812
G.N. Andree6 et al. / Spectrochimica Acta Part A 53 (1997) 811–818
dichroic spectra [5] no reports on the vibra-
tional spectra and assignment of 1,4-dini-
trobenzene isotopic labelled derivatives seem
to have appeared up to now. The normal
coordinate analysis data for the correspond-
ing 1,3-isomer were published quite a long
time ago [6].
On the other hand, ab initio MO calcula-
tions for various classes of organic molecules
[7–15] and ions [7,11–15] have been used
successfully to study their force field and
structure. However, except a short communi-
cation [16], ab initio MO data for 1,4-dini-
trobenzene have not been published up to
now.
with a resolution of 1 cm−1, as well as in
polyethylene on a Grubb-Parsons interferometer
in the far IR range (down to 50 cm−1). The
Raman spectra were measured in the solid state,
and in CCl4, toluene and CH3OH solutions on a
Bruker Raman module FTR 106 with 1 cm−1
resolution. A diode-pumped Nd:YAG laser with
wavelength 1064 nm was used for excitation with
a 180° arrangement [18]. In the 120–20 cm−1
range the spectra were recorded on Jobin-Yvon
Ramanor; the 514.5 nm line of a Spectra Physics
Ar+ laser with a power of about 120 mW was
used for excitation. The data are given in Table 1.
The purpose of the present study is to as-
sign the fundamental vibration frequencies
and to elucidate both the force field and
structure of 1,4-dinitrobenzene molecule on
the basis of IR and Raman spectra, isotope
labelling, and ab initio force field calcula-
tions.
3. Computations
The ab initio quantum chemical computa-
tions were performed using the GAMESS
software [19] at the HF 3-21G, 6-31G and
6-31G** basis set levels.
It is well known that the neat (native, un-
scaled) frequencies, obtained by ab initio
force field calculations, are always higher
than the experimental ones. Thus, for a bet-
ter comparison between these values, the the-
oretical results have usually been modified by
using empirical scaling factors for either the
force constants [10] or resulting frequencies
[8], or by applying empirical correlation
equations [11–15]. The latter method of scal-
ing has been used in this work (Table 2).
The comparison between theoretical and
experimental frequencies of 1,4-dinitrobenzene
has shown that the agreement between them
is improved when the 6-31G basis set is
used instead of the 3-21G one. The imple-
mentation of polarization functions in the
basis set, however, does not lead to a fur-
ther improvement of the results. The mean
deviations, found by comparison between the
calculated and measured frequency values are,
as follows: 44.4 cm−1 for 3-21G, 26.5 cm−1
for 6-31G, and 34.4 cm−1 for 6-31G**.
Thus the polarization basis sets not only
prolonged the processing time but also dete-
riorated the predicted frequencies of 1,4-dini-
trobenzene.
2. Experimental
1,4-dinitrobenzene and its mono-15N la-
belled isotopomer 1,4-dinitrobenzene-14N,15N
were prepared from the corresponding 4-ni-
troaniline and 4-nitroaniline-15N (labelled in
the nitro group, 98 at.% enrichment, Isokom-
merz) and NaNO2, according to ref. [17].
The synthesis of the di-15N labelled 1,4-dini-
trobenzene-15N,15N was carried out using 4-
nitroaniline-15N (a sample as above) and
Na15NO2 (52.8 at.% enrichment, Isokommerz).
Hence, three mixtures of 1,4-dinitrobenzenes
with the following contents were prepared:
1. 52.8% of O125NC6H4NO2 and 47.2% of
O2NC6H4NO2
2. 47.32% of O125NC6H4NO2, 51.74% of
O125NC6H145NO2, and 0.94% O2NC6H4NO2
3. 98% of O125NC6H4NO2 and 2% of
O2NC6H4NO2
The IR spectra of the 1,4-dinitrobenzenes stud-
ied were recorded in the solid state in CsI pellets,
and in CCl4 and CS2 solutions on a Bruker IFS-
113v spectrometer in the 4000 to 160 cm−1 region
Andreev, George N.
