Structural insights into the hexamorphic system of an isoniazid derivative

Article abstract of DOI:10.1039/c5ce00275c

Crystal polymorphism is the capacity of a crystalline solid to exist in more than one structural arrangement. The variation in the crystalline forms often induces different mechanical, thermal, and chemical properties. These changes can markedly influence

Full text of DOI:10.1039/c5ce00275c

CrystEngComm
View Article Online
View Journal
PAPER
Structural insights into the hexamorphic system of
an isoniazid derivative†‡
Cite this: DOI: 10.1039/c5ce00275c
a
b
b
a
a
*
D. Hean, T. Gelbrich, U. J. Griesser, J. P. Michael and A. Lemmerer
Crystal polymorphism is the capacity of a crystalline solid to exist in more than one structural arrangement.
The variation in the crystalline forms often induces different mechanical, thermal, and chemical properties.
These changes can markedly influence the bioavailability, hygroscopicity, stability and other performance
characteristics of the active pharmaceutical ingredient. Isoniazid, a well-known pharmaceutical, is used as
a first-line treatment against Mycobacterium tuberculosis (TB). Derivatives of isoniazid were developed in
response to TB drug resistance. One such derivative synthesized, isonicotinic acid IJE)-IJ1-
phenylethylidene)hydrazide (IPH), was found to exhibit complex polymorphic behaviour. To date, only one
crystal structure of IPH has been reported in the literature. We have discovered and isolated an additional
five polymorphs of IPH from various crystallization techniques, namely slow cooling, rapid evaporation,
sublimation, as well as from hot-stage experiments. All of the polymorphs display hydrogen bonding
through the carbonyl acceptor and hydrazide donor. Structural information about the polymorphs was
obtained by single crystal and powder diffraction, while characterisation included infrared spectroscopy
and Raman spectroscopy. The thermal properties of these polymorphs were also investigated using differ-
ential scanning calorimetry and hot stage microscopy.
Received 5th February 2015,
Accepted 23rd March 2015
DOI: 10.1039/c5ce00275c
www.rsc.org/crystengcomm
points, mechanical processing and shelf-life.6 In addition,
polymorphism may influence every stage of drug compound
Introduction
Since the middle of the 18th century it has been known that
substances can crystallize into more than one form. Further,
the phenomenon of polymorphism of drug compounds has
received extensive attention by academia and industry, includ-
ing much publicised debate by patent lawyers.1 It is well
known that adoption of a particular crystalline structure has a
profound influence over the solid-state properties of a com-
pound, and this was illustrated by early pioneering reports of
Aguiar and colleagues at Parke–Davis, where polymorphism
was shown to influence the bioavailability and dissolution
rates of chloramphenicol palmitate.2,3 The influence on solid-
state properties is not limited to dissolution4 and bioavailabil-
ities5 only, but has a rather extensive influence on melting
development, starting from pre-clinical stages to post market-
ing phase of the drug product, including patent protection.7
Polymorphism can be viewed as the manner in which a
compound can rearrange itself in such a way as to have dif-
ferent packing or conformational arrangements.8–10 Packing
polymorphism results from different packing arrangements
of conformationally rigid molecules, while conformational
polymorphism entails the rearrangement of a conformationally
flexible molecule to produce more than one different confor-
mation in the solid state. However, most often both packing
and conformational differences are observed among two or
more polymorphs.11 Conformational variation of organic com-
pounds is a rather subtle but an often seen feature in
polymorphism.12
Polymorphs can be interconverted by phase transforma-
tions or a solvent-mediated process; phase transformations
can also be induced by heat or mechanical stress.13 When
developing and deciding on which drug form to be marketed
or processed, it is extremely crucial to develop an under-
standing of the phase transformation relationships between
the polymorphic forms. This enables drug development to
focus on either the stable or metastable stable forms, and to
decide whether the desired form with the most beneficial
solid-state properties will remain stable throughout the
lifespan of the drug compound, especially for final
a Molecular Sciences Institute, School of Chemistry, University of the
Witwatersrand, Johannesburg 2050, South Africa.
E-mail: Andreas.lemmerer@wits.ac.za; Fax: +27 11 717 6749;
Tel: +27 11 717 6711
b University of Innsbruck, Institute of Pharmacy, Pharmaceutical Technology,
Josef-Moeller-Haus, Innrain 52c, A-6020 Innsbruck, Austria
† Electronic supplementary information (ESI) available: Details of the XPac
studies, X-ray powder diffraction, RAMAN, FT-IR, NMR, additional DSC traces
and further crystallographic tables and data. CCDC 970533–970537 and
1022894. For ESI and crystallographic data in CIF or other electronic format see
DOI: 10.1039/c5ce00275c
‡ Dedicated to David J. W. Grant on the 10th anniversary of death (died
December 9, 2005 at age 68).
This journal is © The Royal Society of Chemistry 2015
CrystEngComm

View Article Online
Paper
CrystEngComm
formulation. Differential Scanning Calorimetry (DSC) may be
employed to reveal the energetic features of polymorphs of a
particular drug compound.13 However, to determine the
monotropic or enantiotropic relationships between multiple
forms, other data such as solubility and density may be nec-
essary in addition to DSC data.14,15 Moreover, the comprehen-
sive and successful characterisation of polymorphic systems
needs a series of additional techniques such as polarised light
hot-stage microscopy (HSM), X-ray powder diffraction (XRPD),
and methods such as Raman and Fourier transform infrared
(FT-IR) spectroscopy, and solid state NMR amongst others.
Isonicotinic acid hydrazide (isoniazid) was initially pre-
pared by Meyer and Malley in 1912 from ethyl isonicotinate
and hydrazine.16 However, only in 1945 was it used as a first
line treatment against Mycobacterium tuberculosis, when it
was shown that the parent compound, nicotinamide,
expressed biological activity.17 M. tuberculosis subsequently
developed strong resistance against the original isoniazid
drug, which sparked the hunt for other pyridine derivatives
with activity against it.17
a highly polymorphic compound is 5-methyl-2-ij(4-methyl-2-
nitrophenyl)amino]-3-thiophenecarbonitrile (ROY, seven crystal
structures),26 which was recently superseded by flufenamic acid
for the record of solved crystal structures (eight).27 Another
highly polymorphic system is that of triacetone–triperoxide
(TATP), with six known crystal structures.28 In addition, to solv-
ing the six crystal structures for IPH, the phase transformations
and thermal relationships for IPH I–VI have been delineated,
including the determination of some thermodynamic parame-
ters (enthalpies and temperatures of fusion and phase transi-
tions). The use of HSM has been a critical tool in the study of
these polymorphs.
Experimental
Synthesis of isonicotinic acid IJE)-IJ1-
phenylethylidene)hydrazide (IPH)
Isoniazid (1.006 g, 7.335 mmol) (obtained from Sigma-
Aldrich) and acetophenone (1.002 g, 8.343 mmol) were
dissolved in absolute ethanol (75 ml) in a 100 ml round-
bottomed flask. To this mixture a catalytic amount of glacial
acetic acid was added, and the solution was heated at reflux
for 18 hours. A white solid precipitated from the reaction
mixture which was filtered and washed with cold ethanol (3 ×
25 ml) to give desired product (1.660 g 95% yield). Character-
isation of the product was determined by 1H and 13C NMR,
IR and DSC (mp. 171.9 °C) (supplied in ESI†). Recrystalliza-
tion was performed from slow evaporation at ambient condi-
tions using absolute ethanol as a solvent, yielding IPH V
(Fig. 1e).
At the time, a number of derivatized forms of isoniazid,
such as isonicotinic acid IJE)-IJ1-phenylethylidene)hydrazide
(IPH) (Scheme 1), were developed to combat M. tuberculo-
sis.17 However, IPH exhibited poor anti-tuberculosis activity;18
fortunately the molecular backbone comprising the hydraz-
one functional group R1R2CN–NH2, is considered to be an
important
pharmacophore.19,20
This
pharmacophore
expresses a range of bioactivity such as anti-microbial, anti-
convulsant, and antitumor activity.20,21 From an organic solid
state perspective, the pharmacophore hydrazone functional
group has numerous hydrogen bonding sites including a
conformationally flexible molecular backbone. In addition,
IPH possesses an additional pyridine hydrogen bond acceptor
region, and the potential for π⋯π stacking interactions devel-
oping from the aromatic rings on the two extremities of the
molecule.22
Hot Stage Microscopy (HSM)
Hot stage microscopy experiments were performed using an
Olympus SZ61 polarized light microscope, equipped with a
Kofler hot stage. Various heating and cooling procedures
To date, only one crystal structure of isonicotinic acid
IJE)-IJ1-phenylethylidene)hydrazide has been reported in the
literature,23 corresponding to IPH V of the present study. Five
additional forms have been isolated in this study and their
crystal structures have been determined. The total of six
forms has been labelled as IPH I–VI, in a sequence of highest
to lowest melting points. Though it is possible to generate a
high number of polymorphs by using additives as shown for
phenobarbital (eleven forms),24,25 the isolation and subsequent
determination of six different crystal structures for a single com-
ponent molecular compound is rare. A famous example for such
Fig.
1 IPH I, II, III and V recovered from various crystallization
procedures. a–b) Sublimation crystallization performed on a Kofler
hot-bench yielding IPH I. c) IPH II was recovered from slow evapora-
tion of acetonitrile. d–e) IPH III and V were recovered from rapid evap-
oration of ethanol, which were then physically separated under the
polarizing microscope.
Scheme
1 Atomic labelling scheme of isonicotinic acid-IJE)-IJ1-
phenylethylidene)hydrazide (IPH).
CrystEngComm
This journal is © The Royal Society of Chemistry 2015

View Article Online
CrystEngComm
Paper
for temperature induced crystal growth, melt and phase tran-
sitions were applied.
Complementary RAMAN spectra were collected for IPH I–III
and V (see ESI† for spectra) using a Bruker MultiRAM FT-Raman
Spectrometer fitted with a Nd-YAG laser and a Germanium
diode detector. The experiment measured from 4000 cm−1
to 400 cm−1. Infrared and RAMAN experiments were only
performed on the forms with sufficient quantities and
hence there are no spectra available for IPH IV and VI.
Preparation of diffraction crystals of IPH
IPH
I
was recovered from sublimation experiments
performed on a Kofler hot-bench (Fig. 1a–b). IPH II and V
crystals (100 mg each) were recovered from slow cooling solu-
tion (10 ml) experiments using solvents acetonitrile and etha-
nol respectively (Fig. 1c & e). IPH V crystals were also produced
from many different solutions experiments (see ESI† for solu-
tion experiment details). IPH III crystals (100 mg mixture) were
recovered from rapid evaporation; a 100 ml beaker containing
dissolved IPH in a water/ethanol (1 : 1) solution (10 ml) was
boiled in a fume hood until dryness (Fig. 1d). It should be
noted that these slow cooling solution experiments exclusively
produced the respective polymorphs; however the rapid evapo-
ration experiments produced a mixture of IPH III and V. As a
result, diffraction quality crystals of IPH III were manually sep-
arated out under the polarizing microscope. IPH VI was
recrystallized from melt on the hot-stage and IPH IV was
grown on the hot-stage from melt by using a very slow cooling
rate methodology. This was done to provide ample time to
allow sufficient crystal growth for X-ray analysis. Despite that
the crystals were still extremely small and obtained in very
limited amounts.
Powder X-ray diffraction
Powder X-ray diffraction data for all compounds were col-
lected at 293 K on a Bruker D2 Phaser diffractometer which
employs a sealed tube Cu X-ray source (λ = 1.5406 Å), operat-
ing at 30 kV and 10 mA, and LynxEye PSD detector in Bragg–
Brentano geometry. Powder X-ray diffraction confirmed that
the single crystal structures were representative of the bulk
material (see ESI†).
Crystal data and X-ray structural analysis
The Bruker D8 VENTURE PHOTON CMOS area detector dif-
fractometer, equipped with a graphite monochromated Mo
Kα1 radiation (50 kV, 30 mA), was used to collect all the
intensity data. Crystal structures of IPH I–V were collected at
173 K and IPH VI was collected at 100 K to achieve satisfac-
tory thermal ellipsoids and high resolution data. The pro-
gram SAINT+, version 6.02 29 was used to reduce the data and
the program SADABS was used to make corrections to the
empirical absorptions. Space group assignments were made
using XPREP29 on all compounds. In all cases, the structures
were solved in the WinGX30 Suite of programs by direct
methods using SHELXS-97 31 and refined using full-matrix
least-squares/difference Fourier techniques on F2 using
SHELXL-97.31 All non-hydrogen atoms were refined anisotrop-
ically. Thereafter, all hydrogen atoms attached to N atoms
were located in the difference Fourier map and their coordi-
nates refined freely with isotropic parameters 1.5 times those
of the ‘heavy’ atoms to which they are attached. All C–H
hydrogen atoms were placed at idealized positions and
refined as riding atoms with isotropic parameters 1.2 times
those of the ‘heavy’ atoms to which they are attached.
Diagrams and publication material were generated using
Differential Scanning Calorimetry (DSC)
Differential scanning calorimetry data (Table 1) were col-
lected using a Mettler Toledo 822e with aluminium pans
under air purge. Star SW 9.20 was used for instrument con-
trol and data analysis. Exothermic events were shown as
peaks. Various heating and cooling protocols were performed
as quantitative analysis of the polymorphs. The temperature
and energy calibrations were performed using pure indium
(purity 99.99%, m.p 156.6 °C, heat of fusion 28.45 J g−1).
Fourier-Transform Infrared (FT-IR) and Raman spectroscopy
The infrared spectra of IPH I–III and V were recorded (see
ESI† for spectra) using a Bruker Vertex 70 equipped with an
ATR (MVP-Pro). The spectra were measured from 3500 cm−1
to 400 cm−1 and major peaks were assigned using OPUS v6.5.
Table 1 Thermochemical data for polymorphs of IPH I–VI
Form
Property
IPH I
IPH II
IPH III
IPH IV
IPH V
IPH VI
Tfus [°C] DSC
HSM [°C]
171.9 0.2
173
33.1 1.4
—
—
~130
164.7 0.8
165
—
—
—
—
1.328
162
163
—
→I
+3.1 0.2
~130
1.326
—
145
—
—
—
—
→III
—
—
1.337
~156
155–158
—
→III
—
~130
1.303
ΔfusH [kJ mol−1
]
Transition to form
→III
ΔtrsH [kJ mol−1
Tcryst [°C]
]
ρ [g cm−3 a
]
1.287
1.332
a
Calculated from crystal structure determinations at −100 °C.
This journal is © The Royal Society of Chemistry 2015
CrystEngComm

View Article Online
Paper
CrystEngComm
Table 2 Crystallographic data for IPH I–VI
Compound
I
II
III
IV
V
VI
Formula
C14H13N3O1
239.27
0.40 × 0.061 ×
0.020
C14H13N3O1
239.27
0.55 × 0.28 ×
0.10
C14H13N3O1
239.27
0.42 × 0.11 ×
0.050
C14H13N3O1
239.27
0.56 × 0.046 ×
0.040
C14H13N3O1
239.27
0.45 × 0.20 ×
0.060
C14H13N3O1
239.27
0.42 × 0.37 ×
0.040
Mr
Crystal size [mm−3
]
Crystal habit
T [K]
Needle
173(2)
Prism
173(2)
Needle
173(2)
Needle
173(2)
Plate
173(2)
Block
100(2)
System
Space group
a [Å]
b [Å]
c [Å]
Triclinic
P1
Monoclinic
P21/c
10.211(2)
30.331(7)
8.2353(2)
90
110.19(1)
90
2393.9(9)
8/2
Orthorhombic
Pbca
6.3540(4)
7.6624(6)
49.231(3)
90
90
90
2397.0(3)
8/1
1.326
Monoclinic
P21/c
10.621(2)
14.442(2)
8.2589(2)
90
109.62(5)
90
1193.3(3)
4/1
Monoclinic
P21/c
25.899(1)
5.5463(3)
8.3187(4)
90
95.876(4)
90
1188.7(1)
4/1
Monoclinic
P21/c
13.488(2)
9.6611(1)
9.3604(1)
90
90.183(5)
90
1219.7(3)
4/1
¯
9.7360(6)
9.8752(6)
26.154(2)
92.856(4)
100.29(4)
91.291(4)
2469.8
α [°]
β [°]
γ [°]
V [Å3]
Z/Z′
8/4
1.287
ρ [g cm−3
FIJ000)
]
1.328
1.332
1.337
1.303
Scan range (θ)/°
Total reflections
Unique reflections
1.59–25.5
39845
9177 [0.0822]
1.34–28.00
35411
5778 [0.0345]
3.31–25.5
12081
2124 [0.0236]
2.97–23.31
12494
1723 [0.1089]
1.58–28.00
14365
2866 [0.046]
3.03–28.00
20120
2900 [0.069]
ijRIJint)
]
No. data with I ≥ 2σIJI)
4183
669
4579
335
1777
168
1183
167
2245
168
1992
168
Parameters
R1 [I > 2σIJI)]
wR2 (all data)
CCDC
0.0559
0.1223
970537
0.0398
0.1117
970535
0.0425
0.1317
970536
0.0532
0.1241
1022894
0.0605
0.1587
970534
0.0729
0.2468
970533
ORTEP-3,32 PLATON33 and DIAMOND.34 Experimental details
of the X-ray analyses are provided in Table 2.
variations in the rotation angles of aromatic rings about the
C1–C6 and C8–C9 bonds (the dissimilarity indices from this
analysis will be denoted x8). Further details are collected in
section SI4 of the ESI.†
XPac studies
Comparisons of crystal structures were carried out with the
XPac35 software and quantitative dissimilarity indices were
calculated in the previously described manner.36 Two sets of
calculations were performed. The first set was based on geo-
metrical parameters calculated from all 18 non-H atomic
positions matching the IPH template structure, and the dis-
similarity indices obtained from it will be denoted x18. For
the second set of calculations, only a core molecular unit
defined by the positions of eight atoms (C1, C6, O1, N1, N3,
C8, C7, C9; see Scheme 1) was used to minimise the effect of
Results and discussion
Hot-stage microscopy of IPH
The observed phase transitions are illustrated in Fig. 2 to 4
and two microscopic recordings of phase transition cycles in
thin film samples are provided as a video stream in the ESI.†
On heating single crystals of IPH V, recovered from bulk crys-
tallization after the synthesis of the compound, a phase tran-
sition to IPH III is observed (Fig. 2). This transition occurs at
about 129 °C and was found to be irreversible on cooling.
Additional phase transitions are easily recognized in melt
film preparations (between glass slide and cover slip) of IPH.
Such films recrystallize on slow cooling to four concomitant
forms of IPH, namely form I, III, IV and VI (Fig. 3). The
recrystallization process occurs at about 130 °C on cooling
and in repeated experiments, different amounts of the indi-
vidual forms were obtained, ranging from complete coverage
of the thin film preparation (IPH III) to only a small portion
of a specific crystal form beside others. On reheating such
preparations, several phase transitions are observed. The
transitions of IPH IV and VI to IPH III occur over the range
of 145–155 °C. Additionally during this process, IPH IV and
VI is also transformed by adjacent IPH I. However, because
the transition rate of the two metastable forms into IPH I is
Fig. 2 Polarized light HSM photomicrographs of a plate-like crystal of
IPH recovered from bulk recrystallization. a) Shows IPH V before tran-
sition at about 129 °C, whilst b) shows the transition with the proceed-
ing transition interface. The yellow arrows (c–d) show the progress of
the phase transition until the crystal has completely transformed into
IPH III.
CrystEngComm
This journal is © The Royal Society of Chemistry 2015

View Article Online
CrystEngComm
Paper
Fig. 4 Microphotographs of islands of IPH crystallized on silicon oil
coated glass slide on heating. a) Typical appearance of IPH IV (dendrites)
beside droplets recrystallized to IPH III; b) concomitantly recrystallized IPH
III, IV and VI; c) inhomogeneous melting of IPH VI to III at 145 °C (red
arrows in the enlarged insert indicate the recrystallization of III in melted
areas); d) completion of the phase transition VI to III at 155 °C; e–f) end of
melting process of IPH III at about 163 °C.
Fig. 3 Polarized light HSM photomicrographs of a thin film preparation
showing the polymorphs IPH I, III, V and VI formed on slow cooling the
melt and their phase transitions on heating. a) Sample before heating
showing the typical morphology of IPH I (spherulites), IV (mass of fine
needles) and VI (homogeneous plate); b) IPH I grows towards IV and III
appears at the top of the image (dark blue) indicating the phase transitions
IV to I and IV to III starting at about 145 °C (red arrows mark the transition
boundary). c) The phase transitions continue (including that of VI to III)
until c) IPH V and VI are completely transformed into IPH III and I; e) a
further phase transition (155 °C) occurs (IPH III to I) until only IPH I
remains; f) IPH I starts to melt at 173 °C.
and needles (Fig. 4a), which do not quickly transform to IPH
III on heating because they are largely isolated from this
more stable form. Therefore, it was possible to determine the
much smaller than that into IPH III, only a relatively small
portion of IPH I and a large fraction of IPH III is present in
the thin-film samples after the first transition processes are
completed at about 155 °C.
The second phase transition process begins with the end
of the first phase transition range (about 155 °C), and reflects
the transition of IPH III into IPH I. At ca. 165 °C the entire
thin-film is converted into the high temperature form IPH I,
which then melts at 173 °C. When such hot-stage experi-
ments are performed with preparations produced by fusing
spots of IPH powder on a glass slide coated with silicon oil
without a cover slip, additional thermal events can be
observed (Fig. 4). The advantage of this preparation tech-
nique compared to a coherent melt film preparation is the
formation of isolated droplets, in which a single metastable
form may nucleate and grow. If such isolated areas of meta-
stable forms are not seeded or are in contact with more sta-
ble phases, no solid–solid phase transition is induced and
they may survive until their melting point is reached. In fact,
the individual melt droplets of IPH crystallize to different
polymorphic forms, similar to that described above for thin
film preparations (Fig. 3). However, IPH IV forms dendrites
Fig. 5 DSC traces of successive heating/cooling runs beginning with
IPH V. The initial trace shows the first phase transition (trs) from V to
III followed by the inhomogeneous melting (fus III, cryst I) of IPH III
into I, and finally melting (fus) endotherm of IPH I. On cooling, IPH I,
III, IV and VI recrystallize (cryst) concomitantly between about 90 and
80 °C. In the second heating cycle the phase transitions of IV and VI to
III are indicated by a weak endotherm around 157 °C followed by the
imhomogeneous melting of IPH III to I and the melting of IPH I. The
following cool/heat cycle (trace 4 and 5) confirm the reproducibility of
the thermal events. The heating rate is 15 K min−1
.
This journal is © The Royal Society of Chemistry 2015
CrystEngComm

View Article Online
Paper
CrystEngComm
The supercooled melt (or glass) resulting from this fast
cooling process recrystallizes between 60 and 70 °C on
warming and the obtained crystals exclusively consist of IPH
I, melting at 172 °C.
DSC analysis and estimation of the thermodynamic stability
The thermal behaviour of the IPH forms observed with HSM
was additionally verified and complemented by DSC experi-
ments. A series of heating and cooling experiments were
conducted, mostly by starting with powdered IPH V (Fig. 6).
DSC traces of other isolated IPH forms are shown in Fig. S1
to S3 of the ESI.† Reliably measured thermochemical data are
summarised in Table 1.
Fig. 6 DSC heating run of the super-cooled IPH melt. At 65 °C, a broad
exothermic peak is observed, indicating the crystallization of IPH I, followed
by the melting endotherm of IPH I.
In the first heating cycle (Fig. 5), a weak endothermic peak
with an onset temperature of 123 °C indicates the phase tran-
sition of IPH V into III. At 163 °C the inhomogeneous melt-
ing process (melting of III and simultaneous crystallization
of IPH I) of IPH III to I is recorded as endo/exo process. The
final thermal event in the heating cycle is caused by the melt-
ing of IPH I at 172 °C. The subsequent cooling cycle shows
the concomitant crystallization of IPH I, III, IV and VI from
the melt. This crystallization process occurs reproducibly
between 80 and 95 °C (see also cooling cycle 2 and 3 in
homogeneous melting point of IPH IV at 145 °C (Fig. 4b)
although islands of IPH III and VI are present in the close
surrounding. On further heating the inhomogeneous phase
transition of IPH VI into IPH III is observed around 155 °C
(Fig. 4c–d). Owing to the lack of IPH I seeds, no further tran-
sitions occur and these preparations and the melting of
phase pure IPH III can be observed at 163 °C.
A third alternative melt crystallization condition was
applied by quench-cooling melted samples of IPH to 25 °C.
Fig. 7 Asymmetric units and atom numbering schemes of I–VI. Displacement ellipsoids are drawn at the 50% probability level and H atoms are
shown as small spheres of arbitrary size. In the case if I, H atoms not involved in hydrogen bonding are omitted for clarity.
CrystEngComm
This journal is © The Royal Society of Chemistry 2015

View Article Online
CrystEngComm
Paper
The DSC results are largely consistent with the HSM
observations. However, neither technique allowed the deter-
mination of the melting point of IPH V, because of the fast
transition to IPH III at about 123 °C. However, it was possible
to determine the melting temperatures of IPH II (165 °C) and
III (162–163 °C) with both HSM and DSC (see. Fig. S2 and S3
of the ESI†).
As shown in Fig. 6, the supercooled melt resulting from
crash cooling of the melt shows a broad recrystallization exo-
therm at approximately 65 °C upon reheating. Subsequently
only the melting of IPH I is recorded, confirming the HSM
observations (see above).
For a compound with six polymorphs (n = 6) we can calcu-
late fifteen polymorph pairs either being monotropically or
enantiotropically related.37 The available thermal data allow
only a very limited estimation of the thermodynamic relation-
ships of some of the forms in this complex polymorphic sys-
tem. IPH I is the form with the highest melting point and
lowest density. Therefore, we can assume that this form
shows the highest entropy and a low thermodynamic stability
(high free energy) at 0 K. From this, it can be deduced that
IPH I is enantiotropically related to most, if not to all, of the
other forms. IPH V shows the highest density, and our data
suggest that this polymorph is the thermodynamically stable
form at room temperature. The observed endothermic transi-
tion (DSC) of IPH V into IPH III is a clear indication of the
enantiotropic relationship between these two forms (heat-of-
transition rule38). This fact implies the existence of a temperature
window above room temperature where IPH III is thermodynami-
cally more stable than IPH I and V. Moreover, IPH II is the form
with the second highest melting point and its density is slightly
higher than that of IPH III, suggesting a higher stability of IPH II
with respect to IPH III, at least in the higher temperature region.
This finally means that there are three temperature regions, with
different forms being the thermodynamic most stable form:
IPH V at and below room temperature, IPH II in a temperature
window above room temperature and below the melting point,
and finally IPH I in a limited temperature range right below its
melting point. Consequently, IPH III, IV and VI are metastable in
the entire temperature range. However, as already mentioned,
the available thermochemical data do not allow a save estimation
of the thermodynamic stabilities of the IPH forms and more
efforts are needed to resolve the energetic features this polymor-
phic system. Nevertheless, it should be stressed that the IPH
forms exhibit a high kinetic stability, which is indicated by the
observation that I, II and III did not transform to the room tem-
perature stable form V even after a storage period of one year.
Fig. 8 a) Schematic representation of the N–H⋯OC-bonded C(4)
chain motif (IPH: R1 = 4-pyridyl, R2 = phenyl) and b) tree diagram
according to ref. 45 which summarises packing relationships between
crystal structures; c) N–H⋯OC bonded chains of the geometry types
X (III), Y (IV) and Z (VI), viewed along their respective translation vector
(dotted lines indicate an axial glide plane with a glide vector of 1/2 of
the lattice vector normal to the projection plane); d) centrosymmetric
dimer units present in the IPH polymorphs II (molecule B) and VI [sym-
metry operations: (i) 1 − x, − y, 1 − z and (ii) 1 − x, − y, − z].
Fig. 5). Reheating the crystallized melt shows a weak endo-
thermic event around 157 °C followed by an exothermic signal,
well visible in the third heating cycle of Fig. 5. This thermal
event comprises most likely the transformation and inhomoge-
neous melting of untransformed traces of IPH VI to IPH III,
suggesting that the lower melting IPH IV (145 °C, see HSM
experiments) underwent a solid–solid transformation to IPH III
already at lower temperatures. We assume that the transforma-
tion of IV into III is energetically very weak and occurs over a
broad temperature range, which may be why the transition it
is not observable in the DSC signal.
Crystal structures of IPH I–VI
Crystal structure of IPH I
The asymmetric unit of IPH I consists of four molecules,
denoted A–D (Fig. 7). These are pairwise hydrogen bonded to
one another (A + B and C + D) via N1–H1⋯O1C6 interac-
tions which result in two discrete dimers whose graph set
symbol39,40 is R22(8). Form I is the only IPH polymorph in
This journal is © The Royal Society of Chemistry 2015
CrystEngComm

View Article Online
Paper
CrystEngComm
Fig. 9 Two-dimensional packing similarity between the crystal structures of PEH and AHE and in the IPH polymorphs III and V, based on the
stacking of N–H⋯OC-bonded chains of type X. Each structure is viewed perpendicular (top) and parallel (bottom) to the translation vector of its
H-bonded chains. In each diagram one instance of X1 (orange) or X2 (orange + blue) is highlighted and for each chain the direction of translation
is indicated by an arrow in the upper row of diagrams. All hydrogen atoms except those involved in hydrogen bonding are omitted for clarity.
which the position of the amide H atom with respect to the
carbonyl group is syn. (Note: the IPH from the reaction
always comes out with the (E)-geometry about the CN dou-
ble bond and we never observed the IJZ)-isomer).
symbol CIJ4). By contrast, the hydrazide hydrogen and pyri-
dine nitrogen of molecule B are employed in an N1–H1⋯N2
interaction which results in a dimer with a central R22(14)
ring. The A- and B-type molecules are not hydrogen bonded
to one another. IPH II is the only polymorph of IPH in
which the pyridine nitrogen atom is engaged in hydrogen
bonding.
Crystal structure of IPH II
The asymmetric unit of IPH II contains two molecules,
denoted A and B (Fig. 7), which form two distinct hydrogen-
bonded structures. Molecules of type A are linked into an
infinite chain via (amide) N1–H1⋯O1 hydrogen bonds. This
chain propagates parallel to the c-axis and has the graph set
Crystal structures of IPH III–VI
The structure of each of these three polymorphs contains just
one independent molecule (Fig. 7). They all exhibit hydrogen
CrystEngComm
This journal is © The Royal Society of Chemistry 2015

View Article Online
CrystEngComm
Paper
an additional NH2 ring substituent in the latter structure is taken
into account.
By contrast, high dissimilarity indices (x18 = 12.6–14.2) are
obtained for all the X1 comparisons involving IPH III, indi-
cating significant differences both in the geometry of the
individual H-bonded chains and in their stacking mode.
These differences (illustrated in the top row of diagrams in
Fig. 9) also result in relatively large deviations in the lattice
parameters associated with X1 (IPH V, PEH and AHE:
5.55–5.78 and 8.22–8.59 Å vs. III: 6.35 and 7.66 Å).
The structures of IPH V and PEH have a bilayer unit in
common (denoted X2). It consists of two X1 stacks which are
related to one another either by crystallographic (IPH V) or
by local (PEH) inversion symmetry. In the upper four packing
diagrams of Fig. 9, individual X1 units are represented by a
single N–H⋯OC-bonded chain and differences between
the X1 packing sequences of the four structures are
highlighted by arrows indicating the direction of inter-
molecular H→O interactions. Neighbouring X2 bilayers of
IPH V are related to one another by a 21 screw operation
about an axis parallel to the stacking vector of the X1 stacks
(b axis), while neighbouring X2 bilayers of PEH are related by
a two-fold screw operation about an axis which lies perpen-
dicular to both the stacking vector of X1 and the translation
vector of the H-bonded chains (crystallographic c axis). The
X2 units of IPH V and PEH display only small geometrical
differences (x18 = 2.2).
In the polymorphs II and IV, the N–H⋯OC-bonded
chains possess a Y-type geometry (Fig. 8; II: A-type molecules
only, see above). Neighbouring Y chains related by transla-
tion are stacked in the same fashion in IPH II and IV,
resulting in a mutual two-dimensional SC Y1 (Fig. 10). The
Y1 stacks of II lie in ac planes and alternate with stacks of
dimers formed by B-type molecules. In IV, consecutive Y1
stacks (in ac planes) are related by an inversion operation. As
a consequence of these relationships, the unit cells of II and
IV correspond with respect to a, c and β (see Table S4 in ESI†).
However, it should be noted that the Y1 stack geometries
of II and IV also show significant differences, indicated by
the x18 dissimilarity index of 15.0, whereas the corresponding
x8 index of 8.7 is substantially lower. These values indicate
that the common packing arrangement of II and IV is best
maintained within the central section of the Y chains, while
the larger differences arise primarily from the packing of the
aromatic rings, see Fig. 10. An overview of the geometry rela-
tionships arising from the CIJ4) chain motif is given in Fig.
8b, which also includes the geometry type Z found solely in
the IPH polymorph VI (Fig. 8c). Another noteworthy packing
similarity involves the centrosymmetic dimer unit depicted in
Fig. 7d which contains extensive van der Waals contacts and
is present in II (molecule B) and VI (x18 = 8.2).
Fig. 10 Two-dimensional similar packing of molecules, based on
Y-type N–H⋯OC-bonded chains, in the IPH polymorphs II (molecule
type A only) and IV. Each structure is viewed parallel (left) and perpen-
dicular (right) to the direction of translation of its H-bonded chains
(indicated by an arrow), with one instance of the SC Y1 highlighted
(orange). All hydrogen atoms, except for those involved in hydrogen
bonding, are omitted for clarity.
bonded chains which are analogous to the CIJ4) chain formed
by the A-type molecules of II.
Crystal packing relationships
The results discussed below were obtained from pairwise crystal
structure comparisons for the group comprising the IPH forms
I–VI and the crystal structures of 1-IJ2-aminobenzoylhydrazono)-
1-phenylethane41 (AHE) and N′-ij(E)-1-phenylethylidene]-
benzohydrazide42 (PEH), which are two close analogues of IPH.
All IPH polymorphs except I contain an N–H⋯OC-bonded
chain of the CIJ4) type (Fig. 8a) possessing glide symmetry. This
chain can occur in three distinct geometries (denoted X, Y and
Z) (Fig. 8b), which differ from one another in the relative orien-
tation of their molecules with respect to the glide plane (Fig. 8c).
The chain type X is present in III, V, AHE and PEH. Additionally,
the stacking mode of the X chains along the short molecular
dimension and normal to the chain translation vector is the
same in these four crystal structures so that they have two-
dimensional supramolecular construct (SC),35 denoted X1, in
common (Fig. 9). The two-dimensional packing similarity in this
subset is associated with three matching unit cell parameters (two
axis at a 90° angle) (Table S2, ESI†). Geometric differences
between the X1 units of V and PEH are very low. Larger but still
remarkably low (x18 ≤ 5.0) dissimilarity indices are obtained for
comparisons which involve AHE, in particular if the presence of
Conclusions
Polymorphism, although a well-known phenomenon, still requires
a significant consideration during pharmaceutical production
This journal is © The Royal Society of Chemistry 2015
CrystEngComm

View Article Online
Paper
CrystEngComm
owing to its unexpected or occasional appearance. IPH repre-
sents an outstanding example of polymorphism with six poly-
morphs having been identified, characterised and described
in the present study including their crystal structures. This
was possible as a result of the high kinetic stability of the
metastable forms (IPH I, II, III, IV, VI). Hot stage microscopy
proved again to be a powerful and versatile tool in the analy-
sis of meltable polymorphic substances. The functionality of
HSM includes melting point analysis, the qualitative descrip-
tion of thermal phase transitions as well as the controlled
production of single crystals for single crystal structure analy-
sis, as demonstrated in the present study and elsewhere.43,44
There is a competition between a dominating H-bonded
catemer motif (present in forms II–VI) and two distinct dimer
motifs (forms I and [catemer + dimer]: II) in the IPH system.
Further analysis of this compound may yield even more iso-
latable polymorphs, and it could become contender for the
polymorphic crystal structure record.
10 A. Y. Lee, D. Erdemir and A. S. Myerson, Annu. Rev. Chem.
Biomol. Eng., 2011, 2, 259–280.
11 R. J. Davey, N. Blagden, G. D. Potts and R. Docherty, J. Am.
Chem. Soc., 1997, 119, 1767–1772.
12 A. Nangia, Acc. Chem. Res., 2008, 41, 595–604.
13 C. Rustichelli, G. Gamberini, V. Ferioli, M. C. Gamberini, R.
Ficarra and S. Tommasini, J. Pharm. Biomed. Anal., 2000, 23,
41–54.
14 A. Grunenberg, J. O. Henck and H. W. Siesler, Int. J. Pharm.,
1996, 129, 147–158.
15 B. Rodriguez-Spong, C. P. Price, A. Jayasankar, A. J. Matzger
and N. Rodriguez-Hornedo, Adv. Drug Delivery Rev., 2004, 56,
241–274.
16 H. Meyer and J. Mally, Monatsh. Chem., 1912, 33, 393–414.
17 H. H. Fox and J. T. Gibas, J. Org. Chem., 1953, 18, 983–989.
18 M. Malhotra, S. Sharma and A. Deep, Med. Chem. Res.,
2012, 21, 1237–1244.
19 C. G. Wermuth, C. R. Ganellin, P. Lindberg and L. A.
Mitscher, Pure Appl. Chem., 1998, 70, 1129.
20 M. Malhotra, G. Sharma and A. Deep, Acta Pol. Pharm.,
2012, 69, 637–644.
Acknowledgements
21 M. Malhotra, V. Monga, S. Sharma, J. Jain, A. Samad, J.
Stables and A. Deep, Med. Chem. Res., 2012, 21, 2145–2152.
22 T. Steiner, Angew. Chem., Int. Ed., 2002, 41, 48–76.
23 J.-H. Jiang, J. Chen, J. Yang and F.-F. Jian, Acta Crystallogr.,
Sect. E: Struct. Rep. Online, 2009, 65, o3125.
24 N. Zencirci, T. Gelbrich, V. Kahlenberg and U. J. Griesser,
Cryst. Growth Des., 2009, 9, 3444–3456.
25 N. Zencirci, T. Gelbrich, D. C. Apperley, R. K. Harris, V.
Kahlenberg and U. J. Griesser, Cryst. Growth Des., 2009, 10,
302–313.
We would like to thank the Molecular Sciences Institute for
infrastructure support and the National Research Foundation
(grant numbers: SFH1207163079 and 85964) for funding
throughout the duration of this research. The National
Research Foundation National Equipment Programme (UID:
78572) is thanked for financing the purchase of the single-
crystal diffractometer. AL thanks the University of the
Witwatersrand Friedel Sellschop Grant. Additional acknowl-
edgement is given to Prof M. A. Fernandes for crystallographic
assistance; Prof D. G. Billing for powder patterns; and Mr. M.
Smith for RAMAN spectra collection.
26 S. Chen, I. A. Guzei and L. Yu, J. Am. Chem. Soc., 2005, 127,
9881–9885.
27 V. López-Mejías, J. W. Kampf and A. J. Matzger, J. Am. Chem.
Soc., 2012, 134, 9872–9875.
28 O. Reany, M. Kapon, M. Botoshansky and E. Keinan, Cryst.
Growth Des., 2009, 9, 3661–3670.
References
1 W. Cabri, P. Ghetti, G. Pozzi and M. Alpegiani, Org. Process
Res. Dev., 2006, 11, 64–72.
29 Bruker, SAINT+, version 6.02 (including XPREP), Bruker AXS
Inc., Madison, WI, USA, 1999.
2 A. J. Aguiar, J. Krc, A. W. Kinkel and J. C. Samyn, J. Pharm.
Sci., 1967, 56, 847–853.
3 A. J. Aguiar and J. E. Zelmer, J. Pharm. Sci., 1969, 58,
983–987.
4 T. Ueto, N. Takata, N. Muroyama, A. Nedu, A. Sasaki, S.
Tanida and K. Terada, Cryst. Growth Des., 2011, 12, 485–494.
5 R. J. Craven and R. W. Lencki, Chem. Rev., 2013, 113,
7402–7420.
30 L. Farrugia, J. Appl. Crystallogr., 1999, 32, 837–838.
31 G. M. Sheldrick, SHELX, release 97-2 (includes SHELXS and
SHELXL), University of Göttingen, 1997.
32 L. Farrugia, J. Appl. Crystallogr., 1997, 30, 565.
33 A. Spek, J. Appl. Crystallogr., 2003, 36, 7–13.
34 H. Putz and K. Brandenburg, Diamond Version. 3.0e - Crystal
and Molecular Structure Visualization - Crystal Impact GbR,
Kreuzherrenstr. 102, 53227 Bonn, Germany.
6 H. G. Brittain, J. Pharm. Sci., 2002, 91, 1573–1580.
7 S. L. Morissette, Ö. Almarsson, M. L. Peterson, J. F.
Remenar, M. J. Read, A. V. Lemmo, S. Ellis, M. J. Cima
and C. R. Gardner, Adv. Drug Delivery Rev., 2004, 56,
275–300.
35 T. Gelbrich and M. B. Hursthouse, CrystEngComm, 2005, 7,
324–336.
36 T. Gelbrich, T. L. Threlfall and M. B. Hursthouse,
CrystEngComm, 2012, 14, 5454–5464.
37 D. E. Braun, T. Gelbrich, V. Kahlenberg, R. Tessadri, J.
Wieser and U. J. Griesser, J. Pharm. Sci., 2009, 98,
2010–2026.
8 J. D. Dunitz and J. Bernstein, Acc. Chem. Res., 1995, 28,
193–200.
9 A. J. Cruz-Cabeza and J. Bernstein, Chem. Rev., 2013, 114,
2170–2191.
38 A. Burger and R. Ramberger, Microchim. Acta, 1979, 72,
259–271.
CrystEngComm
This journal is © The Royal Society of Chemistry 2015

View Article Online
CrystEngComm
Paper
39 M. C. Etter, Acc. Chem. Res., 1990, 23, 120.
40 J. Bernstein, R. E. Davis, L. Shimoni and N.-L. Chang, Angew.
Chem., Int. Ed. Engl., 1995, 34, 1555–1573.
41 M. F. Simeonov, F. Fülöp, R. Sillanpää and K. Pihlaja, J. Org.
Chem., 1997, 62, 5089–5095.
43 A. Lemmerer, J. Bernstein, U. J. Griesser, V. Kahlenberg,
D. M. Többens, S. H. Lapidus, P. W. Stephens and C.
Esterhuysen, Chem. – Eur. J., 2011, 17, 13445–13460.
44 T. Gelbrich, D. E. Braun, A. Ellern and U. J. Griesser, Cryst.
Growth Des., 2013, 13, 1206–1217.
42 H. K. Fun and S. R. Jebas, Acta Crystallogr., Sect. E: Struct.
Rep. Online, 2008, 64, o1255.
45 T. Gelbrich and M. B. Hursthouse, CrystEngComm, 2006, 8,
448–460.
This journal is © The Royal Society of Chemistry 2015
CrystEngComm