Investigation of hydrogen bonds properties in the terephthalic acid crystal, using molecular dynamics method

Article abstract of DOI:10.1016/j.saa.2014.03.125

The aim of this study was to perform calculations using the method of Car-Parrinello molecular dynamics, leading to the optimized geometry of the molecules of 1,4-benzenedicarboxylic acid (terephthalic acid) in crystals, for the hydrogen form and three variants of substitution of deuterium atoms inside a carboxyl group. Based on the results, trajectories and dipole moments were calculated, what makes possible to simulate vibrations in different systems, and to make calculation of theoretical infrared spectra and atomic power spectra. Theoretical results were compared with the experimental spectra, which verifies the correctness of the method and also was compared with the results obtained by quantum-mechanical calculations using DFT for the isolated dimer. Comparison of the spectra of different forms, allowed for in-depth analysis of the effect of isotopic substitution on the frequency of vibrations and shapes of bands, and confirm the presence of possible coupling effects and intra- and intermolecular interactions. Comparison with the DFT results for the dimer show influence of the crystal structure on the spectra.

Full text of DOI:10.1016/j.saa.2014.03.125

Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 130 (2014) 488–493
Contents lists available at ScienceDirect
Spectrochimica Acta Part A: Molecular and
Biomolecular Spectroscopy
journal homepage: www.elsevier.com/locate/saa
Investigation of hydrogen bonds properties in the terephthalic acid
crystal, using molecular dynamics method
⇑
Ewa Wierzbicka, Marek Boczar, Marek J. Wójcik
Department of Physical Chemistry & Electrochemistry, Faculty of Chemistry, Jagiellonian University in Krakow, Ingardena 3, 30060 Krakow, Poland
h i g h l i g h t s
g r a p h i c a l a b s t r a c t
ꢀ
ꢀ
ꢀ
The acid and its deuterated forms
were studied by Car–Parrinello
molecular dynamics.
The power and theoretical infrared
spectra were calculated for the
different systems.
The theoretical infrared spectra were
compared with the experimental
data.
a r t i c l e i n f o
a b s t r a c t
Article history:
The aim of this study was to perform calculations using the method of Car–Parrinello molecular dynam-
ics, leading to the optimized geometry of the molecules of 1,4-benzenedicarboxylic acid (terephthalic
acid) in crystals, for the hydrogen form and three variants of substitution of deuterium atoms inside a
carboxyl group. Based on the results, trajectories and dipole moments were calculated, what makes pos-
sible to simulate vibrations in different systems, and to make calculation of theoretical infrared spectra
and atomic power spectra. Theoretical results were compared with the experimental spectra, which ver-
ifies the correctness of the method and also was compared with the results obtained by quantum-
mechanical calculations using DFT for the isolated dimer. Comparison of the spectra of different forms,
allowed for in-depth analysis of the effect of isotopic substitution on the frequency of vibrations and
shapes of bands, and confirm the presence of possible coupling effects and intra- and intermolecular
interactions. Comparison with the DFT results for the dimer show influence of the crystal structure on
the spectra.
Received 7 January 2014
Received in revised form 19 March 2014
Accepted 29 March 2014
Available online 16 April 2014
Keywords:
Hydrogen bond
Car–Parrinello molecular dynamics
Terephthalic acid
Ó 2014 Elsevier B.V. All rights reserved.
Introduction
properties of matter [1–4] and also it is involved in stabilization
of the chemical structures and important biological processes
Hydrogen bond (HB) is still one of the most investigated sub-
jects, due to its great importance in many chemical areas. It has
strong effect on the micro- and macroscopic physicochemical
[5–9]. Spectroscopic methods provided many useful information
which made possible to better know HB properties, but only a
combination of spectroscopic and computational methods
[10–12] made possible to obtain complementary information to
understand this very complicated interaction.
⇑
In the cyclic dimers of carboxylic acids, many factors are affected
on the nature of the HB [13]. Calculations of the vibrational spectra
Corresponding author. Tel.: +48 12 663 29 13; fax: +48 12 634 05 15.
E-mail address: wojcik@chemia.uj.edu.pl (M.J. Wójcik).
http://dx.doi.org/10.1016/j.saa.2014.03.125
1386-1425/Ó 2014 Elsevier B.V. All rights reserved.

E. Wierzbicka et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 130 (2014) 488–493
489
¼ 92:75ꢁ
of the HB system are complicated due to large anharmonicity and
a ¼ 7:73 Å b ¼ 6:443 Å c ¼ 3:749 Å
a
b ¼ 109:15ꢁ
c
¼ 95:95ꢁ
proton delocalization [14]. The commonly used models of the cyclic
dimer, describing the shape of the HB band in IR spectroscopy,
assumed that it is the result of: (a) strong anharmonic coupling of
the OAH stretching vibration with the stretching vibration of HB
[15–18], (b) the Fermi resonance [19], appearing between the
OAH stretching vibration and the first overtone of the OAH bending
vibration [20], and (c) Davydov coupling [18,21,22] occurring
between a pair of isoenergetic hydrogen bonds in the cyclic dimer.
The calculations which are presented in this paper, also take into
account the coupling between the vibrations of OAH bonds in the
crystal, which is a factor that strongly influences the HB properties.
Such interactions are very important considering the fact that very
often the hydrogen bonds are most important for holding together
organic crystals. Understanding this is extremely important for the
development of modeling, starting from crystals of small size to
large structures of macromolecules [23]. The progress of computa-
tional methods leads to increasing compatibility of the HB models
with real systems.
Car and Parrinello are the creators of new kind molecular
dynamics (CPMD) [24], where the main idea is the best use of
quantum-mechanical time-scale separation of fast electrons and
slow nuclei motions. This is achieved by converting the quan-
tum-classical problem in a purely classical (in two separable time
scales), on expense of losing information about the physical time in
the quantum network [24–26]. The great advantage of the CPMD
method is the possibility of quantum chemical description of the
system, with effects on the HB factors, such as the crystal period-
icity and thermal fluctuations [27,28], which are not included in
other computational methods, such as DFT [29,30]. In the present
paper, the results of the CPMD calculations for the terephthalic
acid crystal and DFT calculations for its cyclic dimer have been
compared. Also we performed study of the isotopic substitution
effect, on the spectra. CPMD is recently used to simulate IR spectra
of hydrogen-bonded crystals [31,32].
Computational methods
Calculations performed using the CPMD package were aimed
to determine the optimal geometry of the structure of the crystal
with four different isotopic substitutions. Consequently, we calcu-
lated the appropriate trajectories and dipole moments. The
obtained results enabled us to simulate of vibrations of the sys-
tem, leading to theoretical power spectra and infrared spectra.
As a model system four terephthalic acid molecules were used,
which corresponds to 72 atoms. They form two hydrogen-bonded
dimers I and II as shown in Fig. 1. Inside the unit cell are two
pairs of hydrogen bonds. In the two other model systems, hydro-
gen atoms in the carbonyl group were substituted by deuterium
atoms forming isotopomer structures. In the first model system
positions I-H1, H6, H10, H60 and the second I-H1, H60, II-H1, H60
were substituted.
All dynamic calculations were performed using the BLYP func-
tional [35]. To describe the electronic structure of the core we used
Goedecker pseudopotentials [36], while for the valence electrons
Bloch plane wave basis set [37] with the parameter CUTOFF 120
Ry. Optimization of geometry was done with an accuracy 10ꢂ6
,
what permitted to achieve the trajectories. Dipole moment was
calculated at tenth iteration of the optimized geometry. In order
to determine the trajectories, 220,000 iterations of trajectories
were carried for each of the system, with time step equal 2 a.u.
(0.049 fs). Such large number of steps were required because it
was necessary to omit first 70,000 steps in the calculated trajecto-
ries, to stabilize the system. In addition, the geometry optimization
was performed for the isolated dimer using DFT method, with
functional B3LYP and 6-31++G** basis set.
Terephthalic acid is an aromatic organic compound, which is
involved in controlling the processes of some chemical reactions
in organisms, such as being an antioxidant for protection against
damage of internal organs, and myocardial infarction. It is used
widely in industry for the production of high tensile strength poly-
ester fibers. It occurs in nature in the form of white crystals. Its
molecular weight is 166.13 g/mol, and density 1.522 g/cm3. Its
melting point is 573 K, and at temperature 675 K is sublimed. It
is weakly soluble in water and in organic liquids. It crystallizes in
the triclinic system. The unit cell which belongs to the space group
P1, contains one molecule of the acid [33,34]. Terephthalic acid has
two polymorphic forms. Our study concerns one of them because it
is more stable as a result of its higher density and second form
tended to become rarer on storage [34]. Lattice parameters of the
unit cell of the studied crystal acid are:
Experimental section
Infrared spectra were measured for the hydrogen form of tere-
phthalic acid, and for deuterated crystals with deuterium atoms
replacing hydrogen atoms in the carboxylic groups. This substitu-
tion has been obtained by heating a suspension of terephthalic
acid crystals in tetrahydrofuran with the addition of heavy water,
followed by vacuum evaporation. Due to low solubility of tere-
phthalic acid, efficiency of the isotopic exchange process was
low. Only partially isotopically substituted acid was obtained.
ATR technique was used for the spectroscopic measurements,
which were done on Thermo Scientific Nicolet IR200
spectrometer.
Fig. 1. Model system of terephthalic acid crystal.

490
E. Wierzbicka et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 130 (2014) 488–493
Fig. 2. Labeling of atoms used in descriptions of bond lengths and angles.
data (0.108 Å), two times more than for the CPMD calculations per-
Results and discussion
formed for the crystal.
Geometry analysis
Experimental infrared spectra of terephthalic acid
Comparison of crystallographic data with the geometry opti-
mized structures were made. Fig. 2 shows geometry of terephthalic
acid dimer and labeling used in descriptions of bond lengths and
angles.
In the IR spectrum of the hydrogen form the widest range,
located at 2400–3100 cmꢂ1, comes from stretching vibrations of
the OAH groups, which are engaged in the formation of hydrogen
bonds. This is confirmed by the presence of the band shifted to
lower wave numbers compared with the bands of non-hydrogen-
bonded hydroxyl groups. Also large half-width of the band, close
to 700 cmꢂ1, and the presence of fine structure are evidences of
hydrogen bond formation. Obtaining partially isotopically substi-
tuted form of the acid is manifested by a decrease in intensity of
the OAH band, and the appearance at lower wave numbers a
new band, derived from OAD vibration. Spectra of hydrogen-
bonded and partially dueterated crystals are shown in Fig. 3.
In a model system, there are four molecules which are identical
with each other, due to the presence of the inversion. Therefore the
description is limited to the parameters for one of them. Crystallo-
graphic data are simultaneously the output coordinates for the
atoms which geometry has been optimized. The obtained values
of bond lengths and angles for all atoms were averaged due to loss
of identity as a result of optimization. The results are summarized
in Tables 1 and 2, together with the crystallographic data.
The analysis of the results leads to the conclusion that the cal-
culated geometry, which was obtained by optimization with the
CPMD package, does not differ significantly from the crystallo-
graphic data. Theoretical bond lengths are in good agreement with
those experimentally determined. Differences vary between 0 and
0.05 Å. Angles, after geometry optimization using CPMD package,
are in most cases reproduced very well, however, some exceptions
are observed, mainly in the angles formed by the carbon atoms
connected to the hydroxyl groups. It may be caused by limitations
of the method as well as to the uncertainty of determination of the
positions of hydrogen atoms by X-ray diffraction.
Theoretical infrared spectra of terephthalic acid
Fig. 4a shows the experimental spectrum of terephthalic acid
crystal, Fig. 4b the theoretical infrared spectrum of the crystal
Table 2
Angles formed by atoms in terephthalic acid obtained from crystallographic data and
from theoretical calculations.
Description
Crystallographic
data (°) [34]
Average values of
angles from
geometry
optimization using
the CPMD method
(°)
The values of the
angles from
geometry
optimization using
the DFT method (°)
Results of the geometry optimization in the isolated dimer (DFT
method) are much worse. In this case, the calculated lengths of the
OAH bonds are in larger disagreement with the crystallographic
Table 1
O1AH1ꢃ ꢃ ꢃO20 179.97
175.51
110.74
123.21
114.90
121.89
118.85
120.62
120.54
119.68
119.84
120.48
119.79
120.44
119.77
120.53
120.61
118.86
119.68
119.84
120.49
119.79
119.77
120.44
114.92
121.89
123.19
110.76
179.98
110.53
123.51
114.60
121.89
121.23
118.70
120.06
119.86
119.62
120.53
120.08
120.99
118.93
120.07
118.00
121.94
119.84
119.81
120.35
120.09
119.03
120.89
113.12
124.76
122.13
106.76
Bond lengths in terephthalic acid, obtained from crystallographic data and from
theoretical calculations.
H1AO1AC1
O1AC1AO2
O1AC1AC2
O2AC1AC2
C1AC2AC3
C1AC2AC7
C3AC2AC7
C2AC3AC4
C2AC3AH2
H2AC3AC4
C3AC4AC5
C3AC4AH3
H3AC4AC5
C4AC5AC6
C4AC5AC8
C6AC5AC8
C5AC6AC7
C5AC6AH4
H4AC6AC7
C6AC7AC2
C2AC7AH5
C6AC7AH5
C5AC8AO3
C5AC8AO4
O3AC8AO4
C8AO3AH6
115.70
123.52
117.00
119.47
119.39
120.15
120.46
119.40
120.33
120.27
120.13
119.94
119.92
120.46
120.15
119.39
119.40
120.33
120.27
120.13
119.92
119.94
117.00
119.47
123.52
115.70
Description
Crystallographic
data (Å) [34]
Averaged bond
lengths obtained
from geometry
optimization using
Bond lengths
from geometry
optimization
using DFT method
the CPMD method (Å) (Å)
O1AHꢃ ꢃ ꢃO20 2.608
2.582
1.030
1.329
1.250
1.480
1.401
1.386
1.400
1.401
1.386
1.399
1.479
1.330
1.030
1.249
1.082
1.081
1.082
1.081
2.637
O1AH1
O1AC1
O2@C1
C1AC2
C2AC3
C3AC4
C4AC5
C5AC6
C6AC7
C7AC2
C5AC8
O3AC8
O3AH6
C8@O4
C3AH2
C4AH3
C6AH4
C7AH5
1.080
1.272
1.262
1.484
1.401
1.370
1.406
1.401
1.370
1.406
1.484
1.272
1.080
1.262
1.080
1.080
1.080
1.080
1.007
1.322
1.237
1.490
1.404
1.392
1.403
1.403
1.392
1.402
1.491
1.357
0.972
1.216
1.084
1.084
1.084
1.084

E. Wierzbicka et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 130 (2014) 488–493
491
Fig. 3. Experimental spectrum of the hydrogen form (a) and partially deuterated
form (b) of terephthalic acid.
and Fig. 4c the atomic spectrum of hydrogen atoms in the hydro-
xyl groups. Power spectrum and atomic power spectra were cal-
culated using the Fourier transformation (FT) of the trajectory
autocorrelation function, and the vibrational spectrum was
obtained from FT of the dipole autocorrelation function. FT calcu-
lations have been performed with CPMD additional scripts. Theo-
retical spectrum, shown in Fig. 4b, exhibits all calculated bands in
the frequency range 2000–3250 cmꢂ1. Spectrum in Fig. 4c con-
tains all bands calculated for vibrations of hydrogen atoms in
the hydroxyl groups, also these which are inactive in infrared
spectra. Analyzing infrared spectra, similarities between experi-
mental and theoretical spectra can be found in the positions of
the bands and in their shapes. Theoretical IR spectrum does not
reproduce well band at 3050 cmꢂ1. Power spectrum of the OAH
vibrations has low intensity in the range about 3050 cmꢂ1, there-
fore the band at 3050 cmꢂ1 can be attributed to the CAH groups
in aromatic ring.
Fig. 4. Experimental spectrum of terephthalic acid crystal (a), theoretical infrared
spectrum of the crystal (b) and atomic spectrum of hydrogen atoms in the hydroxyl
groups (c).
Although in the theoretical infrared spectra bands positions are
shifted to lower wave numbers compared with the experimental
spectrum, an analogy to the experimental one can be found in their
shapes. Comparing the calculation results with the experimental
data shows that the effect of the revaluation of the oscillator
strength is caused by mass renormalization of vibrating system
by introducing a fictitious mass of the electrons. The fictitious elec-
tron mass makes the nuclei effectively heavier. This is different for
Spectra of isotopically substituted terephthalic acid
Fig. 5a presents experimental IR spectrum of partially deuter-
ated terephthalic acid. Theoretical IR and atomic power spectra
of partially deuterated crystals I-D1, D10/II-H1.H10 and I-H1, D10/
II-D1, H10 are presented in Fig. 5b,c and d,e, respectively.

492
E. Wierzbicka et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 130 (2014) 488–493
Fig. 5. Experimental IR spectrum of partially deuterated terephthalic acid (a); theoretical IR spectrum (b) and atomic power spectrum (c) of the I-D1, D10/II-H1.H10 crystal;
and theoretical IR spectrum (d) and atomic power spectrum (e) of the I-H1, D10/II-D1, H10 crystal. In atomic power spectra of hydrogen atoms highlighted in red color and
deuterium atoms in black color. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
mass-dependent quantities such as velocity or dipole time–corre-
lation functions and the resulting power and infrared spectra,
respectively, are sensitive to mass rescaling. This is particularly
apparent of those modes that involve significant hydrogen or oxy-
calculated for the
frequency values determined from the Born–Oppenheimer approx-
imation. If = 100 a.u. this effect reduces to 5%. This strong effect
contrasted to a covalent solid like crystalline silicon at ambient
conditions [41], where the red shift is of the order of 1% with
l = 400 a.u. was 10–15% compared to reference
l
gen motion [39,40]. Frequency shift induced by finite
l values has
been analyzed numerically for compressed MgO at high tempera-
l
= 800 a.u. The mass renormalization also effects the calculation
ture by Tangney and Scandolo [41]. The red shift effect observed
of physical temperature of the system [41] from the average

E. Wierzbicka et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 130 (2014) 488–493
493
kinetic energy of the nuclei using usual equipartition theorem esti-
mator [42–44]. The temperature is systematically too low if the
mass renormalization effect is not taken into account. It is impor-
tant to note that partially deuterated terephthalic acid crystal
includes homogenous hydrogen or deuterium-bonded dimers H1,
H10 or D1, D10 and mixed dimers H1, D10 or D1, H10.
opportunity of detailed analysis and interpretation of experimental
spectra.
Acknowledgement
This research was supported in part by PL-Grid Infrastructure.
In the area of OAH stretching vibrations band shape is very sim-
ilar to the spectrum of the hydrogen form, which indicates the
presence of homogenous non-deuterated dimers. Two intensive
bands, which come from OAD vibrations, are located at lower fre-
quencies, which is consistent with the quantum-chemical model
References
[1] R.A. Klein, Chem. Phys. Lett. 425 (2006) 128–133.
[2] M. Ziółkowski, S.J. Grabowski, J. Leszczynski, J. Phys. Chem. A 110 (2006) 6514–
6521.
[3] J. Yang, S. Ding, M. Radosz, Y. Shen, Macromolecules 37 (2004) 1728–1734.
[4] M.L. Cowan, B.D. Bruner, N. Huse, J.R. Dwyer, B. Chugh, E.T.J. Nibbering, T.
Elsaesser, R.J.D. Miller, Nature 434 (2005) 199–202.
[5] S. Scheiner, T. Kar, J. Am. Chem. Soc. 117 (1995) 6970–6975.
[6] J. Poater, M. Swart, C.F. Guerra, F.M. Bickelhaupt, Comp. Theor. Chem. 998
(2012) 57–63.
[7] F. Herbst, K. Schröter, I. Gunkel, S. Gröger, T. Thurn-Albrecht, J. Balbach, W.H.
Binder, Macromolecules 43 (2010) 10006–10016.
[8] M. Peyrard, A.R. Bishop, Phys. Rev. Lett. 62 (1989) 2755–2758.
[9] M. Tarek, D.J. Tobias, Phys. Rev. Lett. 88 (2002) 138101-1-4.
[10] D.L. Howard, H.G. Kjaergaard, J. Phys. Chem. A 110 (2006) 9597–9601.
[11] F.X. Coudert, R. Vuilleumier, A. Boutin, ChemPhysChem 7 (2006) 2464–2467.
[12] M. Pagliai, G. Cardini, R. Righini, V. Schettino, J. Chem. Phys. 119 (2003) 6655–
6662.
[13] P. Durlak, Z. Latajka, Chem. Phys. Lett. 477 (2009) 249–254.
[14] G.G. Balint-Kurti, R.N. Dixon, C.C. Marston, Int. Rev. Phys. Chem. 11 (1992)
317–344.
[15] J.J. Peron, C. Sandorfy, J. Chem. Phys. 65 (1976) 3153–3157.
[16] K. Gündog˘du, J. Bandaria, M. Nydegger, W. Rock, C.M. Cheatum, J. Chem. Phys.
127 (2007) 044501-1-9.
[17] S.F. Fischer, G.L. Hofacker, M.A. Ratner, J. Chem. Phys. 52 (1970) 1934–1947.
[18] Y. Maréchal, A. Witkowski, J. Chem. Phys. 48 (1968) 3697–3705.
[19] D.D. Klug, J.S. Tse, Z. Liu, R.J. Hemley, J. Chem. Phys. 125 (2006) 154509-1-6.
[20] A. Witkowski, M. Wójcik, Chem. Phys. 1 (1973) 9–16.
[21] A.M. Yaremko, H. Ratajczak, J. Baran, A.J. Barnes, E.V. Mozdor, B. Silvi, Chem.
Phys. 306 (2004) 57–70.
[22] D. Chamma, O. Henri-Rousseau, Chem. Phys. 248 (1999) 71–89.
[23] R. Taylor, O. Kennard, Acc. Chem. Res. 17 (1984) 320–326.
[24] R. Car, M. Parrinello, Phys. Rev. Lett. 55 (1985) 2471–2474.
[25] D. Marx, J. Hutter, Ab Initio Molecular Dynamics: Basic Theory and Advanced
Methods, Cambridge University Press, Cambridge, 2009.
[26] N.L. Doltsinis, D. Marx, Phys. Rev. Lett. 88 (2002) 166402-1-4.
[27] G. Pirc, J. Stare, J. Mavri, J. Chem. Phys. 132 (2010) 224506-1-7.
[28] P. Durlak, C.A. Morrison, D.S. Middlemiss, Z. Latajka, J. Chem. Phys. 127 (2007)
064304-1-8.
[29] I. Georgieva, D. Binev, N. Trendafilova, G. Bauer, Chem. Phys. 286 (2003) 205–
217.
[30] A.G. Nozad, H. Najafi, S. Meftah, M. Aghazadeh, Biophys. Chem. 139 (2009)
116–122.
ꢀ
ꢁ
pffiffiffi
vOH
ꢄ
2
[18,38]. Recent works show that this ratio is different
vOD
p
from 2 due the anharmonic effect of the slow mode of the hydro-
gen bond. The reasons of the deviations were discussed in a several
papers [46,47,49].
The band at 2250 cmꢂ1 is not reproduced well. This fact may be
caused by imperfection of the calculation method or the presence
of interactions between another types of dimers, not considered in
the present calculations.
The differences between the theoretical and experimental spec-
tra can be caused by various factors. Effect on the location and
shape of the theoretical band has a fictional mass of electrons. This
effect was discussed above. The inverse Fourier transform of the
autocorrelation function of the dipole moment is a product of the
extinction coefficient and refractive index. Therefore, experimental
spectra cannot be directly compared with the theoretical ones. This
is only possible if the refractive index of the sample does not
depend on the wavelength. The basic physical mechanism
responsible for the energy and intensity distributions is an anhar-
monic-type coupling between the high-frequency XAH(D) stretch-
ing vibration and the low-frequency hydrogen bond XꢃꢃꢃY
stretching vibration also Fermi resonance between the XAH(D)
stretching fundamental and the first overtone of XAH(D) in-plane
bending in each H bond of the system. These effects were discussed
in detail by a number of authors [45–48]. Used in the calculations
the Goedecker’s pseudo potential probably not very well describes
near and far-ranged interactions between corresponding vibra-
tions in the crystal. It is also probably one of the reasons of not very
good reproduction of the IR band at 2250 cmꢂ1
.
Power spectrum again contains information about all vibra-
tions, also inactive in IR, related to hydrogen and deuterium atoms
in hydroxyl groups. Comparing them with the IR spectra allows to
separate vibrations coming exclusively from hydrogen bonds, sep-
arately for vibrations OAH and OAD.
[31] J. Kwiendacz, M. Boczar, M.J. Wójcik, Chem. Phys. Lett. 501 (2011) 623–627.
[32] M. Brela, J. Stare, G. Pirc, M. Sollner-Dolenc, M. Boczar, M.J. Wójcik, J. Mavri, J.
Phys. Chem. B 116 (2012) 4510–4518.
´
[33] M. Sledz´, J. Janczak, R. Kubiak, J. Mol. Struct. 595 (2001) 77–82.
[34] B. Bailey, C.J. Brown, Acta Crystallogr. 22 (1967) 387–391.
[35] C.T. Lee, W.T. Yang, R.G. Parr, Phys. Rev. B 37 (1988) 785–789.
[36] S. Goedecker, K. Maschke, Phys. Rev. A 45 (1992) 88–93.
[37] W. Kohn, Phys. Rev. 115 (1959) 809–821.
[38] S. Quillard, G. Louarn, J.P. Buisson, M. Boyer, M. Lapkowski, A. Pron, S. Lefrant,
Synthetic Met. 84 (1997) 805–806.
Conclusions
[39] D. Marx, E. Fois, M. Parrinello, Int. J. Quant. Chem. 57 (1996) 655–662.
[40] D. Marx, J. Hutter, M. Parrinello, Chem. Phys. Lett. 241 (1995) 457–462.
[41] P. Tangney, S. Scandolo, J. Chem. Phys. 116 (2002) 14–24.
[42] M.P. Allen, D. Tildesley, Computer Simulation of Liquids , Clarendon Press,
Oxford, 1987 and 1990.
Analysis of corresponding angles and bonds lengths obtained by
optimizing the geometry of terephthalic acid crystal, leads to the
conclusion that the resulting CPMD geometry is in very good
agreement with the experiment, giving much better results than
for DFT calculation for the isolated dimer.
[43] D. Frenkel, B. Smit, Understanding Molecular Simulation – From Algorithms to
Applications, Academic Press, San Diego, 1996 and 2002.
[44] D.C. Rapaport, The Art of Molecular Dynamic Simulation
University Press, Cambridge, 2001 and 2005.
[45] P. Blaise, M.J. Wójcik, O. Henri-Rousseau, J. Chem. Phys. 122 (2005) 064306.
[46] H. Ghalla, N. Rekik, M. Baazaoui, B. Oujia, M.J. Wójcik, J. Mol. Struct.
(THEOCHEM) 855 (2008) 102.
, Cambridge
Comparison of theoretical infrared spectra of terephthalic acid
with the experimental spectrum shows that most important char-
acteristic features have been preserved. It is possible to approxi-
mately attribute each band, and to find equivalents in their fine
structure, even if they are slightly shifted towards lower
frequencies.
[47] N. Rekik, H. Ghallaa, H.T. Flakus, M. Jabłon´ ska, P. Blaise, B. Oujia, Chem. Phys.
Chem. 10 (17) (2009) 3021–3033.
[48] M. Boczar, Ł. Boda, M.J. Wójcik, J. Chem. Phys. 125 (2006) 084709.
[49] M.J. Wójcik, J. Kwiendacz, M. Boczar, Ł. Boda, Y. Ozaki, Chem. Phys. 372 (2010)
72–81.
The obtained theoretical results are excellent source of informa-
tion, which is the complement to experimental results, giving the