Celecoxib-Nicotinamide Cocrystal Revisited: Can Entropy Control Cocrystal Formation?

Article abstract of DOI:10.1021/acs.cgd.7b00308

The cocrystal of celecoxib-nicotinamide has been revisited to reveal its crystal structure and unusual formation properties. Gibbs free energy and enthalpy of formation at 25 °C were determined by solubility and thermal analysis. The formation of celecoxi

Full text of DOI:10.1021/acs.cgd.7b00308

Subscriber access provided by HACETTEPE UNIVERSITESI KUTUPHANESI
Article
Celecoxib-Nicotinamide Co-Crystal Revisited:
Can Entropy Control Co-Crystal Formation?
Si-Wei Zhang, Andrew P. J. Brunskill, Eric Schwartz, and Shuwen Sun
Cryst. Growth Des., Just Accepted Manuscript • DOI: 10.1021/acs.cgd.7b00308 • Publication Date (Web): 10 Apr 2017
Downloaded from http://pubs.acs.org on April 10, 2017
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Crystal Growth & Design is published by the American Chemical Society. 1155
Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 35
Crystal Growth & Design
1
2
3
4
5
6
7
8
9
Celecoxib-Nicotinamide Co-Crystal Revisited: Can
Entropy Control Co-Crystal Formation?
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Si-Wei Zhang, Andrew P. J. Brunskill, Eric Schwartz, Shuwen Sun
Department of Process Research & Development, Merck & Co., Inc.,
126 East Lincoln Avenue, Rahway, NJ 07065 USA
ABSTRACT: The co-crystal of celecoxib-nicotinamide has been revisited to reveal its crystal
structure and unusual formation properties. Gibbs free energy and enthalpy of formation at 25 °C
were determined by solubility and thermal analysis. The formation of celecoxib-nicotinamide
was found to be an endothermic process and driven by entropy, a mechanism different from
many reported co-crystals but in agreement with previously reported lattice energy calculation.
The co-crystal is stable only above a transition temperature as predicted from thermodynamic
data and confirmed experimentally. This formation mechanism brings new opportunities into co-
crystal research from structural prediction to process development, and enriches our
understanding on the nature of co-crystallization.
KEYWORDS: Co-crystal, celecoxib, nicotinamide, thermodynamics, formation energy
ACS Paragon Plus Environment

Crystal Growth & Design
Page 2 of 35
1
2
3
4
INTRODUCTION
5
6
7
8
Co-crystallization has evolved as a general strategy to expand the number of solid forms
available for pharmaceuticals1-4 or other materials5 and to improve their physicochemical
properties.5-8 The formation of co-crystals does not require the presence of ionizable groups in
co-crystal formers, making co-crystallization a more versatile method in crystal engineering than
salt formation. 9 Accompanying the discovery and structural solution of an increasing number of
co-crystals, more efforts have been focused on their thermodynamic formation properties.10-14 In-
depth understanding of these important characteristics will benefit co-crystal research in both the
discovery and development phases. Acquiring formation enthalpies (ΔHf) of co-crystals by
experimental13 or computational15 methods will aid in the structural prediction and rational
design of co-crystals.15-17 Furthermore, exploration of the phase diagrams of co-crystals,
including binary18-20 and ternary21-23 phase diagrams, provides guidance on their process
development.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The key to understanding the thermodynamics of co-crystals is the formation Gibbs free
energy (ΔGf) which determines the stability of a co-crystal. Although ΔGf can be experimentally
measured for existing co-crystals,10-12 forecasting ΔGf is of great interest for crystal structure
prediction. However, computational calculation of ΔGf is costly.24 As a result, the
computationally inexpensive lattice energy and formation energy (ΔEf) were often used instead
of ΔGf in structural prediction.15,25 It is noted that the difference between ΔEf and ΔGf is the
entropy term, as ΔEf is approximate to ΔHf at ambient conditions.26 Therefore, correlating ΔEf
with co-crystal stability assumes that enthalpy is the major component of free energy and the
main driving force of co-crystallization.
ACS Paragon Plus Environment

Page 3 of 35
Crystal Growth & Design
1
2
3
4
5
6
7
8
9
The validity of the assumption that co-crystallization is enthalpy controlled has been
examined both experimentally and computationally. Indeed, although only limited experimental
data is available, co-crystallization is shown to be a competition between an enthalpy term
favorable for co-crystal formation and an entropy term against the formation.11,12 However, a few
exceptions can be found, especially from computational calculations.27 In a recent study, Chan et
al. demonstrated the ability to calculate, ab initio, the ΔEf of a series of nicotinamide (NIC) co-
crystals based on dispersion-corrected density functional theory (DFT-D). In this study, it was
found that 30 out of 31 NIC co-crystals have negative ΔEf in favor of co-crystallization.15 The
co-crystal formed between NIC and celecoxib (CEL),28 a COX-2 selective nonsteroidal anti-
inflammatory drug (NSAID), was left as an outlier. Such exceptions were often attributed to the
inadequacy of computational methods or
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
simply ignored.
In this work, the CEL-NIC co-crystal was
revisited and its crystal structure was
determined by X-ray crystallography. The
Scheme 1. Structures of celecoxib (CEL) and
nicotinamide (NIC)
structure corrects several errors in the structure
previously determined using X-ray powder data.28 The newly determined structure agrees with
the in silico optimized structure reported by Chan et al.15 which allows the comparison between
experimental and computational studies. ΔGf and ΔHf of the co-crystal were determined from the
solubility and thermal properties of the co-crystal and a physical mixture of component crystals.
At 25 °C, CEL-NIC was found to be energetically stable (ΔGf = -0.7 (0.1) J/g) with an
endothermic formation (ΔHf = +13 (3) J/g) that agrees with lattice energy calculation, indicating
that the co-crystallization is driven by entropy (ΔSf = +0.05 J/g). The existence of a transition
ACS Paragon Plus Environment

Crystal Growth & Design
Page 4 of 35
1
2
3
4
5
6
7
8
9
temperature (Tt) below 25 °C between CEL-NIC and its components was predicted from
thermodynamic data and was confirmed by the inverted stability between the co-crystal and
components observed between 4 °C and 25 °C. CEL-NIC is stable only above this temperature,
unlike other enthalpy-controlled co-crystals which are usually stable below a transition
temperature. The formation of CEL-NIC is different from other co-crystals with reported
formation properties, which challenges the usage of ΔEf or ΔHf as the only indication of co-
crystal formation. This discovery enriches the understanding on the nature of co-crystal
formation, and is relevant to structural prediction and process development of co-crystals.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
MATERIAL AND METHODS
Compounds and solvents. Celecoxib (the most stable polymorph29) was purchased from
AstaTech (Bristol, PA, USA). Nicotinamide (the most stable polymorph30) was purchased from
Acros (Geel, Belgium). Both materials were used as received. All solvents were of HPLC grade
and purchased from Sigma-Aldrich (St. Louis, MO, USA). Analytical standards of nicotinamide
and celecoxib were purchased from Sigma-Aldrich and European Pharmacopoeia Reference
Standards (EDQM, Strasbourg, France), respectively.
X-ray diffraction experiments. Powder X-ray diffraction (PXRD) measurements were carried
out on a Panalytical (Natick, MA, USA) X-pert Pro PW3040 diffractometer with an X’celerator
detector (Cu Kα radiation, voltage 45 kV, and current 40 mA). Approximately 5 mg of powder
was sprinkled on the surface of a zero background silicon (510) sample holder and scanned from
2 to 40° 2θ at a speed of 8.4°/min and a step size of 0.02°. Single-crystal X-ray diffraction
(SCXRD) evaluation and data collection were performed at -173 °C and 25 °C on a Bruker
ACS Paragon Plus Environment

Page 5 of 35
Crystal Growth & Design
1
2
3
4
5
6
7
8
9
(Madison, WI, USA) SMART Apex II diffractometer with Cu Kα radiation, and the crystal
structure was solved using a standard procedure (see the Supporting Information for details).
Thermal analysis. Differential scanning calorimetry (DSC) was conducted with a TA
Instruments (New Castle, DE, USA) Discovery unit under 50 mL/min N2 purge. In a typical run,
5 – 10 mg sample in a Tzero Aluminum pan was heated at 10 °C/min to 185 °C to measure the
temperature and heat of melting, cooled at 10 °C/min, and heated again at 10 °C/min to record
the glass transition temperature of the melt. Results are reported as the average of at least three
replicated experiments.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
High-performance liquid chromatography (HPLC) method. HPLC was performed on an
Agilent (Santa Clara, CA, USA) 1100 HPLC system equipped with a diode array detector. The
separation was carried out on a 4.6 mm × 150 mm, 2.7 μm Ascentis Express fused-core C18
column (Sigma-Aldrich). The mobile phase consisted of (A) 0.1% phosphoric acid and (B)
acetonitrile. The gradient started with 5% B and ramped up to 95% B in 3 min, followed by a 3
min hold and a 3 min re-equilibration. For both CEL and NIC, UV detection was carried out at
wavelength of 220 nm. The external standards were 0.2 mg/ml solutions in acetonitrile. The
injection volume was set as 5 μL. The concentrations were calculated with an established
calibration curve ranging from 0.1 μg/ml to 0.24 mg/ml. The HPLC method was validated by
linearity and recovery.
Preparation of co-crystal and physical mixture. Co-crystals were obtained by dissolving 0.15
M CEL and NIC in chloroform/heptane 5/1 (v/v) at 65 °C. Hot solution was cooled to 50 °C,
seeded with co-crystal, and cooled to room temperature at a rate of 1 °C/min. Single crystals
were acquired by spontaneously cooling a hot chloroform solution of 0.5 M CEL and NIC to
room temperature. To prepare a physical mixture of CEL and NIC crystals for DSC analysis,
ACS Paragon Plus Environment

Crystal Growth & Design
Page 6 of 35
1
2
3
4
5
6
7
8
9
each material was lightly ground in a mortar prior to mixing. The ground powders were mixed at
1:1 molar ratio with a Barnstead Thermolyne (Dubuque, IA, USA) Maxi Mix II vortex mixer at
the maximum speed for 1 minute. The final mixture was analyzed by PXRD to ensure that it
contained only the component crystals and no co-crystals, and was used immediately for
subsequent DSC analysis.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Slurry experiments. To prepare a suspension for solubility measurement or phase conversion,
excess amount of each solid was added into the selected solvent. The suspension was then held
in a circulated metal block and magnetically stirred at 500 rpm. Temperature was controlled
through circulating water by a Julabo circulator (Allentown, PA, USA) with accuracy of ±0.1 °C.
Aliquots of the suspension were withdrawn at predetermined intervals and filtered through a 0.22
μm PTFE syringe filter. Phase of solid residue was immediately determined after filtration by
PXRD. Concentration of CEL or NIC in the filtrate was determined by HPLC. Results are
reported as the average of at least three replicated experiments.
Formation properties of co-crystal. The formation property of a co-crystal AB (1:1
stoichiometry) is defined in reference to equation 1:
A + B → AB (1)
where A and B are the crystal of component A (CEL) and the crystal of component B (NIC),
respectively. The formation volume (ΔVf) of a co-crystal (1:1 stoichiometry) is calculated using
equation 2:31
ΔVf = VAB − [0.5 VA + 0.5 VB] (2)
where VAB, VA, and VB are the volumes of one molecule (including void space) in the cocrystal,
the component crystal A, and the component crystal B, respectively. V is calculated from
crystallographic data using V = Vcell/n, where Vcell is the volume of the unit cell and n the number
ACS Paragon Plus Environment

Page 7 of 35
Crystal Growth & Design
1
2
3
4
5
6
7
8
9
of molecules there in. In this work, one “molecule” of co-crystal AB consists of 0.5 molecules of
A and 0.5 of B. To avoid the ambiguity on the definition of one molecule, ΔVf is also reported in
the unit of cm3/kg per co-crystal.31
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Equation 3 is used to determine the formation enthalpy (ΔHf):13
ΔHf =ΔHm(A+B) (TS → TL) – ΔHm(AB) (TS → TL) (3)
where TS (25 °C) is a temperature at which the co-crystal and the physical mixture of component
crystals are solid and at which ΔHf is evaluated, TL (180 °C) is a temperature at which the co-
crystal and the physical mixture are both melted to the same liquid, ΔHm(A+B) (TS → TL) and
ΔHm(AB) (TS → TL) are the corresponding enthalpies of melting.
The formation energy (ΔEf) is related to ΔHf through equation 4 (P is pressure):
ΔEf = ΔHf + PΔVf (4)
Equation 5 is used to determine the formation Gibbs free energy (ΔGf) at experimental
temperature (25 °C):11
ΔGf = – RTln(CA(A+B) / CA(AB+B)) (5)
where R is the gas constant, CA is the concentration of CEL in a solution in equilibrium with the
solids of physical mixture (A+B), or in equilibrium with the co-crystal and excess NIC (AB+B).
Physical mixture (A+B) was used since the solubility of CEL was found to be related to the
concentration of NIC, which will be discussed in the Results and Discussion section. Co-crystal
and NIC (AB+B) were the solid residue from the incongruent dissolution of co-crystal in the
testing solvent, chloroform. Molar concentration was used instead of activity for simplicity.
Finally, the entropy of formation (ΔSf) can be determined from equation 6:
ΔGf = ΔHf – TΔSf (6)
ACS Paragon Plus Environment

Crystal Growth & Design
Page 8 of 35
1
2
3
4
Determinations of the change of hydrogen-
5
6
bond length upon co-crystallization (ΔRHB)
7
8
9
can be found in supplementary information.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
RESULTS AND DISCUSSION
Crystal structure. Single crystal structure
of CEL-NIC was solved at both -173 °C and
22 °C. Relevant crystallographic data are
Figure 1. Hydrogen bonds in CEL-NIC co-crystal
demonstrated by the 22 °C structure.
2
shown in Table 1. In the 22 °C structure (Figure 1), NIC molecules form amide-amide R2 (8)
dimers.32,33 NIC dimers are linked by CEL molecules (amide H – sulfonyl O; sulfonamide H –
amide O) along the b direction to form infinite ribbons. In addition, these ribbons are stacked
along the a direction and stabilized by the C(4) chain formed between CEL molecules
(sulfonamide H – sulfonyl O). The -173 °C structure has the same hydrogen bond patterns as the
22 °C structure. The crystal volume shrinks by approximately 3% from 22 °C to -173 °C
(thermal expansion coefficient αV = 1.6 × 10-4 °C -1), which is comparable with NIC and other
NIC
co-crystal
formers.31
Disordered
Table 1. Crystallographic information for CEL-NIC co-
crystal
trifluoromethyl groups are observed at 22 °C
but not at -173 °C. It should be noted that the
molecular conformation in our single crystal
structure is different from the reported powder
structure, differing in the orientation of the
sulfonamide group of CEL and the amide
group of NIC.28 The crystal structure is
T, °C
-173 (2)
22 (2)
Wavelength, Å 1.54178
1.54178
monoclinic
P 21/n
0.2×0.04×0.01
5.1908(2)
9.9244(4)
46.3893(16)
90
92.2545(15)
90
2387.92
4
Cryst system
Space group
monoclinic
P 21/n
Cryst size, mm3 0.2×0.04×0.01
a, Å
b, Å
c, Å
5.1077(3)
9.8609(5)
46.021(2)
90
91.1948(19)
90
2317.41
4
1.443
α, deg
β, deg
γ, deg
V, Å3
Z
ρ, cm3/kg
1.401
ACS Paragon Plus Environment

Page 9 of 35
Crystal Growth & Design
1
2
3
4
5
6
consistent with the structure proposed by Chan et al. from a thorough conformational search and
lattice energy minimization starting from the powder structure.15
7
8
9
A remarkable property of NIC co-crystals is their looser molecular packing and shorter
hydrogen bonds relative to the component crystals.31 CEL-NIC follows the same trend. We
demonstrate this from its formation volume (ΔVf) and the change of hydrogen-bond length upon
co-crystallization (ΔRHB). For determination of ΔVf, molecular volumes of CEL-NIC and its
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
components were acquired from room
Table 2. Thermal properties of CEL, NIC, 1:1 physical
mixture and co-crystal
temperature crystal structures (CEL: CSD#
Phase
Tm, °C
ΔHm, J/g
Tg, °C
Celecoxib
Nicotinamide 128.4 (0.1) 196.3 (0.4)
1:1 physical 107.8 (0.1) 111.6 (0.5) 21.9 (0.9)
mixture
162.3 (0.3) 101.9 (0.7) 54.4 (0.1)
DIBBUL;34 NIC: CSD# NICOAM02;35 both
are the most stable polymorph). ΔVf of CEL-
NIC is 18.2 Å3/molecule or 43.6 cm3/kg,
which indicates that the co-crystal occupies a
volume 6.5% larger than individual
components. Compared to other NIC co-
crystals, CEL-NIC has one of the largest
expansions in volume.31 Despite the increase
in volume, the average hydrogen bond length
in CEL-NIC is shorter than the components
(ΔRHB = -0.1 Å; see the Supporting
Information for details), in agreement with
other NIC co-crystals.
a
–
Co-crystal
128.3 (0.1) 101.1 (0.3) 21.1 (0.1)
aTg could not be measured due to crystallization
Figure 2. (a) DSC melting endotherms of CEL-NIC co-
crystal, CEL, NIC, and a physical mixture of CEL and
NIC at 1:1 molar ratio. (b) Relative enthalpies between
co-crystal and the physical mixture. ΔHf is the formation
enthalpy of co-crystal (+13(3) J/g at 25 °C).
ACS Paragon Plus Environment

Crystal Growth & Design
Page 10 of 35
1
2
3
4
5
6
7
8
9
Determination of ΔHf from thermal analysis. DSC traces of the co-crystal, component
crystals, and a physical mixture of component crystals in 1:1 ratio are shown in Figure 2a, while
related thermal properties are recorded in Table 2. Our results are in good agreement with
existing data on CEL,36 NIC,13 and the co-crystal.28 Thermogravimetric analysis shows no
significant weight loss for CEL or NIC up to 180 °C, indicating that the melting data are
unaffected by thermal decomposition. The Tg values in Table 2 are the glass transition
temperatures observed during the second heating of the liquids produced by melting the crystals.
The similar Tg values between co-crystal and physical mixture are consistent with similar
chemical compositions of the crystalline samples before melting.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 2b demonstrates the determination of the formation enthalpy (ΔHf) of CEL-NIC co-
crystal.13 To obtain the formation enthalpy, heat flow data of the co-crystal and the physical
mixture in Figure 2a were integrated from a common liquid-state temperature (180 °C) down to a
common solid-state temperature (25 °C). Given that both the co-crystal and the physical mixture
melted to the same liquid at 180 °C, the enthalpy difference of the co-crystal relative to the
physical mixture is its formation enthalpy (equation 3). For CEL-NIC, ΔHf was determined to be
+13 (3) J/g at 25 °C, which indicates that the co-crystal has higher enthalpy compared with the
physical mixture. Although ΔHf was determined at a specific temperature, it is expected to be
insensitive to temperature change around that temperature. This is supported by the similarity in
the slope of relative enthalpies around 25 °C (Figure 2b). Note that both CEL and NIC used in
this study are the most stable polymorph at ambient conditions; ΔHf will change accordingly if a
different polymorph is used as the starting material.11
It is of interest to compare the ΔHf acquired from experimental and computational methods.
ΔHf is comparable to formation energy (ΔEf) for a typical NIC co-crystal under ambient
ACS Paragon Plus Environment

Page 11 of 35
Crystal Growth & Design
1
2
conditions.26 As in the case of CEL-NIC, the difference (PΔVf) is about 4 mJ/g at 1 bar (equation
4), which is negligible compared with the magnitude of ΔHf. Chan et al. reported the ΔEf of
CEL-NIC to be +0.36 kcal/mol, or +3.0 J/g, from lattice energy calculation.15 The optimized co-
crystal structure used in the calculation was the same as our single crystal structure. Therefore,
both analyses agreed that the formation process of CEL-NIC is endothermic, different from other
NIC co-crystals (34 co-crystals with calculated ΔEf data and 4 co-crystals with experimental ΔHf
data15,26). It is noteworthy that most experimentally analyzed co-crystals have negative ΔHf and
ΔGf, as we will discuss later.
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Determination of ΔGf from solubility
measurements. An endothermic ΔHf would
raise a question of whether the co-crystal
formation is thermodynamically preferred
(ΔGf < 0). We measured the solubility of
CEL, NIC, a physical mixture of 1:1 ratio, and
the co-crystal in chloroform at 25 °C (Table
3). We found the solubility of co-crystal is
slightly lower than the physical mixture,
indicating that the co-crystal is more stable.
On the other hand, the solubility of the
physical mixture is significantly higher than
individual components, showing the solubility
enhancement effect which had previously
Table 3. Solubility of CEL, NIC, 1:1 physical mixture,
and co-crystal in chloroform at 25 °C
Phase
CEL
NIC
CEL, mg/ml
32.2 (0.5)
-
NIC, mg/ml
-
6.6 (0.1)
1:1 physical
mixturea
Co-crystalb
64.4 (0.9)
55.5 (0.8)
14.2 (0.3)
13.6 (0.2)
aConcentration of CEL and NIC after 1 day of
dissolution when co-crystal was not yet formed in the
excess solids (Figure S1a)
bExcess solids contained co-crystal and NIC (Figure S1b)
Figure 3. Dissolution profiles of a 1:1 physical mixture
of CEL and NIC in chloroform at 25 °C. Equilibrium
solubility of pure CEL (32.2 mg/ml) or pure NIC (6.6
mg/ml) is indicated by dotted lines. Equilibrium
solubility of CEL-NIC co-crystal (CEL, 55.5 mg/ml;
NIC, 13.6 mg/ml) is indicated by dashed lines. CEL-NIC
co-crystal was found in the solid phase after 192 hour (8
day) of dissolution.
ACS Paragon Plus Environment

Crystal Growth & Design
Page 12 of 35
1
2
been reported for NIC.37 No phase change was observed through the experiment except for the
co-crystal, where excess NIC was formed along with the remaining co-crystal as a result of
incongruent dissolution (Figure S1b). ΔGf of the co-crystal was determined to be -0.7 (0.1) J/g at
25 °C (equation 5).
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The negative ΔGf suggests that at 25 °C, formation of the co-crystal is thermodynamically
favored. This was further examined by continuing agitation of the physical mixture in
chloroform at the same temperature. As shown in Figure 3, a slight decrease in concentration
towards the equilibrium solubility of the co-crystal was observed after the initial plateau. Phase
change in excess solids was also found during the experiment (Figure S1a). Co-crystal started to
form in the excess solids between 160-192 hours of agitation, while excess CEL disappeared at t
= 336 hours showing the co-crystallization was complete. The decrease of concentration
occurred simultaneously with the formation of co-crystal. Only the co-crystal and NIC were
found in the excess solids at the end of the experiment. The observation of spontaneous co-
crystallization supports that the co-crystal is thermodynamically stable at 25 °C.
Entropy-governed co-crystallization process. The relationship between ΔHf, ΔSf, and ΔGf is
reviewed in Table S1. The combination of a negative ΔGf and a positive ΔHf suggests that co-
crystallization is entropically favored and dominated by the -TΔSf term. In this case, temperature
plays an unusual role in co-crystal formation, as higher temperature would favor co-
crystallization. Assuming ΔHf and ΔSf do not change drastically around 25 °C, a transition
temperature (Tt) below 25 °C can be expected. At T < Tt, co-crystal will dissociate because the
physical mixture is thermodynamically stable (ΔGf > 0). Tt is estimated to be around 10 °C from
equation 6 based on the formation properties of CEL-NIC.
ACS Paragon Plus Environment

Page 13 of 35
Crystal Growth & Design
1
2
3
4
5
6
7
8
9
To verify this hypothesis, stability of CEL-NIC was examined in multiple conditions (Table
4). Excess amount of the co-crystal was agitated in chloroform/heptane 5/1 (v/v) in the range of
4 – 25 °C. The co-crystal was stable at 25 °C and 20 °C, but dissociated into anhydrous CEL and
NIC at 15 °C and 4 °C. Therefore, the existence of a Tt between 15 – 20 °C was demonstrated. Tt
is expected to be solvent independent if no solvates can be formed. Slurry competition between
the co-crystal and its physical mixture was carried out in saturated solvents of Toluene and
IPA/heptane 1/3 (v/v). In both solvents, co-crystal was stable at 20°C but not at 15 °C. It is
therefore concluded that the relative stability between CEL-NIC and its physical mixture
inverted between 15 °C and 20 °C, and the co-crystallization is controlled by entropy.
The formation of CEL-NIC is compared with other co-crystals with experimental formation
properties. It should be noticed that these properties are only available for a limited number of
co-crystals. As surveyed in Table 5, all co-crystals are thermodynamically favored. The CEL-
NIC co-crystal, bearing a marginal favorable ΔGf, is consistent with this trend. On the other
hand, most of co-crystals are enthalpically favored except for Mefenamic acid – 4,4'-bipyridine,
however the positive ΔHf is within the experimental error.38 Therefore, the positive ΔHf of CEL-
NIC is different from the other co-crystals.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table 4. Stability test of CEL-NIC in multiple conditions
Starting phase
Slurry condition
Ending phase
CEL-NIC co-crystala
CEL-NIC co-crystala
CEL-NIC co-crystala
CEL-NIC co-crystala
chloroform:heptane 5:1, 25 °C, 1 day
chloroform:heptane 5:1, 20 °C, 7 days
chloroform:heptane 5:1, 15 °C, 7 days
chloroform:heptane 5:1, 4 °C, 5 days
CEL-NIC, with excess NIC
CEL-NIC, with excess NIC
CEL and NIC
CEL and NIC
CEL-NIC
CEL and NIC
CEL-NIC
CEL and NIC
CEL-NIC co-crystal, CEL, NICb toluene, 20 °C, 1 day
CEL-NIC co-crystal, CEL, NICb toluene, 15 °C, 1 day
CEL-NIC co-crystal, CEL, NICb IPA:heptane 1:3, 20 °C, 1 day
CEL-NIC co-crystal, CEL, NICb IPA:heptane 1:3, 15 °C, 1 day
aFor experiments in chloroform:heptane, slurry was formed by adding excess amount of CEL-NIC into the solvent.
bFor experiments in toluene or IPA:heptane, saturated solvents with CEL and NIC were used.
ACS Paragon Plus Environment

Crystal Growth & Design
Page 14 of 35
1
2
3
4
5
6
7
8
9
Table 5. Formation properties of co-crystals available in the literature
Co-crystal
Ratio
1:1
1:1
1:1
1:1
1:1
1:1
2:1
T, °C
25
33
25
25
25
20
20
30
30
30
20
25
20
20
20
20
25
25
25
25
25
25
25
25
25
25
25
50
60
60
ΔGf, J/ga
ΔHf, J/ga
13.0 (3.0)
Reference
Celecoxib – nicotinamide
Carbamazepine – saccharin
Bicalutamide – benzamide
Bicalutamide – salicylamide
Carbamazepine – nicotinamide
Isonicotinamide – benzoic acid
Theophylline – oxalic acid
-0.7 (0.1)
This work
-10.5 (0.5)b -14.1 (2.1)
11
12
-6.2 (0.5)
-3.9 (0.5)
-35.5 (1.5)
-20.1 (0.9)
-15.3 (1.8)b n/a
-44.4 (0.7)b n/a
10
39
40
41
23
41
42
43
13
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
-13.5
-12.5
-1.2
n/a
n/a
c
Theophylline – glutaric acid
Theophylline – salicylic acid
Adefovir dipivoxil – saccharin
Proline – myricetin
1:1
1:1
1:1
2:1
1:1 form 1
1:1 form 2
4:1
1:2
1:2
1:1
1:1
1:1 form 1
1:1 form 2
2:1
c
-15.5
-18.9 (1.3)b n/a
-29.0
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
Nicotinamide – R-mandelic acid
-23.0 (3.0)
-18.0 (3.0)
-8.0 (3.0)
-20.0 (3.0)
-40.2 (3.2)
-10.1 (1.9)
-26.0 (1.6)
-9.6 (1.1)
-5.4 (0.9)
-4.2 (0.6)
-9.6 (2.9)
-6.4 (0.6)
-5.7 (0.9)
0.3 (1.3)
-11.7 (0.7)
-1.2
31
44
45
Nicotinamide – diflunisal
Theophylline – diclofenac
Theophylline – diflunisal
Felodipine – 4,4'-bipyridine
46
38
N-phenylanthranilic acid – 4,4'-bipyridine 2:1
Niflumic acid – 4,4'-bipyridine
Tolfenamic acid – 4,4'-bipyridine
Mefenamic acid – 4,4'-bipyridine
Flufenamic acid – 4,4'-bipyridine
Melatonin – pimelic acid
2:1
2:1
2:1
2:1
1:1
1:1 form 1
1:1 form 2
47
48
Vitamin D2 – vitamin D3
-19.1
-5.9
aExperimental error is only recorded when available in the reference unless otherwise noted.
bIf ΔGf was acquired from multiple experiments or solvent conditions, the average and standard deviation of all
reported values are recorded.
cIt was reported that ΔHf is 98.4 J/g for theophylline – glutaric acid and 275.3 J/g for theophylline – salicylic acid.41
The data are of doubtful reliability as they were derived from DSC thermograms with evidenced sample
decomposition.
ACS Paragon Plus Environment

Page 15 of 35
Crystal Growth & Design
1
2
3
4
It is challenging to rationalize why the co-
5
6
7
8
crystal is entropically favored from its structure.
However, we propose that the entropy gain is
related to the volume expansion upon co-
crystallization. The formation of CEL-NIC is
accompanied with one of the largest ΔVf among
all NIC co-crystals despite shorter hydrogen
bonds. As the hydrophilic part of CEL molecule
is reoriented to form tighter hydrogen bonds, its
hydrophobic part will have to pack loosely and
enjoy more freedom. In addition, the pyrazole
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 4. Packing diagrams of (a) CEL-NIC co-crystal
(22 °C structure) and (b) CEL crystal (CSD#: DIBBUL,
solved at room temperature). Disordered trifluoromethyl
group is observed only in the co-crystal structure.
part in the co-crystal is not hydrogen-bonded (Figure 4). We used the program Mercury (version
3.9) to analyze the occupied volume of part of CEL in CEL-NIC and in pure celecoxib (CSD#:
DIBBUL34). In CEL-NIC, each benzyl sulfonamide group occupied 165 Å3 of space, slightly
smaller than that in pure celecoxib (175 Å3). The reduced mobility of the hydrophilic part is in
agreement with tighter hydrogen bonds. The rest of the CEL molecule, however, occupied a
space of 275 Å3 in the co-crystal, 42 Å3 (18 %) larger than that in pure celecoxib. The increased
mobility in the hydrophobic part of CEL is also evidenced by the observation of trifluoromethyl
disordering in the co-crystal structure but not in pure celecoxib. It’s of interest to further
investigate the ΔSf of CEL-NIC and other co-crystals with a large volume expansion upon
formation.
The fact that co-crystallization can be controlled by entropy may affect the current practices
in co-crystal screening and synthesis. First of all, lattice energy calculation has been widely used
ACS Paragon Plus Environment

Crystal Growth & Design
Page 16 of 35
1
2
3
4
5
6
7
8
9
in virtual co-crystal screening and structure prediction, with the hypothesis that formation energy
(ΔEf, ≈ ΔHf) dominates the co-crystal formation.15,17 However, this work indicates that pairs of
components with unfavorable ΔEf should not be automatically eliminated from further
consideration, especially when the energy difference is marginal. Wet-lab screening should be
applied to those cases for verification. On the other hand, consistent efforts are needed for the
development of ΔGf-based computational methods to better evaluate co-crystal formation, as ΔEf
term is found to be inconclusive and insufficient in the study of CEL-NIC.15
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Secondly, thermodynamic properties are vital in the selection of crystallization condition as
the crystallization outcome is affected. Direct crystallization from super-saturated solution and
slurry conversion are widely used in the isolation of organic solids. To induce super-saturation
and to increase the yield, temperature reduction and anti-solvent addition are preferred over
solvent evaporation at the industrial scale. However, the existence of a Tt below the co-crystal
screening temperature – which is usually room temperature – will bring uncertainty to the
cooling crystallization outcomes. This is similar to the existence of enantiotropically related
polymorphs narrows down the operational temperature range for acquiring a particular form.
Therefore, anti-solvent addition may be a better process for entropically favored co-crystals, as
the formation properties of a co-crystal are independent of solvent, as long as no solvates of
components or co-crystal can be formed. It is therefore suggested that the stability or solubility
of a co-crystal should be thoroughly studied at various temperatures before designing the
synthetic process.12
This work also highlights that ΔGf can be acquired from a single dissolution experiment with
excess amount of co-crystal components (Figure 3). Dissolution will first produce a solution
saturated with both components, which is also suitable for co-crystal screening.49,50 Should the
ACS Paragon Plus Environment

Page 17 of 35
Crystal Growth & Design
1
2
3
4
5
6
7
8
9
co-crystal be more stable than the components, the concentration of both components will
decrease as co-crystal forms. The transition can be accelerated by seeding and monitored with
online methods to capture the concentration change.51 The decrease in concentration should not
be confused with the “parachute-like” profile which is commonly observed for co-crystal
dissolution due to recrystallization.50 The “parachute-like” dissolution could only be observed for
thermodynamically unstable phases or for a stable co-crystal in an environment that one
component (usually the more soluble one – the co-former) is undersaturated.12 A stable co-
crystal will not dissociate into a physical mixture in non-solvating solvents or in the solid state.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
CONCLUSION
This study revisited the co-crystal containing celecoxib and nicotinamide and identified its
remarkable formation process which is driven by entropy. A single crystal structure is reported,
which agrees with the structural optimization of Chan et al.15 and allows the comparison between
experimental and computational investigations. The ΔHf of the co-crystal was determined from
its melting enthalpy and that of the physical mixture of component crystals. The ΔGf of the co-
crystal was determined from solubility data. We find that the formation of CEL-NIC co-crystal
increases enthalpy, in agreement with computational simulation but distinct from other NIC co-
crystals and co-crystals with reported formation enthalpies. This peculiarity is further supported
by the fact that co-crystal formation is thermodynamically favorable at 25 °C, resulting in the
conclusion that the co-crystallization is entropically driven. The existence of a transition
temperature (Tt) below 25 °C which inverts the stability between the co-crystal and physical
mixture was predicted by the thermodynamic formation properties and confirmed
experimentally, further supporting this conclusion.
ACS Paragon Plus Environment

Crystal Growth & Design
Page 18 of 35
1
2
3
4
5
6
7
8
9
The entropy-driven formation process of CEL-NIC not only enriches our understanding on
the mechanism of co-crystallization but also brings new opportunities into co-crystal research
and development. It challenges the usage of ΔEf or ΔHf as a primary indicator of co-crystal
formation during computational simulation, and encourages new methods or algorithms with
incorporation of the entropy effect. It questions the usage of temperature reduction during
process development as a way to form co-crystals, as the co-crystal phase may not be favored at
lower temperature. It is suggested that the thermodynamic properties of a co-crystal should be
thoroughly studied at process temperatures to avoid unexpected outcomes. This study
emphasizes the importance of thermodynamic studies as a valuable complement to structural and
kinetic investigations in understanding the formation and stability of co-crystals.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ASSOCIATED CONTENT
Supporting Information
XRPD patterns of excess solids during the dissolution of physical mixture or co-crystal; the
correlation between ΔGf, ΔHf, and ΔSf; calculation method of the change of hydrogen-bond
length upon co-crystallization (ΔRHB) for CEL-NIC co-crystal; crystallographic data and
structure refinement for CEL-NIC co-crystal (100K and 293K)
Accession Codes
CCDC 1532378 and 1532379 contain the supplementary crystallographic 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.
ACS Paragon Plus Environment

Page 19 of 35
Crystal Growth & Design
1
2
3
4
5
6
AUTHOR INFORMATION
Corresponding Author
7
8
9
*E-mail: siwei.zhang@merck.com. Phone: +1 (732) 594 9127.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ORCID
Siwei Zhang: 0000-0001-8677-3722
Notes
The authors declare no competing financial interest.
ACKNOWLEDGEMENTS
S.-W. Z. thanks C. Scott Shultz for providing the optimized synthetic route of CEL-NIC co-
crystal. S-W. Z. thanks Alfred Lee and Narayan Variankaval for helpful discussions.
ACS Paragon Plus Environment

Crystal Growth & Design
Page 20 of 35
1
2
3
4
REFERENCES
5
6
7
8
1.
2.
Aakeroy, C. B.; Forbes, S.; Desper, J. J. Am. Chem. Soc. 2009, 131, 17048-17049.
Karki, S.; Friscic, T.; Fabian, L.; Laity, P. R.; Day, G. M.; Jones, W. Adv. Mater. 2009,
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
21, 3905-3909.
3.
Ferretti, V.; Dalpiaz, A.; Bertolasi, V.; Ferraro, L.; Beggiato, S.; Spizzo, F.; Spisni, E.;
Pavan, B. Mol. Pharmaceut. 2015, 12, 1501-1511.
4.
5.
6.
Li, Z.; Matzger, A. J. Mol. Pharmaceut. 2016, 13, 990-995.
Bolton, O.; Matzger, A. J. Angew. Chem. Int. Edit. 2011, 50, 8960-8963.
Childs, S. L.; Chyall, L. J.; Dunlap, J. T.; Smolenskaya, V. N.; Stahly, B. C.; Stahly, G.
P. J. Am. Chem. Soc. 2004, 126, 13335-13342.
7. McNamara, D. P.; Childs, S. L.; Giordano, J.; Iarriccio, A.; Cassidy, J.; Shet, M. S.;
Mannion, R.; O'Donnell, E.; Park, A. Pharm. Res. 2006, 23, 1888-1897.
8.
9.
Cherukuvada, S.; Kaur, R.; Row, T. N. G. CrystEngComm 2016, 18, 8528-8555.
Li, Z. B.; Yang, B. S.; Jiang, M.; Eriksson, M.; Spinelli, E.; Yee, N.; Senanayake, C. Org.
Process Res. Dev. 2009, 13, 1307-1314.
10.
11.
12.
Schartman, R. R. Int. J. Pharmaceut. 2009, 365, 77-80.
Oliveira, M. A.; Peterson, M. L.; Davey, R. J. Cryst. Growth Des. 2011, 11, 449-457.
Surov, A. O.; Solanko, K. A.; Bond, A. D.; Bauer-Brandl, A.; Perlovich, G. L.
CrystEngComm 2016, 18, 4818-4829.
13. Zhang, S.-W.; Guzei, I. A.; de Villiers, M. M.; Yu, L.; Krzyzaniak, J. F. Cryst. Growth
Des. 2012, 12, 4090-4097.
14. Perlovich, G. L. CrystEngComm 2015, 17, 7019-7028.
ACS Paragon Plus Environment

Page 21 of 35
Crystal Growth & Design
1
2
3
4
15.
Chan, H. C. S.; Kendrick, J.; Neumann, M. A.; Leusen, F. J. J. CrystEngComm 2013, 15,
5
6
3799-3807.
7
8
9
16.
Chan, H. C. S.; Woollam, G. R.; Wagner, T.; Schmidt, M. U.; Lewis, R. A.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
CrystEngComm 2014, 16, 4365-4368.
17.
18.
Abramov, Y. A.; Loschen, C.; Klamt, A. J. Pharm. Sci. 2012, 101, 3687-3697.
Berry, D. J.; Seaton, C. C.; Clegg, W.; Harrington, R. W.; Coles, S. J.; Horton, P. N.;
Hursthouse, M. B.; Storey, R.; Jones, W.; Friscic, T.; Blagden, N. Cryst. Growth Des. 2008, 8,
1697-1712.
19.
20.
Cherukuvada, S.; Row, T. N. G. Cryst. Growth Des. 2014, 14, 4187-4198.
Xu, J.; Huang, Y.; Ruan, S.; Chi, Z.; Qin, K.; Cai, B.; Cai, T. CrystEngComm 2016, 18,
8776-8786.
21.
Jayasankar, A.; Reddy, L. S.; Bethune, S. J.; Rodriguez-Hornedo, N. Cryst. Growth Des.
2009, 9, 889-897.
22.
273.
23.
24.
25.
Guo, K.; Sadiq, G.; Seaton, C.; Davey, R.; Yin, Q. X. Cryst. Growth Des. 2010, 10, 268-
Zhang, S.; Rasmuson, A. C. Cryst. Growth Des. 2013, 13, 1153-1161.
Price, S. L. Chem. Soc. Rev. 2014, 43, 2098-2111.
Braun, D. E.; Ardid-Candel, M.; D’Oria, E.; Karamertzanis, P. G.; Arlin, J.-B.; Florence,
A. J.; Jones, A. G.; Price, S. L. Cryst. Growth Des. 2011, 11, 5659-5669.
26.
27.
28.
Zhang, S. W.; Kendrick, J.; Leusen, F. J. J.; Yu, L. J. Pharm. Sci. 2014, 103, 2896-2903.
Seaton, C. C.; Parkin, A. Cryst. Growth Des. 2011, 11, 1502-1511.
Remenar, J. F.; Peterson, M. L.; Stephens, P. W.; Zhang, Z.; Zimenkov, Y.; Hickey, M.
B. Mol. Pharmaceut. 2007, 4, 386-400.
ACS Paragon Plus Environment

Crystal Growth & Design
Page 22 of 35
1
2
3
4
29.
Lu, G. W.; Hawley, M.; Smith, M.; Geiger, B. M.; Pfund, W. J. Pharm. Sci. 2006, 95,
5
6
305-317.
7
8
9
30.
31.
Li, J. J.; Bourne, S. A.; Caira, M. R. Chem. Commun. 2011, 47, 1530-1532.
Zhang, S.-W.; Harasimowicz, M. T.; de Villiers, M. M.; Yu, L. J. Am. Chem. Soc. 2013,
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
135, 18981-18989.
32.
Etter, M. C.; Macdonald, J. C.; Bernstein, J. Acta Crystallogr. B-Struct. Sci. 1990, 46,
256-262.
33.
Bernstein, J.; Davis, R. E.; Shimoni, L.; Chang, N. L. Angew. Chem. Int. Edit. 1995, 34,
1555-1573.
34.
Vasu Dev, R.; Shashi Rekha, K.; Vyas, K.; Mohanti, S. B.; Rajender Kumar, P.; Om
Reddy, G. Acta Crystallogr. C 1999, 55, IUC9900161.
35.
Miwa, Y.; Mizuno, T.; Tsuchida, K.; Taga, T.; Iwata, Y. Acta Crystallogr. B 1999, 55,
78-84.
36.
Chawla, G.; Gupta, P.; Thilagavathi, R.; Chakraborti, A. K.; Bansal, A. K. Eur. J. Pharm.
Sci. 2003, 20, 305-317.
37.
38.
Handa, Y. P.; Benson, G. C. Fluid Phase Equilibr. 1979, 3, 185-249.
Surov, A. O.; Simagina, A. A.; Manin, N. G.; Kuzmina, L. G.; Churakov, A. V.;
Perlovich, G. L. Cryst. Growth Des. 2015, 15, 228-238.
39. Boyd, S.; Back, K.; Chadwick, K.; Davey, R. J.; Seaton, C. C. J. Pharm. Sci. 2010, 99,
3779-3786.
40.
41.
Zhang, S.; Rasmuson, A. C. CrystEngComm 2012, 14, 4644-4655.
Zhang, S.; Chen, H.; Rasmuson, A. C. CrystEngComm 2015, 17, 4125-4135.
ACS Paragon Plus Environment

Page 23 of 35
Crystal Growth & Design
1
2
3
4
42.
Ma, K.; Zhang, Y.; Kan, H. L.; Cheng, L. F.; Luo, L.; Su, Q.; Gao, J.; Gao, Y.; Zhang, J.
5
6
7
8
J. Pharm. Res. 2014, 31, 1766-1778.
43. Liu, M. Y.; Hong, C.; Yao, Y. S.; Shen, H. Y.; Ji, G.; Li, G. W.; Xie, Y. Eur. J. Pharm.
Biopharm. 2016, 107, 151-159.
44. Evora, A. O. L.; Castro, R. A. E.; Maria, T. M. R.; Silva, M. R.; ter Horst, J. H.;
Canotilho, J.; Eusebio, M. E. S. Int. J. Pharmaceut. 2014, 466, 68-75.
45. Surov, A. O.; Voronin, A. P.; Manin, A. N.; Manin, N. G.; Kuzmina, L. G.; Churakov, A.
V.; Perlovich, G. L. Mol. Pharmaceut. 2014, 11, 3707-3715.
46. Surov, A. O.; Solanko, K. A.; Bond, A. D.; Bauer-Brandl, A.; Perlovich, G. L.
CrystEngComm 2014, 16, 6603-6611.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
47.
48.
49.
Yan, Y.; Chen, J. M.; Lu, T. B. CrystEngComm 2015, 17, 612-620.
Wang, J. R.; Yu, Q. H.; Dai, W. J.; Mei, X. F. Chem. Commun. 2016, 52, 3572-3575.
Nehm, S. J.; Rodriguez-Spong, B.; Rodriguez-Hornedo, N. Cryst. Growth Des. 2006, 6,
592-600.
50.
51.
Sun, C. C. Expert Opin. Drug Del. 2013, 10, 201-213.
Sarraguca, M. C.; Ribeiro, P. R. S.; Dos Santos, A. O.; Lopes, J. A. J. Pharm. Sci. 2015,
104, 4099-4108.
ACS Paragon Plus Environment

Crystal Growth & Design
Page 24 of 35
1
2
3
4
For Table of Contents Use Only
5
6
7
8
Celecoxib-Nicotinamide Co-Crystal Revisited: Can
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Entropy Control Co-Crystal Formation?
Si-Wei Zhang, Andrew P. J. Brunskill, Eric Schwartz, Shuwen Sun
The co-crystal of celecoxib-nicotinamide has been revisited to reveal its crystal structure and
unusual formation properties. Solubility and thermal analyses revealed that the formation of
celecoxib-nicotinamide is endothermic and driven by entropy, a mechanism different from many
reported co-crystals. The co-crystal is stable only above a transition temperature as predicted
from thermodynamic data and confirmed experimentally.
ACS Paragon Plus Environment

Page 25 of 35
Crystal Growth & Design
1
2
3
4
Scheme 1. Structures of celecoxib (CEL) and nicotinamide (NIC)
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment

Crystal Growth & Design
Page 26 of 35
1
2
3
4
Figure 1. Hydrogen bonds in CEL-NIC co-crystal demonstrated by the 22 °C structure.
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment

Page 27 of 35
Crystal Growth & Design
1
2
3
4
5
6
Figure 2. (a) DSC melting endotherms of CEL-NIC co-crystal, CEL, NIC, and a physical mixture of CEL and NIC
at 1:1 molar ratio. (b) Relative enthalpies between co-crystal and the physical mixture. ΔHf is the formation enthalpy
of co-crystal (+13(3) J/g at 25 °C).
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment

Crystal Growth & Design
Page 28 of 35
1
2
3
4
5
6
Figure 3. Dissolution profiles of a 1:1 physical mixture of CEL and NIC in chloroform at 25 °C. Equilibrium
solubility of pure CEL (32.2 mg/ml) or pure NIC (6.6 mg/ml) is indicated by dotted lines. Equilibrium solubility of
CEL-NIC co-crystal (CEL, 55.5 mg/ml; NIC, 13.6 mg/ml) is indicated by dashed lines. CEL-NIC co-crystal was
found in the solid phase after 192 hour (8 day) of dissolution.
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment

Page 29 of 35
Crystal Growth & Design
1
2
3
4
5
Figure 4. Packing diagrams of (a) CEL-NIC co-crystal (22 °C structure) and (b) CEL crystal (CSD#: DIBBUL,
solved at room temperature). Disordered trifluoromethyl group is observed only in the co-crystal structure.
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment