X-ray crystal structures, molecular mechanics and quantum chemical calculations, and calculations of the nonlinear optical polarizabilities in the series of monohalogen substituted derivatives of dicyanovinylbenzene

Article abstract of DOI:10.1016/S0022-2860(02)00468-4

The molecular and crystal structure of a series of potential non-linear optical (NLO) compounds - o-, m- and p-halogen-substituted derivatives of dicyanovinylbenzene, has been studied. The molecular geometry was investigated with an X-ray diffraction anal

Full text of DOI:10.1016/S0022-2860(02)00468-4

Journal of Molecular Structure 650 (2003) 1–20
www.elsevier.com/locate/molstruc
X-ray crystal structures, molecular mechanics and quantum
chemical calculations, and calculations of the nonlinear optical
polarizabilities in the series of monohalogen substituted derivatives
of dicyanovinylbenzene
*
Mikhail Yu. Antipin , Vladimir N. Nesterov, Shan Jiang, Oleg Ya. Borbulevych,
David M. Sammeth, Ekateriva V. Sevostianova, Tatiana V. Timofeeva
Department of Chemistry, New Mexico Highlands University, Las Vegas, NM 87701, USA
Received 24 July 2002; accepted 12 September 2002
Abstract
The molecular and crystal structure of a series of potential non-linear optical (NLO) compounds — o-, m- and p-halogen-
substituted derivatives of dicyanovinylbenzene, has been studied. The molecular geometry was investigated with an X-ray
diffraction analysis and discussed along with results of the molecular mechanics (MM3) and ab initio quantum calculations
(MP2 and DFT). The influence of the crystal packing on the molecular planarity has been revealed. Calculations of the dipole
moments and molecular hyperpolarizabilities for molecules in question were carried out in the finite field approach. In addition
to a several acentric compounds of this series found before, in the present investigation one more acentric crystal of the p-I
substituted dicyanovinylbenzene was found (space group Pna21, Z ¼ 4). Second harmonic generation (SHG) of the laser light
for this compound was confirmed by the Kurtz powder test.
q 2002 Elsevier Science B.V. All rights reserved.
Keywords: Non-linear optical properties; Derivatives of dicyanovinylbenzene; X-Ray analysis; Molecular mechanics; Quantum chemical
calculations
1. Introduction
is that in many respects these compounds might be
more efficient materials (for second harmonic
generation (SHG), frequency mixing, electrooptical
modulation, photorefractive and optical limiting
applications, etc.) than the well-known inorganic
compounds like KDP, BaTiO3, LiNbO3. In many
cases the efficiency of organic NLO materials is at
least one or more order of magnitude above that of
the best inorganic ones due to ultrafast response
times and low dielectric constants. An additional
Organic chromophore compounds with non-linear
optical (NLO) properties have been under intensive
investigation during last years because of their
potential applicability in image processing and
optical communications. The main reason for this
*
Corresponding author. Tel.: 505-454-3208.
E-mail address: m_antipin@yahoo.com (M.Yu. Antipin).
0022-2860/03/$ - see front matter q 2002 Elsevier Science B.V. All rights reserved.
PII: S0022-2860(02)00468-4

2
M.Yu. Antipin et al. / Journal of Molecular Structure 650 (2003) 1–20
advantage of organic NLO compounds is high
optical damage thresholds and straightforward
methods of preparation and fabrication, especially
preparation of the thin crystalline films.
molecular mechanics calculations of the series of
monohalogen-substituted derivatives of dicyanovi-
nylbenzene (1). We considered series of twelve
monosubstituted dicianovinylbenzenes (2–13) with
the F, Cl, Br, and I substituents in the ortho-, meta-
and para-positions, respectively (see Scheme
below).
It is well recognized today that NLO-active
organic molecules must have a polar and highly
conjugated p-electron system terminated with elec-
tron donor and acceptor groups. Big progress has
been achieved during last years to create new NLO-
active compounds possessing very high values of
the second-order polarizability (b ) using molecular
design and crystal engineering methods. Progress in
this field is summarized in a series of modern
monographs [1–3] and recent original publications
[4–10].
Many important and new structure/property
relations for parent series of organic NLO com-
pounds were established and described in literature
as a result of systematical studies of simple model
chromophore compounds (see, for example [4–10]).
This allowed one to conduct a directional chemical
synthesis and preparation of new compounds
possessing necessary NLO characteristics. A big
progress in this field is related to a prior quantum
calculations of the molecular hyperpolarizabilities
and to estimation (and/or measurements) of their
optical properties and NLO crystalline responses.
Single crystal X-ray data provide in many respects
necessary information for understanding NLO struc-
ture/properties relations and for design of new NLO
active materials.
Substituent Position
Substituent
Ortho
Meta
Para
H
F
1
2
3
4
5
–
6
7
8
9
–
10
11
12
13
Cl
Br
I
Five new compounds of this series (4–6, 10 and
12) were first synthesized and investigated in details
by methods mentioned above. These data are
compared with results of our previous structural
investigations of the compounds 1 [6], 2, 3 [8], and
compound 11 which was described earlier by other
authors [15,16].
However, not all features in design of organic
NLO-active materials and crystals are completely
understood at present. In particular, it is known that
only non-centrosymmetric compounds may exhibit
NLO effects (for example, second harmonic
generation,) in the solid state [11,12], but a prior
prediction of the centric/acentric crystal structure of
a new organic compound is still a problem of the
modern computational and structural chemistry [13,
14]. Therefore, further search for new structure/
property relations in series of model compounds
and their systematical X-ray structural analysis are
important fields of research directed to under-
standing of the ‘rules’ governing formation of the
crystals of a given symmetry.
Results of the quantum calculations of the
molecules 1–13 in the isolated state and molecular
hyperpolarizabilities (b ) are also presented and
compared. Crystals of compounds 2, 5 and 11 are
acentric. Therefore they manifest second harmonic
(SH) of the laser light. In the present study we
measured UV spectra and powder second harmonic
generation activity for the compound 5 (space group
Pna21), and compared these data with results for other
acentric compounds in this series, namely 2 (space
group Pc) and 11 (space group P21) described before
[8,15].
In the present study we carried out synthesis,
X-ray structural analysis, quantum chemical and

M.Yu. Antipin et al. / Journal of Molecular Structure 650 (2003) 1–20
3
2. Experimental part and computational details
mechanics (MM) MM3 program [18,19] and
quantum chemical approaches. Using the MM3
program we found optimal conformation of 1–13,
and barriers of internal rotation of the dicyanovinyl
group around the single C(1)–C(7) bond (see
Figs. 3–5). These data were compared with two
sets of ab initio calculations of molecules 1–13. We
fully optimized molecular geometries with the
density functional method and using Moeller–
Plesset approximation. Molecules with the F, Cl
and Br substituents were optimized at the MP2/6-
31G** and B3LYP/6-31G** levels of theory, and
molecules with the iodine substituents were opti-
mized at the MP2/3-21G** and B3LYP/3-21G**
levels, respectively. All calculations were carried
out with the GAUSSIAN 94 programs [20]. For
unsubstituted molecule 1 the barrier of internal
rotation around the C(1)–C(7) bond was evaluated
using the MP2 approximation.
2.1. Materials
Synthesis of the 2-(2-bromophenyl)-1,1-dicyano-
vinyl (4), 2-(2-iodophenyl)-1,1-dicyanovinyl (5),
2-(3-fluorophenyl)-1,1-dicyanovinyl (6), 2-(4-fluoro-
phenyl)-1,1-dicyanovinyl (10), and 2-(4-bromophe-
nyl)-1,1-dicyanovinyl (12) was carried out using the
Knoevenagel reaction from the malononitrile and
corresponding F-, Br- or I-substituted benzaldehydes.
The same synthetic procedure we used before for the
compounds 2 and 3 [8]. All materials were recrys-
tallized from the ethanol and showed no impurities by
the GC-MS before use. Compounds were obtained
with the yields 71–79% and characterized with the H1
and C13 NMR and UV-visible spectra.
2.2. X-Ray diffraction analysis
Results of the quantum calculations (MP2) and X-
ray data were used further for calculations of the
molecular dipole moments m and molecular hyperpo-
larizabilities b and g in the finite field approach. For
calculations we used modified MOPAC program, AM1
approximation [21,22], and HYPER program [23].
Single crystals of 4–6, 10, and 12 for X-ray
analysis were grown by slow evaporation from the
ethanol solutions. Details of the data collection and
structure refinement are given in Table 1. Absorption
correction (SADABS program) was applied for com-
pounds 4 and 5.
The structures were solved by direct methods and
refined by a full-matrix least squares in anisotropic
approximation for non-hydrogen atoms. The hydro-
gen atoms were localized on difference Fourier maps.
Hydrogen atoms for structures 5, 10 and 12 were
refined with the ‘riding’ model, while for structures 4
and 6 they were refined isotropically. All calculations
were performed using the SHELXTL program [17],
Ver. 5.10 DOS/WIN95/NT. Bond lengths, bond
angles, and important torsion angles are summarized
in Tables 2–4 along with corresponding available
data for the compounds 1–3 [6,8] and 11 [15], and
with data obtained by quantum calculations. The
general view of molecules studied with the atomic
numbering schemes is presented in Figs. 1 and 2.
2.4. Second harmonic measurements
Since crystals of 5 are acentric, the Kurtz
technique [24] was applied to measure nonlinear
optical response for this compound in the powder
sample. Green SH (532 nm) signals were measured
with 1 Hz Q-switched Nd:YAG laser (1064 nm,
10 mJ per 7 ns pulse). In order to collect more
effectively the SH light generated and scattered in
the sample, the ellipsoid mirror was set up with the
sample in the primary focus, and the detector
(Hamamatsu HC-120 photomultiplier assembly) in
the secondary focus. The SH signals were averaged
by a digital storage scope to suppress noise. The
sample was loaded into a 250 mm thickness plastic
cell between two glass plates. The P(2v ) / P2(v )
second harmonic square dependence was
obtained for the compound 5 and for the reference
compound (urea).
2.3. Molecular mechanics and quantum chemical
calculations of the molecular geometry, rotational
barriers, and molecular hyperpolarizability
According to the Kurtz method [24], the phase-
matchable materials give SH intensity, independent
of, or increasing with, particles size, whereas SH
All molecular geometries of the isolated mol-
ecules 1–13 were calculated using the molecular

4
M.Yu. Antipin et al. / Journal of Molecular Structure 650 (2003) 1–20
Table 1
Crystal determination summary for compounds 4–6, 10 and 12
Molecule
4
5
6
10
12
Empirical formula
Formula weight
Temperature (K)
C
10H5BrN2
C10H5IN2
280.06
C10H5FN2
172.16
298(2)
0.71073
Monoclinic
4
C10H5FN2
172.16
298(2)
0.71073
Triclinic
2
C10H5BrN2
233.07
298(2)
0.71073
Triclinic
2
233.07
110(2)
0.71073
Monoclinic
4
298(2)
˚
Wavelength (A)
0.71073
Orthorhombic
4
Crystal system
Z
Space group
P21=c
Pna21
P21=c
P-1
P-1
Unit cell dimensions
˚
a (A)
˚
b (A)
3.8747(8)
21.379(4)
11.061(2)
90
7.8096(16)
17.109(3)
7.2817(15)
90
3.8678(15)
20.036(9)
11.046(5)
90
6.9617(16)
7.366(2)
9.151(3)
107.19(2)
99.01(2)
102.85(2)
424.4(2)
1.347
6.6019(13)
7.7472(15)
9.2558(19)
103.99(3)
97.58(3)
˚
c (A)
a (deg)
b (deg)
g (deg)
95.50(3)
90
90
95.27(3)
90
90
92.17(3)
3
˚
Volume (A )
912.1
972.9(3)
1.912
852.4(6)
1.342
454.15(16)
1.704
Density (calculated)
(Mg/m3)
1.697
Diffractometer
Smart CCD
4.455
Smart CCD
3.243
Siemens P3/PC
0.098
Siemens P3/PC
0.098
Siemens P3/PC
4.473
Absorption coefficient
(mm21
)
F(000)
456
528
352
176
228
u range for data collection
(deg)
1.9–30.0
2.4–30.1
2.8–26.0
2.4–29.1
2.3–26.1
Data/restraints/
2639/0/118
2844/1/118
1640/0/138
2267/0/118
1722/0/118
parameters
Goodness-of-fit on F 2
Reflections collected
Independent reflections
0.876
0.995
0.998
1.021
1.046
7378
11,092
1888
2443
1885
2639
2844
1640
2267
1722
[R(int) ¼ 0.0288]
R1 ¼ 0:0302;
wR2 ¼ 0:0560
R1 ¼ 0:0595;
wR2 ¼ 0:0628
0.864
[R(int) ¼ 0.0405]
R1 ¼ 0:0398;
wR2 ¼ 0:1085
R1 ¼ 0:0494;
wR2 ¼ 0:1139
1.886
[R(int) ¼ 0.0362]
R1 ¼ 0:0532;
wR2 ¼ 0:1058
R1 ¼ 0:1532;
wR2 ¼ 0:1284
0.144
[R(int) ¼ 0.0239]
R1 ¼ 0:0675;
wR2 ¼ 0:1066
R1 ¼ 0:1847;
wR2 ¼ 0:1323
0.126
[R(int) ¼ 0.0249]
R1 ¼ 0:0429;
wR2 ¼ 0:1050
R1 ¼ 0:0621;
wR2 ¼ 0:1129
0.462
Final R indices
[with I . 2sðIÞ]
R indices (all data)
Largest diff. peak
23
˚
(e A
)
Largest diff. hole
23
20.446
20.574
20.152
20.252
20.874
˚
(e A
)
from non-phase-matchable materials first increases,
reaching a maximum for particle size comparable to
the coherence length, then decreases. Due to small
amount of the sample 5 the test was carried out
using only available powder (with particle size 90–
180 mm). In this region the phase-matchable
materials attain their maximum SH intensity inde-
pendent of particle size, and non-phase-matchable
materials still varying inversely with the particle
size but very close to zero. Considerable intensity of
the SH signal was detected for the compound 5 (five
times greater then that for urea with the same
particle size), that assumes the phase-matchability of
the material studied.
Luminescence spectrum was measured for com-
pound 5. It showed no luminescent light that could
interfere with the SH signal at 532 nm. Absorption
spectra of solution (ethanol) and suspended powder
(in glycerol) were measured. The absorption bands
due to p ! p* transitions appeared at the UV region
of the spectrum (lmax 302 nm). No absorption was
observed at the 1064 nm and at 532 nm regions.

M.Yu. Antipin et al. / Journal of Molecular Structure 650 (2003) 1–20
5
Table 2
Bond length (A), bond angles(8) and important torsion angles(8) in ortho-substituted molecules 1–5
˚
Molecule
Method
1
2
3
4
5
Bond lengths
Hal–C
X-ray
MP2
–
–
–
1.363(3)
1.358
1.736(2)
1.740
1.900(2)
1.906
2.106(5)
2.157
B3LYP
X-ray
MP2
1.348
1.759
1.915
2.153
N(1)–C(9)
N(2)–C(10)
C(1)–C(6)
C(1)–C(2)
C(2)–C(3)
C(3)–C(4)
C(4)–C(5)
C(5)–C(6)
C(1)–C(7)
C(7)–C(8)
C(8)–C(9)
C(8)–C(10)
1.140(3)
1.184
1.149(3)
1.184
1.139(2)
1.184
1.154(3)
1.184
1.151(8)
1.191
B3LYP
X-ray
MP2
1.164
1.164
1.164
1.164
1.167
1.149(3)
1.184
1.144(4)
1.184
1.141(2)
1.184
1.138(3)
1.184
1.139(9)
1.191
B3LYP
X-ray
MP2
1.164
1.163
1.163
1.163
1.166
1.401(3)
1.407
1.413(3)
1.408
1.403(2)
1.407
1.400(3)
1.408
1.408(9)
1.416
B3LYP
X-ray
MP2
1.412
1.414
1.414
1.414
1.415
1.395(3)
1.408
1.398(3)
1.403
1.407(2)
1.408
1.404(3)
1.408
1.399(8)
1.415
B3LYP
X-ray
MP2
1.412
1.411
1.419
1.419
1.422
1.382(4)
1.392
1.375(4)
1.386
1.382(2)
1.394
1.382(3)
1.394
1.374(8)
1.406
B3LYP
X-ray
MP2
1.390
1.386
1.392
1.391
1.395
1.376(4)
1.396
1.389(4)
1.394
1.376(3)
1.394
1.382(3)
1.394
1.38(1)
1.406
B3LYP
X-ray
MP2
1.396
1.394
1.393
1.394
1.395
1.375(4)
1.397
1.400(4)
1.398
1.381(3)
1.397
1.383(3)
1.397
1.34(1)
1.407
B3LYP
X-ray
MP2
1.398
1.399
1.396
1.396
1.396
1.375(4)
1.393
1.387(4)
1.392
1.377(2)
1.391
1.376(3)
1.391
1.36(1)
1.403
B3LYP
X-ray
MP2
1.390
1.389
1.388
1.388
1.388
1.450(3)
1.457
1.456(3)
1.456
1.450(2)
1.459
1.457(3)
1.460
1.453(8)
1.480
B3LYP
X-ray
MP2
1.453
1.450
1.453
1.453
1.455
1.350(3)
1.360
1.356(3)
1.359
1.349(2)
1.358
1.352(3)
1.358
1.350(8)
1.358
B3LYP
X-ray
MP2
1.367
1.366
1.366
1.366
1.363
1.441(3)
1.435
1.437(3)
1.435
1.435(2)
1.436
1.431(3)
1.436
1.44(1)
1.442
B3LYP
X-ray
MP2
1.432
1.432
1.432
1.432
1.424
1.438(3)
1.437
1.448(3)
1.437
1.437(2)
1.436
1.448(3)
1.436
1.44(1)
1.441
B3LYP
1.435
1.435
1.436
1.436
1.429
Bond angles
C(6)–C(1)–C(2)
X-ray
MP2
118.5(2)
119.1
116.5(2)
117.2
116.6(1)
118.1
117.2(2)
116.6(5)
117.8
118.4
B3LYP
X-ray
MP2
118.4
116.6
116.7
116.7
117.2
C(6)–C(1)–C(7)
C(2)–C(1)–C(7)
124.8(2)
123.9
125.4(2)
125.1
124.1(1)
121.7
122.5(2)
121.6
122.9(5)
120.6
B3LYP
X-ray
MP2
125.0
126.1
123.6
123.3
122.9
116.7(2)
117.0
118.1(2)
117.7
119.3(1)
120.2
120.3(2)
120.5
120.5(5)
121.0
B3LYP
116.6
117.3
119.7
120.0
119.9
(continued on next page)

6
M.Yu. Antipin et al. / Journal of Molecular Structure 650 (2003) 1–20
Table 2 (continued)
Molecule
Method
1
2
3
4
5
C(3)–C(2)–C(1)
Hal–C(2)–C(1)
Hal–C(2)–C(3)
C(2)–C(3)–C(4)
C(3)–C(4)–C(5)
C(4)–C(5)–C(6)
C(5)–C (6)–C(1)
C(8)–C(7)–C(1)
C(7)–C(8)–C(9)
C(7)–C(8)–C(10)
C(9)–C(8)–C(10)
N(1)–C(9)–C(8)
N(2)–C(10)–C(8)
X-ray
MP2
120.8(2)
120.6
121.1
–
123.7(2)
123.1
121.5(4)
121.1
121.6(2)
121.5(6)
121.1
121.5
B3LYP
X-ray
MP2
123.2
121.8
121.8
121.0
118.4(2)
118.4
120.7(1)
120.4
120.6(2)
120.2
122.4(4)
120.3
–
B3LYP
X-ray
MP2
–
118.7
121.4
121.9
122.9
–
117.9(2)
118.5
117.8(1)
118.5
117.8(2)
118.2
116.1(4)
118.7
–
B3LYP
X-ray
MP2
–
118.1
116.8
116.4
116.1
119.6(2)
120.0
119.8
120.5(2)
119.8
119.9
120.6(2)
120.7
120.6
120.1(2)
119.9
120.2
130.9(2)
128.9
131.6
126.3(2)
124.7
125.4
119.0(2)
119.7
119.2
114.6(2)
115.6
115.4
176.7(2)
178.4
178.9
179.4(2)
179.6
179.7
118.7(2)
118.6
119.9(2)
119.7
119.6(2)
119.5
118.9(6)
119.7
B3LYP
X-ray
MP2
118.6
119.8
119.8
120.2
120.0(3)
120.1
120.2(2)
120.1
120.0(2)
120.1
121.3(7)
120.1
B3LYP
X-ray
MP2
120.2
120.0
119.9
119.9
120.3(2)
120.5
119.9(2)
120.0
120.3(2)
120.1
120.4(7)
120.0
B3LYP
X-ray
MP2
120.4
120.1
120.0
120.1
120.8(2)
120.6
121.8(2)
120.9
121.3(2)
120.9
121.2(6)
120.8
B3LYP
X-ray
MP2
121.0
121.7
121.8
121.6
129.6(2)
127.9
129.6(1)
126.0
126.9(2)
125.8
130.0(6)
124.2
B3LYP
X-ray
MP2
130.8
131.2
131.3
131.1
127.0(2)
124.6
125.7(1)
123.9
124.5(2)
123.8
126.9(8)
122.9
B3LYP
X-ray
MP2
125.5
126.2
126.3
126.0
119.1(2)
119.7
120.1(1)
120.1
120.6(2)
120.1
118.1(6)
120.8
B3LYP
X-ray
MP2
119.2
118.9
118.8
118.9
113.9(2)
115.6
114.2(1)
115.9
114.8(2)
116.0
115.0(5)
116.3
B3LYP
X-ray
MP2
115.3
114.9
114.9
115.1
177.8(3)
178.7
178.4(2)
178.5
177.4(2)
178.5
178.9(8)
178.5
B3LYP
X-ray
MP2
179.1
178.6
178.6
179.2
178.9(3)
179.8
179.0(2)
179.7
177.9(2)
179.8
179.4(8)
179.8
B3LYP
180.0
180.0
180.0
179.7
Torsion angles
C(6)–C(1)–C(7)–C(8)
X-ray
MP2
9.7(3)
23.1
5.4(2)
24.1
11.8(1)
37.5
230.9(3)
38.4
222(1)
49.8
B3LYP
X-ray
MP2
0.0
0.0
3.2
0.0
0.0
C(2)–C(1)–C(7)–C(8)
C(1)–C(7)–C(8)–C(9)
C(1)–C(7)–C(8)–C(10)
171.1(2)
157.9
180
174.9(2)
158.1
180
169.1(1)
145.1
177.2
0.8(1)
3.3
152.7(2)
144.4
180
159.7(6)
131.9
B3LYP
X-ray
MP2
180.0
1.1(3)
0.8
2.7(2)
1.7
26.8(3)
3.2
24(1)
2.8
0.0
B3LYP
X-ray
MP2
0.0
0.0
0.4
0.0
179.0(3)
179.9
180.0
179.0(2)
179.2
180.0
178.9(1)
178.1
179.8
175.3(2)
178.2
180.0
178.9(6)
177.9
180.0
B3LYP

M.Yu. Antipin et al. / Journal of Molecular Structure 650 (2003) 1–20
7
Table 3
Bond length (A), bond angles(8) and important torsion angles(8) in meta-substituted molecules 6–9
˚
Molecule
Method
6
7
8
9
Bond lengths
Hal–C
X-ray
MP2
1.359(3)
1.353
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
1.736
1.754
1.901
1.907
2.153
2.140
B3LYP
X-ray
MP2
1.346
N(1)–C(9)
N(2)–C(10)
C(1)–C(6)
C(1)–C(2)
C(2)–C(3)
C(3)–C(4)
C(4)–C(5)
C(5)–C(6)
C(1)–C(7)
C(7)–C(8)
C(8)–C(9)
C(8)–C(10)
1.137(3)
1.185
1.184
1.164
1.184
1.164
1.191
1.167
B3LYP
X-ray
MP2
1.164
1.143(3)
1.184
1.184
1.163
1.184
1.163
1.190
1.166
B3LYP
X-ray
MP2
1.163
1.396(3)
1.407
1.406
1.411
1.407
1.411
1.415
1.411
B3LYP
X-ray
MP2
1.412
1.397(3)
1.407
1.406
1.411
1.407
1.411
1.417
1.414
B3LYP
X-ray
MP2
1.411
1.361(3)
1.387
1.391
1.389
1.391
1.389
1.403
1.391
B3LYP
X-ray
MP2
1.385
1.363(4)
1.391
1.396
1.395
1.395
1.394
1.407
1.398
B3LYP
X-ray
MP2
1.391
1.369(3)
1.396
1.396
1.397
1.396
1.397
1.407
1.398
B3LYP
X-ray
MP2
1.397
1.378(3)
1.393
1.393
1.390
1.393
1.390
1.404
1.391
B3LYP
X-ray
MP2
1.391
1.446(3)
1.458
1.459
1.455
1.459
1.455
1.478
1.458
B3LYP
X-ray
MP2
1.455
1.345(3)
1.360
1.360
1.366
1.360
1.365
1.360
1.363
B3LYP
X-ray
MP2
1.366
1.433(3)
1.435
1.435
1.432
1.435
1.432
1.440
1.424
B3LYP
X-ray
MP2
1.432
1.441(3)
1.437
1.437
1.435
1.437
1.435
1.442
1.428
B3LYP
1.435
Bond angles
C(6)–C(1)–C(2)
X-ray
MP2
118.3(2)
119.4
–
–
–
119.5
118.8
–
119.4
118.7
–
119.4
B3LYP
X-ray
MP2
118.8
118.6
C(6)–C(1)–C(7)
C(2)–C(1)–C(7)
124.9(2)
123.8
–
123.9
125.1
–
123.9
125.1
–
123.3
B3LYP
X-ray
MP2
125.0
125.2
116.9(2)
116.8
–
117.3
116.6
116.1
116.6
116.2
B3LYP
116.2
116.2
(continued on next page)

8
M.Yu. Antipin et al. / Journal of Molecular Structure 650 (2003) 1–20
Table 3 (continued)
Molecule
Method
6
7
8
9
C(3)–C(2)–C(1)
Hal–C(3)–C(2)
Hal–C(3)–C(4)
C(2)–C(3)–C(4)
C(3)–C(4)–C(5)
C(4)–C(5)–C(6)
C(5)–C(6)–C(1)
C(8)–C(7)–C(1)
C(7)–C(8)–C(9)
C(7)–C(8)–C(10)
C(9)–C(8)–C(10)
N(1)–C(9)–C(8)
N(2)–C(10)–C(8)
X-ray
MP2
119.0(2)
119.1
–
–
–
119.8
120.3
–
119.6
120.3
–
119.8
120.7
–
B3LYP
X-ray
MP2
119.7
118.3(2)
118.7
119.4
119.4
–
119.2
119.4
–
119.3
119.6
–
B3LYP
X-ray
MP2
118.9
118.5(2)
119.0
119.7
119.7
–
119.5
119.6
–
119.7
120.0
–
B3LYP
X-ray
MP2
119.2
123.2(2)
122.3
121.0
120.9
–
121.3
121.0
–
121.0
120.4
–
B3LYP
X-ray
MP2
122.0
118.3(3)
118.2
118.9
119.0
–
118.7
118.9
–
118.9
119.3
–
B3LYP
X-ray
MP2
118.4
120.7(2)
121.1
121.1
121.1
–
121.1
121.1
–
120.9
121.0
–
B3LYP
X-ray
MP2
121.1
120.5(2)
119.9
119.6
120.0
–
119.8
120.1
–
119.9
120.0
–
B3LYP
X-ray
MP2
120.1
130.7(2)
128.5
128.4
131.3
–
128.4
131.3
–
127.3
130.8
–
B3LYP
X-ray
MP2
131.4
126.6(2)
124.6
124.6
125.4
–
124.6
125.4
–
124.0
125.0
–
B3LYP
X-ray
MP2
125.5
119.4(2)
119.7
119.7
119.2
–
119.7
119.2
–
120.2
119.4
–
B3LYP
X-ray
MP2
119.2
114.0(2)
115.7
115.7
115.4
–
115.7
115.4
–
115.8
115.6
–
B3LYP
X-ray
MP2
115.3
177.6(3)
178.5
178.5
178.8
–
178.5
178.9
–
178.6
179.5
–
B3LYP
X-ray
MP2
178.8
179.4(3)
179.6
179.6
179.6
179.6
179.6
179.8
179.8
B3LYP
179.6
Torsion angles
C(6)–C(1)–C(7)–C(8)
X-ray
MP2
26.4(4)
24.8
–
–
–
24.7
0.0
24.8
0.0
32.2
0.0
B3LYP
X-ray
MP2
0.0
C(2)–C(1)–C(7)–C(8)
C(1)–C(7)–C(8)–C(9)
C(1)–C(7)–C(8)–C(10)
173.6(2)
156.2
180
–
–
–
156.4
180.0
–
156.2
180.0
–
148.8
180.0
–
B3LYP
X-ray
MP2
–1.2(4)
0.9
0.9
0.8
0.7
B3LYP
X-ray
MP2
0.0
0.0
0.0
0.0
179.5(2)
179.9
180.0
–
–
–
179.9
180.0
179.9
180.0
180.0
180.0
B3LYP

M.Yu. Antipin et al. / Journal of Molecular Structure 650 (2003) 1–20
9
Table 4
Bond length (A), bond angles(8) and important torsion angles(8) in para-substituted molecules 10–13
˚
Molecule
Method
10
11
12
13
Bond lengths
Hal–C
X-ray
MP2
1.356(2)
1.352
1.718
1.734
1.748
1.123
1.185
1.164
1.112
1.184
1.163
1.403
1.408
1.412
1.387
1.408
1.412
1.396
1.390
1.388
1.370
1.396
1.396
1.376
1.397
1.398
1.396
1.390
1.388
1.459
1.456
1.452
1.333
1.361
1.367
1.439
1.434
1.432
1.440
1.436
1.435
1.889(4)
1.899
–
–
–
–
–
–
–
–
–
–
–
–
–
2.150
2.135
B3LYP
X-ray
MP2
1.341
1.901
N(1)–C(9)
N(2)–C(10)
C(1)–C(6)
C(1)–C(2)
C(2)–C(3)
C(3)–C(4)
C(4)–C(5)
C(5)–C(6)
C(1)–C(7)
C(7)–C(8)
C(8)–C(9)
C(8)–C(10)
1.130(3)
1.185
1.146(5)
1.185
1.191
1.167
B3LYP
X-ray
MP2
1.164
1.164
1.135(3)
1.184
1.135(5)
1.184
1.191
1.166
B3LYP
X-ray
MP2
1.164
1.163
1.397(3)
1.409
1.402(5)
1.408
1.419
1.412
B3lyp
X-ray
MP2
1.414
1.412
1.387(3)
1.408
1.384(6)
1.408
1.420
1.413
B3LYP
X-ray
MP2
1.413
1.412
1.378(3)
1.391
1.381(6)
1.391
1.401
1.390
B3LYP
X-ray
MP2
1.388
1.389
1.364(3)
1.390
1.382(6)
1.396
1.407
1.398
B3LYP
X-ray
MP2
1.392
1.395
1.359(3)
1.391
1.376(6)
1.397
1.408
1.400
B3LYP
X-ray
MP2
1.394
1.397
1.382(3)
1.391
1.381(5)
1.391
1.402
1.390
B3LYP
X-ray
MP2
1.388
1.388
1.454(3)
1.456
1.450(5)
1.456
1.474
1.455
B3LYP
X-ray
MP2
1.451
1.452
1.341(3)
1.361
1.344(6)
1.361
1.363
1.364
B3LYP
X-ray
MP2
1.367
1.367
1.437(3)
1.434
1.425(6)
1.434
1.438
1.423
B3LYP
X-ray
MP2
1.432
1.432
1.437(3)
1.436
1.450(5)
1.436
1.442
1.428
B3LYP
1.434
1.435
Bond angles
C(6)–C(1)–C(2)
X-ray
MP2
118.6(2)
118.9
120
118.5(3)
–
118.7
118.1
124
118.7
118.4
118.1
–
B3LYP
X-ray
MP2
118.2
118.1
C(6)–C(1)–C(7)
C(2)–C(1)–C(7)
124.4(2)
124.3
124.2(4)
124.3
124.4
125.2
117
125.1
125.1
–
B3LYP
X-ray
MP2
125.1
125.1
117.0(2)
116.9
117.2(3)
116.9
117.0
116.8
116.5
116.9
B3LYP
116.7
116.8
(continued on next page)

10
M.Yu. Antipin et al. / Journal of Molecular Structure 650 (2003) 1–20
Table 4 (continued)
Molecule
Method
10
11
12
121.7(3)
13
C(3)–C(2)–C(1)
Hal–C(4)–C(3)
Hal–C(4)–C(5)
C(2)–C(3)–C(4)
C(3)–C(4)–C(5)
C(4)–C(5)–C(6)
C(5)–C (6)–C(1)
C(8)–C(7)–C(1)
C(7)–C(8)–C(9)
C(7)–C(8)–C(10)
C(9)–C(8)–C(10)
N(1)–C(9)–C(8)
N(2)–C(10)–C(8)
X-ray
MP2
121.4(2)
121.1
121
–
121.2
121.6
119
121.2
121.3
121.5
–
B3LYP
X-ray
MP2
121.6
121.6
118.2(2)
119.0
119.1(3)
119.5
119.7
119.5
119
119.8
119.9
–
B3LYP
X-ray
MP2
119.0
119.5
118.4(2)
118.9
119.3(3)
119.5
119.6
119.5
118
119.8
119.8
–
B3LYP
X-ray
MP2
118.9
119.4
117.8(2)
118.4
118.4(4)
119.0
119.2
118.9
122
119.3
119.4
–
B3LYP
X-ray
MP2
118.3
118.9
123.3(2)
122.1
121.6(3)
121.0
120.7
121.0
120
120.4
120.3
–
B3LYP
X-ray
MP2
122.1
121.1
118.8(2)
119.1
119.5(4)
119.7
119.9
119.7
119
120.1
120.1
–
B3LYP
X-ray
MP2
119.1
119.7
120.1(2)
120.4
120.3(4)
120.4
120.4
120.7
131
120.5
120.7
–
B3LYP
X-ray
MP2
120.7
120.7
131.6(2)
129.3
131.1(3)
129.2
129.3
131.5
125
130.3
130.8
–
B3LYP
X-ray
MP2
131.6
131.4
125.5(2)
124.7
125.5(3)
124.8
124.8
125.3
121
124.9
125.0
–
B3LYP
X-ray
MP2
125.2
125.3
119.8(2)
119.6
119.6(4)
119.6
119.6
119.2
114
119.6
119.4
–
B3LYP
X-ray
MP2
119.2
119.2
114.8(2)
115.7
114.9(4)
115.7
115.7
115.5
177
115.5
115.6
–
B3LYP
X-ray
MP2
115.5
115.5
178.1(2)
178.8
177.6(4)
178.7
178.8
179.0
177
179.5
B3LYP
X-ray
MP2
179.0
179.6
179.4(3)
179.6
178.5(5)
179.6
–
179.6
179.6
180.0
179.8
B3LYP
179.7
179.6
Torsion angles
C(6)–C(1)–C(7)–C(8)
X-ray
MP2
25.2(3)
19.7
1
19.3
0.0
3.3(7)
19.7
–
0.0
0.0
–
B3LYP
X-ray
MP2
0.0
0.0
C(2)–C(1)–C(7)–C(8)
C(1)–C(7)–C(8)–C(9)
C(1)–C(7)–C(8)–C(10)
175.6(2)
161.4
180
179
161.7
180
5
2178.9(4)
161.3
180.0
180.0
–
B3LYP
X-ray
MP2
180.0
20.6(4)
0.6
1.3(7)
0.6
0.6
0.0
0.0
–
B3LYP
X-ray
MP2
0.0
0.0
0.0
179.3(2)
179.9
180.0
176
179.9
180.0
2178.1(4)
179.8
180.0
180.0
B3LYP
180.0

M.Yu. Antipin et al. / Journal of Molecular Structure 650 (2003) 1–20
11
Fig. 2. General view of molecules 10 and 12 with atomic numbering
scheme.
nine of them crystal structures are determined, and all
thirteen compounds are characterized by the MM and
quantum calculations. This allows one to carry out a
detailed comparison of the molecular geometry,
crystal packing features and NLO properties in this
representative series.
3.1. Molecular geometry and crystal packing
Molecular geometry parameters obtained from the
X-ray data for compounds 4–6 and 10, 12 at present
work along with experimental results published
before (molecules 1–3, and 11), and the results of
ab initio calculations for the whole series 1–13 are
presented in Tables 2–4 separately for the ortho, meta
and para-substituted compounds, respectively. Corre-
sponding bond lengths and bond angles for most of
molecules are similar (except for Hal–C distances)
and have expected values. Nevertheless, some
geometrical features are worth to mention.
Fig. 1. General view of molecules 4–6 with atomic numbering
scheme.
3. Results and discussion
The presented series of the compounds is quite
unique because it contains very similar structures
differing only by nature and position of the halogen
substituents (F, Cl, Br, and I) at the aromatic ring. For
In particular, there is a small tendency in the C–C
bond alternation in the aromatic rings: bond lengths

12
M.Yu. Antipin et al. / Journal of Molecular Structure 650 (2003) 1–20
Fig. 3. Relative energy of molecule 1 vs torsion angle C(6)–C(1)–C(7)–C(8) (8).
C(2)–C(3) and C(5)–C(6) are somewhat shorter (by
˚
0.01–0.02 A) than those for the four other ring bonds.
cases. Earlier, the same trend was found to be more
pronounced for some other substituted derivatives of
dicyanovinylbenzene [6–8].
This may reflect some contribution of the quinoid
structure to electronic structure of these molecules,
but this trend is not very distinct in all investigated
According to both experimental data and quantum
calculations, slight elongation of the phenyl rings
Fig. 4. Energy of molecule 10, 11, 12 and 13 vs torsion angle C(6)–C(1)–C(7)–C(8) (8).

M.Yu. Antipin et al. / Journal of Molecular Structure 650 (2003) 1–20
13
Fig. 5. Energy of molecule 2, 3, 4 and 5 vs torsion angle C(6)–C(1)–C(7)–C(8 (8).
bonds C(1)–C(2) and C(1)–C(6) attached to the
acceptor dicyanovinyl substituent has been found
(Tables 2–4). Halogen substituents in the ring do not
influence geometry of dicyanovinyl fragment, namely
the C(1)–C(7) and C(7)–C(8) bonds, see Tables 2–4.
The values of these bonds vary in the rather narrow
angle. This tendency withstands only in part,
however, because in the Cl and Br derivatives
corresponding angles are almost equal. In the latter
case steric effects are partly compensated by the
difference of the Cl–C and Br–C bond lengths. The
same trend is reproduced by quantum calculations.
Moreover, for the related bond angles C(2)–C(1)–
C(7) in the series 1–5, corresponding value for the
ortho-I-derivative (5) is even smaller than that for
ortho-Br derivative (4). We can explain this by the
weak intramolecular electrostatic attractive inter-
action of the large and polarizable iodine atom with
the positively charged H(7) atom that shrinks the
C(2)–C(1)–C(7) bond angle.
˚
˚
intervals (1.44–1.45 A and 1.34–1.36 A, respect-
ively) according to both experimental and theoretical
data, as for most other derivatives of the dicyanovi-
nylbenzene studied before [6–8].
According to the X-ray data significant differences
in the molecular geometry were found only for some
bond and torsion angles. In particular, a decrease of
the ipso-bond angles C(1)–C(2)–C(3) for the o-
substituted 2-5 (in the sequence F-Cl-Br-I), and
especially C(3)–C(4)–C(5) bond angles for the p-
substituted molecules 10–12 (F, Cl, Br) reflects a
gradual decrease of the electronegativity of the
halogen substituents. This effect is well documented
in literature [25]. For instance, for all F-substituted
compounds studied (2, 6, and 10) corresponding ipso-
angles are very close to the value of 123.48 - tabulated
in [26] for F-substituted benzenes. Another significant
difference was found for the Hal–C(2)–C(1) angles in
molecules 2–5 which may result from steric demands
of the o-halogen atoms: the larger is atomic volume of
the substituent, the larger is corresponding bond
As we may expect, the most varying values in the
series of the compounds studied should be torsion
angles, namely angles about the single C(1)–C(7)
bond which result to molecular non-planarity. In
particular, the s-cis torsion angles C(6)–C(1)–C(7)–
C(8) vary in the interval from 21.08 to 230.98,
reaching the maximum values 230.9 and 222.48 for
the bulky o-Br and o-I substituents in the structures 4
and 5, respectively (for all other molecules this angle
is less than 128). Nevertheless, there is no direct
correlation between the nature of the o- and other
substituents and molecular planarity degree: all p-
substituted molecules 10–12, as well as m-F (6) and

14
M.Yu. Antipin et al. / Journal of Molecular Structure 650 (2003) 1–20
Table 5
Some important intramolecular short contacts in structures 4–6 and 10–12
Contact
1
2
3
4
5
6
10
11
12
H(6)…C(9)
C(6)…C(9)
2.52
3.11
2.28
2.52
3.10
2.35
2.42
3.04
2.54
2.50
3.03
2.81
2.55
3.11
2.81
2.05
3.09
–
2.45
3.08
–
2.35
3.02
–
2.42
3.05
–
C(2)–X…H(7)
˚
The sum of the van der Waals radii (A) of the contacting atoms: H…C ¼ 2.87; C…C ¼ 3.54; H…H ¼ 2.20; F…C ¼ 2.56; Cl…C ¼ 2.86;
Br…C ¼ 2.97; I…C ¼ 3.13 [26].
o- F (2) are almost planar, while non-substituted
molecule 1 in not planar and corresponding torsion
angle (9.78) is almost the same as in the o-Cl (3)
molecule (11.88). Similar tendency was observed for
some other derivatives of dicyanovinylbenzene stu-
died before [6–8]. These results may reflect both the
influence of the intramolecular interactions and the
influence of crystal packing on the molecular
conformation.
For all these molecules anti-orientation of the
dicyanovinyl group to the halogen-substituent is
optimal. This is in a agreement with experimental
conformations of molecules 2–5 in crystals and
results of the gas phase investigation of the 2-
bromostyrene [27] where the torsion angle C(2)–
C(1)–C(7)–C(8) was found to be 1558 according to
the gas electron diffraction data, and to 1388 according
to the ab initio calculations at the MP2/6-31G(d)
level. For the ortho-substituted molecules the degree
of non-planarity depends on sterical and electronic
requirements of the halogen-substituent but it also
may be influenced by the molecular packing in
crystal.
The curves characterizing internal rotation of the
dicyanovinyl group in the molecule 1 and obtained by
ab initio (MP2) and MM calculations are compared in
Fig. 3. The most pronounced dissimilarity of the MP2
and MM curves is located near the planar molecular
conformation. If MM calculation gives energy
difference ,0.8 kcal/mol in favor of non-planar
conformation, according to MP2 calculation this
difference is negligibly small (,0.1 kcal/mol).
Nevertheless, qualitative ab initio and MM results
are in good agreement, therefore for other molecules
2–13 only MM calculations of the conformational
energy versus the torsion angle C(2)–C(1)–C(7)–
C(8) have been done (Figs. 4 and 5).
Tables 5 and 6 summarize important intra- and
intermolecular shortened contacts in the structures
studied, in particular the H(6)…C(9), C(6)…C(9), and
H(7)…Hal (at the C(2) atom) intramolecular contacts,
and intermolecular contacts of the halogens and H(7)
atoms. It is obvious that all these contacts are shorter
than the sum of the van der Waals radii of
corresponding atoms [28]. The halogen atoms form
a small number of intermolecular shortened contacts,
while the most important contacts are H(7)…N(1,2)
ones which may be considered as week hydrogen
bonds C–H…N. These H-bonds were found in all
structures studied except for compound 5 where the
ortho-iodine atom probably prevents from formation
of such a bond. Due to formation of these weak H-
bonds molecules in crystals form infinite chains (4, 6),
centrosymmetric dimers (12), or both intermolecular
associates (10). It was noted earlier [6] that ‘shielding’
of the acidic H(7) atom participating in the H-bond
formation (due to a presence of o-substituent) may
result in formation of the acentric crystal structure.
Thus, crystal structures with o-F and o-I substituents
(2 and 5) as well as with o-methoxy-group [6] are non-
centrosymmetric and active in SHG in the solid state.
Both quantum and MM calculations allow one
to conclude that for the meta- and para-substituted
molecules there are no serious intramolecular
restrictions on rotation of two planar molecular
fragments (benzene ring and dicyanovinyl frag-
ment) with respect to the C(1)–C(2) bond in the
vicinity ^308 of planar conformation. Correspond-
ing potential curves are almost identical, therefore
we present only data for the para-substituted
molecules 10–13 (Fig. 4). It is most likely that
in the crystals of 6–13 molecular conformation is
defined by intermolecular interactions. For the
ortho-substituted molecules 2–5 the halogen’s
influence on the molecular conformation is more
pronounced (Fig. 5).

M.Yu. Antipin et al. / Journal of Molecular Structure 650 (2003) 1–20
15
Table 6
Some important short intermolecular contacts and hydrogen bonds in structures 4–6, 10, and 12
˚
Contact distance (A)
Crystal
Contact atoms
Symm. code of primed atom
Angle C–H…N (8)
Sum of the van der
˚
Waals radii (A) [26]
4
Br(1)…C(10)0
C(7)…N(1)0
H(7)…N(1)0
I(1)…N(1)0
C(9)…C(10)0
C(7)…N(1)0
H(7)…N(1)0
C(6)…C(7)0
C(7)…N(1)0
H(7)…N(1)0
C(7)…N(2)0
H(7)…N(2)0
Br(1)…N(1)0
Br(1)…C(9)0
C(7)…N(2)0
H(7)…N(2)0
x, 0.5 2 y, z 2 0.5
x 2 1, 0.5 2 y, z 2 0.5
x 2 1, 0.5 2 y, z 2 0.5
1.5 2 x, 0.5 þ y, z 2 0.5
21 þ x, y, z
3.58
3.30
2.40
3.34
3.38
3.30
2.57
3.41
3.58
2.75
3.37
2.71
3.27
3.50
3.54
2.63
–
3.64
3.41
2.74
3.67
3.54
3.41
2.74
3.54
3.41
2.74
3.41
2.74
3.51
3.64
3.41
2.74
–
160
–
5
6
–
1 þ x, 0.5 2 y, 0.5 þ z
1 þ x, 0.5 2 y, 0.5 þ z
2x, 2y, 2 2 z
–
162
–
10
12
1 þ x, y, z
–
1 þ x, y, z
149
–
2x, 2y, 1 2 z
2x, 2y, 1 2 z
129
–
1 þ x, y, 1 þ z
1 þ x, y, 1 þ z
–
2x, 2y, 2z
2x, 2y, 2z
–
166
On the other hand, the o-Cl, o-Br (3, 4) and o-NO2 [8]
derivatives form centrosymmetric crystal structures.
Therefore, search for polymorphism and theoretical
modeling of other crystal packing arrays in this series
may be important for understanding the ‘rules’
dictating formation of a given crystal.
the C–H…N hydrogen bonds in these structures is
different (angle at the H atom), and in addition, the
second C(7)–H(7)…N(1) H-bond appears in the
structure of 10 which connects molecules into infinite
chains.
Molecular packing diagrams for structures studied
are presented in Fig. 6. Several interesting features of
the crystal packing mode may be noted. In particular,
unit cell parameters of crystals 4 (o-Br) and 6 (m-F)
are very similar (with the same space group), and in
both structures molecules form layers along the
3.2. Comparison results of quantum MP2 and DFT
calculations
Good agreement between the calculated (MP2
and DFT) and experimental (X-ray) values was
obtained for most molecular geometry parameters
except the angles of rotation around the C(1)–C(7)
bond. The DFT method gives slightly better results
for bond angles but very often the MP2/DFT
difference is close to the corresponding experimental
s-values. The most significant difference between
the two methods was found in description of
molecular planarity. For each molecule 1–13 we
optimized molecular geometry starting from two
models: planar with the symmetry Cs, and non-
planar with torsion angle C(2)–C(1)–C(7)–C(8)
equal to 208. Some results of computations are listed
in Table 7.
˚
shortest unit cell parameter a (about 3.9 A, see
Table 1). In both crystal structures the C–H…N
hydrogen bonds join molecules in the infinite chains
along the c-crystal axes. The same crystal packing
array and unit cell parameters were found earlier for
the o-Cl derivative 3 [8], so we may consider these
compounds as isostructural ones.
For the pair p-F and p-Br (10 and 12), the
parameters of the triclinic unit cells are also rather
close (but unit cell angles differ more significantly),
and molecules in the layer form centrosymmetric
dimers via the same C–H…N hydrogen bonds
(Fig. 6). On the other hand, molecular superposition
in neighboring layers is slightly different: there is no
molecular overlap between rings in 10, while some
small overlap exists in 12. As a result, geometry of
Presented data show that MP2 predicts non-planar
structures in the case of ortho-substituted molecules
with noticeable energy difference, while DFT calcu-
lations from both starting points result into planar

16
M.Yu. Antipin et al. / Journal of Molecular Structure 650 (2003) 1–20
Fig. 6. Molecular packing diagrams for compounds 4–6, 10 and 12.
molecular geometry. It is also obvious that for the
meta- and para-substituted molecules the energy
difference between the planar and non-planar struc-
tures is negligible (MP2), so there it is no reason to
discuss different results of two methods in the two
latter cases. Nevertheless, inconsistency of the DFT
method for description of conjugated molecules that
was emphasized before [29] should be also pointed
out for ortho-substituted molecules.
3.3. Evaluation of NLO properties
Fig. 7 and Table 8 present evaluation of the
molecular hyperpolarizability (b) for the compounds
of interest calculated with molecular geometry
parameters obtained by the MM (for planar and non-
planar conformations), ab initio (MP2), and X-ray
methods. Table 8 contains also calculated dipole
moments (m ) and third-order polarizabilities (g )

M.Yu. Antipin et al. / Journal of Molecular Structure 650 (2003) 1–20
17
Table 7
together with available experimental values for the
b’s. All methods give quite similar results: in accord
with the substituent position molecular hyperpolariz-
ability slightly increases in the sequence of the
ortho ! meta ! para substituted molecules. This
may be explained by the lower molecular planarity
and shortest possible length of conjugation in the case
of ortho-substituted compounds. The similar effect
was described before for the series of methoxy-
substituted dicianovinylbenzenes by theoretical cal-
culations and experimental EFISH measurements [6,
7], but in our case this tendency is much weaker
because of electron-acceptor properties of the halogen
substituents (in contrast to the donor properties of the
OMe-group).
Energy difference for planar (pl) and non-planar (npl) confor-
mations of molecules 1–13
Molecule
Method
t C(6)–C(1)–
C(7)–C(8)(8)
DE(kcal/mol)
1
MP2
23.1
0.0
0.10
–
B3LYP
MP2
2
24.1
0.0
0.13
–
B3LYP
MP2
3
37.5
3.2
0.96
0.01
1.08
–
B3LYP
MP2
4
38.4
0.0
B3LYP
MP2
5
49.8
0.0
1.51
–
B3LYP
MP2
6
24.8
0.0
0.13
–
B3LYP
MP2
Molecules 4 and 5 have the lowest hyperpolar-
izabilities due to their largest non-planarity. Intro-
duction of the halogen substituent in the 4-position
of the aromatic ring (molecules 10–13) results in
higher hyperpolarizability. Planar molecular struc-
ture and better conditions for electron conjugation
and charge transfer result in larger values of b and g
in this series. Meta-substituted molecules (6–9)
have the same magnitudes of b and g as
unsubstituted compound (1).
7
24.7
0.0
0.12
–
B3LYP
MP2
8
24.8
0.0
0.12
–
B3LYP
MP2
9
32.2
0.0
0.06
–
B3LYP
MP2
10
11
12
13
19.7
0.0
0.05
–
B3LYP
MP2
19.3
0.0
0.04
–
B3LYP
MP2
19.7
0.0
0.04
–
B3LYP
MP2
There are only four experimental b values
available for this series (Table 8) which were
obtained in solution by EFISH technique [8].
0.0
–
B3LYP
0.0
–
Fig. 7. Calculated second order polarizabilities b (10251 cm3 V22) of molecule 1–13.

18
M.Yu. Antipin et al. / Journal of Molecular Structure 650 (2003) 1–20
Table 8
in correction factors between 2.0 and 3.4. Although
the present data are not sufficient, they give some
clue to estimate approximately the order of the
molecular b values in solutions. In present case, the
correction factors (experimental b/calculated b) are
in the vicinity of 1.5 (molecule 1), 3.5 (molecule 2
and 3) and 2.0 (molecule 11). Some recent
approaches to estimate solvent effects on the
molecular hyperpolarizabilities are considered, in
particular, in [31].
Calculated dipole moments (m, D), second-order polarizabilities b
(10251 Cm3 V22) and g (10261 Cm4 V23) for molecules 1–13
Moleculea
Method
m
b
g
Exptl bb
1
MM3-npl
MP2
4.396
4.641
4.646
4.371
4.639
4.320
4.197
4.357
4.151
4.117
4.285
3.932
4.080
4.213
3.903
2.933
3.098
2.892
3.150
3.322
3.130
3.301
3.162
3.251
3.210
3.441
3.071
3.310
3.528
3.242
3.241
3.430
3.189
3.241
3.416
10.961
13.831
12.889
8.186
3.060
3.765
3.505
2.966
3.753
3.625
2.761
3.325
3.588
2.639
3.265
3.076
2.441
2.383
3.176
3.061
3.755
3.509
3.352
4.121
3.404
4.189
3.374
3.930
3.645
4.689
3.944
4.320
5.715
4.492
4.495
5.924
5.231
4.320
6.032
X-ray
22.75
34.16
36.38
2
3
4
5
6
MM3-npl
MP2
10.596
10.030
6.933
X-ray
MM3-npl
MP2
As we noted earlier, compound 5 forms acentric
crystals with the space group Pna21 and therefore it
is active in the SHG in the solid (powder) state.
Our measurements have shown that the efficiency
of the compound 5 (despite rather small molecular
NLO response) is about 5 times more than that for
urea, and this compound is phase-matchable. Using
the locally written NLOP program [32], the
angle between the crystal polar axis c and vectorial
part of the b value was calculated and it was found
to be equal to 58.88 which is very close to the
optimal value 54.78 [11,12]. The fragment of
8.629
X-ray
9.915
MM3-npl
MP2
6.282
8.019
X-ray
7.871
MM3-npl
MP2
5.938
5.660
X-ray
8.363
MM3-npl
MP2
10.186
12.062
10.530
11.394
13.661
11.392
13.648
11.884
13.295
15.393
20.042
16.616
17.509
22.859
18.706
16.903
21.808
19.055
15.908
21.297
X-ray
7
MM3-npl
MP2
8
MM3-npl
MP2
the crystal packing of
5
along with the
orientation of the b’s vectorial part is presented
in Fig. 8.
9
MM3-npl
MP2
10
MM3-npl
MP2
So, three compounds of the series studied (2, 5
and 11) were found to be active in the SHG in the
solid state with the efficiency about 12, 5 and 20
times more than that of urea as a reference
compound. The data of Table 8 along with crystal
data for these compounds allow one to understand
and explain this difference. Compound 11 has the
largest calculated molecular b value in this series
(about 3 times more than that for 5) and molecular
packing array (space group P21) is close to the
optimal one [8]. Compound 2 (space group Pc) is
less active than 11 and more active than 5 (because
of larger b value). Corresponding EFISH data for 11
and 2 are in line with this observation.
X-ray
11
12
MM3-npl
MP2
X-ray
46.40
MM3-npl
MP2
X-ray
13
MM3-npl
MP2
a
Obtained from MM3, MP2 calculations and X-ray geometry (X-
ray); npl: non-planar.
b
Experimental data were obtained by the EFISH method in
solution (chloroform), see Refs. [6,8] for details.
The difference between the calculated and exper-
imental b values might be explained by the solvent
and dispersion effects. A comparison has been done
earlier between the calculated and experimental b
values for different series of organic compounds (in
particular, methoxy-derivatives of dicyanovinylben-
zene [6,7], Schiff’s bases [30], p-nitroaniline
derivatives [23], and some others) which resulted
4. Supporting information available
Tables of the X-ray structure determination
summary, tables of the non-hydrogen atoms coordi-
nates and their anisotropic displacement parameters,
tables of hydrogen atoms coordinates and their

M.Yu. Antipin et al. / Journal of Molecular Structure 650 (2003) 1–20
19
Fig. 8. Fragment of crystal packing of 5 along with the orientation of vectorial part of molecular hyperpolarizability.
[5] F. Meyers, S.R. Marder, B.M.N. Pierce, J.L. Bredas, J. Amer.
Chem. Soc. 116 (1994) 10703–10714.
isotropic displacement parameters, bond lengths,
bond angles, and torsion angles.
[6] M.Y. Antipin, T.A. Barr, B. Cardelino, R.D. Clark, C.E.
Moore, T. Myers, B. Penn, M. Romero, M. Sangadasa, T.V.
Timofeeva, J. Phys. Chem. B 101 (1997) 2770.
Acknowledgements
[7] M.Y. Antipin, T.V. Timofeeva, R.D. Clark, V.N. Nesterov, M.
Sanghadasa, L. Romero, M. Romero, B. Penn, J. Phys. Chem.
A 102 (1998) 7222.
We gratefully acknowledge financial support of
this research by NASA grant NAG8-1708, and NASA
cooperative agreement NCC8-195-A. We also thank
K.Suponitsky for his help with quantum chemical
calculations.
[8] T.V. Timofeeva, V.N. Nesterov, M.Y. Antipin, R.D. Clark, M.
Sanghadasa, B.H. Cardelino, C.E. Moore, D.O. Frazier,
J. Molec. Struct. 519 (2000) 225–241.
[9] O.Y. Borbulevych, R.D. Clark, A. Romero, L. Tan, M.Y.
Antipin, V.N. Nesterov, B.H. Cardelino, C.E. Moore, M.
Sanghadasa, T.V. Timofeeva, J. Molec. Struct. 604 (2002)
73–86.
[10] G. Chen, S. Mukamel, J. Phys. Chem. 100 (1996)
11080–11085.
References
[11] J.L. Oudar, J. Zyss, Phys. Rev. A 26 (1982) 2016–2027.
[12] J. Zyss, J.L. Oudar, Phys. Rev. A 26 (1982) 2028–2048.
[13] A. Gavezzotti, Acc. Chem. Res. 27 (1994) 309–314.
[14] J.J. Wolff, Angew. Chem. Int. Ed. 35 (1996) 2195–2197.
[15] Y. Delugeard, Cryst. Struct. Commun. 4 (1975) 289.
[16] F. Messager, Bull. Soc. Fr. Miner. Crist. 89 (1966) 134.
[17] Sheldrick, G.M., SHELXTL PLUS. PC Version, A system of
computer programs for the determination of crystal structure
from X-ray diffraction data, Rev. 1994 502.
[1] H.S. Nalwa, S. Miyata (Eds.), Nonlinear Optics of Organic
Molecules and Polymers, CRC Press, Boca Raton, 1997.
[2] M.C. Kuzyk, C.W. Dirk (Eds.), Characterization Techniques
and Tabulations for Organic Nonlinear Optical Materials,
Marcel Dekker, inc.,, New York, 1998.
[3] J. Zyss (Eds.), Molecular Nonlinear Optics: Materials,
Physics, and Devices, Physics, and Devices, Academic
Press, 1994.
[4] D.R. Kanis, M.A. Ratner, T.J. Marks, Chem. Rev. 94 (1994)
195–242.
[18] N.L. Allinger, Y.H. Yuh, J.-H. Lii, J. Amer. Chem. Soc. 111
(1989) 8551.

20
M.Yu. Antipin et al. / Journal of Molecular Structure 650 (2003) 1–20
[19] J.-H. Lii, N.L. Allinger, J. Amer. Chem. Soc. 111 (1989) 8566.
[20] GAUSS94 W: Gaussian-94, Rev. A1; Gaussian, Inc., Pitts-
burg, PA, 1995.
Importance, Oxford University Press, Oxford, 1992, chapter
18.
[27] Q. Shen, J. Kuhns, K. Hagen, A.D. Richardson, J. Molec.
Struct. 567-568 (2001) 73.
[21] M.J.S. Dewar, E.G. Zoebisch, E.F. Healy, J. Amer. Chem.
Soc. 107 (1985) 3902.
[28] R.S. Rowland, R. Taylor, J. Phys. Chem. 100 (1996) 7384.
[29] B. Champagne, E. Perpete, D. Jacquemin, S. Van Gisbergen,
E. Baerends, C. Soubra-Ghaoui, K. Robins, B. Kirtman,
J. Phys. Chem. A 104 (2000) 4755.
[22] B.H. Cardelino, C.E. Moore, D.O. Frazier, J. Phys. Chem. A
101 (1997) 2207.
[23] B.H. Cardelino, C.E. Moore, R.E. Stickel, J. Phys. Chem. 95
(1991) 8645–8652.
[30] K.N. Bhat, M. Aggarwal, B.H. Cardelino, C.E. Moore, B.
Penn, D. Fraizer, M. Sanghadasa, T.A. Barr, Comput. Mater.
Sci. 8 (1997) 309.
[24] S.K. Kurtz, T.T. Perry, J. Appl. Phys. 39 (1968) 3798.
[25] A. Domenicano, I. Hargittai, G. Schultz, The 8th Austin
Symposium on Molecular Structure, Austin, Texas, USA,
1980, pp. 17–21.
[31] W. Bartkowiak, R. Zalesny, W. Niewodniczanski, J. Leszc-
zynski, J. Phys. Chem. A 105 (2001) 10702–10710.
[32] K.Yu. Suponitsky, T.V. Timofeeva, M.Yu. Antipin, to be
published, 2003.
[26] A. Domenocano, in: A. Domenicano, I. Hargittai (Eds.),
Accurate Molecular Structures: Their Determination and