T. Bardakçı et al. / Journal of Molecular Structure 1054–1055 (2013) 76–82
77
study, we investigate IR and Raman spectra of L and its [CuCl2L2]
complex experimentally. To assign the experimental vibrational
bands in terms of internal vibrations in the modes, we performed
density functional theory (DFT) calculations. We have already
shown that the hybrid DFT functional B3LYP is superior to the HF
and MP2 methods in investigating vibrational spectra of L [3].
The effects of coordination on the vibrational modes have also been
discussed. Moreover, we have investigated the structure of the
[CuCl2L2] complex based on the experimental vibrational and EPR
spectra as well as theoretical DFT studies.
was used to record the first-derivative absorption signals at the
room temperature. A cylindrical quartz tube was used to place
the powder sample in the cavity. Operating conditions were cho-
sen as 0.2 mW microwave power, 1 Oe modulation amplitude,
3300 Oe centerfield, 600 Oe sweep width, 81.92 ms time constant
and 83.89 s sweep time with multiple accumulations to enhance
the signal-to-noise ratio.
3. Quantum chemical calculations
Quantum chemical calculations were performed with Gaussian
03 software package [18] on a linux server cluster. The structures
and normal modes were visualized by Gausview 03 program [19].
The geometry optimizations, vibrational spectra calculations, and
percent potential energy distribution (PED) of free L and [CuCl2L2]
complex in the gas phase were performed by the DFT exchange-cor-
relation functionals BVP86 and B3LYP employing the 6-311G+(d,p)
basis set. BVP86 is a combination of Becke’s 1988 exchange func-
tional [20], Vosko’s correlation functional V for local spin density
part [21] and Perdew’s 1986 gradient correlation functional [22]
while B3LYP consists of Becke’s three-parameter exchange and
Lee–Yang–Parr’s correlation functional [23–25].
2. Experimental
2.1. Material and synthesis
The ligand, L and CuCl2 at the reagent grade were purchased
from Aldrich Chemical Company Inc., and used without further
purification.
[CuCl2L2] was synthesized by adding two moles of L to one mole
of CuCl2 in ethyl alcohol with constant stirring. The precipitated
complex was filtered, washed with diethylether and dried in vac-
uum. All these processes were carried out at room temperature.
The experimental C, H, N and Cu contents of the synthesis complex
agree very well with the theoretically expected contents of the
[CuCl2L2] (see Table 1).
4. Results and discussion
4.1. Molecular structure
2.2. FT-IR measurements
The optimized molecular structures of free L and [CuCl2L2] com-
plex in the gas phase at the room temperature (see Fig. 1 and Ta-
ble 2) have been obtained in terms of BVP86 and B3LYP density
functionals with the 6-311G+(d,p) basis set. The standard error S
and the mean absolute error MAE in the bond lengths relative to
the previous X-ray data of L [26] are respectively 0.183 and 0.014
with B3LYP, and 0.184 and 0.018 with BVP86. S and MAE for the
bond angles are respectively 0.464 and 1.213 with B3LYP, and
0.470 and 1.125 with BVP86. Therefore, B3LYP performs slightly
better than BVP86 both for the bond lengths and the bond angles.
DFT calculations predict some changes in the geometry param-
eters of L with coordination. For instance, C–N bond has been in-
creased from 1.404 Å to 1.439 Å with BVP86, and from 1.401 Å to
1.442 Å with B3LYP. N–H bond has also been increased 0.007 Å
with two of the methods. C–C bonds generally remain unchanged;
however bonds close to Nitrogen atom i.e. C2–C3 and C3–C4 have
been decreased because of coordination.
The 3500–500 cmꢁ1 region FT-IR spectra of L and [CuCl2L2] were
recorded on a Nicolet 6700 Fourier Transform IR (FT-IR) spectrom-
eter with attenuated total reflectance (ATR) sampling technique at
room temperature. The far region (650–50 cmꢁ1) FT-IR (FT-FIR)
spectra of L and [CuCl2L2] were recorded on the same spectrometer
but by preparing polyethylene (PE) pellets. PE pellets were pre-
pared as described previously taking the sample/PE ratio as 1/10
[17], i.e., ꢂ8 mg of PE and 0.8 mg of sample were mixed and
ground in a mortar. The metalic die was heated to 230 °C, and
the temperature of the die surface was expected to be 120–
140 °C (since the melting point of PE is 120–130 °C). The mixture
of PE and samples were added to the die, and put under 6-ton pres-
sure for 2 min. Then the die was disassembled and the PE pellet
was put in the pellet holder of FT-IR spectrometer to record the
far IR region of the spectrum. Since the expansion is big for far re-
gion, the spectra of the title molecules appear discontinuous.
[CuCl2L2] complex can in principle be found in doublet, quartet
and sextet spin states with one, three and five unpaired electrons,
respectively. The present DFT calculations exclusively predict dou-
blet ground state with one unpaired electron for the complex at the
room temperature (Table 3). The quartet and sextet spin states lie
ꢂ80 and ꢂ160 kcal/mol above the ground-state doublet, respec-
tively. Thus these two higher energy spin states are not accessible
at the room temperature. DFT calculations predict the Cu–N bonds
by 0.17 Å shorter than the Cu–Cl bonds (Table 4).
2.3. Dispersive Raman measurements
The room temperature 3500–50 cmꢁ1 region Raman spectra of L
and [CuCl2L2] were recorded with Dispersive Raman Microscope
(DXR) by using a laser of 532 nm wavelength. At the laser powers
that do not burn the present dark-colored complex, a good-re-
solved dispersive Raman spectrum could not be recorded.
2.4. EPR measurement
4.2. Vibrational assignments
A conventional X-Band Bruker EMX model EPR spectrometer
employing an AC magneticfield (5 kHz) modulation technique
All observed IR and Raman normal modes of L and [CuCl2L2]
(Figs. 2–4) have been assigned with the aid of DFT calculations (Ta-
ble 5). The DFT calculations overestimate generally the vibrational
frequencies due to the neglect of crystal packing effects and anhar-
monicity as well as the basis set and electronic correlation incom-
pleteness. Therefore, we derived scaling factors on L and applied
them to [CuCl2L2] to approximately account for the computational
errors. For the frequencies above and below 1623 cmꢁ1, the BVP86
scaling factors are 0.974 (R2 = 0.79) and 1.005 (R2 = 0.996) and, the
Table 1
Elemental analysis of the [CuCl2L2] as percentage.
Comp.
Found (cal.)%
Cu
C
N
H
[CuCl2L2]
(Theoretical)
18.49
(18.22)
48.16
(48.21)
7.84
(8.03)
5.17
(5.21)