Substituent Electronegativity and Isostructurality in the Polymorphism of Clonixin Analogues

Article abstract of DOI:10.1021/acs.cgd.8b01180

To gain understanding of how the variation in local weak interactions influences intermolecular interactions and subsequent crystal packing, four analogues of 2-(3-chloro-2-methyl-phenylamino)-nicotinic acid [clonixin] were synthesized by replacing the ch

Full text of DOI:10.1021/acs.cgd.8b01180

Article
pubs.acs.org/crystal
Cite This: Cryst. Growth Des. XXXX, XXX, XXX−XXX
Substituent Electronegativity and Isostructurality in the
Polymorphism of Clonixin Analogues
Meng Liu,† Guowen Shen,‡ Zhong Yuan,§ Sean Parkin,∥ Faquan Yu,† Mingtao Zhang,‡ Sihui Long,*,†
and Tonglei Li*,⊥
†Key Laboratory for Green Chemical Process of Ministry of Education, Hubei Key Laboratory of Novel Reactor and Green
Chemical Technology, School of Chemical Engineering and Pharmacy, Wuhan Institute of Technology, Wuhan, Hubei, China
‡Computational Center for Molecular Science, College of Chemistry, Nankai University, Tianjin, China
§Department of Medicinal Chemistry, School of Pharmaceutical Sciences, Central South University, Changsha, Hunan, China
∥Department of Chemistry, University of Kentucky, Lexington, Kentucky 40506, United States
⊥Department of Industrial and Physical Pharmacy, Purdue University, West Lafayette, Indiana, United States
S
* Supporting Information
ABSTRACT: To gain understanding of how the variation in
local weak interactions influences intermolecular interactions
and subsequent crystal packing, four analogues of 2-(3-chloro-
2-methyl-phenylamino)-nicotinic acid [clonixin] were synthe-
sized by replacing the chlorine with fluorine (1), bromine (2),
iodine (3), and hydrogen (4). Compounds 1, 2, and 3 were
found to be polymorphic as shown by the discovery of two
forms for each (1-I, 1-II; 2-I, 2-II; and 3-I, 3-II, respectively),
while compound 4 has only a single identified form. Similar to
clonixin, the analogue molecules in their crystals are
associated either by the acid−acid dimer homosynthon or
the acid−pyridine heterosynthon, depending on the dihedral
angle between the two aromatic rings. Moreover, forms 2-I, 2-
II, 3-I, and 3-II are isostructural to clonixin forms I and IV, compound 4 is isostructural to compound 1 form I, and 2-II is
structurally similar to clonixin form II. The phase behaviors of these polymorphic systems were studied by differential scanning
calorimetry, and it was found that 2-II converted into 2-I when heated, 3-II transformed into 3-I upon heating, and 1-II
underwent transition into 1-I. Quantum mechanical calculations including conformational energy, hydrogen-bonding strength,
lattice energy, and molecular contact were performed to provide further insight into the polymorphism. Given that all halogen
derivatives are polymorphic, while the hydrogen analogue (4) is not, our study highlights the importance of intermolecular
interactions in determining crystal packing.
platelet-inhibitory effects,2,3 and recently it was found to be
1. INTRODUCTION
effective for the treatment of acute migraine.4,5 Like most
[2-(2-Methyl-3-chloroanilino)nicotinic acid [clonixin or CLX]
classic NSAIDs, it is a nonselective cyclooxygenase (COX)
inhibitor and causes side effects such as ulcer.
CLX is a conformationally flexible multifunctional molecule
with two aromatic rings (one nicotinic acid, and the other
(Scheme 1) is a nonsteroidal anti-inflammatory drug (NSAID)
used to treat inflammation in rheumatism, osteoporosis,
collagen diseases, bursitis, and gout.1 It also exhibited
disubstituted benzene) bridged by NH. Because of conforma-
Scheme 1
tional flexibility and participation of hydrogen bonding by
different functional groups, the molecule forms different
polymorphs in the solid state.6,7 Its polymorphic forms and
one solvate were fully characterized by a series of spectroscopic
and spectrometric methods such as powder X-ray diffraction
(PXRD), Fourier transform infrared (FT-IR), Raman, and
solid-state NMR, as well as theoretical approaches.8,9
Received: August 6, 2018
Revised: September 16, 2018
Published: October 1, 2018
© XXXX American Chemical Society
A
DOI: 10.1021/acs.cgd.8b01180
Cryst. Growth Des. XXXX, XXX, XXX−XXX

Crystal Growth & Design
Article
We studied several 2-anilinonicotinic acid compounds that
are structurally similar to CLX. Many of these systems are
highly polymorphic: for example, 2-(phenylamino)nicotinic
acid [2-PNA] and 2-[methyl(phenyl)amino]nicotinic acid (2-
MPNA) are, since each has four forms obtained so far.10,11 We
showed that, by chemically conjugating substituent groups to
the benzene ring or alkylating the H on the bridging N, we
could induce a diarylamine to take on a nonplanar
conformation between the two aromatic rings leading to the
acid−pyridine hydrogen-bonding chain motif12,13 or to
maintain a planar conformation leading to the acid−acid
dimer motif.14 In addition, a sandwich conformation between
the twisted phenyl ring and carboxyl group leads to both acid−
acid and acid−pyridine hydrogen-bonding motifs in the
crystal.15 These studies demonstrated that intermolecular
interactions, both strength and pattern, are mutually influenced
by the molecular conformation. Because these molecules have
both carboxyl and pyridinyl N, they can associate by either the
acid−acid or the acid−pyridine hydrogen bonding. Formation
of a specific motif is mainly determined by the conformation
and the diversity in conformational change, and hydrogen-
bonding motif contributes to the polymorphism of these
molecules. In addition, we studied the polymorphism of
tolfenamic acid, in which the molecule only has the carboxyl
but no pyridinyl N, and we found that the polymorphism is
still determined by a mutual regulation of the hydrogen-
bonding strength and molecular conformation.16
Scheme 2. Synthesis of CLX Analogues
compound was dissolved in a given solvent to form a saturated
solution at room temperature. Then the solution was set for slow
evaporation, until single crystals were harvested with or without
solvent remaining.19 All crystallization experiments were conducted in
the ambient atmosphere. An example is given by the following: 50 mg
of compound 1 was dissolved in 10 mL of high-performance liquid
chromatography (HPLC)-grade methanol in a glass vial at room
temperature. The vial was covered with perforated Parafilm. Crystals
were obtained as yellow plates in approximately one week. Two forms
were produced for each of compounds 1, 2, and 3. Only one crystal
form was obtained for compound 4 in the solvents tested.
2.4. Crystal Structure Determination. Crystal structures of the
synthesized analogues of CLX were determined by single-crystal X-ray
diffraction. Data collection was performed at 90 K on a Nonius
kappaCCD diffractometer with Mo Kα radiation (λ = 0.710 73 Å).20
Cell refinement and data reduction were done using SCALEPACK
and DENZO-SMN. Structure solution and refinement were
performed using the SHELXS and SHELXL programs, respec-
tively.21,22
In this study, we prepared several new analogues of CLX by
replacing Cl with other halogens (F, Br, and I; Scheme 1). We
also compared them with the crystal structure of the hydrogen
analogue (4).12 Because the halogens differ subtly in
electronegativity and atomic size, it would be interesting to
further explore the impact of molecular interactions on crystal
packing. As a diarylamine, CLX attributes its polymorphism to
the interplay between energetically different planar and
nonplanar conformers and relative strengths of the acid−acid
and acid−pyridine hydrogen-bonding motifs. It is expected
that the same tug-of-war should apply in the new systems. In
addition, isostructurality and electronegativity of the halogen/
hydrogen substituents could tilt the balance of forces, resulting
in particular polymorphism of each compound. A variety of
solvents with apolar, protic, and polar aprotic characteristics
was used for the crystal growth of the CLX analogues to obtain
possible polymorphs/solvatomorphs. The solid forms were
fully characterized by single-crystal X-ray diffraction, FT-IR
and Raman spectroscopies, and differential scanning calorim-
etry (DSC). The molecular structure-polymorphism relation-
ship was analyzed both experimentally and theoretically,
providing insight on the impact of each substituent.
PXRD data for each sample were collected on a Rigaku X-ray
diffractometer with Cu Kα radiation (40 kV, 40 mA, λ = 1.5406 Å)
between 5.0−50.0° (2θ) at ambient temperature. The finely ground
sample was placed on a quartz plate in an aluminum holder prior to
measurement.
2.5. Thermal Analyses. Phase behavior of the solid forms was
studied by DSC. The experiments were performed on TA Instruments
DSCQ20-1250. Tzero pans and aluminum hermetic lids were used at
a heating rate of 10 °C/min to measure a few milligrams of sample.
2.6. Spectroscopic Measurement. IR spectra were recorded of a
sample dispersed in KBr pellets using a PerkinElmer FT-IR
spectrometer, while Raman spectra were recorded of a sample
compressed in a gold-coated sample holder using a Thermo Raman
confocal microscope.
2.7. Computation. To shed light on the polymorphism of the
derivatives, we performed a series of theoretical calculations, including
energy difference between planar and nonplanar conformers of each
compound, strength of the acid−acid dimer and the acid−pyridine
hydrogen bonds, and lattice energies of CLX analogues. In addition,
conformational energy scans were conducted for each compound to
identify low-energy conformers. Molecular contacts in each crystal
form were examined by Hirshfeld surface analyses.23,24
The planar and nonplanar conformations were optimized at
B3LYP/6-311g(d,p)25,26 level based on the molecular structures in
crystals. All optimizations were performed without any geometrical
constraints. Frequency calculations were performed for all stationary
points at the same level to identify the minima (zero imaginary
frequency). The hydrogen-bond strength was evaluated of the exact
structures in the crystals at m062x27/def2tzvp28 level with
consideration of dispersion energy. The basis set superposition
error (BSSE) was accounted for by the counterpoise method.29 The
energy scan of the dihedral angle between the aromatic rings was
conducted at B3LYP/6-311g(d,p) level in steps of 10°.
2. EXPERIMENTAL SECTION
2.1. Materials. All chemicals were purchased from commercial
sources. 2-Chloronicotinic acid and p-toluenesulfonic acid were from
Aladdin Industrial Corporation; pyridine and other solvents were
from Sinopharm Chemical Reagent Co., Ltd., and were used as
received; o-toluidine was from Sinopharm Chemical Reagent Co.,
Ltd.; 3-bromo-2-methylaniline and 3-fluoro-2-methylaniline were
from Energy Chemical Co., Ltd.; and 3-iodo-2-methylaniline was
from Shuya Chemical Science and Technology.
2.2. Synthesis. Clonixin analogues were synthesized according to
a reported method (Scheme 2).17,18 Details of synthesis steps can be
found in the Supporting Information.
2.3. Crystal Growth. Crystallization of each compound was
performed in selected organic/aqueous solvents. Typically, a
To calculate lattice energy, a crystal structure was first optimized at
the PW1PW/POB-TZVP28 level with the lattice parameters kept the
same as the experimental values using Crystal14.30 Dispersion energy
contributions were included by the DFT-D3 program of Grimme with
Becke-Johnson damping.31,32 The BSSE was considered. The energy
convergence was set to 1 × 10−7 hartree, and the root-mean-square
(RMS) values of convergence were set to 0.0003 and 0.0012 au for
B
DOI: 10.1021/acs.cgd.8b01180
Cryst. Growth Des. XXXX, XXX, XXX−XXX

Crystal Growth & Design
Article
the energy gradient and atomic displacement, respectively. All
calculations were conducted on a Linux cluster.
I, 1-II, 2-I, 3-I, and 4 to be of monoclinic, space group P21/c
(Z = 4); 2-II and 3-II were triclinic, space group P1 (Z = 2).
̅
Crystallographic data of the forms are listed in Table 2; for
complete CIF files, see the Supporting Information. The two
crystallographically independent molecules in the two forms of
1 are conformers that differ in the dihedral angle between the
two aromatic rings (7.28(5)° in 1-I and 38.93(7)° in 1-II).
The same conformational polymorphism can be observed in
compounds 2 and 3 as well (67.24(3)° in 2-I and 1.69(8)° in
2-II; 68.45(8)° in 3-I and 1.97(8)° in 3-II). Superposition of
the molecules in the corresponding forms clearly reveals the
main conformational difference (Figure 2). The molecule in
the crystal of compound 4 is nearly flat, as shown by the
dihedral angle of 2.51(3)°.
3. RESULTS AND DISCUSSION
3.1. Crystal Structures. Two forms (1-I and 1-II; 2-I and
2-II; 3-I and 3-II) were discovered for each of compounds 1−
3, and one form was found for 4 (Figure 1). Form 1-I crystals
All crystal forms have one molecule in the asymmetric unit
(Z′ = 1). On the one hand, for 1, 2, and 3, the polymorphism
clearly results from the interplay between intramolecular
interactions, manifested by the conformation, and intermo-
lecular interactions. In 1-I, the molecule takes a nearly planar
conformation, forming the acid−acid hydrogen-bonding
33−35
2
homosynthon (R2 (8) in graph set notation,
Figure 3a)
with the bond length and angle being 1.794 Å and 175.85°,
respectively. The molecule in 1-II, on the other hand, bears a
nonplanar conformation, and the proton transfers from the
carboxyl to pyridine N and causes subsequent intermolecular
carboxylate-pyridinium NH hydrogen bonding with the bond
length and angle being 1.868 Å and 153.63°, respectively.
Molecules in 1-II are zwitterionic and form one-dimensional
(1-D) hydrogen-bonded chains (Figure 3b). The link between
conformation and intermolecular hydrogen-bonding motif
agrees with the general rule established in our recent study
regarding the formation of either the acid−acid homosynthon
or the acid−pyridine heterosynthon in 2-PNA analogues, that
is, when the dihedral angle between the pyridine ring and the
phenyl ring is less than 30°, the acid−acid homosynthon
forms; otherwise, the acid−pyridine heterosynthon appears.14
The dihedral angles in 1-I and 1-II are 7.28 (5)° and 38.93
(7)°, respectively. In addition, both forms bear an intra-
molecular hydrogen bond between the NH bridging the two
aromatic rings and the carbonyl O (S6). The bond distance
and angle are 1.940 Å and 139.97° in 1-I and 1.827 Å and
142.19° for 1-II. Interestingly, neither form seems to resemble
any forms of CLX.
Figure 1. Crystals of the polymorphs of clonixin analogues. Scale bar:
0.2 mm.
were obtained as yellow plates from methanol (MeOH), ethyl
acetate (EtOAc), ethanol (EtOH), dimethyl sulfoxide
(DMSO), N,N-dimethylformamide (DMF), dichloromethane
(DCM), acetonitrile (CH3CN), chloroform (CHCl3), tetrahy-
drofuran (THF), ether (Et2O), or acetic acid (CH3COOH);
form 1-II crystals were grown as light yellow needles from
water (H2O); 2-I crystals were harvested as colorless blocks
from ethanol, isopropyl alcohol (i-PrOH), ether, DMF,
DMSO, DCM, THF, or MeOH, and 2-II crystals were
obtained as light yellow plates from ethyl acetate or acetone
(CH3COCH3); 3-I formed as colorless blocks from ethanol,
and 3-II formed as colorless plates from acetone; 4 grew as
yellow elongated blocks from all the solvents used. The
crystallization results are summarized in Table 1. Structure
determination by single-crystal X-ray diffraction found forms 1-
Table 1. Crystal Growth of Compounds 1−4 in Different
Solvents
Similar observations can be made for compounds 2 and 3 as
well. In 2-I, the only molecule in the asymmetric unit is highly
twisted, and the aforementioned dihedral angle is 67.24(3)°.
The crystal structure is disordered at the phenyl ring. Not
surprisingly, the molecules are connected through acid−
pyridine heterosynthon (C6), forming one-dimensional chains
parallel to the b axis. The hydrogen bond parameters are 1.870
Å and 171.28°, and those of the intramolecular hydrogen bond
are 1.938 Å and 134.85° (Figure 4). 2-I is similar to the form I
of CLX. Unlike 2-I, the molecule of 2-II is almost flat, and the
dihedral angle is 1.69(8)°. As expected, the molecules pair up
via the acid−acid homosynthon. The parameters of the inter-
and intramolecular hydrogen bond are 1.856 Å and 171.81°,
and 1.924 Å and 140.59°, respectively (Figure 4). 2-II
resembles the form IV of CLX. In 3-I, the molecule is highly
twisted, as indicated by its dihedral angle of 68.45(8)°. Both
aromatic rings are disordered. Because of the nonplanarity, a 1-
D acid−pyridine hydrogen-bonding chain is formed (Figure
5). The parameters of the intermolecular and intramolecular
hydrogen bonds are, respectively, 1.863 Å and 167.59°, and
solvent
acetone
chloroform
ethyl acetate
methanol
dichloromethane
ethanol
acetonitrile
ether
isopropyl alcohol
dimethyl sulfoxide
tetrahydrofuran
acetic acid
DMF
method
form
3-II
slow evaporation
slow evaporation
slow evaporation
slow evaporation
slow evaporation
slow evaporation
slow evaporation
slow evaporation
slow evaporation
slow evaporation
slow evaporation
slow evaporation
slow evaporation
slow evaporation
slow evaporation
slow evaporation
slow evaporation
slow evaporation
1-I
1-I
1-I
1-I
1-I
1-I
1-I
1-I
1-I
1-I
1-I
1-I
1-I
1-II
2-II
2-I
2-II
2-I
2-I
2-I
4
4
4
4
4
4
4
4
4
4
4
4
4
4
3-I &-3-II
3-I &-3-II
3-I
2-I
2-I
2-I
2-I
3-I &-3-II
2-I
water
benzene
toluene
pet ether
3-I
hexane
C
DOI: 10.1021/acs.cgd.8b01180
Cryst. Growth Des. XXXX, XXX, XXX−XXX

Crystal Growth & Design
Article
Table 2. Crystallographic Data of Compounds 1−4
1-I
1-II
2-I
2-II
3-I
3-II
4
formula
formula weight
crystal size
C13H11FN2O2
246.24
0.40 × 0.15 ×
0.02
C13H11FN2O2
246.24
0.40 × 0.10 ×
0.05
C13H11BrN2O2
307.15
0.30 × 0.25 ×
0.20
C13H11BrN2O2
307.15
0.35 × 0.30 ×
0.02
C13H11IN2O2
354.14
0.30 × 0.20 ×
0.10
C13H11IN2O2
354.14
0.40 × 0.20 ×
0.05
C13H12N2O2
228.25
0.45 × 0.10 ×
0.10
crystal system
space group
a, Å
b, Å
c, Å
α, deg
β, deg
γ, deg
monoclinic
P21/c
3.821(1)
21.056(4)
13.791(3)
90
96.01(1)
90
4, 1
1103.5(4)
1.482
90(2)
0.113
512.0
monoclinic
P21/c
3.85100(10)
22.3090(6)
12.5920(4)
90
92.4090(10)
90
4, 1
1080.85(5)
1.513
90(2)
0.115
512.0
monoclinic
P21/c
7.605(1)
14.106(2)
11.762(2)
90
101.82(1)
90
4, 1
1235.0(3)
1.652
90 (2)
3.323
616.0
triclinic
monoclinic
P21/c
7.8162(16)
14.053(3)
12.023(2)
90
103.05(3)
90
4,1
1286.5(4)
1.828
90(2)
2.484
688
triclinic
P1
̅
monoclinic
P21/c
4.933(1)
21.146(4)
10.378(2)
90.00
97.10(1)
90.00
4,1
1074.3(4)
1.411
90(2)
0.097
480
P1
̅
7.646(1)
7.098(1)
11.050(2)
100.32(1)
101.36(1)
87.46(1)
2, 1
578.41(15)
1.764
90 (2)
7.10120(10)
7.69820(10)
11.5625(2)
78.6401(7)
79.5043(7)
86.5626(7)
2,1
609.140(16)
1.931
90(2)
Z, Z′
V, Å3
Dcal, g cm−3
T, K
abs coeff (mm−1
F(000)
)
3.548
308.0
2.624
344
θ range, deg
limiting indices
1.77−27.41
−4 ≤ h ≤ 4
−27 ≤ k ≤ 27
−17 ≤ l ≤ 17
100%
1.83−27.45
−4 ≤ h ≤ 4
−28 ≤ k ≤ 28
−16 ≤ l ≤ 16
100%
2.28−27.49
−9 ≤ h ≤ 9
−18 ≤ k ≤ 16
−15 ≤ l ≤ 15
100%
1.91−27.50
−9 ≤ h ≤ 9
−9 ≤ k ≤ 9
−14 ≤ l ≤ 14
99.8%
2.26−27.47
−10 ≤ h ≤ 10
−18 ≤ k ≤ 18
−15 ≤ l ≤ 15
99.7%
1.82−27.48
−9 ≤ h ≤ 9
−9 ≤ k ≤ 9
−14 ≤ l ≤ 15
99.7%
2.20−27.49
−6 ≤ h ≤ 6
−27 ≤ k ≤ 27
−13 ≤ l ≤ 13
99.8%
completeness to
2θ
unique reflections 1701
1504
2310
2124
2420
2651
1852
R1 [I > 2σ(I)]
wR2 (all data)
CCDC code
0.0498
0.1289
1860257
0.0526
0.1660
1860258
0.0436
0.0980
1860259
0.0396
0.0964
1860260
0.0300
0.0652
1860261
0.0280
0.0741
1860262
0.0467
0.1414
CUNKOM
Figure 2. Superposition of molecules in the asymmetric units of the
polymorphs of compounds 1, 2, and 3.
1.944 Å and 135.54°. 3-I resembles the form I of CLX, while in
3-II, the molecule is nearly flat with the dihedral angle of
1.97(8)°. Acid−acid dimers are formed as anticipated (Figure
5). The parameters are 1.862 Å and 173.27° for the
intermolecular hydrogen bond, and 1.921 Å and 140.87° for
the intramolecular bond. 3-II resembles form IV of CLX.
Only one crystal form was obtained for compound 4. The
structure is similar to 1-I and unlike any known form of CLX.
The similarity may stem from the similar atomic size of F and
H. The conformation is planar with a dihedral of 2.51(3)°. The
crystal is based on dimers built on the acid−acid homosynthon
(Figure 3c). The hydrogen-bond parameters are 1.801 Å and
177.71°, and 1.945 Å and 140.56° for inter- and intramolecular
bonds, respectively.
Figure 3. Crystal packing of 1-I (a), 1-II (b), and 4. For clarity, only
hydrogens participating hydrogen bonds are shown.
Figure 6 shows PXRD patterns of individual forms of each
system collected at room temperature, along with PXRD
patterns calculated from the single-crystal structures deter-
mined at 90 K. The similarity between the experimental and
simulated patterns indicates the thermodynamic stability at
room temperature. The comparison also suggests that each
batch of crystallization product had a relatively high phase
purity.
3.2. Thermal Properties. Melting point and heat of fusion
were measured by DSC for each system (Figure 7). 1-I showed
a single thermal event with an onset temperature of 184.3 °C,
D
DOI: 10.1021/acs.cgd.8b01180
Cryst. Growth Des. XXXX, XXX, XXX−XXX

Crystal Growth & Design
Article
Figure 4. Crystal packing of 2-I (a) and 2-II (b). For clarity, only
hydrogens participating hydrogen bonds are shown.
Figure 7. DSC of polymorphs of compounds 1−3 and the only form
of compound 4.
the first one at 141 °C, which is believed to be solid−solid
phase transition into a new solid form. The new form melted at
252.0 °C (Figure 7). It seems that the new forms generated
from heating of 2-I and 2-II could be of the same crystal form,
but more experimentation is needed to characterize this new
high-temperature form. The thermal behaviors of 3-I and 3-II
also involved multiple events. 3-I melted at 229.8 °C (heat of
fusion of 11.7 J/g), followed by recrystallization to a new form,
which melted at 244.5 °C. There seems to be a small peak at
∼170 °C (heat of fusion of 3.5 J/g) as well. 3-II melted at
234.4 °C, which recrystallized into the same new form (Figure
7). Lastly, compound 4 gave only one melting peak with an
onset temperature of 168.0 °C (Figure 7).
Figure 5. Crystal packing of 3-I (a) and 3-II (b). For clarity, only
hydrogens participating hydrogen bonds are shown.
which corresponds to melting. 1-II displayed at least two
thermal events. The first, at the onset temperature of 170.4 °C,
corresponds to the melting of the form, followed by
crystallization of the melt, as suggested by the exothermic
peak. This new form is expected to be 1-I, because its melting
point matches that of 1-I (Figure 7). The shoulder of the first
melting peak may also suggest that it crystallized into another
form, which then transitioned into 1-I. 2-I revealed two
thermal events by DSC. The first had an onset temperature of
237.9 °C with a small heat of fusion of 2.4 J/g, which could be
a solid−solid phase transition to a new form. This new form
melted at ∼250 °C. 2-II also showed two thermal events with
3.3. Computational Analyses. Optimization of single
molecules of the compounds indicates that the neutral planar
conformers are more stable than the nonplanar ones to various
degrees; the energy differences span 1.5−2.2 kcal/mol (Table
Figure 6. PXRD of polymorphs of compounds 1−3 and the one form of 4.
E
DOI: 10.1021/acs.cgd.8b01180
Cryst. Growth Des. XXXX, XXX, XXX−XXX

Crystal Growth & Design
Article
a
Table 3. Energy Difference between Planar and Nonplanar/Zwitterion Conformations of Compounds 1−4
CLX
1
2
3
4
E (hartree)
E0 (hartree)
G (hartree)
planar
nonplanar
zwitterion
planar
nonplanar
zwitterion
planar
nonplanar
−1222.414 954
−1222.412 646
−862.060 186
−862.057 174
−862.042 685
−861.839 094
−861.836 616
−861.822 422
−861.882 541
−861.879 885
−861.866 384
1.67
−3336.334 895
−3336.332 655
−7681.714 188
−7681.712 187
−762.794 924
−762.792 084
−1222.195 247
−1222.193 504
−3336.115 946
−3336.114 178
−7681.495 705
−7681.494 045
−762.565 777
−762.563 477
−1222.239 784
−1222.237 647
−3336.162 566
−3336.159 436
−7681.543 637
−7681.540 089
−762.608 556
−762.605 590
zwitterion
nonplanar soft-enter replaced as-planar
zwitterion-planar
ΔG (kcal/mol)
ΔG (kcal/mol)
1.34
1.96
2.23
1.86
10.14
a
Gibbs free energies were calculated under 298.15 K and 1 atm.
3). The largest difference is observed for the zwitterionic and
planar species of 1, with the planar conformer 10.14 kcal/mol
lower in energy than the zwitterion. Given the size similarity
between F and H of 1 and 4, the electronegative F appears to
strengthen the planar conjugation between phenyl and amino
groups. Our previous study did demonstrate that the presence
of highly electronegative substituents on the phenyl ring could
enforce a planar conformation in the diarylamine com-
pounds.13 The major difference in conformation stems from
the dihedral angle τ, as illustrated by Figure 2. From the
conformation energy scan of τ (Figure 8), similar results can be
ing motifs are summarized in Table 4. Thus, the loss of
stability due to molecular nonplanarity can be compensated by
Table 4. Hydrogen Bonding Strength of Various Synthons
intermolecular number of
hydrogen
interactions
(kcal/mol)
hydrogen bond strength
bond
1
(kcal/mol)
description
NP
P
−10.94
−13.39
−12.95
−22.89
−10.94
nonplanar, acid−pyridine
heterosynthon
2
2
1
−6.7
−6.47
−22.89
planar, acid−acid
homosynthon
1-I
1-II
planar, acid−acid
homosynthon
nonplanar zwitterion,
carboxylate−pyridinium
NH heterosynthon
2-I
2-II
3-I
3-II
4
−10.84
−13.06
−10.8
1
2
1
2
2
−10.84
−6.53
−10.8
−6.41
−6.67
twisted, acid−pyridine
heterosynthon
planar, acid−acid
homosynthon
nonplanar, acid−pyridine
heterosynthon
−12.82
−13.35
planar, acid−acid
homosynthon
planar, acid−acid
homosynthon
the gain from hydrogen bonding. Lattice energies for the new
crystals are calculated to be −48.62 and −52.15 kcal/mol for
1-I and 1-II, −54.17 and −52.60 kcal/mol for 2-I and 2-II,
−55.89 and −56.00 kcal/mol for 3-I and 3-II, and −47.82
kcal/mol for 4. The overall strengths of intermolecular
interactions of the polymorphs of each compound and
among the compounds are generally within a few Kcal/mol,
and the differences apparently result from the substitution by
the halogens and hydrogen. The calculated densities are 1.482
and 1.513 g/cm3 for 1-I and 1-II, 1.652 and 1.764 g/cm3 for 2-
I and 2-II, 1.828 and 1.931 g/cm3 for 3-I and 3-II, and 1.411
g/cm3 for 4. For the polymorphic systems of 1 and 3, the
ordering in lattice energy is in agreement with the densities,
while for 2-I and 2-II, the lattice energies do not match with
the densities. Yet this is not a total surprise, since packing is
not the only factor contributing to the lattice energy. The DSC
measurement suggests that 1-I and 2-I are the more stable
forms at higher temperature, compared with respective 1-II
and 2-II, with the lattice energy values only echoing the two
forms of 2. It could be the zwitterionic form of 1-II becomes
less stable at elevated temperature. Interestingly, the lattice
energies of 3-I and 3-II are almost identical, and the DSC
Figure 8. Relaxed potential energy surface scan for the rotation of the
aniline bond (C2N7C8C13).
observed with the global minimum located at 0° and two local
minima (stereochemically identical) at 120°, 119°, 115°,
and 110° for 1−4, with the energy barriers between the
global and local minima of 3.2, 2.6, 2.1, and 2.8 kcal/mol,
respectively.
Our earlier work and other related studies showed that the
heterogeneous acid−pyridine hydrogen bond is energetically
favored relative to the homogeneous acid−acid hydrogen
bond.36−40 In both CLX and the three compounds (1, 2, and
3) investigated, the same observation is made. The energy
strengths of the acid−pyridine hydrogen bond (10.94, 10.84,
and 10.80 kcal/mol of CLX-NP, 2-I, and 3-I) are considerably
higher than those of the acid−acid hydrogen bond (6.70, 6.47,
6.53, 6.41, and 6.67 kcal/mol for CLX-P, 1-I, 2-II, 3-II, and 4).
The hydrogen bond between the pyridinium NH and
carboxylate in 1-II is much stronger than the acid−pyridine
hydrogen bond due to the electrostatic interaction (22.89 kcal/
mol). The hydrogen-bonding strength values and correspond-
F
DOI: 10.1021/acs.cgd.8b01180
Cryst. Growth Des. XXXX, XXX, XXX−XXX

Crystal Growth & Design
Article
Figure 9. Fingerprint plots and relative contribution of intermolecular contacts on the base of Hirshfeld surface analysis: (a) 1-I, (b) 1-II, (c) 2-I,
(d) 2-II, (e) 3-I, (f) 3-II, and (g) 4.
results suggest a high-temperature form that is different from
either.
of 4 is the smallest (Table 2). It seems intuitive that 4 should
be polymorphic, given that the acid−pyridine hydrogen
bonding is much stronger, and a form bearing such a motif
needs to overcome a smaller conformational energy barrier. In
addition, kinetic factors might be ruled out, because a variety of
different solvents were utilized in crystallization of these
compounds. All these details suggest that secondary inter-
actions such as π−π stacking, as compared to hydrogen
bonding, may further guide the molecular packing during
nucleation. When the molecule remains planar, it can form
much stronger π−π stacking, as compared with a twisted
conformation. Nuclei formed by planar conformers become
more favorable energetically and continue to grow. For the
halogenated clonixin derivatives (1, 2, and 3), conversely,
either hydrogen-bonding motif can be realized in practice,
resulting in respective polymorphs, likely because of similar
energy states achieved jointly by conformation and π−π
stacking.
Hirshfeld surface analyses further reveal intermolecular
contacts of these compounds. Two-dimensional (2-D) finger-
print plots and relative contributions of various interactions by
Hirshfeld surfaces were calculated for 1−4 (Figure 9).
Contributions by the H···O contacts show significant changes
from 12.8%, 9.5%, and 9.4% in 1-I, 2-II, 3-II (all with planar
conformations) to 20.1%, 10.2%, and 10.3% in 1-II, 2-I, and 3-
I (all with nonplanar conformations), respectively. These
changes obviously stem from the change in the hydrogen-
bonding motif, with 1-I and 1-II showing the largest increase,
corroborating the hydrogen bonding in 1-II being much
stronger. This is further revealed by the decrease in
contribution of the H···N contact from 3.7% of 1-I to 1.3%
of 1-II, while increases are seen from 4.5% and 4.4% of 2-II
and 3-II to 6.2% and 5.4% of 2-I and 3-I, respectively.
Overall, the most interesting finding of the study lies in the
polymorphism of the compounds. All of the halogen
derivatives are polymorphic, whereas for compound 4 only
one form has been obtained. Its DSC measurement also
revealed no high-temperature form. It is also intriguing that the
energy difference between the planar and twisted conformers
4. CONCLUSIONS
Four analogues of CLX were synthesized by replacing the
chlorine with fluorine (1), bromine (2), iodine (3), and
hydrogen (4), and their polymorphism was investigated.
G
DOI: 10.1021/acs.cgd.8b01180
Cryst. Growth Des. XXXX, XXX, XXX−XXX

Crystal Growth & Design
Article
nicotinic acid) on platelet function and fibrinolytic activity in vitro.
Thromb. Res. 1975, 6, 9.
(3) Arkel, Y. S.; Schrogie, J. J.; Williams, R. Effect of clonixin and
aspirin on platelet aggregation in human volunteers. J. Clin.
Pharmacol. 1976, 16, 30.
(4) Morse, E. E. Effects of drugs on platelet function. Ann. Clin. Lab.
Sci. 1977, 7, 68.
(5) Krymchantowski, A. V.; Carneiro, H.; Barbosa, J.; Jevoux, C.
Lysine clonixinate versus dipyrone (metamizole) for the acute
treatment of severe migraine attacks: a single-blind, randomized
study. Arq. Neuro-Psiquiatr. 2008, 66, 216.
Compounds 1, 2, and 3 were found to be polymorphic, as two
forms were discovered for each, while for compound 4, only
one form was identified. Similar to CLX, the molecules in their
crystals are connected either by the acid−acid dimer
homosynthon or the acid−pyridine heterosynthon, depending
on the dihedral angle between the two aromatic rings.
Isostructurality was observed between 2-I, 3-I and form I of
CLX and between 2-II, 3-II and form IV of CLX, likely due to
the similar electronegativity and not significant size difference.
4 is isostructural to 1-I, but not to the other crystals, possibly
because of the size similarity between F and H and
electronegativity dissimilarity between F/H and Cl, Br, and I.
Phase behaviors of the polymorphic systems were studied by
DSC, which showed phase transitions between the different
forms in each system. Conformational energy, hydrogen-bond
strength calculations, and conformation scans suggested both
conformational flexibility and the possibility of multiple
hydrogen-bonding motifs lead to the polymorphism of the
compounds and suggest additional forms might be found for
each compound. Lattice energy evaluation and Hirshfeld
analysis further provided insight on the relative stability of the
polymorphs of each system and contributions to the stability of
each form. The lack of polymorphs of compound 4 deserves
further investigation.
(6) Nangia, A. Conformational polymorphism in organic crystals.
Acc. Chem. Res. 2008, 41, 595.
(7) Cruzcabeza, A. J.; Bernstein, J. Conformational polymorphism.
Chem. Rev. 2014, 114, 2170−2191.
(8) Takasuka, M.; Nakai, H.; Shiro, M. Infrared spectral and X-Ray
studies of polymorphic forms of 2-(2-methyl-3-chloroanilino)-
nicotinic acid. J. Chem. Soc., Perkin Trans. 2 1982, 2, 1061.
(9) Long, S.; Mao, T. F.; Chen, P.; Liu, M.; Parkin, S.; Zhang, M.; Li,
T.; Zhou, P.; Yu, F. Strong hydrogen bond leads to a fifth crystalline
form and polymorphism of clonixin. ChemistrySelect 2017, 2, 4942.
(10) Long, S.; Parkin, S.; Siegler, M. A.; Cammers, A.; Li, T.
Polymorphism and phase behaviors of 2-(Phenylamino)nicotinic acid.
Cryst. Growth Des. 2008, 8, 4006.
(11) Long, S.; Parkin, S.; Siegler, A.; Brock, C.; Cammers, A.; Li, T.
Polymorphism of an organic system effected by the directionality of
hydrogen-bonding chains. Cryst. Growth Des. 2008, 8, 3137.
(12) Long, S.; Li, T. Controlled Formation of the Acid-Pyridine
Heterosynthon over the Acid-Acid Homosynthon in 2-Anilinonico-
tinic Acids. Cryst. Growth Des. 2009, 9, 4993.
(13) Long, S.; Li, T. Enforcing molecule’s π-conjugation and
consequent formation of the acid-acid homosynthon over the acid-
pyridine heterosynthon in 2-anilinonicotinic acids. Cryst. Growth Des.
2010, 10, 2465.
ASSOCIATED CONTENT
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.cgd.8b01180.
■
S
Synthesis of CLX analogues; IR and Raman spectra
(PDF)
(14) Chen, P.; Zhang, Z.; Parkin, S.; Zhou, P.; Cheng, K.; Li, C.; Yu,
F.; Long, S. Preferred formation of the carboxylic acid−pyridine
heterosynthon in 2-anilinonicotinic acids. RSC Adv. 2016, 6, 81101.
(15) Long, S.; Zhou, P.; Parkin, S.; Li, T. From competition to
commensuration by two major hydrogen-bonding motifs. Cryst.
Growth Des. 2014, 14, 27.
(16) Mattei, A.; Li, T. Interplay between molecular conformation
and intermolecular interactions in conformational polymorphism: A
molecular perspective from electronic calculations of tolfenamic acid.
Int. J. Pharm. 2011, 418, 179.
(17) Kermack, W. O.; Weatherhead, A. P. Some anilinopyridine
derivatives. J. Chem. Soc. 1942, 726.
(18) Ting, P. C.; Kaminski, J. J.; Sherlock, M. H.; Tom, W. C.; Lee,
J. F.; Bryant, R. W.; Watnick, A. S.; McPhail, A. T. Substituted 1,3-
dihydro-2H-pyrrolo[2,3-b]pyridin-2-ones as potential antiinflamma-
tory agents. J. Med. Chem. 1990, 33, 2697.
Accession Codes
CCDC 1860257−1860262 contain the supplementary crys-
tallographic data for this paper. These data can be obtained
free of charge via www.ccdc.cam.ac.uk/data_request/cif, or by
emailing data_request@ccdc.cam.ac.uk, or by contacting The
Cambridge Crystallographic Data Centre, 12 Union Road,
Cambridge CB2 1EZ, UK; fax: +44 1223 336033.
AUTHOR INFORMATION
Corresponding Authors
*Phone: (027) 87194980. E-mail: Sihuilong@wit.edu.cn;
longsihui@yahoo.com. (S. Long)
■
*Phone: (765) 494-1451. E-mail:tonglei@purdue.edu. (T. Li)
ORCID
(19) Chen, P.; Jin, Y.; Zhou, P.; Parkin, S.; Zhang, Z.; Long, S. Solid-
state characterization of 2-[(2,6-dichlorophenyl)amino]-benzalde-
hyde: an experimental and theoretical investigation. J. Chin. Chem.
Soc. 2017, 64, 531.
(20) Nonius Collect; Nonius BV: Delft, the Netherlands, 2002.
(21) Sheldrick, G. M. A short history of SHELX. Acta Crystallogr.,
Sect. A: Found. Crystallogr. 2008, 64, 112.
Faquan Yu: 0000-0002-5062-8731
Sihui Long: 0000-0002-4424-6374
Tonglei Li: 0000-0003-2491-0263
Notes
The authors declare no competing financial interest.
(22) Sheldrick, G. M. Crystal structure refinement with SHELXL.
Acta Crystallogr., Sect. C: Struct. Chem. 2015, 71, 3.
(23) Hirshfeld, F. L. Bonded-atom fragments for describing
molecular charge densities. Theor. Chim. Acta 1977, 44, 129.
(24) Spackman, M. A.; Jayatilaka, D. Hirshfeld surface analysis.
CrystEngComm 2009, 11, 19.
(25) Becke, A. D. Density-functional thermochemistry. III. The role
of exact exchange. J. Chem. Phys. 1993, 98, 5648.
(26) Krishnan, R.; Binkley, J. S.; Seeger, R.; Pople, J. A. Self-
consistent molecular orbital methods. XX. A basis set for correlated
wave functions. J. Chem. Phys. 1980, 72, 650.
ACKNOWLEDGMENTS
■
S.L. thanks Natural Science Foundation of Hubei Province for
financial support (2014CFB787). T.L. is grateful to NSF for
supporting the work (DMR1006364).
REFERENCES
■
(1) Sherlock, M. H.; Sperber, N. Antiinflammatory 2-Anilinonico-
tinic acids. Br. Pat. GB 1147701, 1969.
(2) Widmer, K.; Reed, R. C.; Freireich, E. J. The influence of a new
analgetic-antiphlogistic compound (2-(2-methyl-3′chloro)-anilino-
H
DOI: 10.1021/acs.cgd.8b01180
Cryst. Growth Des. XXXX, XXX, XXX−XXX

Crystal Growth & Design
Article
(27) Zhao, Y.; Truhlar, D. G. The M06 suite of density functionals
for main group thermochemistry, thermochemical kinetics, non-
covalent interactions, excited states, and transition elements: two new
functionals and systematic testing of four M06-class functionals and
12 other functionals. Theor. Chem. Acc. 2008, 120, 215−241.
(28) Weigend, F.; Ahlrichs, R. Balanced basis sets of split valence,
triple zeta valence and quadruple zeta valence quality for H to Rn:
Design and assessment of accuracy. Phys. Chem. Chem. Phys. 2005, 7,
3297−3305.
(29) Boys, S. F.; Bernardi, F. The calculation of small molecular
interactions by the differences of separate total energies. Some
procedures with reduced errors. Mol. Phys. 1970, 19, 553.
(30) Dovesi, R.; Orlando, R.; Erba, A.; Zicovich-Wilson, C. M.;
Civalleri, B.; Casassa, S.; Maschio, L.; Ferrabone, M.; De La Pierre,
M.; D’Arco, P.; Noel, Y.; Causa, M.; Rerat, M.; Kirtman, B.
CRYSTAL14: A program for the ab initio investigation of crystalline
solids. Int. J. Quantum Chem. 2014, 114, 1287.
(31) Grimme, S.; Antony, J.; Ehrlich, S.; Krieg, H. A consistent and
accurate ab intio parameterization of density functional dispersion
correction (DFT-D) for the 94 elements H-Pu. J. Chem. Phys. 2010,
132, 154014.
(32) Grimme, S.; Ehrlich, S.; Goerigk, L. Effect of the damping
function in dispersion corrected density functional theory. J. Comput.
Chem. 2011, 32, 1456−1465.
(33) Etter, M. C. Encoding and decoding hydrogen-bond patterns of
organic compounds. Acc. Chem. Res. 1990, 23, 120.
(34) Etter, M. C.; Macdonald, J. C.; Bernstein, J. Graph-set analysis
of hydrogen-bond patterns in organic crystals. Acta Crystallogr., Sect.
B: Struct. Sci. 1990, 46, 256.
(35) Bernstein, J.; Davis, R. E.; Shimoni, L.; Chang, N. L. Patterns in
hydrogen bonding: functionality and graph set analysis in crystals.
Angew. Chem., Int. Ed. Engl. 1995, 34, 1555.
(36) Li, T.; Zhou, P.; Mattei, A. Electronic origin of pyridinyl N as a
better hydrogen-bonding acceptor than carbonyl O. CrystEngComm
2011, 13, 6356.
(37) Steiner, T. Competition of hydrogen-bond acceptors for the
strong carboxyl donor. Acta Crystallogr., Sect. B: Struct. Sci. 2001, 57,
103−106.
̈
(38) Aakeroy, C. B.; Beatty, A. M.; Helfrich, B. A. ″Total synthesis″
supramolecular style: design and hydrogen-bond-directed assembly of
ternary supermolecules. Angew. Chem., Int. Ed. 2001, 40, 3240−3242.
(39) Vishweshwar, P.; Nangia, A.; Lynch, V. M. Molecular
complexes of homologous alkanedicarboxylic acids with isonicotina-
mide: X-ray crystal structures, hydrogen bond synthons, and melting
point alternation. Cryst. Growth Des. 2003, 3, 783−790.
(40) Shattock, T. R.; Arora, K. K.; Vishweshwar, P.; Zaworotko, M.
J. Hierarchy of Supramolecular Synthons: Persistent Carboxylic
Acid···Pyridine Hydrogen Bonds in Cocrystals That also Contain a
Hydroxyl Moiety. Cryst. Growth Des. 2008, 8, 4533−4545.
I
DOI: 10.1021/acs.cgd.8b01180
Cryst. Growth Des. XXXX, XXX, XXX−XXX