Crystal structure, ab initio calculations and fingerprint plots of a new polymorph of N′,N″,N?-triphenylbiuret

Article abstract of DOI:10.1016/j.molstruc.2011.03.057

A new polymorph of N′,N″,N?-triphenylbiuret, C ₂₀H₁₇N₃O₂ (form II), has been synthesized and the structure has been solved by X-ray diffraction. The crystals are monoclinic, space group P2₁/c, with a

Full text of DOI:10.1016/j.molstruc.2011.03.057

Journal of Molecular Structure 995 (2011) 66–71
Contents lists available at ScienceDirect
Journal of Molecular Structure
journal homepage: www.elsevier.com/locate/molstruc
Crystal structure, ab initio calculations and fingerprint plots of a new polymorph
of N0,N00,N000-triphenylbiuret
Pedro S. Pereira Silva a,*, Raza Murad Ghalib b, Sayed Hasan Mehdi b, Rokiah Hashim b, Othman Sulaiman b,
Manuela Ramos Silva a
a CEMDRX Physics Department, University of Coimbra, P-3004-516 Coimbra, Portugal
b School of Industrial Technology, Universiti Sains Malaysia, 11800 Pulau Pinang, Malaysia
a r t i c l e i n f o
a b s t r a c t
A new polymorph of N0,N00,N000-triphenylbiuret, C20H17N3O2 (form II), has been synthesized and the struc-
ture has been solved by X-ray diffraction. The crystals are monoclinic, space group P21/c, with a = 7.6966
(3) Å, b = 12.5490 (4) Å, c = 18.5996 (6) Å, b = 107.632(2)°, Mr = 331.37, V = 1712.04 (10) Å3, Z = 4 and
R = 0.0454. The hydrogen bonding of this polymorph is considerably different from that of the previously
known structure. The molecules are linked in infinite chains, via C–Hꢀ ꢀ ꢀO hydrogen bonds and there is
also an intramolecular N–Hꢀ ꢀ ꢀO hydrogen bond.
Article history:
Received 26 February 2011
Received in revised form 24 March 2011
Accepted 24 March 2011
Available online 31 March 2011
Keywords:
Triphenylbiuret
Polymorphism
Ab initio calculations
Relative conformational energies
Fingerprint plots
The intermolecular interactions present in this polymorph, and on the previously reported polymorph,
were analysed by means of the fingerprint plots derived from the Hirshfeld surfaces. The fingerprint plots
evidenced the different packing modes of the two structures.
Quantum–mechanical ab initio calculations for the free molecule were performed using the Hartree–
Fock and DFT/B3LYP methods with the 6-31G(d,p) basis set of wave functions. The solid-state conforma-
tions compared with those obtained theoretically from DFT calculations for the isolated molecules show
significant differences.
Some difficulties of using quantum–mechanical calculations for the determination of relative confor-
mational energies are also discussed.
Ó 2011 Elsevier B.V. All rights reserved.
1. Introduction
In this work we used the CrystalExplorer software [3] to calcu-
late the 2D fingerprint plots, to elucidate the different crystal envi-
The interest in polymorphism is rapidly growing because of its
importance in industrial processes, namely in the pharmaceutical
industry, since different physical properties of polymorphs can
substantially alter the durability, solubility and bioavailability of
the pharmaceutical drugs [1]. Conformationally flexible molecules
have more degrees of freedom than rigid molecules, so a greater
scope for polymorphism might be expected. The energies involved
in rotating parts of the molecule about single bonds are compara-
ble to the energy differences observed between polymorphs and so
it is not surprising that molecules with torsional degrees of free-
dom, such as the title compound, can exhibit different conforma-
tional polymorphs.
ronments of the two polymorphs. Such graphical tools based on
Hirshfeld surfaces [4,5] and on the derived two-dimensional (2D)
fingerprint plots [6,7] are a valuable tool for visualizing and analyz-
ing intermolecular interactions in polymorphs and make the task
of polymorph discrimination considerably easier.
We have also calculated the energies of the two solid state con-
formations of triphenylbiuret known so far and several methods of
calculating the relative energy were compared.
2. Experimental and computational methods
2.1. Preparation of the title compound
In this article we present the crystal structure of a new poly-
morph of N0,N00,N000-triphenylbiuret and we compare it with the pre-
viously known structure [2] using several computational tools.
A mixture of Phenyl isocyanate (1.08 ml, 10 mmol) and diethyl
malonate (1.51 ml, 10 mmol) in molar ratio 1:1 were refluxed in
THF on a water bath for 6 h in the presence of Na2CO3 (Scheme
1). The reaction mixture was dried on rota vapour at low pressure
and further fractionated with chloroform.
⇑
Corresponding author. Tel.: +351 239 410 652; fax: +351 239 829 158.
The chloroform soluble part was further dried on rotavapor and
chromatographed over silica gel (Merck, 0.04–0.063 mm, 230–400
E-mail
addresses:
psidonio@pollux.fis.uc.pt,
pedro.sidonio@gmail.com
(P.S. Pereira Silva).
0022-2860/$ - see front matter Ó 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.molstruc.2011.03.057

P.S. Pereira Silva et al. / Journal of Molecular Structure 995 (2011) 66–71
67
Scheme 1. Synthesis route.
mesh ASTM) column loaded in light petroleum ether. The column
Table 1
was eluted successively with light petroleum ether and light petro-
leum ether–diethylether (9:1–1:1). The elutants obtained in sol-
vent system light petroleum ether–diethylether (9:1–8:2), on
crystallization with chloroform–n-hexane (1:1), gave the transpar-
ent crystals of N0,N00,N000-triphenylbiuret (550 mg) (mp. 148 °C). The
melting point was determined on a Thermo Fisher digital melting
point apparatus of IA9000 series. The polymorph reported by
Carugo et al. [2] was obtained with a reaction of triethylindium
with an excess of phenyl isocyanate, as described by Tada & Okaw-
ara [8].
Crystallographic data and structure refinement of the title compound (form II).
Empirical formula
Formula weight
Temperature (K)
Wavelength (Å)
Crystal system
C20H17N3O2
331.37
293(2)
0.71073
Monoclinic
Space group
P21/c
a (Å)
7.6966 (3)
b (Å)
12.5490 (4)
c (Å)
18.5996 (6)
b (°)
107.632(2)
1712.04 (10)
Volume (Å3)
Z
4
2.2. Crystal structure determination
Calculated density (g/cm3)
Absorption coefficient (mmꢂ1
F(0 0 0)
1.286
0.085
696
)
A crystal of (form II) with a block shape and having approximate
dimensions of 0.40 mm ꢁ 0.32 mm ꢁ 0.25 mm was glued on a
glass fibre and mounted on a Bruker Apex II diffractometer. Diffrac-
tion data were collected at room temperature 293(2) K using
Crystal size (mm)
h range for data collection (°)
Index ranges
Reflections collected/unique
Completeness to h = 25.00°
Refinement method
Data/restraints/parameters
Goodness-of-fit on F2
0.40 ꢁ 0.32 ꢁ 0.25
2.78 –26.82
ꢂ10 < h < 10, ꢂ16 < k < 17, ꢂ24 < l < 24
34706/4521 [R(int) = 0.0260]
99.9%
graphite monochromated Mo K
a (k = 0.71073 Å). Data reduction
Full-matrix least-squares on F2
4521/0/226
was performed with APEX II [9]. Lorentz and polarization correc-
tions were applied. Absorption correction was applied using SAD-
ABS [10]. The crystallographic structure was solved by direct
methods (SHELXS-97) [11]. Refinements were carried out with
SHELXL-97 package [11]. All refinements were made by full-matrix
least-squares on F2, with anisotropic displacement parameters for
all non-hydrogen atoms. All the hydrogen atoms could be located
in a difference Fourier synthesis but they were placed at calculated
positions and then, included in the structure factor calculation in a
riding model using SHELXL-97 defaults. The final least-squares cy-
1.003
Final R indices [I > 2
r
(I)]
R1 = 0.0454 wR2 = 0.1096
R1 = 0.0796 wR2 = 0.1312
0.168 and ꢂ0.208
R indices (all data)
Largest diff. peak and hole (e Å–3
)
Table 2
Comparison of selected geometrical parameters for form II and form I as determined
by X-ray diffraction and from DFT calculations (Å, °).
cle was based on 4521 observed reflections [I > 2r(I)], 226 variable
parameters, converged with R = 0.0454 and wR = 0.1096. Addi-
tional information to the structure determination is given in Table
1. Selected structural parameters can be seen in Table 2. Hydrogen
bond geometric data is given on Table 3. Supplementary data have
been deposited at the Cambridge Crystallographic Data Centre
(CCDC No. 795370).
Experimental
Form II
DFT
Form I
Form II
Form I
O1–C1
O2–C2
N1–C1
N1–C2
N1–C3
N2–C1
N2–C9
N3–C2
1.2201(16)
1.2055(17)
1.3965(17)
1.4255(17)
1.4463(17)
1.3540(17)
1.4190(18)
1.3404(17)
1.4136(17)
124.94(11)
120.65(11)
123.80(12)
126.67(12)
114.75(12)
115.90(12)
ꢂ88.22(18)
50.5(2)
1.2197(15)
1.2134(17)
1.4122(18)
1.4275(17)
1.4472(16)
1.3505(17)
1.4230(18)
1.3400(16)
1.4184(19)
123.92(12)
120.53(12)
126.11(12)
126.96(14)
113.94(12)
116.24(14)
ꢂ68.26(19)
ꢂ6.6(2)
1.2334
1.2209
1.4106
1.4491
1.4458
1.3695
1.4144
1.3589
1.4099
125.66
120.56
128.52
127.31
114.24
115.29
ꢂ90.33
2.38
1.2333
1.2206
1.4097
1.4483
1.4468
1.3699
1.4140
1.3596
1.4098
125.69
120.56
128.57
127.37
114.30
115.29
ꢂ89.59
ꢂ1.51
2.3. Ab initio calculations
Ab initio studies using the Firefly QC package [12], which is par-
tially based on the GAMESS (US) source code [13], were performed
to calculate the energy differences between conformers. Various
constraints were imposed during the calculations to determine
the best methodology for calculating the relative energies.
Usually, single point calculations on each polymorph are con-
sidered the best measure of relative energy. However, small differ-
ences in the experimental bond lengths and angles between the
conformers will lead to large energy differences [14].
The different methods used in the ab initio calculations were the
following
N3–C15
C1–N1–C2
C1–N1–C3
C1–N2–C9
C2–N3–C15
N2–C1–N1
N3–C2–N1
C1–N1–C3–C4
C1–N2–C9–C14
C2–N1–C3–C8
C2–N3–C15–C16
N1–C1–N2–C9
N1–C2–N3–C15
N2–C1–N1–C2
N3–C2–N1–C1
ꢂ89.06(17)
ꢂ154.59(15)
ꢂ179.83(14)
167.26(14)
ꢂ172.99(13)
ꢂ4.3(2)
ꢂ72.33(19)
ꢂ167.13(17)
177.07(13)
175.63(15)
163.15(13)
10.3(2)
ꢂ90.90
ꢂ179.67
179.90
179.54
ꢂ178.99
ꢂ0.52
ꢂ90.38
179.67
ꢂ179.91
ꢂ179.88
179.90
ꢂ0.19
(1) The single point energy for the experimental X-ray geome-
tries (DE
crys);

68
P.S. Pereira Silva et al. / Journal of Molecular Structure 995 (2011) 66–71
Table 3
H-bond geometry (Å, °).
there is also an intramolecular N–Hꢀ ꢀ ꢀO hydrogen bond with
geometry similar to that found in form I [see Table 3], forming
six-membered rings with graph-set symbol S(6) [19,20]. Of the
two intramolecular C–Hꢀ ꢀ ꢀO hydrogen bonds present in form I
(C13–H13ꢀ ꢀ ꢀO2; C20–H20ꢀ ꢀ ꢀO2) only the C20–H20ꢀ ꢀ ꢀO2 is present
in this polymorph since the ring C9–C14 is considerably out of the
biuret plane, not allowing a conventional hydrogen bond between
the atoms C13 and O2.
D–H
Hꢀ ꢀ ꢀA
2.55
1.91
2.37
Dꢀ ꢀ ꢀA
3.162(2)
2.593(2)
2.903(2)
D–Hꢀ ꢀ ꢀA
124
135
116
C16–H16ꢀ ꢀ ꢀO2i
0.93
0.86
0.93
N3–H3ꢀ ꢀ ꢀO1 (intra)
C20–H20ꢀ ꢀ ꢀO2 (intra)
Symmetry code i: 1 ꢂ x, y ꢂ 1/2, 3/2 ꢂ z.
The use of available strong hydrogen bond donors and acceptors
is almost axiomatic [19], so an unusual structural feature in form II
is the absence of the expected N2–H2ꢀ ꢀ ꢀO1 hydrogen bond, that is
present in form I.
(2) The single point energy for the X-ray geometries with the
positions of the hydrogen atoms in form I modified to be like
those in form II (D
E1), i.e., with the same distances to the
parent atoms as in form II;
The molecular packing is also influenced by several C–Hꢀ ꢀ ꢀ
p
(3) Energy for the structures with all parameters optimised
except the torsion angles chosen to define each conformer
interactions. The strongest C–Hꢀ ꢀ ꢀ
p interaction is of the type II as
described by Malone et al. [21], with a Hꢀ ꢀ ꢀpi distance of 2.81 Å,
(D
E2).
and a C–Hꢀ ꢀ ꢀ
p angle of 162° [symmetry code: (i) 1 + x, y, z]; the
H8 atom is being attracted in the direction of the centre of the aro-
These calculations were performed with the molecular orbital
HF method and within DFT using B3LYP (Becke three-parameter
Lee–Yang–Parr) for exchange and correlation, which combines
the hybrid exchange functional of Becke [15,16] with the correla-
tion functional of Lee, Yang and Parr [17]. The calculations were
matic ring C15–C20. In the other two relevant C–Hꢀ ꢀ ꢀ
p
interac-
tions, the hydrogen atoms H17 and H12 are above the centre of
the phenyl ring C3–C8 but the C–H bonds point towards the ring
edge. These two interactions are of the type III according to the
classification of Malone et al. [21], the first with a H17ꢀ ꢀ ꢀpii dis-
performed with a 6-31G(d,p) basis set. For the method
D
E1 we also
tance of 2.83 Å, and a C17–H17ꢀ ꢀ ꢀ
p angle of 149° [symmetry code:
performed HF and DFT calculations with the augmented basis set
6-311+G(d,p).
(ii) 1 ꢂ x, ꢂ1/2 + y, 3/2 ꢂ z] and the second with a H12ꢀ ꢀ ꢀpiii dis-
tance of 3.02 Å, and a C12–H12ꢀ ꢀ ꢀ
(iii) 3 ꢂ x, 1 ꢂ y, 2 ꢂ z].
p angle of 143° [symmetry code:
We performed also geometry optimizations starting from the
experimental X-ray geometries of the two conformers. These opti-
mizations were performed only within DFT (B3LYP for exchange
and correlation and 6-31G(d,p) basis set). Tight conditions for con-
vergence of both the self-consistent field cycles and the maximum
density and energy gradient variations were imposed (10ꢂ5 atomic
units) in both calculations. At the end of these geometry optimiza-
tions we conducted Hessian calculations to guarantee that the final
structures correspond to true minima, using the same level of the-
ory as in the geometry optimizations.
3.2. Fingerprint plots
When comparing the same molecule in different crystal envi-
ronments, Hirshfeld surfaces and fingerprint plots [4–7] have been
shown to be a powerful tool for elucidating and comparing inter-
molecular interactions, complementing other tools currently avail-
able for the visualization of crystal structures and for their
systematic description and analysis, e.g. graph-set analysis [19]
and topological analysis [22].
3. Results and discussion
The intermolecular interactions of this new polymorph and of
the previously reported [2] were analyzed using the two-
dimensional fingerprint plots [6,7] derived from Hirshfeld surfaces
[4,5], using the software CrystalExplorer, version 2.1 [3]. 2D-finger-
print plots were generated by using the di and de pairs measured on
each individual spot of the calculated Hirshfeld surface. Fig. 3
shows unambiguously that different intermolecular interactions
are present in the polymorphs form II and form I, which result in
different packing modes. In the plot of form II the ‘wings’ are more
3.1. Crystal structure
In an attempt to synthesize a totally different compound we ob-
tained serendipitously a second polymorph, form II, of N0,N00,N000-
triphenylbiuret. It is monoclinic, like the already known form [2],
form I, and has the same space group (P21/c) but a less extensive
hydrogen-bonding arrangement. In the biuret moiety the individ-
ual –NH–CO–N– groups are planar within the standard error and
the angle between these groups is considerably smaller [5.8(1)°]
than the corresponding angle in form I [12.8(2)°]. The N1–C3 bond
is significantly longer [1.4463(17) Å] than the N2–C9 and N3–C15
bonds [1.4190(18) and 1.4136(17) Å] (see Fig. 1).
pronounced which indicate aromatic interactions, namely C–Hꢀ ꢀ ꢀ
p
contacts. Only the 2D-fingerprint plot of the form I shows the pres-
ence of a pair of long sharp spikes, a characteristic of the strong
hydrogen bonds, in this case the N–Hꢀ ꢀ ꢀO hydrogen bonds. There
is evidence of close intermolecular Hꢀ ꢀ ꢀH contacts in both poly-
The N3–C2 [1.3404(17) Å] and N2–C1 [1.3540(17) Å] bond
lengths are in the expected range for delocalized C–N bonds [18].
The geometry around the N1 atom is the expected for a hybridized
sp2 atom.
morphs, that occur as a characteristic hump in the region
di = de = 1.2 Å of the plot.
3.3. Results of the ab initio calculations
While in the form I the phenyl rings C9–C14 and C15–C20 are
roughly coplanar with the least squares plane of the biuret moiety
[with dihedral angles of 11.6 (2)° and 10.9 (2)°, respectively], in
this new polymorph they are considerably out of this plane, with
the dihedral angles being 50.31(5)° and 21.58(6)° respectively.
The ring C3–C8 is almost perpendicular to the biuret fragment
[dihedral angle of 87.98(5)° between the two planes] and this
can be explained by the steric hindrance of N2–H2.
The molecules are linked in infinite chains running parallel to
the b axis, via C–Hꢀ ꢀ ꢀO hydrogen bonds [Fig. 2; Table 3]. These
chains have a periodicity of six atoms, graph-set symbol C(6)
according to Etter’s graph-set theory [19,20]. In this polymorph
Using the three methods described previously we have per-
formed the calculation of the energy differences between the two
conformers of the title compound. The results are presented on
Table 4. As expected, the results obtained using method
DE
crys (sin-
gle point energy for the experimental X-ray geometries) lead to
very large differences of energy between the two conformations,
and the major source of these differences seems to be the hydrogen
atomic positions, since in our structure the hydrogen atoms were
positioned geometrically and in the previously reported poly-
morph they were refined freely, leading to large differences in
the X–H bond lengths between the two structures. The energy dif-

P.S. Pereira Silva et al. / Journal of Molecular Structure 995 (2011) 66–71
69
Fig. 1. Diagram of the title compound with the ellipsoids drawn at the 50% probability level, with the atomic labelling scheme (Mercury, version 2.3 [24]).
optimized structure we can see that the conformation of form I
is closer to the equilibrium geometry: RMSD(form II-opt) = RMSD(-
form I-opt)= I-opt) ce:hsp sp="0.25"/>= ce:hsp sp="0.25"/>0.3803.
According to this, we expect that the conformation of form I should
be more stable than the form II conformation and thus, the method
D
E2 (energy for the structures with all parameters optimised ex-
cept torsion angles) seems to be more sensible than the other
two methods used.
In order to gain some insight of the influence of the intermolec-
ular interactions on the molecular geometry, namely the relative
orientation of the phenyl rings, we have performed DFT calcula-
tions of the equilibrium geometry of the free molecule starting
from the experimental X-ray geometries of the two conformers
of both forms II and I. The two equilibrium geometries obtained
are almost equal (see Table 2) and it can be seen that the DFT cal-
culations reproduce well the observed experimental bond lengths
and valency angles of the two conformers but some of the calcu-
lated dihedral angles in the free molecule differ significantly from
the experimental (Table 2).
The differences between the calculated and experimental bond
lengths are smaller than 0.0236 Å and 0.0208 Å, for the form II and
the form I conformers, respectively. The largest disagreement be-
tween the experimental and calculated angle values correspond
to the angle C1–N2–C9, which has a considerably smaller value
in both crystals.
In the optimized structure, the rings C9–C14 and C15–C20 are
almost in the plane of the biuret moiety and the ring C3–C8 is per-
pendicular to this plane (Table 5). Looking at the geometries of the
two polymorphs (Fig. 4), the most obvious deviation from the equi-
librium geometry of the free molecule is the large rotation of the
ring C9–C14 from the biuret plane, in form II, and this cannot be
explained by the hydrogen bonds since this ring does not partici-
Fig. 2. Packing diagram of form II, with the H-bonds shown as dashed lines.
ferences obtained with methods
typical literature values (<5 kcal molꢂ1) [23], especially for method
D D
E1 and E2 are much closer to
D
E2. The rotations about single bonds (intramolecular torsions) are
worth 1–3 kcal molꢂ1 but can be as high as 8 kcal molꢂ1 due to ste-
ric factors or restricted rotation [14], so in the title compound we
can expect energy differences of this order of magnitude.
In method
D
E1 the hydrogen atomic positions are changed in
the conformation of form I to match exactly those of conformation
of form II. However, other small differences in the remaining bond
lengths and angles can still be a source of errors.
Calculating the Root Mean Square Deviations (Mercury software
[24]) between the X-ray geometries of both polymorphs and the
pate in any hydrogen bond. However when looking for
actions, one finds that form II has nine of these interactions with a
p p inter-
ꢀ ꢀ ꢀ
distance between ring centroids, dc–c, smaller than 6 Å and an angle

70
P.S. Pereira Silva et al. / Journal of Molecular Structure 995 (2011) 66–71
Fig. 3. Two-dimensional fingerprint plot for the forms II and I.
Table 4
Results of ab initio calculations, with energies (aua) and relative conformational energies (kcal molꢂ1).
Method
HF/6-31G(d,p)
Form II
B3LYP/6-31G(d,p)
Form II
Form I
D
E
Form I
DE
D
D
D
Ecrys
E1
ꢂ1080.09944
ꢂ1080.09944
ꢂ1080.42104
ꢂ1080.31151
ꢂ1080.07680
ꢂ1080.42448
133.07
ꢂ14.21
2.16
ꢂ1086.75497
ꢂ1086.75497
ꢂ1087.12204
ꢂ1086.98430
ꢂ1086.73690
ꢂ1087.12840
143.90
ꢂ11.34
3.99
E2
HF/6-311+G(d,p)
B3LYP/6-311+G(d,p)
D
E1
ꢂ1076.87632
ꢂ1076.85727
ꢂ11.95
ꢂ1083.65124
ꢂ1083.63724
ꢂ8.78
a
1 au = 627.47237 kcal molꢂ1
.
b smaller than 60° (b is the slipping angle defined by the vector c1–
c2, from the first ring centroid to the second and the normal to the
plane of the first ring) and the phenyl ring C9–C14 participates in
six of these interactions with the strongest [dc–c = 3.929(1) Å] being
between C9–C14 rings of two neighbouring molecules. In the form
Table 5
Dihedral angles between the least-square planes of the phenyl rings and the plane of
the biuret moiety for the experimental and optimized geometries.
Phenyl ring
C3–C8
C9–C14
C15–C20
Form II
Form I
Opt. form II
Opt. form I
88.0(1)
72.9(1)
89.5
50.3(1)
11.6(1)
2.4
21.6(1)
10.9(1)
0.8
I the
pꢀ ꢀ ꢀp interactions are generally weaker.
It is difficult to quantify the influence of each interaction but it
is known that in typical cases, a crystal form with favourable single
bond torsion finds an alternative polymorph where a slightly disfa-
voured torsional geometry is compensated by better interactions
[23]. The latter situation is that of form II where the rotations of
the phenyl rings from the equilibrium positions may be explained
90.0
1.5
0.4
by the aromatic interactions (C–Hꢀ ꢀ ꢀ
p
and
pꢀ ꢀ ꢀp).
The conformational energy difference between form II and form
I is small and can be balanced by improved crystal packing as may
be the case of the new form II which has a slightly more efficient
close packing. Such packing corresponds to more and stronger
short non-bonding intermolecular contacts than in the previously
reported form.
4. Conclusions
A new polymorph of N0,N00,N000-triphenylbiuret was crystallized
and the structure was determined using single crystal X-ray dif-
fraction. This structure was compared to the previously reported
polymorph and the fingerprint plots of the two structures were de-
rived from the Hirshfeld surfaces to analyse the intermolecular
interactions. The major difference between the two polymorphs
is the presence/absence of the N–Hꢀ ꢀ ꢀO intermolecular hydrogen
bond. The DFT geometry optimizations suggest that the supramo-
lecular aggregation plays an important role in stabilizing the
Fig. 4. Comparison of the molecular conformation of form II, as established from
the X-ray study (red) with the X-ray geometry of form I (blue) and with the
optimized geometry (green) (Software used for visualization: VMD, version 1.8.6,
April 7, 2007 [25]). (For interpretation of the references to colour in this figure
legend, the reader is referred to the web version of this article.)

P.S. Pereira Silva et al. / Journal of Molecular Structure 995 (2011) 66–71
71
[2] O. Carugo, G. Poli, L. Manzoni, Acta Cryst. C 48 (1992) 2013.
observed geometries of the two polymorphs. Compared ab initio
calculations of conformational energies show that the positioning
of the hydrogen atoms may be a source of large errors, and that ex-
treme care should be taken in order to obtain significant results.
[3] S.K. Wolff, D.J. Grimwood, J.J. McKinnon, D. Jayatilaka, M.A. Spackman,
CrystalExplorer 2.1, University of Western Australia, Perth, 2007.
[4] M.A. Spackman, P.G. Byrom, Chem. Phys. Lett. 267 (1997) 215.
[5] J.J. McKinnon, A.S. Mitchell, M.A. Spackman, Chem. – Eur. J. 4 (1998) 2136.
[6] J.J. McKinnon, M.A. Spackman, A.S. Mitchell, Acta Cryst. B 60 (2004) 627.
[7] M.A. Spackman, D. Jayatilaka, CrystEngComm 11 (2009) 19.
[8] H. Tada, R. Okawara, J. Org. Chem. 35 (1970) 1666.
[9] Bruker, APEX2 and SAINT, Bruker AXS Inc., Madison, Wisconsin, USA, 2003.
[10] G.M. Sheldrick, SADABS. University of Göttingen, Germany, 2003.
[11] G.M. Sheldrick, Acta Cryst. A 64 (2008) 112.
[12] A.A. Granovsky, PC GAMESS/Firefly version 7.1.F, 2009. <http://
classic.chem.msu.su/gran/gamess/index.html>.
[13] M.W. Schmidt, K.K. Baldrige, J.A. Boatz, S.T. Elbert, M.S. Gordon, J.J. Jensen, K.
Koseki, N. Matsunaga, K.A. Nguyen, S. Sue, T.L. Windus, M. Dupuis, J.A.
Montgomery, J. Comput. Chem. 14 (1993) 1347.
5. Supplementary material
Crystallographic data for the structure reported in this paper
have been deposited with the Cambridge Crystallographic Data
Center number, CCDC 795370. Copies of this information may be
obtained free of charge on application to CCDC, 12 Union Road,
Cambridge CB2 1EZ, UK (fax: +44 1223 336 033; e-mail: depos-
it@ccdc.cam.ac.uk or http://www.ccdc.cam.ac.uk).
[14] D. Buttar, M.H. Charlton, R. Docherty, J. Starbuck, J. Chem. Soc. Perkin Trans. 2
(1998) 763.
[15] A.D. Becke, Phys. Rev. A 38 (1988) 3098.
[16] A.D. Becke, J. Chem. Phys. 98 (1993) 5648.
Acknowledgements
[17] C. Lee, W. Yang, R.G. Parr, Phys. Rev. B 37 (1988) 785.
[18] G. Gilli, V. Bertolasi, F. Bellucci, B. Fabrizio, V. Ferretti, J. Am. Chem. Soc. 108
(1986) 2420.
[19] M.C. Etter, Acc. Chem. Res. 23 (1990) 120.
[20] J. Bernstein, R.E. Davis, L. Shimoni, N.-L. Chang, Angew. Chem. Int. Ed. Engl. 34
(1995) 1555.
We would like to acknowledge Universiti Sains Malaysia (USM)
for the University Grant 1001/PTEKIND/8140152 to Dr. Raza Murad
Ghalib and Dr. Sayed Hasan Mehdi. P.S. Pereira Silva acknowledges
the support by Fundação para a Ciência e a Tecnologia, under the
scholarship SFRH/BD/38387/2008.
[21] J.F. Malone, C.M. Murray, M.H. Charlton, R. Docherty, A.J. Lavery, J. Chem. Soc.
Faraday. Trans. 93 (1997) 3429.
[22] V.A. Blatov, Acta Cryst. A 62 (2006) 356.
[23] A. Nangia, Acc. Chem. Res. 41 (2008) 595.
References
[24] C.F. Macrae, P.R. Edgington, P. McCabe, E. Pidcock, G.P. Shields, R. Taylor, M.
Towler, J. van de Streek, J. Appl. Crystallogr. 39 (2006) 453.
[25] W. Humphrey, A. Dalke, K. Schulten, J. Mol. Graph. 14 (1) (1996) 33.
[1] J. Bernstein, Polymorphism in Molecular Crystals. IUCr Monographs on
Crystallography, Oxford University Press, Oxford, 2002.