Seed-triggered solid-to-solid transformation between color polymorphs: striking differences between quasi-isomorphous crystals of dichloro-substituted salicylideneaniline regioisomers

Article abstract of DOI:10.1039/d0ce00679c

Two regioisomers, N-(4′-chlorosalicylidene)-4-chloroaniline (4CC) and N-(5′-chlorosalicylidene)-4-chloroaniline (5CC), were prepared and were found to exhibit color polymorphism with yellow and orange forms. Microscopy observations revealed that only 4CC exhibited a solid-to-solid transition from the yellow to orange form during heating, and that contact with orange seed crystals could trigger the transformation. The transformation was tracked by thermal analysis and variable-temperature infrared spectroscopy. When polymorphs of the same color were compared, the 4CC and 5CC crystal structures were similar, whereas the molecular arrangement of 4CC crystals was significantly disordered with respect to the molecular orientation. Assuming that the disorder results from stacking faults, crystal models were constructed free of disorder and their molecular environment was analyzed in detail. Finally, it was proposed that certain types of stacking faults could have promoted the solid-to-solid transition. This journal is

Full text of DOI:10.1039/d0ce00679c

CrystEngComm
View Article Online
View Journal
PAPER
Seed-triggered solid-to-solid transformation
between color polymorphs: striking differences
between quasi-isomorphous crystals of dichloro-
substituted salicylideneaniline regioisomers†
Cite this: DOI: 10.1039/d0ce00679c
c
Zaixiang Zhang,ab Masahiro Suzuki,c Yu Yang,c Isao Yoshikawa,
ab
c
*
*
Qiuxiang Yin
and Hirohiko Houjou
Two regioisomers, N-(4′-chlorosalicylidene)-4-chloroaniline (4CC) and N-(5′-chlorosalicylidene)-4-
chloroaniline (5CC), were prepared and were found to exhibit color polymorphism with yellow and orange
forms. Microscopy observations revealed that only 4CC exhibited a solid-to-solid transition from the yellow
to orange form during heating, and that contact with orange seed crystals could trigger the transformation.
The transformation was tracked by thermal analysis and variable-temperature infrared spectroscopy. When
polymorphs of the same color were compared, the 4CC and 5CC crystal structures were similar, whereas
the molecular arrangement of 4CC crystals was significantly disordered with respect to the molecular
orientation. Assuming that the disorder results from stacking faults, crystal models were constructed free
of disorder and their molecular environment was analyzed in detail. Finally, it was proposed that certain
types of stacking faults could have promoted the solid-to-solid transition.
Received 7th May 2020,
Accepted 2nd July 2020
DOI: 10.1039/d0ce00679c
rsc.li/crystengcomm
thermodynamic properties and the links between crystal
structures and the kinetic factors controlling polymorphic
Introduction
Material properties are governed by both molecular structure
and the packing of molecular species in the solid state. Thus,
the design, synthesis, and modification of novel materials
with specialized physicochemical properties is a significant
pursuit for crystal engineering in many fields.1–3 Crystal
polymorphs possess identical chemical compositions but exist
in different crystalline forms, due to different molecular
arrangements and/or conformations. Polymorphs display
distinct physical and chemical properties, including energetic
properties and bioavailability.4–7 Polymorphism has been an
important consideration in many fields, including
pharmaceuticals,8 foods,9 dyes/pigments,10 and organic
electronics,11 for decades. Knowledge of the polymorphic
landscape and the mechanism of phase transformation
systems.12,13 The mechanism of transformation has drawn
particular interest, and two main mechanisms have been
proposed.14 The nucleation and growth mechanism describes
the collapse of crystal packing structure followed by the
growth of a new crystal from the amorphous or melted
components. In this mechanism, polymorph transformation
is discontinuous, with a clear and immediate transition.15,16
In the concerted, or martensitic, mechanism, crystals undergo
a rapid transition without diffusion of molecules or atoms.17
Recent approaches to crystal engineering have utilized
halogen–halogen and/or CH–halogen interactions to control
molecular arrangement in crystalline states.18–23 Owing to
their moderate strengths and bimodal (type I and type II)
directionality, halogen interactions can help generate various
polymorphs and trigger transformations among them by
stimuli such as heat and stress. In addition, moderate
compatibility among halogen atoms often affords ideal solid
solutions.24,25 Several studies26–28 of halogen-bearing
N-salicylideneanilines have provided insight on the effects of
halogen interactions on their thermochromic and
photochromic properties, i.e., reversible color-changing
abilities triggered by temperature or photo-irradiation.29,30
The mechanism governing the relationship between color
change and crystal structure has been investigated.31–33 In
the present work, we evaluated N-(4′-chlorosalicylidene)-4-
provides
a
fundamental
understanding
of
these
a School of Chemical Engineering and Technology, State Key Laboratory of
Chemical Engineering, Tianjin University, Tianjin 300072, People's Republic of
China. E-mail: qxyin@tju.edu.cn
b The Co-Innovation Center of Chemistry and Chemical Engineering of Tianjin,
Tianjin University, Tianjin 300072, People's Republic of China
c Institute of industrial Science, The University of Tokyo, 4-6-1 Komaba Meguro-ku,
Tokyo 153-8505, Japan. E-mail: houjou@iis.u-tokyo.ac.jp
† Electronic supplementary information (ESI) available. CCDC 2001051–2001053.
For ESI and crystallographic data in CIF or other electronic format see DOI:
10.1039/d0ce00679c
This journal is © The Royal Society of Chemistry 2020
CrystEngComm

View Article Online
Paper
CrystEngComm
chloroaniline and N-(5′-chlorosalicylidene)-4-chloroaniline
(denoted by 4CC and 5CC, respectively) (Chart 1), which
differ only by substituents at the 4′- and 5′-positions of the
salicylidene group. We expected that this subtle difference
might influence their solid-state properties. While both
compounds displayed yellow and orange polymorphs, only
4CC exhibited solution-mediated and seed-triggered
transformation from the yellow to the orange form. We
compared 4CC and 5CC to elucidate the factors affecting
their crystal structures and phase-transition behaviors.
Results and discussion
Description of polymorphism
Two target compounds, 4CC and 5CC, were prepared from
the corresponding aldehydes and amine, as described in the
Experimental section. Crystals of 4CC displayed different
colors and habits depending on the solvent used for
recrystallization (Table S1†). For example, acetonitrile
produced orange prisms, while pyridine gave yellow plates.
Acetone afforded both types of crystals, and their population
appeared to change over the course of recrystallization
(Fig. 1). This phenomenon suggests that the yellow crystal is
a
kinetic form and that acetone promotes a solution-
mediated phase transition to the orange form.34,35 The as-
prepared crystals showed no changes within 2 days when
they were exposed to dichloromethane or acetone vapor.
Thus, we conclude that the phase transition in solution
involves dissolution and recrystallization processes. However,
the 4CC-y sample crashed in a mortar turned to 4CC-o when
exposed to acetone in 1 hour, implying that the surface
roughness promotes the transition. Similarly, 5CC crystals
were mostly orange when precipitated from acetone,
acetonitrile, and dichloromethane, while mostly yellow when
deposited from methanol and ethanol. In contrast, the
populations of yellow and orange 5CC crystals did not
change, even after 1 day. The like-colored (orange or yellow)
crystals of 4CC obtained from various solvents exhibited the
same thermal behavior (vide infra) and so did those of 5CC,
indicating their crystallographic identities. Hereafter, we call
the yellow polymorphs 4CC-y and 5CC-y, and the orange
polymorphs 4CC-o and 5CC-o.
Fig.
1
Crystals of 4CC obtained from various solvents and
recrystallization times: (a) acetonitrile (24 h), (b) acetone (1 h), (c)
acetone (24 h), (d) pyridine (24 h).
(Fig. 2).36,37 The spectra of 4CC-y exhibit a broad absorption
band from 400 to 500 nm; absorbance at approximately 440 nm
increases gradually with temperature, thus demonstrating
thermochromism. In contrast, the spectra of 4CC-o display a
distinct valley at approximately 420 nm, which highlights the
structured peak at approximately 460 nm. These features are
also observed for 5CC-y and 5CC-o, but the peak and valley were
redshifted by 20–30 nm compared to the corresponding
polymorphs of 4CC, respectively. Based on previously reported
analogs of salicylideneaniline, the observed thermochromism of
4CC and 5CC could be explained by a shift in tautomeric
equilibrium between the enol and keto forms.29 However, the
observed spectra clearly indicate that a thermochromic response
will not produce the yellow-to-orange crystal transition. Thus,
these microscopy observations provide a useful method to
distinguish color polymorphs of 4CC and 5CC over a wide
temperature range, from ambient to nearly-melting conditions.
Microscopic transmission absorption spectra of the single
crystals revealed that their color variations did not originate
from differences in habit, thickness, or surface conditions, but
from the intrinsic nature of their constituent molecules
Phase transition studies
The thermal behavior of the crystals was examined under a
microscope using a temperature-controlled stage. The 4CC-y
Chart
1 N-(4′-Chlorosalicylidene)-4-chloroaniline (4CC) and N-(5′-
chlorosalicylidene)-4-chloroaniline (5CC).
CrystEngComm
This journal is © The Royal Society of Chemistry 2020

View Article Online
CrystEngComm
Paper
Fig. 2 Solid-state microscopic absorption spectra of (a) 4CC-y, (b) 4CC-o, (c) 5CC-y, and (d) 5CC-o.
and 4CC-o crystals normally melted at 144 °C and 154 °C,
respectively. However, the 4CC-y sample converted to 4CC-o
crystals, caused in some cases by a solid-to-solid transition at
approximately 100–140 °C, and in other cases by
recrystallization from the melt at approximately 140–150 °C.
Assuming that these phenomena were caused by
contamination with trace amounts of 4CC-o crystals in the
4CC-y sample, we placed a tiny 4CC-o crystal onto a 4CC-y
sample. We observed an yellow-to-orange transition
spreading over the entire 4CC-y crystal, confirming that the
solid-to-solid transition was triggered by the orange crystal
(Fig. 3).38,39 This suggests that 4CC-y undergoes a seed-
triggered phase transition to 4CC-o.40–42 When melted, the
4CC-o crystals consistently crystallized in the yellow phase
(4CC-y) during cooling. Thus-obtained yellow crystals,
presumably a pure phase sample, never converted to the
orange form. In contrast to 4CC, the thermal behavior of 5CC
crystals was relatively simple. The 5CC-y sample started to
melt and then vaporize (or sublime) at approximately 140–
148 °C, with orange crystals developing gradually. We
suspected that this behavior resulted from contamination
with trace amounts of 5CC-o. The 5CC-y crystals completely
disappeared at 149 °C, and the 5CC-o crystals melted
immediately thereafter (<150 °C). The melt recrystallized as
the yellow form during cooling. Unlike 4CC, no solid-to-solid
transition was observed, even with the contact mixture
containing both 5CC-y and 5CC-o crystals.
The transition behaviors described above were reproduced
using differential scanning calorimetry (DSC) profiles (Fig.
S1, Table S2†). Starting from an as-prepared 4CC-y sample,
the first heat profile exhibited an endothermic anomaly
(128.3 J g−1) at 154.0 °C (onset), attributed to melting of the
4CC-o phase. No obvious anomalies were found
corresponding to the solid-to-solid transition; rather, jagged
dents at the baseline in the 100–140 °C region were observed.
The first cooling profile exhibited an exothermic peak (−110.2
J
g−1) at 123.7 °C and the second heating profile
demonstrated only an endothermic anomaly (115.1 J g−1) at
144.5 °C (onset), attributed to melting of 4CC-y, indicating
that the melt had recrystallized into the 4CC-y phase during
cooling (however, persistent repetition of this cycle revealed
that 4CC-y and 4CC-o melt seemingly at random, inconsistent
with observations from microscopy). This contradiction may
result from the different sample holders for DSC and
microscopy: during DSC, vaporized sample deposited on the
wall of the aluminum pan to produce 4CC-o seed crystals,
which then triggered the transition to 4CC-o. This
assumption is supported by a comparison that a higher
heating rate (10 °C min−1) reduced the frequency of the
occurrence of 4CC-o's melt. When starting from 4CC-o,
This journal is © The Royal Society of Chemistry 2020
CrystEngComm

View Article Online
Paper
CrystEngComm
Fig. 3 Images of 4CC crystals during heating at approximately 145 °C. The frame count is indicated at the top left of each image (33.37 ms per
frame). A movie version is available on ESI.†
profiles similar to those of 4CC-y were observed. In contrast,
the DSC profiles of 5CC-y exhibited a simple repetition of an
endothermic anomaly (105–106 J g−1) from 147.4 (onset) to
150 °C (peak), attributed to unresolved melting points of
5CC-y and 5CC-o, and exothermic peaks at 132 and 115 °C,
attributed to recrystallization into the 5CC-y and 5CC-o
phases, respectively. When the cycle started with 5CC-o, the
first endothermic anomaly (127–129 J g−1) appeared from
148.0 (onset) to 150 °C (peak), attributed to melting of 5CC-o;
after the second cycle, the onset shifted slightly to 147.0 °C.
FT-IR spectra of the yellow and orange polymorphs of 4CC
and 5CC were clearly distinctive; the changes observed with
thermal stimulus were consistent with observations from
microscopy and DSC (Fig. S2 and S3†). The gradual
conversion from 4CC-y to 4CC-o was also tracked by variable
temperature FT-IR measurements (Fig. S4†).
It should be noted that the structures of 4CC-y and 4CC-o
were solved assuming a static disorder ratio of 1 : 1 for
opposite molecular orientations, which resulted in a pseudo
C2-axis and a pseudo inversion point at the center of the
CN bond.44 As a result, Z′ values for those crystals are 0.5.
Like-colored 4CC and 5CC crystal cells share several
common features in their molecular arrangements. In the
yellow crystals, 4CC-y and 5CC-y (Fig. 4), the molecules have
a twisted conformation (angles between the normal vectors
of two benzene rings are 46.1° and 44.6°, respectively) and
form a layered structure via stacking along the b-axis. Each
layer involves two C–H⋯Cl contacts18 with H–Cl distances of
3.06 Å (for both contacts in 4CC-y) or 3.00 and 3.11 Å (5CC-y).
In 4CC-y, the arrangement is a mixture of head (salicylidene
ring)-to-tail (aniline ring) and head-to-head (tail-to-tail), due
Table 1 Crystallographic data
Crystal structure and modeling
4CC-y
4CC-o
5CC-y43
5CC-o
Of the crystals studied, only the crystal structure for 5CC-y
has been published (CCDC 136648).43 The crystal structures
of 4CC-y, 4CC-o, and 5CC-o, were elucidated in this study.
The phase purity of all the polymorphs were verified and the
phase transition from 4CC-y to 4CC-o was successfully
tracked by powder XRD analysis (Fig. S5–S7†). Even after
4CC-y was milled in a mortar, neither the color nor the
powder XRD pattern changed. From the crystallographic data
summarized in Table 1, it is clear that the cell dimensions
are similar between 4CC and 5CC crystals of the same color.
Space group
I2/c
P21/c
P21/c
P21/c
a/Å
b/Å
c/Å
β/°
V/Å3
Dcalcd
Z
Z′
T/K
R
6.140(5)
6.848(6)
27.39(2)
95.26(2)
1146.8(16)
1.541
4
0.5
93
0.0447
0.1002
10.004(6)
4.688(3)
11.919(7)
93.2393IJ10)
558.1(6)
1.584
2
0.5
93
0.0504
0.1208
27.419(7)
6.901(2)
6.137(1)
95.57(2)
1155.75
1.529
4
1
190
0.0340
0.0905
20.427(6)
4.5728(13)
12.101(4)
92.301(5)
1129.4(6)
1.565
4
1
93
0.0333
0.0830
Rw
CrystEngComm
This journal is © The Royal Society of Chemistry 2020

View Article Online
CrystEngComm
Paper
Fig. 4 Molecular arrangement in a (a) 4CC-y crystal: viewed along the b-axis (top) and a-axis (bottom); and in a (b) 5CC-y crystal: viewed along
the b-axis (top) and c-axis (bottom). Symbols denoting symmetry elements are overlaid.
This journal is © The Royal Society of Chemistry 2020
CrystEngComm

View Article Online
Paper
CrystEngComm
to disorder with respect to molecular orientation (Fig. 5).
However, molecular packing in the 5CC-y crystal suggests
that the head-to-head (tail-to-tail) arrangement is favorable.
Within each layer, type-I Cl⋯Cl contacts are observed, with
interatomic distances of 3.38 Å (4CC-y) or 3.39 Å (5CC). The
inter-layer distance is equal to half that of the b-axis: 3.42 Å
for 4CC-y and 3.45 Å for 5CC-y. Phenyl rings from adjacent
layers are tilted 46.1° from each other, suggesting T-shaped
CH/π interactions.
In the orange crystals, 4CC-o and 5CC-o (Fig. 6), the
molecules exhibit planar conformations (angles between the
normal vectors of two benzene rings are 0.00° and 2.85°,
respectively) and the crystal comprises two ribbon structures
forming an orthogonal cross pattern. This conformational
difference indicates packing effects.45 Each ribbon has two
C–H⋯Cl contacts, for which the distance between H and Cl
atoms is 2.97 Å (for both contacts in 4CC-o) or 2.96 and 3.00
Interestingly, polymorphs of the same color share similar
crystal structures, suggesting that their photophysical
properties are controlled by molecular conformations and
arrangements more effectively than by molecular structure.
Hirshfeld surface analysis and subsequent crystal fingerprint
plots would help to quantify similarities in the molecular
arrangements in the crystals,46,47 but these methods do not
provide adequate analysis for crystal data involving structural
disorders. Therefore, we attempted to remove disorder in the
4CC crystals by modeling ordered structures referenced to the
5CC crystals.
First, we presumed that molecules should be placed such
that the double C–H⋯Cl contacts adopted a head-to-head (tail-
to-tail) arrangement in 4CC-y. This arrangement was achieved
by removing the pseudo-C2-axis from the center of the CN
bond and by choosing the P21/a space group (Fig. S8†). Analysis
of packing similarity using Mercury software produced an RMS
score of 0.341 (15 out of 15 molecules in common). The
fingerprint plot of this model was quite similar to that of 5CC-y
(Fig. 8), especially for the CH/π wings around (di, de) = {(1.8,
1.2), (1.2, 1.8)}. It is also possible that the layer is inverted in
the middle of the b-axis. Such an arrangement can be
Å
(5CC-o), similar to those observed for the yellow
polymorphs. As with 4CC-y, due to disorder, the contacts are
arranged as a mixture of head-to-tail and head-to-head (tail-
to-tail). However, molecular packing in the 5CC-o crystal
suggests that the head-to-head (tail-to-tail) arrangement is
favorable. In both 4CC-o and 5CC-o, the ribbons form a
slipped stack along the b-axis, resulting in inter-ribbon
distances of 3.42 Å and 3.41 Å, respectively, within the range
of normal π-stacking distance. Among the crossing ribbons,
type-II Cl⋯Cl contacts are observed with interatomic
distances of 3.44 Å (4CC-o) or 3.46 and 3.51 Å (5CC-o).
Double C–H⋯Cl and Cl⋯Cl contacts play an important role
in forming the packing structure in each of the crystals
studied.18–23 The relevant geometrical parameters are
summarized in Table S3.† Similar interactions have been
observed for γ-phase 1,4-dichlorobenzene and many other
halogenated benzenes. Because 4CC and 5CC are regioisomers
in which H and Cl atoms are permutated at the 4- and
5-positions, they are likely to have similar crystal structures.
¯
reproduced by selecting the I1 space group (Fig. S9†). Because
the inter-layer interactions are relatively weak, this change in
arrangement has a minimal effect on the lattice energy. Indeed,
¯
the crystal fingerprint plot of the I1 model was hardly distinct
from that of the P21/a model (Fig. 7).
For 4CC-o, we again presumed that the molecules should be
placed in the head-to-head arrangement. This arrangement
was achieved by removing the pseudo inversion point and by
doubling the a-axis with keeping the space group P21/c (Fig.
S10†). Analysis of packing similarity gave an RMS score of
0.444 (15 out of 15 molecules in common). The fingerprint plot
of this model was very similar to that of 5CC-o (Fig. 7). Similar
to 4CC-y, a stacking fault along the b-axis could account for the
disorder observed. We also verified another model comprising
an elongated (two-fold) b-axis with two independent units
arranged in opposing directions (Fig. S11†). This model has a
unit cell four times larger than the original and belongs to the
Cc space group. Again, the crystal fingerprint plot of the Cc
model was comparable that of the P21/c model. Fig. S12 and
S13† compare the partial fingerprints contributed by
ClIJinterior)⋯any atomIJexterior) contacts.
Based on Hirshfeld surface analysis, we calculated area ratios
for interatomic contacts in each modeled crystal structure. For
example, the ratio of the H–C pair is the sum of the
HIJinterior)⋯C(exterior) contact and the C(interior)⋯HIJexterior)
contact. The results, shown in Fig. 8, reveal structural
similarities in like-colored forms. These results provide
quantitative evidence for their structural similarities. They also
imply similar energies between the “ordered” models for 4CC-y
and 4CC-o and confirm that disorder was caused by stacking
fault along the b-axis. For 4CC-y and 5CC-y, H–H, H–C, and C–C
interactions total approximately 50% of all contacts, but their
compositions are markedly different, as a result of different
stacking modes for the aromatic rings. It is notable that there is
Fig. 5 Schematic representation of double CH⋯Cl contacts (shown
by dashed lines): (a) head-to-head arrangement of 4CC; (b) head-to-
tail arrangement of 4CC; (c) head-to-head arrangement of 5CC.
CrystEngComm
This journal is © The Royal Society of Chemistry 2020

View Article Online
CrystEngComm
Paper
Fig. 6 Molecular arrangement in a (a) 4CC-o crystal: viewed along the b-axis (top) and c-axis (bottom); and in a (b) 5CC-o crystal: viewed along
the b-axis (top) and c-axis (bottom). Symbols denoting symmetry elements are overlaid.
This journal is © The Royal Society of Chemistry 2020
CrystEngComm

View Article Online
Paper
CrystEngComm
Fig. 7 Crystal fingerprint plots for (a) 4CC-y (P21/a model), (b) 4CC-y (I1¯ model), (c) 5CC-y, (d) 4CC-o (P21/c model), (e) 4CC-o (Cc model), and (f)
5CC-o.
CrystEngComm
This journal is © The Royal Society of Chemistry 2020

View Article Online
CrystEngComm
Paper
point, hence the polymorph pair is monotropic (Fig. S14†). In
contrast, 5CC-y and 5CC-o have substantially the same
melting point (147.6–148.7 °C) and similar enthalpies of
fusion (33–34 kJ mol−1), suggesting that they possess
comparable thermodynamic stabilities. It is not clear why the
yellow and orange forms have significantly different
differences in energy (3 kJ mol−1 for 4CC and <1 kJ mol−1 for
5CC), since they exhibit similar intermolecular interactions.
Theoretical validation of conformational and lattice energies
may be possible via various quantum chemical methods, but
these are outside the scope of this study. The proposed
stability ranking accounts for the solution-mediated 4CC-y to
4CC-o transformation observed in certain solvents (no such
transition was observed with 5CC-y). The transition suggests
that the yellow forms are favored kinetically in both the
solution and melt crystallization processes. This assumption
also factors in faster crystallization for yellow forms
compared to orange forms upon cooling the melt.
Fig. 8 Comparison of the area ratio of atom–atom contacts based on
Hirshfeld analysis.
a stark contrast in the ratio of C–Cl and Cl–Cl contacts between
the yellow and orange crystal structures.
The discussion of energy order is not sufficient to explain
why seed-triggered phase transitions are observed only for
4CC. Here, insight into activation energy is required. We
propose that there is some mechanism by which activation
energy is lowered only for the 4CC-y to 4CC-o path. Crystal
structure data suggests that the conversion from yellow to
orange forms would require a drastic change in molecular
Crystal structures and subsequent numerical analyses
reveal that like-colored crystals (4CC-y/5CC-y and 4CC-o/5CC-
o) exhibit similar molecular environments, though
differences in molecular structure influences their
arrangements. The differences in molecular arrangement
between 4CC and 5CC can be illustrated by the positions of
the H and Cl atoms, which approximate parallelogram and
trapezoid geometries, respectively (Fig. 9a and b,
respectively). The pseudo symmetry of 4CC gives rise to
pseudo translational symmetry, hence each layer can stack
on another, while slipping longitudinally (gray line in
Fig. 9c). This type of stacking fault is consistent with the
disorder we observed. This assumption is supported by
similar crystal fingerprints for the two 4CC-y models and the
two 4CC-o models (Fig. 8), since these models differ only
with respect to translation of the layer in the middle of the
b-axis. In contrast, this type of stacking fault is unlikely to
occur with the 5CC molecule (Fig. 9d).
arrangement, not just
a simple sliding or small-angle
rotation. Hence, molecular packing constraints would
provide a high activation barrier, especially for a perfect
crystal lattice. A notable difference in the 4CC and 5CC
crystal structures is that the former allows for significant
disorder with respect to orientation, which can be explained
by a stacking fault along the b-axis, likely caused by its
pseudo-symmetry. In light of the disorder in molecular
orientation observed for both 4CC-y and 4CC-o, this putative
stacking fault may play a role in lowering the activation
energy for the transition from the yellow to the orange form.
DSC results indicated that 4CC-o has a higher melting
point (154 °C) and larger enthalpies of fusion (34 kJ mol−1)
than those of 4CC-y (144 °C and 31 kJ mol−1), suggesting that
4CC-o is more stable at all temperatures below the melting
Conclusions
In summary, we examined two regioisomers of dichloro-
substituted salicylideneaniline, 4CC and 5CC, and found that
both compounds exhibited polymorphism with yellow and
orange form. Of the two compounds, only 4CC exhibited a
solid-to-solid transition from yellow to orange forms during
heating. We observed that contact with orange seed crystals
could trigger the transformation. When polymorphs of the
same color were compared, the crystal structures of 4CC and
5CC were quite similar to each other, with the
supramolecular synthons involving the same halogen–
halogen and CH–halogen contacts. The molecular
arrangements of both the 4CC-y and 4CC-o crystals were
significantly disordered with respect to molecular
orientation, probably due to pseudo-symmetry. We
considered crystal models that were free of disorder and
proposed stacking faults that could promote the solid-to-
solid transition. Further study is needed to validate this
Fig. 9 Simplified models of the molecular arrangement in (a) 4CC and
(b) 5CC crystals. The head and tail of each arrow represent the
salicylidene and aniline rings, respectively. The dotted lines indicate
CH⋯Cl contacts. Putative stacking faults in (c) 4CC and (d) 5CC
crystals are depicted in gray lines.
This journal is © The Royal Society of Chemistry 2020
CrystEngComm

View Article Online
Paper
CrystEngComm
theory; in particular,
thermodynamics of the process and a wider variety of
halogen-substituted salicylideneanilines should be
considered. The present study provides insight regarding the
kinetic stability of polymorphs, which will contribute a
diverse range of fields including pharmaceuticals, foods,
dyes/pigments, and organic electronics.
a
theoretical evaluation of the
ArH, 1H), 7.37 (d, ArH, 1H), 7.41 (m, ArH, 2H), δ 8.54
(sH, –NCH–, 1H), 13.02 (br, OH, 1H), 13C NMR (CDCl3)
δ 118.88, 119.73, 122.46, 123.78, 129.63, 131.30, 132.99,
133.19, 146.44, 159.61, 161.58.
Crystallography
For X-ray diffraction of single crystals, data were collected on
a VariMax DW with Saturn CCD, λ (Mo Kα) = 0.71075 Å. The
structures were solved by direct method (SHELXS-2013)48 and
refined on F2 by full-matrix least-square techniques (SHELXL-
2018).49 Crystallographic data have been deposited with
Cambridge Crystallographic Data Centre: Deposition number
CCDC-2001051, 2001052, 2001053 for 4CC-o, 4CC-y, 5CC-o.
Experimental section
Measurements
The thermal behavior of the solid samples were observed
with a conventional optical microscope equipped with a
temperature-controlled stage (Linkam THMS600). The
ultraviolet-visible (UV-vis) absorption spectra of the solid
Definitions
samples were acquired using
a
conventional optical
microscope equipped with an optical fiber connected to a
spectrometer (Ocean Optics USB4000) and the above-
mentioned temperature-controlled stage. The measurement
light was guided from a halogen lamp (100 W). Differential
scanning calorimetry was carried out in Perkin-Elmer Pyris 1
DSC at a heating rate of 10 K min−1 under the protection of
nitrogen. Fourier transform infrared spectroscopy (FT-IR)
spectra were measured by a JASCO FT/IR-420 spectrometer
and the temperature was controlled by ES100P digital
4CC-y Yellow plate recrystallized with acetone
4CC-o Orange plate recrystallized with acetone
5CC-o Orange needle recrystallized with dichloromethane
and hexane
Conflicts of interest
There are no conflicts to declare.
controller. The samples were hold at a temperature for 2 min Acknowledgements
before measurement.
Z. Z. and Q. Y. acknowledge the scholarship of Chinese
Scholarship Committee for the support. A part of this work
(X-ray single crystallographic analysis) was supported by
Synthesis
The Schiff base compounds were prepared by the by
condensation of 4-chlorosalicyl- or 5-chlorosalicylaldehyde (2
mmol) and 4-chloroaniline (2 mmol) in methanol. The
products were characterized by elemental analysis, FAB MS
spectrometry, IR, and 1H and 13C NMR spectroscopy. The
NMR data were acquired for samples in chloroform-d, using
a JEOL ECS-400 spectrometer (400 MHz for H). All chemicals
were purchased from Tokyo Chemical Industry Co., Ltd. and
used without further purification.
Nanotechnology-Platform-Project by the Ministry
of
Education, Culture, Sports, Science and Technology Japan,
Grant Number JPMXP09A19UT0047.
Notes and references
1
1 G. R. Desiraju, J. Am. Chem. Soc., 2013, 135, 9952–9967.
2 D. Braga, F. Grepioni and L. Maini, Chem. Commun.,
2010, 46, 6232–6242.
4CC. To a solution of 4-chlorosalicylaldehyde (0.155 g,
0.990 mmol) dissolved in 6 ml methanol, 4-chloroaniline
(0.127 g, 0.996 mmol) was added. After 24 hours, the reaction
solution were filtrated and yellow crystals were obtained
(0.221 g, 84%).
3 P. Erk, H. Hengelsberg, M. F. Haddow and R. van Gelder,
CrystEngComm, 2004, 6, 474.
4 L. Yu, Acc. Chem. Res., 2010, 43, 1257–1266.
5 J. P. Reddy, D. Swain and V. R. Pedireddi, Cryst. Growth Des.,
2014, 14, 5064–5071.
Analysis for C13H9Cl2NO, calcd. C 58.67, H 3.41, N 5.26,
found C 58.59, H 3.40, N 5.36; FAB-MS m/z 265.0 calcd. For
6 A. Zimmermann, B. Frøstrup and A. D. Bond, Cryst. Growth
Des., 2012, 12, 2961–2968.
7 Q. Zhang and X. Mei, Cryst. Growth Des., 2016, 16,
3535–3542.
8 R. Docherty, K. Pencheva and Y. A. Abramov, J. Pharm.
Pharmacol., 2015, 67, 847–856.
9 P. R. Ramel and A. G. Marangoni, J. Dairy Sci., 2017, 100,
6930–6937.
10 E. F. Paulus, F. J. J. Leusen and M. U. Schmidt,
CrystEngComm, 2007, 9, 131–143.
11 H. Chung and Y. Diao, J. Mater. Chem. C, 2016, 4, 3915–3933.
12 G. S. Nichol and W. Clegg, Acta Crystallogr., Sect. B: Struct.
Sci., 2005, 61, 464–472.
1
[M]+ found 265.1; H NMR (CDCl3) δ 6.93 (dd, ArH, 1H), 7.04
(d, ArH, 1H), 7.22 (m, ArH, 2H), 7.32 (d, ArH, 1H), 7.40 (m,
ArH, 2H), 8.27 (s, –NCH–, 1H), 13.34 (br, OH, 1H), 13C
NMR (CDCl3) δ 117.76, 117.79, 119.93, 122.59, 129.77, 132.95,
133.30, 139.41, 146.69, 161.98, 162.03.
5CC. To a solution of 5-chlorosalicylaldehyde (0.785 g, 5.0
mmol) dissolved in 20 ml methanol, 4-chloroaniline (0.639 g,
5.0 mmol) was added. After 24 hours, the reaction solution
were filtrated and yellow crystals were obtained (1.189 g, 89%).
FAB-MS m/z 265.0 calcd. For [M]+ found 265.2; 1H NMR
(CDCl3) 6.98 (d ArH, 1H), 7.22 (m, ArH, 2H), 7.33 (dd,
CrystEngComm
This journal is © The Royal Society of Chemistry 2020

View Article Online
CrystEngComm
Paper
13 S. Takahashi, H. Miura, H. Kasai, S. Okada, H. Oikawa
and H. Nakanishi, J. Am. Chem. Soc., 2002, 124,
10944–10945.
14 J. Anwar, S. C. Tuble and J. Kendrick, J. Am. Chem. Soc.,
2007, 129, 2542–2547.
32 F. Arod, P. Pattison, K. J. Schenk and G. Chapuis, Cryst.
Growth Des., 2007, 7, 1679–1685.
33 H. Sugiyama and H. Uekusa, CrystEngComm, 2018, 20,
2144–2151.
34 W. Du, Q. Yin, J. Gong, Y. Bao, X. Zhang, X. Sun, S. Ding, C.
Xie, M. Zhang and H. Hao, Cryst. Growth Des., 2014, 14,
4519–4525.
15 G. T. Beckham, B. Peters and B. L. Trout, J. Phys. Chem. B,
2008, 112, 7460–7466.
16 G. T. Beckham, B. Peters, C. Starbuck, N. Varíankaval and
B. L. Trout, J. Am. Chem. Soc., 2007, 129, 4714–4723.
17 M. T. Ruggiero, J. Axel Zeitler and T. M. Korter, Phys. Chem.
Chem. Phys., 2017, 19, 28502–28506.
18 M. Podsiadło, A. Olejniczak and A. Katrusiak,
CrystEngComm, 2014, 16, 8279–8285.
19 A. Mukherjee and G. R. Desiraju, IUCrJ, 2014, 1, 49–60.
20 D. Braga, L. Brammer and N. R. Champness, CrystEngComm,
2005, 7, 1–19.
21 P. Politzer, J. S. Murray and M. C. Concha, J. Mol. Model.,
2007, 13, 643–650.
22 K. E. Riley, C. L. Ford and K. Demouchet, Chem. Phys. Lett.,
2015, 621, 165–170.
23 F. F. Awwadi, R. D. Willett, K. A. Peterson and B. Twamley,
Chem. – Eur. J., 2006, 12, 8952–8960.
24 M. A. Khoj, C. E. Hughes, K. D. M. Harris and B. M. Kariuki,
Cryst. Growth Des., 2017, 17, 1276–1284.
25 A. K. S. Romasanta, D. Braga, M. T. Duarte and F. Grepioni,
CrystEngComm, 2017, 19, 653–660.
35 J. H. ter Horst and P. W. Cains, Cryst. Growth Des., 2008, 8,
2537–2542.
36 H. Houjou, T. Kato, H. Huang, Y. Suzuki, I. Yoshikawa and
T. Mutai, Cryst. Growth Des., 2019, 19, 1384–1390.
37 Y. Suzuki, T. Kato, H. Huang, I. Yoshikawa, T. Mutai and H.
Houjou, J. Photochem. Photobiol., A, 2019, 385, 112096.
38 F. H. Herbstein, Acta Crystallogr., Sect. B: Struct. Sci.,
2006, 62, 341–383.
39 B. P. Krishnan and K. M. Sureshan, J. Am. Chem. Soc.,
2015, 137, 1692–1696.
40 H. Ito, M. Muromoto, S. Kurenuma, S. Ishizaka, N.
Kitamura, H. Sato and T. Seki, Nat. Commun., 2013, 4, 1–5.
41 T. Seki, K. Sakurada and H. Ito, Angew. Chem., Int. Ed.,
2013, 52, 12828–12832.
42 B. A. Zakharov, S. V. Goryainov and E. V. Boldyreva,
CrystEngComm, 2016, 18, 5423–5428.
43 J. Burgess, J. Fawcett, D. R. Russell, S. R. Gilani and V.
Palma, Acta Crystallogr., Sect. C: Cryst. Struct. Commun.,
1999, 55, 1707–1710.
26 D. A. Safin, K. Robeyns and Y. Garcia, RSC Adv., 2012, 2,
11379.
27 D. A. Safin, M. G. Babashkina, K. Robeyns and Y. Garcia,
RSC Adv., 2016, 6, 53669–53678.
44 I. O. Staehle, B. Rodríguez-Molina, S. I. Khan and M. A.
Garcia-Garibay, Cryst. Growth Des., 2014, 14, 3667–3673.
45 A. J. Cruz-Cabeza, J. W. Liebeschuetz and F. H. Allen,
CrystEngComm, 2012, 14, 6797–6811.
28 H. Houjou, H. Ikedo and I. Yoshikawa, Chem. Commun.,
2017, 53, 10898–10901.
46 J. J. Mckinnon, A. S. Mitchell and M. A. Spackman, Chem. –
Eur. J., 1998, 4, 2136–2141.
29 E. Hadjoudis and I. M. Mavridis, Chem. Soc. Rev., 2004, 33,
579–588.
47 M. A. Spackman and J. J. McKinnon, CrystEngComm, 2002, 4,
378–392.
30 T. Fujiwara, J. Harada and O. Keiichiro, J. Phys. Chem. B,
2004, 108, 4035–4038.
48 G. M. Sheldrick, Acta Crystallogr., Sect. A: Found. Crystallogr.,
2008, 64, 112–122.
31 K. Ogawa, Y. Kasahara, Y. Ohtani and J. Harada, J. Am.
Chem. Soc., 1998, 177, 7107–7108.
49 G. M. Sheldrick, Acta Crystallogr., Sect. C: Struct. Chem.,
2015, 71, 3–8.
This journal is © The Royal Society of Chemistry 2020
CrystEngComm