6
2.5
1.8
2.39
2.31
2.31
2.29
2.28
Water
the B3LYP/6-311G** results deviated in the range from 0 Å to
0.027 Å for the bond lengths of (1) and from 0 Å to 0.025 Å for the
bond lengths and from 1.4ꢀ to 4.0ꢀ for the torsion angles of (2). It
appears that the B3LYP/6-311G** method correctly reproduces the
signs of the bond lengths, bond angles and torsion angles, which is
useful for investigating the characteristics of some structurally-
related dyes.
1 ꢁ z]. It was also found that N(1)eH(1)/O(2) hydrogen bonds
linked the molecules, forming parallel, infinite sheets along the
b axis. The strong NeH/O interaction between H(1) and O(2) was
found to have a distance of 1.99 Å (Table 2), which was effective in
stabilizing the crystal structure.
The molecular electrostatic potential (MEP) was investigated
using B3LYP/6-311G**-optimized geometries, as shown in Fig. 6.
This method gives information about the proper region by which
compounds have intermolecular interactions between their units.
In MEP analysis, the negative (red and yellow) and the positive
(blue) regions correspond to nucleophilic and electrophilic
centers, respectively. The positive (blue) regions of dyes (1) and
(2) were localized on H1 and had electrophilic reactivity.
Conversely, the negative (red and yellow) regions were observed
at the isoxazolone ring around O1, O2 and N2 for (1) and around
O2, O3 and N2 for (2) and had nucleophilic reactivity. Hence, the
MEP maps clearly showed that the intermolecular hydrogen
bonds in the crystal structures were between the positive and
negative regions, indicating possible sites for intermolecular
interactions.
3.3. Crystal packing and MEP analysis
The crystal packing of dyes (1) and (2) along their axes is pre-
sented in Figs. 4 and 5. Dyes (1) and (2) crystallized in the ortho-
rhombic Cmca and monoclinic P21/n space groups, respectively. In
the packing of (1), there were two offset
p/p aromatic, intermo-
lecular interactions that occurred between the pyrrole rings and
isoxazolone rings [Cg1/Cg1 ¼ 3.565(9) Å, symmetry code: ꢁ1/
2 þ x,1/2 ꢁ y, ꢁz; Cg2/Cg2 ¼ 3.489(8) Å, symmetry code: ꢁ1/2 ꢁ x,
y, 1/2 ꢁ z]. The molecule also contained hydrogen bond acceptor
and donor sites, and thus, the classical strong hydrogen bonds
N(1)eH(1)/N(2) [symmetry code: ꢁx, 1/2 ꢁ y, z þ 1/2] and weak
hydrogen bonds C(4)eH(4)/O(1) [symmetry code: ꢁx, ꢁ1/2 þ y, 1/
2 ꢁ z] existed between molecules (Table 2), which also played
crucial roles in the crystal packing. Two types of hydrogen bonds
linked neighbouring molecules, in head-to-tail fashion, leading to
the formation of hydrogen-bonded chains along the c and b axes,
respectively. However, adjacent molecules of (2) were not stacked
3.4. The UVeVis spectra and the dominant electronic excitations of
dyes
The absorption spectral data of dyes (1) and (2) in different
solvents are listed in Table 3, and a sample UVeVis spectrum (dye
(1)) is shown in Fig. 7. All concentrations followed Beer’s law, and
no changes in the shape of the absorption spectra were observed.
Dyes (1) and (2) absorbed in the range of 412e439 nm and had
through classical
the benzyloxy group, which distorted the molecular planarity. Two
types of weak CeH/ intermolecular interactions existed in (2),
C(7)eH(7B)/ (in the isoxazolone ring) [C/ 3.721 Å,
(in the
¼ 3.517 Å, symmetry code: 1 ꢁ x, ꢁy,
p/p interactions, due to steric hindrance from
p
p
p
¼
molar extinction coefficients of 1.8 ꢂ 104e3.1 ꢂ 104 L molꢁ1 cmꢁ1
.
symmetry code: 2 ꢁ x, ꢁy, 1 ꢁ z] and C(20)eH(20B)/
p
The electronic absorption spectra of the dyes in different solvents
benzene ring of indole) [C/
p
exhibited an intense absorption with high extinction coefficients,
0.4
2.44
Ethyl acetate
Ethanol
Methanol
(1): y= -0.3332x + 2.70 R2= 0.9753
(2): y= -0.3114x + 2.64 R2= 0.9728
2.40
2.36
2.32
2.28
DMSO
Water
0.3
0.2
0.1
0.0
0.8
0.9
1.0
1.1
1.2
350
400
450
500
f (n, ε)
Wavelength/nm
Fig. 8. Correlation of wavenumbers (1/
(2).
l) in various solvents vs. f (n, ε) of dyes (1) and
Fig. 7. UVeVis absorption spectra of (1) in different solvents.

X.-H. Zhang et al. / Dyes and Pigments 93 (2012) 1408e1415
1413
Table 4
Main calculated orbital transitions.
Dyes
Excited state
Gas/TD-B3LYP
Methanol/TD-B3LYP (PCM)
ΔE/ev
fa
3.2803 0.5175
ΔE/ev
ΔE/ev
fa
Transition characterb
Transition characterb
(1)
1
2
3
1
2
3
3.4210
3.6527
3.7969
3.2065
3.5453
3.6665
0.4421
0.0003
0.0291
0.3086
0.2018
0.0003
HOMO / LUMO (98.24%)
HOMO-3 / LUMO (98.64%)
HOMO-1 / LUMO (95.95%)
HOMO / LUMO (76.21%)
HOMO-1 / LUMO (73.68%)
HOMO-4 / LUMO (98.52%)
HOMO / LUMO (98.56%)
HOMO-1 / LUMO (98.08%)
HOMO-3 / LUMO (98.61%)
HOMO / LUMO (83.46%)
HOMO-1 / LUMO (81.21%)
HOMO-5 / LUMO (98.62%)
2.9077
2.8564
3.6199
3.8923
3.0193
3.3953
3.9086
0.0324
0.0003
0.3112
0.2725
0.0003
(2)
a
Oscillator strength.
The proportion of the main transition is given in parentheses.
b
due to
p
e
p
* electronic transitions. The lmax of dye (2) was
for the dyes. The lowest energy absorption of the dyes corre-
sponded to the first dipole-allowed * electronic transition from
4.4e7.2 nm greater than that of dye (1), which was due to
increased conjugation in the dyes. It was also found that the
absorption maxima showed a clear red shift when passing from
the low-polar solvents to the high-polar solvents. It is suggested
pep
HOMO to LUMO orbitals, which could be attributed to intra-
molecular charge transfer (ICT) in the dyes. The sketches of the
molecular orbitals (MOs) for dyes (1) and (2) are illustrated in Fig. 9.
Nearly all of the MOs in the HOMO of (1) and (2) were substantially
localized on the indole and isoxazolone rings, and only a few
contributions from the OCH2 fragment for (2) were observed. The
MOs in the LUMO were mainly distributed at the pyrrole ring,
isoxazolone nucleus and methylidyne fragments. The excitation
involved the intramolecular electron displacement from the indole
groups to the isoxazolone moiety, whereas indole and isoxazolone
groups acted as electron donors and electron acceptors, respec-
tively. The HOMO and LUMO energy values were ꢁ6.7561 eV
and ꢁ2.5288 eV for (1), and ꢁ5.90873 eV and ꢁ2.4891 eV for (2),
respectively, at the DFT level, and the calculated energy gaps were
4.23 eV and 3.42 eV for (1) and (2), respectively. Dye (2), which has
a smaller frontier orbital gap, was more polarizable than dye (1)
[30].
that dyes with
pep* transitions are more polar in their excited
states than in their ground states. In another study, a spectral red
shift in high-polar solvents was concluded to be mainly due to
a greater degree of energy reduction of the excited states than that
of ground states [29]. As a preliminary study of the soluteesolvent
interactions, the Bayliss function [f (n, ε)
¼
(n2
ꢁ
1)/
(2n2 þ 1) þ (ε ꢁ 1)/(ε þ 2)] was used, where n was the refractive
index and ε was the static dielectric constant of the solvents.
Calculated f values ranged from 0.811, for ethyl acetate, to 1.156, for
DMSO. Plotting the results of 1/
the correlation between 1/
l with f (n, ε) is shown in Fig. 8, and
l
and f (n, ε) provided an adjusted R2
value of 0.97. This function indicated that the spectral shifts were
essentially due to dispersion, induced electronic polarization and
an induced dipole moment. However, a limitation of the function is
that it does not include intermolecular interactions and hydrogen-
bonding interactions, and this is the main reason for deviation of
the linear correlation.
3.5. IR spectra
The experimental excitation energies, calculated main orbital
compositions of the computed lower-lying singlet excited states,
transition features and oscillator strengths (f) of the dyes are given
in Table 4. The TDDFT calculations predicted three transition states
and only one dominant transition by the largest oscillator strength
Since IR spectra can give important information on the identi-
fication of compounds, the vibrational analysis of dyes (1) and (2)
was also performed at both the experimental and theoretical
calculation B3LYP/6-31G* levels. For comparison, the observed and
simulated IR spectra are presented in Fig. 10. From the theoretical
Fig. 9. Molecular orbitals of dyes (1) (right) and (2) (left).

1414
X.-H. Zhang et al. / Dyes and Pigments 93 (2012) 1408e1415
calculations, 3n ꢁ 6 frequencies (75 and 117 frequencies for dyes (1)
1.0
0.5
and (2), respectively) were found. However, the intensities of the
majority of the frequencies were low; therefore, the typical
frequencies are reported in Table 5. The calculated spectra were
found to be higher than the experimental values, possibly because
the electron correlation and vibrational anharmonicity were
neglected by this calculation. For reducing the systematic errors,
the scaling factor 0.9613 [27] was used for the calculated harmonic
vibrational wavenumbers at the B3LYP/6-31G* level. The adjusted
R2 values were 0.9997 and 0.9996 for dyes (1) and (2), which
indicated good linearity between the scaled and experimental
wavenumbers, as shown in Fig. 11.
(1) Exp.
a
b
c
d
0.0
400
(1) Scal.
(2) Exp.
(2) Scal.
Based on the good agreement between the scaled and observed
IR spectra and the structural similarity between dyes (1) and (2),
further studies on the assignment of the dyes were carried out. In
Fig.10 and Table 5, it is shown that the typical vibrations of (1) were
very similar to that of (2). However, the wavenumbers of (1) were
slightly higher than that of (2), and this could be because of the
lower energy of (2) in a larger conjugated system, as was discussed
in section 3.1. Obviously, the experimental vibrational frequencies
at 3458 cmꢁ1 and 3454 cmꢁ1 for dye (1) and (2), respectively, were
classified as the mode 1# yNeH, and the frequencies at 3157 cmꢁ1
for (1) and at 3119 cmꢁ1 for (2) were classified as the mode 2# CeH
stretching vibrations of the pyrrole rings. Two CeH stretching
vibrations were detected in the benzene rings, with wavenumbers
falling in the region of 3063e3125 cmꢁ1. The CeH stretching
vibrations of the methyl and methylene groups (mode 5#) were
detected at lower wavenumbers (2939e22 922 cmꢁ1), in compar-
ison to the aromatic CeH stretching modes. The strong bands at
about 1700 cmꢁ1 were attributed to the C]O stretching vibrational
mode. The dyes within whole, molecular conjugate systems were
predicted to have significant CeC and CeN bond stretching modes
ranging from 1598 to 1117 cmꢁ1, which was in excellent agreement
with the experimental observation ranging from 1132 to
1599 cmꢁ1. The normal vibration 16,17, and 19# modes and the 20#
mode were classified as CeH in-plane bending and out-of-plane
200
0
1.0
0.5
0.0
400
200
0
500
1000
1500
2000
2500
3000
3500
Wavenumber (cm-1)
Fig. 10. The experimental IR spectra (a: (1); c: (2)) and scaled simulated IR spectra at
the B3LYP/6-31G* level (b: (1); d: (2)).
vibrations in the range of 1109e789 cmꢁ1 and 756e727 cmꢁ1
,
respectively. There were also medium bands (mode 18#) ranging
from 874 to 878 cmꢁ1, and these were referred to as NeO stretching
vibration.
Table 5
The measured IR, calculated vibrational wavenumbers at the B3LYP/6-31G* level and assignments for dyes (1) and (2).
Mode
Vibratory feature
(1)
(2)
Exp.a
Cal.
Scal.b
Int.
Exp.a
Cal.
Scal.b
Int.
1#
2#
3#
4#
5#
6#
7#
8#
yNeH (pyrrole)
y]CeH (pyrrole)
y]CeH (benzene)
y]CeH (benzene)
yCeH (CH3, CH2)
yC]O
3458 (m)
3157 (m)
3125 (w)
3063 (w)
2939 (m)
1701 (s)
1599 (vs)
1582 (vs)
1481 (s)
1452 (vs)
1377 (s)
1348 (m)
1298 (vs)
1219 (m)
1132 (w)
1109 (m)
995 (m)
3650
3287
3215
3203
3052
1811
1662
1635
1547
1539
1421
1400
1334
1259
1166
1135
1020
914
3508
3160
3091
3079
2934
1741
1598
1572
1487
1479
1366
1346
1282
1211
1121
1091
981
102.74
37.47
25.74
25.55
11.06
246.03
419.27
54.13
167.21
181.28
233.48
90.41
73.80
119.05
125.50
52.09
62.14
67.46
45.84
50.25
3454 (m)
3119 (m)
3094 (m)
3063 (m)
2922 (m)
1703 (s)
1595 (s)
1580 (vs)
1472 (w)
1439 (s)
1375 (vs)
1337 (w)
1288 (s)
1200 (m)
1132 (m)
1107 (m)
990 (m)
874 (w)
3651
3288
3207
3196
3044
1810
1660
1638
1538
1517
1421
1374
1325
1246
1162
1134
1020
914
3509
3161
3083
3073
2926
1740
1596
1574
1479
1458
1366
1321
1274
1197
1117
1090
981
108.40
47.77
34.63
23.58
30.89
248.54
501.95
78.83
170.01
182.82
189.21
136.69
122.97
208.29
103.67
95.98
68.48
63.42
57.97
25.29
yC]C
yC]C
yC]C
yC]C
yC]C
yC]C
yC]C
yC]C
yC]C
dCH
,
,
,
,
yC]N
yC]N
yC]N
yC]N
9#
10#
11#
12#
13#
14#
15#
16#
17#
18#
19#
20#
,
yC]N
dCH
yNeO
d]CH
d]CH
878 (m)
812 (m)
756 (m)
878
793
730
879
773
725
825
759
789 (m)
727 (m)
804
755
a
vs: very strong, s: strong, m: middle, w: weak for IR intensity.
The scaling factor 0.9613 was used for the wavenumbers obtained at the B3LYP/6-31G* level.
b

X.-H. Zhang et al. / Dyes and Pigments 93 (2012) 1408e1415
1415
[3] Ishioka T, Tanatani A, Nagasawa K, Hashimoto Y. Anti-androgens with full
antagonistic activity toward human prostate tumor LNCaP cells with mutated
androgen receptor. Bioorg Med Chem Lett 2003;13:2655e8.
[4] Laughlin SK, Clark MP, Djung JF, Golebiowski A, Brugel TA, Sabat M, et al. The
development of new isoxazolone based inhibitors of tumor necrosis factor-
2
3000
2000
1000
(2)
R =0.9996
b
alpha (TNF-a) production. Bioorg Med Chem Lett 2005;15:2399e403.
[5] Lowe DB, Magnuson S, Qi N, Campbell A-M, Cook J, Hong Z, et al. In vitro SAR
of (5-(2H)-isoxazolonyl) ureas, potent inhibitors of hormone-sensitive lipase.
Bioorg Med Chem Lett 2004;14:3155e9.
[6] Dayan FE, Duke SO, Reddy KN, Hamper BC, Leschinsky KL. Effects of isoxazole
herbicides on protoporphyrinogen oxidase and porphyrin physiology. J Agric
Food Chem 1997;45(3):967e75.
[7] Hess FD. Light-dependent herbicides: an overview. Weed Sci 2000;48(2):
160e70.
[8] Aret E, Meekes H, Vlieg E, Deroover G. Polymorphic behavior of a yellow
isoxazolone dye. Dyes Pigments 2007;72:339e44.
[9] Deij MA, Horst JHt, Meekes H, Jansens P, Vlieg E. Polymorph formation studied
by 3D nucleation simulations. Application to a yellow isoxazolone dye, par-
(1) R2=0.9997
a
3000
2000
1000
500
1000
1500
2000
Exp.
2500
3000
3500
acetamol, and
[10] Alías S, Andreu R, Blesa MJ, Cerdán MA, Franco S, Garín J, et al. Iminium salts of
-dithiafulvenylpolyenals: an easy entry to the corresponding aldehydes and
L-glutamic acid. J Phys Chem B 2007;111:1523e30.
u
doubly proaromatic nonlinear optic-phores. J Org Chem 2008;73:5890e8.
[11] Lee J-E, Kim HJ, Han MR, Lee SY, Jo WJ, Lee SS, et al. Crystal structures of C.I.
disperse red 65 and C.I. disperse red 73. Dyes Pigments 2009;80:181e6.
[12] Laufer SA, Margutti S. Isoxazolone based inhibitors of p38 MAP kinases. J Med
Chem 2008;51(8):2580e4.
Fig. 11. Graphic correlation between the experimental and the scaled theoretical
frequencies obtained by the B3LYP/6-31G* level (a: (1); b: (2)).
[13] Kim K-H, Jung DH, Kim D, Lee A, Choi K, Kim Y, et al. Crystal structure
prediction of organic materials: tests on the 1,4-diketo-3,6-
diphenylpyrrolo(3,4-c)pyrrole and 1,4-diketo-3,6-bis(40- dipyridyl)-pyrrolo-
[3,4-c]pyrrole. Dyes Pigments 2011;89:37e43.
[14] Ebead YH. Spectrophotometric investigations and computational calculations
of prototropic tautomerism and acid-base properties of some new azo dyes.
Dyes Pigments 2011;92(1):705e13.
[15] Kim S, Mor GK, Paulose M, Varghese OK, Baik C, Grimes CA. Molecular design
of near-IR harvesting unsymmetrical squaraine dyes. Langmuir 2010;26(16):
13486e92.
[16] Kolev T, Tsanev T, Kotov S, Mayer-Figge H, Spiteller M, Sheldrick WS, et al.
Anyles of 4-benzoylpyridine e crystal structure and spectroscopic properties.
Dyes Pigments 2009;82:95e101.
[17] Casanova D, Rotzinger FP, Grätzel M. Computational study of promising
organic dyes for high-performance sensitized solar cells. J Chem Theory
Comput 2010;6(4):1219e27.
[18] Preat J. Photoinduced energy-transfer and electron-transfer processes in dye-
sensitized solar cells: TDDFT insights for triphenylamine dyes. J Phys Chem C
2010;114(39):16716e25.
4. Conclusion
The results of X-ray and theoretical calculation analyses indi-
cated that merocyanine dyes containing an isoxazolone nucleus
prefer to be in the ketonic form when in a solid state. Dyes (1) and
(2) crystallized in the orthorhombic Cmca space group and
monoclinic P21/n space group, respectively. In dye (1), intermo-
lecular N(1)eH(1)/N(2) and C(4)eH(4)/O(1) hydrogen bonds
and offset
p
/p
aromatic interactions stabilized the crystal struc-
intermo-
ture. In dye (2), there were two types of weak CeH/
p
lecular interactions that existed, and N(1)eH(1)/O(2) hydrogen
bonds linked the molecules into parallel, infinite sheets along the
b axis. The molecular electrostatic potential (MEP) obtained by
B3LYP/6-311G** clearly showed the possible sites for nucleophilic
attack among the crystal packing. Using the scaling factor 0.9613 to
remove the major systematic errors of B3LYP/6-31G*, the IR spectra
could be predicted with high accuracy.
[19] Meguellatia K, Ladamea S, Spichty M.
A conceptually improved TD-DFT
approach for predicting the maximum absorption wavelength of cyanine
dyes. Dyes Pigments 2011;90(2):114e8.
[20] Fu YL, Huang W, Li CL, Wang LY, Wei YS, Huang Y, et al. Monomethine cyanine
dyes with an indole nucleus: microwave-assisted solvent-free synthesis,
spectral properties and theoretical studies. Dyes Pigments 2009;82:409e15.
[21] Sajan D, Vijayan N, Safakath K, Philip R, Joe IH. Intramolecular charge transfer
and Z-Scan studies of a semiorganic nonlinear optical material sodium acid
phthalate hemihydrate: a vibrational spectroscopic study. J Phys Chem A
2011;115(29):8216e26.
[22] Rieff B, Bauer S, Mathias G, Tavan P. IR spectra of flavins in solution: DFT/MM
description of redox effects. J Phys Chem B 2011;115(9):2117e23.
[23] Zhang YQ, Ma JJ, Wang C, Li JC, Zhang DN, Zang XH, et al. One-pot synthesis of
3-methyl-4-arylmethylene-isoxazol-5(4H)-ones. Chin J Org Chem 2008;28:
141e4.
[24] Bruker. SHELXTL, version 5.1. Madison, WI: Bruker AXS Inc.; 1999.
[25] Spek AL. Structure validation in chemical crystallography. Acta Crystallogr
2009;D65:148e55.
[26] Tomasi J, Mennucci B, Cammi R. Quantum mechanical continuum solvation
models. Chem Rev 2005;105:2999e3093.
Acknowledgements
We appreciate the financial support for this research by the
Program of the National Basic Research and Development Program
of China (973) under Grant No. 2011CB707702; the National Natural
Science Foundation of China under Grant Nos. 81090272,
81000632, 30900334, 81101083, 81101084, 81101100; the Shaanxi
Provincial Natural Science Foundation Research Project under
Grant No. 2009JQ8018, and the Science Research Startup Founda-
tion of Xidian University (K50510100002, K50510100004,
K50510100005).
[27] Scott AP, Radom L. Harmonic vibrational frequencies: an evaluation of
Hartree-Fock, Møller-Plesset, quadratic configuration interaction, density
functional theory, and semiempirical scale factors. J Phys Chem 1996;100:
16502e13.
References
[28] Frisch MJ, Trucks GW, Schlegel HB, Scuseria GE, Robb MA, Cheeseman JR, et al.
Gaussian 03. Revision B.03. Pittsburgh PA: Gaussian, Inc.; 2003.
[29] Airinei A, Homocianu M, Dorohoi DO. Changes induced by solvent polarity in
electronic absorption spectra of some azo disperse dyes. J Mol Liq 2010;157:
13e7.
[30] Powell BJ, Baruah T, Bernstein N, Brake K, McKenzie RH, Meredith P, et al.
A first-principles density-functional calculation of the electronic and vibra-
tional structure of the key melanin monomers. J Chem Phys 2004;120(18):
8608e15.
[1] Deng B-L, Hartman TL, Robert W, Buckheit J, Pannecouque C, Clercq ED, et al.
Replacement of the metabolically labile methyl esters in the alkenyldiaryl-
methane series of non-nucleoside reverse transcriptase inhibitors with iso-
xazolone, soxazole, oxazolone, or cyano substituents. J Med Chem 2006;49:
5316e23.
[2] Ishioka T, Kubo A, Koiso Y, Nagasawa K, Itaib A, Hashimoto Y. Novel non-
steroidal/non-anilide type androgen antagonists with an isoxazolone
moiety. Bioorg Med Chem 2002;10:1555e66.