Characterization of a new non-centrosymmetric polymorph of diphenyl-1,3,4-oxadiazole

Article abstract of DOI:10.1016/S0022-2860(02)00569-0

Diphenyl-1,3,4-oxadiazole (DPO) crystallization experiments from solutions clearly reveal the polymorphism of the substance. Besides the formerly known centrosymmetric monoclinic structure with space group P2₁/c (DPO I) a new monoclinic structure with the non-centrosymmetric space group Cc is found (DPO II): a = 2.4134(4) nm, b = 2.4099(3) nm, c = 1.2879(2) nm, β = 110.048(3)° and V = 7.0363(17) nm³. The asymmetric unit contains six independent molecules in a complex packing motif. A re-determination of the crystal structure of DPO I at room temperature gives lattice parameters a = 0.51885(6) nm, b = 1.8078(2) nm, β = 1.21435(14) nm, β = 93.193(3)°, and V = 1.1373(2) nm³. X-ray measurements at 363 K show a significant increase of the unit cell volume by 1.6%. Differences between both structures concerning morphology and characteristic Raman bands are outlined in detail. DSC investigations show an irreversible transition from DPO I to DPO II at 97°C. DPO II does not show any transition in the temperature range up to the melting point at 141°C. The non-centrosymmetric DPO II structure shows triboluminescence.

Full text of DOI:10.1016/S0022-2860(02)00569-0

Journal of Molecular Structure 649 (2003) 219–230
www.elsevier.com/locate/molstruc
Characterization of a new non-centrosymmetric polymorph
of diphenyl-1,3,4-oxadiazole
a
Olga Franco , G u¨ nter Reck , Ingo Orgzall , Burkhard W. Schulz , Burkhard Schulz
b
c,
b
c
*
a
Institut f u¨ r Physik, Universit a¨ t Potsdam, Physik Kondensierter Materie, IFZ-DOBS, Am Neuen Palais 10, D-14469 Potsdam, Germany
b
Bundesanstalt f u¨ r Materialforschung und-pr u¨ fung (BAM), Richard-Willst a¨ tter-Straße 11, D-12489 Berlin, Germany
c
Universit a¨ t Potsdam, Interdisziplin a¨ res Forschungszentrum ‘D u¨ nne Organische und Biochemische Schichten’ Am Neuen Palais 10,
D-14469 Potsdam, Germany
Received 1 July 2002; revised 18 October 2002; accepted 18 October 2002
Abstract
Diphenyl-1,3,4-oxadiazole (DPO) crystallization experiments from solutions clearly reveal the polymorphism of the
substance. Besides the formerly known centrosymmetric monoclinic structure with space group P2 /c (DPO I) a new
1
monoclinic structure with the non-centrosymmetric space group Cc is found (DPO II): a ¼ 2.4134(4) nm, b ¼ 2.4099(3) nm,
3
c ¼ 1.2879(2) nm, b ¼ 110.048(3)8, and V ¼ 7.0363(17) nm . The asymmetric unit contains six independent molecules in a
complex packing motif. A re-determination of the crystal structure of DPO I at room temperature gives lattice parameters
3
a ¼ 0.51885(6) nm, b ¼ 1.8078(2) nm, c ¼ 1.21435(14) nm, b ¼ 93.193(3)8, and V ¼ 1.1373(2) nm . X-ray measurements at
3
63 K show a significant increase of the unit cell volume by 1.6%. Differences between both structures concerning morphology
and characteristic Raman bands are outlined in detail. DSC investigations show an irreversible transition from DPO I to DPO II
at 97 8C. DPO II does not show any transition in the temperature range up to the melting point at 141 8C. The non-
centrosymmetric DPO II structure shows triboluminescence.
q 2002 Elsevier Science B.V. All rights reserved.
Keywords: Diphenyl-1,3,4-oxadiazole; X-ray structure; Polymorphism; Non-centrosymmetric structure
1
. Introduction
applications, optical and display techniques, new
electronic devices, batteries etc. [1,2]. But, tailoring
of the properties requires the exact knowledge of the
relationship between structures and properties of the
materials and their mutual dependence. Based on a
model where complex structures are built from small
and relatively simple units like the synthonic
approach [3–5] it is assumed that some of the
properties of a basic unit are preserved in the final
product while others may change in a systematic
manner. Therefore, the knowledge of structure and
properties of more or less simple starting materials is
The rapid technological development requires the
search for advanced materials with tailor-made
properties. Organic substances may easily be modified
by chemical syntheses thus tuning the properties and
therefore have attracted increasing attention as high
performance materials for instance in sensoric
*
Corresponding author. Tel.: þ49-331-9772915; fax: þ49-331-
772985.
E-mail address: orgzall@rz.uni-potsdam.de (I. Orgzall).
9
0022-2860/03/$ - see front matter q 2002 Elsevier Science B.V. All rights reserved.
PII: S0022-2860(02)00569-0

220
O. Franco et al. / Journal of Molecular Structure 649 (2003) 219–230
required to determine an effective and successful
combination procedure. Structural investigations of
crystalline organic compounds are the starting point
for several different other research topics. Many
efforts are directed to the understanding of the
relations between molecular structure and super-
molecular arrangement.
monoclinic (space group P2 /c ) with cell parameters
1
a ¼ 0.52032 nm, b ¼ 1.81436 nm, c ¼ 1.21844 nm
and b ¼ 93.2438 (DPO I). The packing motif shows
infinite stacks and a herring bone configuration, a
general principle known from many different organic
materials, where the solid is formed by p–p
interactions and van der Waals contacts between the
different molecules (cf. for instance [14]). The nearly
planar molecules of adjacent stacks show opposite
orientation given by the O-atom and the center of the
oxadiazole ring and also an opposite inclination
against the stack axis a by ^418 ([13], see also Ref.
[12], where the high pressure behavior of this DPO
structure is discussed in detail, for a structure
description). A re-determination of the crystal struc-
ture of DPO I gave very similar structural parameters.
Additionally, structural investigations at higher tem-
peratures were carried out. In order to ensure a better
comparison between structural parameters obtained at
different temperatures the newly obtained data for
DPO I will be presented additionally.
Especially organic molecules tend to show a
distinct polymorphism that complicates all assertions
about the crystal structure resulting under real
conditions and therefore the properties what is of
special importance [6–9]. This polymorphism is due
to very small differences in the free energy between
the different polymorphic structures, usually in the
range of approximately 4–8 kJ/mol. The crystal-
lization process is influenced by a variety of different
factors like pressure and temperature, kind of
solvents, cooling rates, mechanical treatment, impu-
rities etc. and their complicated interplay. Thus, a way
to obtain the desired supermolecular architecture for
molecules with already optimized functional structure
could also be possible by the application of various
crystal growth techniques and conditions before
changing the chemical structure again.
But, during the preparation of the starting DPO
material for the high pressure investigations [12]
using solution crystallization crystals with different
morphology were found. The careful analysis by X-
ray and Raman techniques showed that DPO
crystallizes at least in two different polymorphic
forms from solution. So, in the following we will
describe the newly found second DPO crystal
structure and discuss the main differences and
structural relations.
Compounds containing the 1,3,4-oxadiazole ring as
a basic building block are known as scintillator
materials or as biologically active agents. Modifi-
cations of their chemical structures open up possibi-
lities for new technical applications for instance as
potential electroluminescent materials or as active
sensoric substances. Also other optical properties like
possibilities for non-linear processes should be taken
into account.
First results of structural investigations on differ-
ent compounds containing the diphenyl-1,3,4-oxa-
diazole (DPO) moiety as central building block have
been published recently [10–12] (or see the Cam-
bridge Structural Database). The structures found up
to now for the differently in para-position substituted
DPO basic units were mostly monoclinic or orthor-
hombic but always centrosymmetric. In the following
we discuss the polymorphism of the basic DPO unit
with a newly determined crystal structure (DPO II).
These results may serve as starting point for
investigations of a family of DPO with different
substitution scheme.
2. Sample preparation and experimental setup
Synthesis and crystallization. 2,5-Di(phenyl)-
1,3,4-oxadiazole was synthesized by direct conden-
sation of benzoic acid with hydrazine hydrate in
polyphosphoric acid [15]. 2.38 grams (0.02 mol)
benzoic acid and 0.55 g (0.011 mol) hydrazine
hydrate were stirred in 50 g polyphosphoric acid
(84.6% P O ) at 160 8C for 3 h. After cooling the clear
2
5
solution was poured into water, filtered and the
product was dried in vacuum. The product was
recrystallized three times from ethanol.
Melting point: 139 8C (determined by melting point
1
3
A previous structure solution for DPO was
presented by Kuznetsov et al. [13]. The structure is
microscope, lit. 138–140 8C [26]).Solid– C-NMR:
O–CyN: 163 ppm. Yield: 95 wt%.

O. Franco et al. / Journal of Molecular Structure 649 (2003) 219–230
221
To investigate the crystallization behavior of DPO
and to determine the appropriate conditions for the
precipitation of a single phase, either DPO I or DPO
II, several different experiments have been carried out
carefully. The synthesized DPO was recrystallized
from different solvents like petrolether or ethanol with
varying dilution at ambient or decreased temperature.
Additionally, the products resulting from a distinct
solvent were dissolved in the other solvent and
recrystallized again. The resulting crystals were first
microscopically inspected for their morphology and in
such cases were only ill crystallized material resulted
by Raman spectroscopy to determine the phase
content of the sample. But, in most cases the visual
inspection of well crystallized samples is sufficient to
differentiate between both structures having a unique
habit (needle, plate) as could be concluded from
extensive Raman investigations (see below).
For the high temperature measurement a heated air
stream was directed onto the same crystal also used for
theroomtemperaturemeasurement. ThedataofDPOII
were recorded at 293 K. The structures were solved by
direct methods and refined by full-matrix least-squares
calculations using SHELX-97 [16]. All hydrogen atoms
of both DPO I structures were located by difference
Fourier syntheses and refined isotropically. In the case
of DPO II only small single crystals with relatively
weak diffraction power were available. Therefore, the
atomic coordinates were refined using restraints
demanding nearly equal lengths for corresponding
bonds of the six symmetry independent molecules. The
existence of the space group Cc instead of C2/c was
confirmed by the results of the structure analysis. A
careful inspection of the molecular packing using
CELLGRAF [29] indicated that there are no twofold
axes in the crystal structure. The hydrogen atoms of
DPO II were calculated according to their geometrical
conditions and included in the structure refinement
using the riding model. Atomic coordinates, thermal
parameters, bond lengths and angles have been
deposited at the Cambridge Crystallographic Data
Centre (CCDC) and allocated the deposition number
CCDC 190263 (DPO I at 293 K), CCDC 190264 (DPO
I at 353 K), and CCDC 160022 (DPO II).
In general, it has to be stated that it was rather
impossible to obtain a fully crystallized sample
containing only one single unique DPO phase from
the solution experiments. Always both phases were
present in different ratios at least in traces like small
admixtures, especially few needles of phase I
(
preferentially grown at impurities) in larger amounts
of phase II or very small crystallites grown on the
surface of larger crystals. This is mainly observed if
phase I crystallizes in excess. Then some needles with
very small crystallites of phase II on their surface are
found. In the latter two cases the admixtures are well
below the X-ray detection limit and may best be
detected visually. Therefore, it is concluded, that the
final structure is independent on the DPO synthesis
route as could be the case for instance by the existence
of very small and nearly inextractable traces of
synthesis residues or molecular fragments. So, only
the influence of the solvent and the ambient conditions
have to be taken into account.
Crystal data of DPO I (at 293 K). A colorless
needle shaped crystal with dimensions 1.00 £
3
0.075 £ 0.070 mm was used for data collection.
Crystal system: monoclinic, space group P2 /c (No
1
14), a ¼ 0.51885(6) nm, b ¼ 1.8078(2) nm, c ¼
1.21435(14) nm, b ¼ 93.193(3)8, V ¼ 1.1373(2)
3
3
nm , Z ¼ 4, d(calc.) ¼ 1.298 Mg/m , linear absorp-
2
1
tion coefficient m ¼ 0.083 mm , and Fw ¼ 222.24
for C H N O, F(000) ¼ 464. The total number of
1
4 10 2
reflections in the range 4.08 , 2Q , 46.68 was
10468, 1595 independent, 1154 observed ½I .
2sðIÞꢀ: The number of refined parameters was 195.
Final R (on F) ¼ 0.0486.
The same holds also for sublimated DPO. The
product predominately consists of DPO I needles that
are stronger agglomerated but also contains smaller
admixtures of DPO II stuck together in polycrystalline
grains. Smaller crystallites of the second structure are
also found on the surface of the needles.
Crystal data of DPO I (at 353 K). The same crystal
described above was used for data collection. Crystal
system: monoclinic, space group P2 /c (No 14),
1
a ¼ 0.52131(12) nm, b ¼ 1.8116(5) nm, c ¼ 1.2155(3) nm,
3
b ¼ 92.933(9)8, V ¼ 1.1464(5) nm , Z ¼ 4, d(calc.) ¼
3
X-ray structure analysis. The X-ray data of DPO I
were collected on a SIEMENS SMART diffractometer at
1.288 Mg/m , linear absorption coefficient m ¼
2
1
0.083 mm , and Fw ¼ 222.24 for C H N O,
1
4
10
2
2
0
93 and 353 K using Mo Ka radiation (l ¼
F(000) ¼ 464. The total number of reflections in
the range 4.08 , 2Q , 46.68 was 10685, 1580
˚
.71073 A) monochromatized by a graphite crystal.

222
O. Franco et al. / Journal of Molecular Structure 649 (2003) 219–230
independent, 1110 observed ½I . 2sðIÞꢀ: The number
of refined parameters was 195. Final R (on
F ) ¼ 0.0500.
was always around of 5 mg. Closed aluminum
crucibles were used. The system was calibrated
using the usual calibration substances like indium or
zinc. Heating rates of 10 8C/min were chosen. An
experimental error of 1 8C for the determination of the
phase transition temperatures (peak value) resulted.
Crystal data of DPO II (at 293 K). A colorless plate
shaped crystal with dimensions 0.23 £ 0.12 £
3
.03 mm was used for data collection. Crystal system:
0
monoclinic, space group Cc (No 9), a ¼ 2.4134(4) nm,
b ¼ 2.4099(3) nm, c ¼ 1.2879(2) nm, b ¼ 110.048(3)8,
3
V ¼ 7.0363(17) nm , Z ¼ 24 (6 symmetry indepen-
3. Results and discussion
dent molecules having different conformations),
3
d(calc.) ¼ 1.259 Mg/m , linear absorption coefficient
3.1. DPO—new polymorphic structure DPO II and
comparison to DPO I
2
1
m ¼ 0.083 mm , and Fw ¼ 222.24 for C H N O,
1
4
10
2
F(000) ¼ 2784. The total number of reflections in the
range 2.48 , 2Q , 40.08 was 9639, 4254 indepen-
dent, 3007 observed ½I . 2sðIÞꢀ: The number of
refined parameters was 920. Final R (on F) ¼ 0.0630.
Raman investigations. A Raman triple spectro-
graph ‘T64000’ (INSTRUMENTS S.A.) with a liquid
nitrogen cooled CCD detector the was applied for the
recording procedure. The resolution of the system is
From the solution crystallization experiments two
morphologically different forms of DPO crystals
resulted. In general, both forms consist of white
colorless opaque to slightly transparent crystals. One
form shows a needle-like habit (Fig. 1(a)) with a
length up to several millimeters for the long needle
axis. The width may amount to maximum 1 mm while
the thickness is much smaller. The second poly-
morphic form exhibits plate-like up to prismatic
shapes (Fig. 1(b)) with all dimensions usually smaller
than 0.5–1 mm. Depending on the crystallization
conditions both crystal types form larger aggregates,
in the case of the needles this are fan-shaped
polycrystals while the second polymorph forms larger
clusters of impinged grains.
2
1
about 1 cm . The Raman spectra were excited by a
He–Ne laser (632.8 nm) focussed to a spot of 10 mm
with a reduced power output of less than 4 mW to
avoid thermal heating effects of the sample. All
spectra were calibrated using the spectral lines of Ne
from a pen ray calibration lamp.
The selected crystals were mounted on usual micro
slides either as single crystals or as powder to avoid
orientational effects. An additional camera system
allowed the visual inspection of the sample and the
precise localization of the laser spot on the sample.
DSC. The DSC studies were performed applying a
DSC 7 device (PERKIN –ELMER). The sample mass
For the structure determination well grown crystals
of both polymorphs were selected. The structure
determined for the needle like crystals—DPO I—
corresponds at room temperature to that published by
Kuznetsov et al. [13]. The monoclinic unit cell (P2 /c)
1
Fig. 1. Example for the morphology of DPO I (a) and DPO II (b).

O. Franco et al. / Journal of Molecular Structure 649 (2003) 219–230
223
Fig. 2. Molecular packing of DPO I (a) and DPO II (b).
contains four DPO molecules (one symmetry inde-
pendent). DPO I crystallizes in a herringbone
structure, characterized by stacks of parallel mol-
ecules (see Fig. 2(a)). The molecules of adjacent
stacks have opposite inclination to the stack axis, in
this case the a-axis. Within the stacks, the molecules
are oriented in the same direction, given by the line
joining the center and the O atom of the oxadiazole
ring. Neighboring stacks are translated in b-direction,
avoiding an aligned arrangement of the oxadiazole
rings along the c-axis. The molecules of adjacent
stacks show an inclination angle of ^418 to the stack
axis, respectively. The most intense inter-molecular
interactions in DPO I occur along a stack between the
p systems of the oxadiazole unit of one molecule and
the phenyl groups of the adjacent molecules (Fig. 3).
The strong electron acceptor character of the oxadia-
zole moiety induces the formation of p-complexes
resulting from the interaction between one p-donor
a p-complex amounts to 0.3352(2) nm, a usual
spacing for such complexes [8]. The forces between
different stacks are described by van der Waals
interactions.
The structure determined for polymorph DPO II
forming plates or small prisms is different to that
published by Kuznetsov et al. [13]. It is also
monoclinic, space group Cc but with a rather complex
packing motif compared to DPO I. The structure of
DPO II contains six symmetry independent molecules
(A, B, C, D, E, F). A DPO molecule from this
structure together with the average bond distances is
depicted in Fig. 4. The conformations of the six
symmetry independent molecules are different as
indicated by the dihedral angles between the central
oxadiazole ring and the phenyl rings. These angles are
listed in Table 1. The molecular arrangement of DPO
II is illustrated in Fig. 2(b). The crystal packing is
(
phenyl ring) and a p-acceptor (oxadiazole ring) [8].
Therefore, a sandwich arrangement along the stacks,
where the oxadiazole ring of one molecule is located
between two phenyl rings of the neighboring
molecules, is the result of the optimization of the
p–p interactions between p-donors and p-acceptors.
In this way, the oxadiazole group of every molecule
interacts with two phenyl rings of the adjacent
molecules and the phenyl ring of a molecule interacts
with one oxadiazole ring of the adjacent molecule.
The distance between the molecular planes of
Fig. 3. Molecular pair in the crystal structure of DPO I. (The e.s.d. is
0.0002 nm).

224
O. Franco et al. / Journal of Molecular Structure 649 (2003) 219–230
Analyzing a layer the shortest spacing between
adjacent stacks of the same layer occurs between C
atoms of phenyl rings and amounts to 0.383(2) nm.
The relatively large value of this distance together
with the perpendicular orientation between molecules
of different stacks leads to the conclusion that the
interactions between stacks are almost van der Waals
interactions. Additionally, between different layers
weak hydrogen bonds could exist because the shortest
distance between layers is found between one of the
phenyl ring C atoms of a molecule and a nitrogen
atom of the oxadiazole ring of other molecule. The
C–N spacing amounts to 0.345(1) nm, well in the
range 0.3–0.4 nm expected for weak hydrogen bonds
[17] and also the angle between the three involved
atoms corresponds to this assumption with 1438
(range 100–1808).
Although the molecules of both polymorphs are
almost planar the conformation shows small but
significant differences. In DPO I the three rings are
perfectly planar with endocyclic torsion angles lower
than 18. The deviation of the phenyl groups from the
molecular plane, defined by the oxadiazole unit,
amounts to 2.0 and 21.88. Both phenyl rings are
inclined in the same direction. The molecular
conformation of DPO II is not as planar as in the
first polymorph. The rings themselves are nearly
planar again with endocyclic torsion angles up to 38,
they do not lie in the same plane. The amount of the
dihedral angles between the rings varies among the six
molecules constituting the asymmetric unit with a
maximum angle of 11.5(2)8 and a mean value of 68.
Except in molecule D the phenyl rings are rotated in
opposite directions to the oxadiazole group. A
detailed comparison of the conformations is given in
Table 1.
Fig. 4. DPO molecule. Average bond distances are given for the six
symmetry independent molecules of DPO II. The standard
deviations are between 0.0010 and 0.0016 nm.
characterized by layers parallel to the ab-plane These
layers are built by stacks of molecules, where the
molecular planes are perpendicular to the layer plane.
Inside a layer, alternating stacks of parallel molecules
produce a zig-zag arrangement. The molecules of
adjacent stacks are oriented nearly perpendicular to
each other. Along a stack parallel molecules are
placed in such a way that the oxadiazole unit of a
molecule is sandwiched between the phenyl rings of
the adjacent molecules, leading to the formation of p-
complexes. In contrast to the first polymorph every
third molecule in these stacks is oriented in opposite
direction in relation to its neighbors. Therefore, the
formation of perfect stacks is prevented and two
principally different kinds of p-complexes are
formed, one between molecules with the same
orientation and another between those with opposite
orientation. All occurring stacks are depicted in Fig. 5.
The spacing between averaged molecular planes
range from 0.342(1) to 0.353(1) nm.
Table 1
Dihedral angles in DPO II, DPO I (at 293 K), and DPO I (at 353 K)
(signs of angle values were taken from pertinent torsion angles)
For both polymorphs the high planarity induces at
least a partial conjugation of the p system along the
whole molecule confirmed by the shortness of the
inter-ring bonds. In both polymorphs the mean value
amounts to 0.146 nm and is shorter than 0.154 nm,
expected for a C–C single bond and such bond
shortening is usually considered to result in a partial
conjugation of a single bond (for instance Ref. [18]).
Especially for DPO II these values also indicate that
the deviation of the phenyl rings from the molecular
plane is not large enough to prevent a partial
conjugation within the molecule.
Polymorph
P1(O1,C2,N3,
N4,C5,C6,C12)–
P2(C5,C6,C7,C8,
C9,C10,C11)
P1(O1,C2,N3,N4,
C5,C6,C12)–
P3(C2,C12,C13,C14,
C15,C16,C17)
DPO II molecule A
Molecule B
22.37(13)
23.74(14)
5.03(19)
26.83(18)
23.15(14)
9.45(16)
Molecule C
Molecule D
25.31(14)
24.68(15)
11.46(18)
2.01(8)
3.75(12)
Molecule E
28.51(16)
3.37(12)
Molecule F
DPO I (at 293 K)
DPO I (at 353 K)
21.82(9)
21.76(10)
2.17(9)

O. Franco et al. / Journal of Molecular Structure 649 (2003) 219–230
225
Fig. 5. Molecular pairs in the crystal structure of DPO II. (The e.s.d’s are between 0.001 and 0.002 nm).
3.2. Raman spectroscopic investigations of DPO I and
DPO II
by the presence of several independent molecules
with slightly different geometry in the unit cell. The
assignment of the different modes is done tentatively
using literature data for other organic compounds,
especially those containing phenyl and 1,3,4-oxa-
diazole rings (a general review is found in Refs.
[19,20]). Additional data for 1,3,4-oxadiazoles are
taken from Refs. [21–23] as well as from Raman
investigations of different diphenyl-1,3,4-oxadia-
zoles with varying functional groups and different
chain length up to the phenyloxadiazole polymer
[6]. This assignment may be compared with
molecular modeling calculations using GAUSSIAN 94
Fig. 6(a) shows the Raman spectra of both phases
DPO I and DPO II, respectively, in the finger print
2
1
region between 15 and 1700 cm . The spectra are
characterized by well resolved internal modes and
2
1
the external lattice vibrations below 200 cm . The
internal modes are dominated by two groups, one
2
around 1000 cm and a second between approxi-
1
2
1
mately 1500 and 1600 cm . Several smaller bands
may be found between these two groups but these
shall not be discussed here in detail. Looking at
these two groups separately it is obvious that there
exist no remarkable differences between DPO I and
DPO II in the first region within the resolution limits
where a first attempt showed rather good
agreement.
The bands result from both building blocks, the
two monosubstituted phenyl rings and a diaryl
substituted 1,3,4-oxadiazole ring. The most promi-
2
1
of the system (^1 cm ). In the region around
1
2
1
000 cm the main band at 992 cm is approxi-
21
2
1
mately twice broader for DPO II compared to DPO I
while the smaller surrounding lines are more intense
in DPO II. The broadening may be explained
nent internal vibrations at 1607 and 992 cm are
attributed to ring vibrations of the phenyl rings,
2
1
while the oxadiazole ring marks at 1554 cm
.

226
O. Franco et al. / Journal of Molecular Structure 649 (2003) 219–230
2
Fig. 6. Raman spectra of DPO I and II in the complete finger print region (a) and in the region of the external modes below 400 cm (b).
1
The literature reports three oxadiazole ring modes
also be assigned to ring vibrations mainly of
the phenyl rings. Generally, it has to be mentioned,
that the combination of different building blocks
with known Raman frequencies may lead to larger
deviations due to superposition and interaction of
different ring vibrations and possibly with C–H
motions. So, frequency shift or mode splitting may
occur, some lines may be missing or new ones
appear so that the assignment for the resulting
compound’s spectrum not necessarily must be
unique. Several experimental data in literature
were also obtained in solution. But here the
molecules are investigated in the solid crystalline
state and therefore the interaction with neighbored
molecules within the crystal has to be taken into
account.
2
in the region around 1500 cm : 1542, 1479, and
1
2
456 cm , so that the first vibration may be
1
1
identified with the experimental value. The other
two bands may also be found in the spectra but
with lower intensity. The phenyl band at
2
1
1
607 cm
and an additional smaller line around
580 cm result from the splitting of the quadrant
2
1
1
stretch mode that is degenerated for the pure
benzene ring due to the substitution. The phenyl
semicircle stretch bands are not identified with
security, but these bands are always weak in
2
1
Raman. The strong mode at 992 cm
should be
due to the diaryl substituted oxadiazole having two
additional characteristic frequencies at 1115 and
2
1
1
022 cm
according to Ref. [24]. This is also
concluded from the comparison with Raman spectra
of other diphenyl oxadiazoles [25] and from first
rough molecular modeling results. Monosubstituted
phenyl rings should exhibit a band usually showing
high Raman intensities in the range between 990
Remarkable differences in the Raman spectra are
found in the region of the external modes. Both
phases clearly may be distinguished by their
2
1
spectra below approximately 400 cm as illus-
trated in Fig. 6(b). While DPO I shows at least
21
four well resolved bands below 50 cm DPO II
2
1
and 1010 cm
deformation or ring breathing vibration. A ring
resulting from the in-plane ring
has only three but rather broad lines in that region.
Nevertheless, it is expected that the bands consist
of superimposed vibrations and not of a single one.
Adjacent to this for DPO II a very broad shoulder-
like structure is found also consisting of a super-
position of several lines seen in several slight
maxima. These could only be resolved by a
deconvolution procedure for an appropriate number
of bands. This structure is stronger orientation
dependent than the lines at lower wave numbers.
2
1
stretching vibration should occur below 1000 cm
.
Both modes may interact and mix due to the same
symmetry species [19] giving rise to an intense
2
1
Raman band around 1000 ^ 5 cm . The slightly
asymmetric shape of the bands with a shoulder at
lower wavelengths may indicate a convolution of
both bands resulting either from the oxadiazole and
the monosubstituted phenyl rings. The other lines
with minor intensities found in the spectra may

O. Franco et al. / Journal of Molecular Structure 649 (2003) 219–230
227
Compared to this DPO I is characterized by two
broad but separate bands between 50 and
3.3. Thermal investigations and phase transition from
DPO I to DPO II
2
00 cm . Another remarkable difference is a
1
1
band found at 164 cm
2
1
for DPO I while it is
shifted to about 154 cm in DPO II. Also minor
The thermal behavior of DPO was characterized
using the usual DSC technique. For the DSC study
three successive heating and cooling cycles were
performed. During the first heating, DPO I shows an
endothermic peak at 97 8C (value at peak maximum)
with an enthalpy increment of DH ¼ 2.7 kJ/mol.
Further heating induces the melting of the crystal at
141 8C (DH ¼ 24 kJ/mol). The melting point is in
accordance with Ref. [27]. By cooling, the liquid
recrystallizes at 115 8C (DH ¼ 220 kJ/mol). The
subsequent heating and cooling cycles show only
the peak corresponding to the solid-liquid phase
transition, no indications for further processes within
the sample are found (see Fig. 7). Therefore it may be
concluded that the first peak describes an irreversible
phase transition.
2
1
21
lines around 250 and 340 cm show some smaller
shifts. These smaller variations are indicated by
arrows in the figure. Summarizing it has to be
stated that it is of crucial importance to investigate
the low frequency Raman modes to investigate the
phase behavior of DPO because the internal modes
are not sufficient to discriminate between the
different structures at least under ambient pressure
(
the high pressure behavior and the resulting
differences for both structures will be discussed
elsewhere [26]). But as also may be concluded
micro Raman spectroscopy turns out to be a
valuable tool for the distinction between different
crystal structures due to lower requirements for the
crystal perfection compared to single crystal X-ray
investigations and to the faster procedure. In the
different crystallization attempts Raman spec-
troscopy was successfully applied to determine
the phase content.
The DSC analysis of DPO II indicates that this
crystal does not undergo any solid-solid phase
transitions. The melting temperature found for the
three heating cycles was 141 8C also corresponding to
the DPO melting temperature [27] (DH ¼ 24 kJ/mol).
Fig. 7. DSC study of the transition of DPO I into DPO II. The inserts show the Raman spectra in the region of the lattice modes for the
identification of the phase content.

228
O. Franco et al. / Journal of Molecular Structure 649 (2003) 219–230
Thus both compounds show the same melting
temperature and heat of fusion. The substance shows
the analogous recrystallization behavior like DPO I
the a-axis. This corresponds to a change of the
stacking interactions in DPO I.
An intensity data collection for a complete
structure determination could be carried out at
353 K. But none significant deviations of correspond-
ing molecular parameters were observed. Only the
spacing within the molecular stacking pairs increased
from 0.3352(2) to 0.3375(2) nm (Fig. 3). An attempt
to determine the complete crystal structure at 363 K
failed because the single crystal was destroyed by
sublimation. Before this destruction the formerly clear
crystal became turbid.
(
temperature: 113–114 8C, DH ¼ 220 kJ/mol).
To investigate the nature of the solid-solid
transition in DPO I an additional DSC run with
DPO I was stopped at temperatures slightly above the
transition. Here, also during cooling of the sample no
peaks appear. All recovered DSC samples were then
investigated by means of Raman spectroscopy. All
samples, those that recrystallized from the melt after
three measuring cycles as well as that from the
stopped DSC run showed clearly the Raman charac-
teristics of DPO II. So, due to missing hints on further
transitions it is concluded that DPO I converts to DPO
II with increasing temperature in a solid-solid phase
transition. This structure also recrystallizes from the
melt under normal conditions. Accompanying powder
X-ray investigations under increasing temperature
also confirmed the transition process although the
limited precision of the temperature determination
prevented a detailed insight into the transition process
i.e. if there exist possibly intermediate stages and
structures or if it proceeds directly from one to the
other structure.
Because of the weak diffraction power of the
available crystals high temperature X-ray measure-
ment for DPO II gave no very significant results.
But nevertheless it was found that the volume of
the unit cell increased at 363 K by 0.8% and at
393 K by 1.4% compared with that at the room
temperature.
Although some relations between crystallographic
axes exist there is no greater evidence for a topotactic
reaction between both phases. Appropriate relations
between the individual molecules and therefore
possible motions that could lead from structure I to
II could not be found. This mechanism would require
larger molecular motions and especially the change to
an opposite orientation of every third molecule within
the solid state.
Fig. 2 shows the packing of molecules in the
corresponding unit cells of both polymorphs. The
lattice of DPO II can nearly be obtained from that of
DPO I by the following matrix: (3 1 0) (23 1 0)
From the temperature treatment only DPO II
results as final product. No back transformation to
DPO I could be detected. Additional high pressure
X-ray and Raman investigations showed that neither
DPO I nor DPO II exhibit any pressure induced
transition in the pressure range up to 7 GPa (for an
equation of state for DPO I see Ref. [12], that for
DPO II is found together with extensive Raman
(
20.5 0 1). To obtain some further insights into a
possible topotactic transition mechanism X-ray single
crystal investigations at higher temperatures were
carried out. Table 2 lists the lattice parameters of DPO
I obtained by measurements between 293 and
3
1
63 K. The cell volume significantly increases up to
.6%. The relatively largest increase was observed for
Table 2
Lattice parameters of DPO I at different temperatures
3
V (nm )
T (K)
a (nm)
b (nm)
c (nm)
a (8)
b (8)
g (8)
2
3
3
3
3
3
96
03
13
23
53
63
0.51885(14)
0.51947(16)
0.51979(18)
0.52049(17)
0.52108(16)
0.52382(19)
1.80883(42)
1.80946(45)
1.80975(46)
1.81178(46)
1.81205(50)
1.80898(68)
1.21435(29)
1.21509(27)
1.21481(32)
1.21650(35)
1.21639(34)
1.22081(46)
90.000(24)
90.026(27)
89.991(27)
90.042(28)
90.020(31)
90.265(31)
93.193(34)
93.129(36)
93.098(36)
93.006(38)
93.019(35)
92.712(46)
90.003(28)
90.015(39)
90.025(31)
90.034(33)
90.000(33)
90.115(43)
1.13728(53)
1.14043(55)
1.14108(57)
1.14561(56)
1.14695(59)
1.15552(91)

O. Franco et al. / Journal of Molecular Structure 649 (2003) 219–230
229
investigations of both structures in Ref. [26]). That
means that no way was found to transform DPO II into
DPO I. Consequently, DPO II could be the thermo-
dynamically stable phase but it generally cannot be
fully ruled out, that the transformation is only
kinetically hindered.
As in general polymorphism is a common
phenomenon especially for organic molecular crystals
thermodynamic and kinetic aspects shall not be
discussed here. It should only be remarked that
summarizing all observations the crystallization
process of DPO from the solution should be an
example for Ostwald’s law of stages. DPO I mostly
forms first from the solution (if it is not prevented due
to the specific conditions) due to the faster kinetics
(lower nucleation activation energy, higher growth
rate). This structure remains metastable. The crystal-
lization of DPO II sets in later and seems to be slower.
But, generally a deeper insight into the whole
processes resulting either in DPO I or DPO II or
their mixture is still missing due to the complex
interactions and influences of the variety of different
thermodynamic and kinetic factors.
Concluding it shall be stated that the free energy of
both structures is nearly the same as expected from
several examples of organic polymorphism and
indicated by the low energy barrier to induce the
transition of DPO I into DPO II. Only a small energy
of 2.7 kJ/mol is necessary. The inter-molecular
interactions in both crystals are nearly similar. In
both cases, they are principally strong p–p inter-
actions along the infinite p-complexes, van der Waals
forces between nearly perpendicular phenyl rings and
weak H-bonds between one N atom of the oxadiazole
unit and the C–H group of the phenyl ring of a nearly
perpendicular molecule.
The newly discovered structure of DPO is a non-
centrosymmetric one thus offering possibilities for
non-linear optical effects. Additionally, it is worth
noting that the crystals of DPO II show a blue
triboluminescence signal if they are mechanically
treated or crushed while the DPO I crystals do not
exhibit such features. As known from other tribolu-
minescent materials it is always a non-centrosym-
metric structure that is active in contrast to
centrosymmetric polymorphs [28]. Therefore it was
concluded that a piezoelectric charge may be
responsible for this effect. The non-centrosymmetric
space group Cc belonging to the crystal class m offers
the possibilities for both, piezo-and pyroelectricity. In
a first rough experiment both coefficients were
measured at a compacted non-oriented powder sample
4
. Conclusion
For a rather simple molecule, the 2,5-diphenyl-
,3,4-oxadiazole molecule as a root molecule for a
1
huge class of differently substituted compounds with
many different potential applications it could be
shown that at least two different polymorphs exist.
These two forms already differ by their morphology,
having either needle-like or plate-like habit. Although
both structures are monoclinic the packing motif is
completely different. Nevertheless, some features of
the molecular arrangement like stacks of molecules
connected by van der Waals interactions or by the
formation of p-complexes within the stack are
encountered for DPO I as well as for DPO II. This
leads to similarities of the intermolecular interactions.
The simpler structure DPO I transforms into the more
complex structure II under heating with low transition
enthalpy as expected. To distinguish between both
structures extensive X-ray investigations are not
always required, it could be shown, that Raman
spectroscopy offers a fast and easy-to-use tool for
discrimination between different polymorphs due to
characteristic differences of the lattice modes.
Additionally, at least in this case already the careful
investigation of the varying morphology is sufficient
for this purpose.
of DPO II. A piezocoefficient of d33 ¼ 3 pC/N and a
2
pyroelectric coefficient of p ¼ 0.01 nC/Kcm
3
resulted. The pyroelectric measurements showed a
stronger hysteresis. Compared to other substances
especially like polymers these parameters are rather
small but the results give further evidence for the
importance
triboluminescence.
of
piezoelectric
charges
in
So the experiences obtained with this molecule
may serve as starting points for future investigations
of differently substituted DPO molecules with the aim
of a higher perfection of the grown crystals and
possibly new application fields for this group
of compounds. Additionally these investigations
substantially contribute to the understanding of

230
O. Franco et al. / Journal of Molecular Structure 649 (2003) 219–230
structure–property relationships, particularly in con-
nection with high pressure experiments that may give
a deeper insight into the different molecular
interactions.
[11] S. Stockhause, M.S. Wickleder, G. Meyer, I. Orgzall, B.
Schulz, J. Mol. Struct. 561 (2001) 175.
[
[
[
12] O. Franco, G. Reck, I. Orgzall, B. Schulz, J. Phys. Chem.
Solids 63 (2002) 1805.
13] V.P. Kuznetsov, L.D. Patsenker, A.I. Lokshin, A.V. Tolma-
chev, Crystallogr. Rep. 43 (1998) 430.
14] B.K. Vainshtein, V.M. Fridkin, V.L. Indenbom, Structure of
Crystals, second ed., Springer, Berlin, 1995.
Acknowledgements
[
[
[
15] W. Neugebauer, JP 256946, 1959.
16] G. Sheldrick, Universit a¨ t G o¨ ttingen, Germany, 1997.
17] R.E. Mel e´ ndez, Hydrogen-bonded ribbons, tapes and sheets as
motifs for crystal engineering, in: H. Weber (Ed.), Design of
Organic Solids, Springer, Berlin, 1998, pp. 97–129, Topics in
Current Chemistry 198.
This work was supported by the Deutsche
Forschungsgemeinschaft under Grants No. OR
7
8/3-2 and SCHU 842/12-3.
[
18] B. K o¨ nig, H. Butensch o¨ n, Memofix Organische Chemie,
second ed., VCH, Weinheim, 1999.
References
[
19] N.B. Colthup, L.H. Daly, S.E. Wiberley, Introduction to
Infrared and Raman Spectroscopy, second ed., Academic
Press, New York, 1975.
[
1] K. M u¨ llen, G. Wegener, Electronic Materials, Wiley, New
York, 1998.
[
[
20] D. Lin-Vien, N.B. Colthup, W.G. Fateley, J.G. Grasselli, The
Handbook of Infrared and Raman Characteristic Frequencies
of Organic Molecules, Academic Press, New York, 1991.
21] J.W. Anthonsen, D.H. Christensen, J.T. Nielsen, O.F. Nielsen,
Spectrochim. Acta 32A (1976) 971.
[2] J.M. Lehn, Supramolecular Chemistry—Concepts and Per-
spectives, VCH, Weinheim, 1995.
[
[
[
3] G.R. Desiraju, Angew. Chemie 107 (1995) 2541.
4] J. Maddox, Nature 335 (1988) 201.
5] G.R. Desiraju, Crystal Engineering: The Design of Organic
Solids, Elsevier, Amsterdam, 1989.
[
[
[
[
[
[
22] A.A. El-Azhary, Spectrochim. Acta Part A 51 (1995) 995.
23] A.A. El-Azhary, Spectrochim. Acta Part A 52 (1996) 33.
24] M. Milone, Gazz. Chim. Ital. 63 (1933) 456.
25] I. Orgzall, O. Franco, unpublished results.
[
[
6] M.J. Zaworotko, Chem. Commun. (2001) 1.
7] M.R. Caira, Design of organic solids, in: H. Weber (Ed.),
Crystalline Polymorphism of Organic Compounds, Springer,
Berlin, 1998, pp. 162–208, Topics in Current Chemistry 198.
8] J.P. Glusker, M. Lewis, M. Rossi, Crystal Structure Analysis
for Chemists and Biologists, VCH, Weinheim, 1994.
9] J.D. Dunitz, X-Ray Analysis and the Structure of Organic
Molecules, Cornell University Press, Ithaca, NY, 1979.
26] O. Franco, Thesis, Universit a¨ t Potsdam, 2002.
27] Beilsteins Handbuch der Organischen Chemie, vol. 27, fourth
ed., Berlin: Julius Springer, 1937, p. 591.
[
[
[28] I. Sage, G. Bourhill, J. Mater. Chem. 11 (2001) 231.
[29] G. Reck, G. Walther, R.-G. Kretschmer, CELLGRAF,
Bundesanstalt f u¨ r Materialforschung und-pr u¨ fung, Berlin,
1996.
[10] I. Orgzall, B. Lorenz, J. Mikat, G. Reck, G. Knochenhauer, B.
Schulz, J. Phys. Chem. Solids 60 (1999) 1949.