Synthon hierarchies in crystal forms composed of theophylline and hydroxybenzoic acids: Cocrystal screening via solution-mediated phase transformation

Article abstract of DOI:10.1021/cg501204k

Theophylline, a pharmaceutical model compound, was screened for cocrystal formation using nine (di)hydroxybenzoic acids as cocrystal formers, namely, 2-, 3-, 4-hydroxybenzoic- and 2,3-, 2,4-, 2,5-, 2,6-, 3,4-, and 3,5-dihydroxybenzoic acids. The experiments were conducted using an efficient, reliable, and fast screening method based on thermodynamically driven solution-mediated phase transformation and revealed the formation of eight cocrystals and one salt. The theophylline-based solids were identified using powder and structurally characterized via single-crystal X-ray diffraction. The obtained solids are sustained by theopylline-carboxylic acid heterosynthons, in addition to both theophylline and carboxylic acid homosynthons. The results are a part of a systematic study of xanthine-alkaloid cocrystals that aims to establish synthon hierarchies in cocrystals comprised of molecules with multiple hydrogen-bonding functional groups. Structural features and synthon hierachies are discussed and compared to those in a series of published caffeine cocrystals.

Full text of DOI:10.1021/cg501204k

Subscriber access provided by BRIGHAM YOUNG UNIV
Article
Synthon Hierarchies in Crystal Forms Composed of
Theophylline and Hydroxybenzoic acids: Cocrystal
Screening via Solution-mediated Phase Transformation
Dejan-Krešimir Bu#ar, R Henry, G Zhang, and Leonard R. MacGillivray
Cryst. Growth Des., Just Accepted Manuscript • DOI: 10.1021/cg501204k • Publication Date (Web): 19 Aug 2014
Downloaded from http://pubs.acs.org on August 23, 2014
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 13
Crystal Growth & Design
1
2
3
4
5
6
7
8
9
Synthon Hierarchies in Crystal Forms Composed of
Theophylline and Hydroxybenzoic acids: Cocrystal Screening
via Solution-mediated Phase Transformation§
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
DejanꢀKrešimir Bučar,a† Rodger F. Henry,b Geoff G. Z. Zhang,b* and Leonard R. MacGillivray,a*
a Department of Chemistry, University of Iowa, IA 52242, USA
b Drug Product Development, Research and Development, AbbVie Inc., North Chicago, IL 60064, USA
KEYWORDS. supramolecular synthons, synthon hierarchies, selfꢀassembly, crystal engineering, pharmaceutical cocrystals,
pharmaceutical salts, cocrystal screening, solutionꢀmediated phase transformation, theophylline.
ABSTRACT: Theophylline, a pharmaceutical model compound, was screened for cocrystal formation using nine
(di)hydroxybenzoic acids as cocrystal formers; namely 2ꢀ, 3ꢀ, 4ꢀhydroxybenzoicꢀ and 2,3ꢀ, 2,4ꢀ, 2,5ꢀ, 2,6ꢀ 3,4ꢀ and 3,5ꢀ
dihydroxybenzoic acids. The experiments were conducted using an efficient, reliable and fast screening method based on
thermodynamicallyꢀdriven solutionꢀmediated phase transformation, and revealed the formation of eight cocrystals and one
salt. The theophyllineꢀbased solids were identified using powder and structurallyꢀcharacterized via singleꢀcrystal Xꢀray
diffraction. The obtained solids are sustained by theopyllineꢀcarboxylic acid heterosynthons, in addition to both theophylline
and carboxylic acid homosynthons. The results are a part of a systematic study of xanthineꢀalkaloid cocrystals that aims to
establish synthon hierarchies in cocrystals comprised of molecules with multiple hydrogenꢀbonding functional groups.
Structural features and synthon hierachies are discussed and compared to those in a series of published caffeine cocrystals.
1. INTRODUCTION
and additives as being safe for human consumption, whereof a
major portion is suitable for crystal engineering of
pharmaceutical solids.18 Many studies have now demonstrated
that the cocrystallization of PAs can afford solids that exhibit
physicochemical properties (e.g. solubility/dissolution rate,19ꢀ21
The lack of required physicochemical properties for drug
candidates (e.g. solubility/dissolution rate, mechanical
properties, stability) represents a serious problem in drug
development.1 To boost property performance, drug candidates
are subjected to saltꢀ and polymorph screening2ꢀ4 to identify
solids with improved physicochemical properties. However,
the saltꢀ and polymorph approaches are somewhat limited in
scope.5ꢀ7 Specifically, salt screening can only be performed on
pharmaceutical agents (PAs) that are considered to be
constituted of fullyꢀionizable functional groups. In addition,
only a relatively few number of salt formers (i.e. ~120 acids
and bases) are recognized as safe for use in drug formulation.8
Polymorph formation, another common approach to enhance
physicochemical properties of drugs, also exhibits serious
bioavailability,22 hygroscopicity,7, 23 mechanical properties24ꢀ26
superior to those of the parent PA.27ꢀ28
)
Herein, we report eight cocrystals of theophylline (theo) based
on 2ꢀhydroxyꢀ (2HBA), 3ꢀhydroxyꢀ (3HBA), 4ꢀhydroxyꢀ
(4HBA), 2,3ꢀdihydroxyꢀ (23DBA), 2,4ꢀdihydroxyꢀ (24DBA),
2,5ꢀdihydroxyꢀ (25DBA), 3,4ꢀdihydroxyꢀ (34DBA) and 3,5ꢀ
dihydroxybenzoic acid (35DBA), as well as a monohydrate salt
involving the carboxylate anion of 2,6ꢀdihydroxybenzoic acid
(26DBA) (Scheme 1). All multiꢀcomponent solids were
discovered using solutionꢀmediated phase transformation
(SMPT) – an efficient, reliable and fast technique suitable for
cocrystal screening.29ꢀ37 The results are a part of extensive and
ongoing structural studies of cocrystals of xanthines comprised
of multiple supramolecular synthons31 that aim to establish
hierarchies of supramolecular synthons in molecular
cocrystals.31, 38ꢀ44
limitations. In particular, polymorph formation is
a
serendipitous event,9ꢀ10 and can, nowadays, in many cases be
neither predicted11 nor designed. Polymorphs are also known to
spontaneously undergo phase transformations12ꢀ13 accompanied
by undesired and unacceptable changes of physical
properties.14ꢀ15 With these limitations in mind, drug formulators
have recently turned to crystal engineering and
cocrystallization.16ꢀ17 Interests in crystal engineering stem from
opportunities to form molecular complexes involving PAs that
do not contain ionizable functional groups.5ꢀ6 The cocrystal
approach also provides drug formulators with more synthetic
possibilities. Specifically, the US Food and Drug
Administration recognizes several thousand food ingredients
Supramolecular synthons
The ability to fully control and predict the formation of
multiple supramolecular synthons in cocrystals (and thereby
physicochemical properties) remains limited owing to a lack of
knowledge on hierarchies of supramolecular synthons in
45
organic solids.38ꢀ40,
An understanding of synthon
ACS Paragon Plus Environment

Crystal Growth & Design
Page 2 of 13
1
2
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
Scheme 1. Theo and (di)hydroxybenzoic acids.
hierarchies in cocrystals is expected to provide more control in
the design of pharmaceutical solids with desired properties, and
facilitate strategies to develop difficultꢀtoꢀobtain ternary46 and
quaternary47 cocrystals (i.e. cocrystals of more than two
components). An ability to design cocrystals based on more
than two components can also be expected to enable the
eventual development of solids that function as drug cocktails
with advanced performance.48ꢀ55
allow theo to participate in selfꢀassembly processes more
diverse and likely more difficult to predict than those involving
caf. A survey of the Cambridge Structural Database57 (CSD)
reveals theo to selfꢀassemble with hydroxylated carboxylic
acids leading to supramolecular synthons other than the O–
H(carboxyl)····N(imidazole), O–H(hydroxyl)···O=C(urea) and
O–H(hydroxyl)···O=C(amide) synthon, as seen in caf. The
variety of interactions observed in cocrystals composed of theo
and hydroxylated carboxylic acids is outlined in Scheme 2. The
CSD survey also revealed theo molecules to interact with each
other via cyclic arrays of N–H(imidazole)···O=C(amide)
hydrogen bonds, cyclic arrays of N–H(imidazole)···O=C(urea)
and C–H(amide)···O=C(urea) hydrogen bonds, and N–
H(imidazole)···O=C(amide) hydrogen bonds. Lastly, hydroxyl
and carboxyl groups of the CCFs form cyclic arrays of Oꢀ
Xanthineꢀbased coꢀcrystals
To study the hierarchy of supramolecular synthons in
cocrystals composed of multiple synthons, we selected a series
of xanthine alkaloids involving caffeine (caf), theophylline
(theo), theobromine, and related compounds. Xanthine
alkaloids are a class of common pharmaceutical model
compounds that comprise a purine ring derivatized with Nꢀ
methyl groups, whereby the position of the Nꢀmethyl groups
defines the pharmacological profile of each alkaloid.56 The
purine moiety of xanthines involves structural fragments that
commonly occur in PAs. The CCFs that we describe are
regioisomers of hydroxyꢀ and dihydroxybenzoic acids. The
CCFs, hence, provide an opportunity to study the impact of
structural isomerism on selfꢀassembly and synthon hierarchies.
Importantly, the CCFs are comprised of functional groups
common in PAs, as well as in both CCFs and salt formers
generally used to alter physicochemical properties of PAs.
H(hydroxyl)···O–H(hydroxyl)···O=C(carboxyl)
hydrogen
bonds, as well as discrete O–H(hydroxyl)···O=C(carboxyl)
hydrogen bonds.
We show here that the O–H(carboxyl)···N(imidazole)
heterosynthon (i.e. synthon I, Scheme 2) is not as robust as in
the related caf cocrystal series.31 While caf invariably formed
cocrystals with (di)hydroxybenzoic acids through the expected
synthon I, theo less predictably interacts with the CCF
carboxyl group either via synthon I, or a cyclic array of O–
H(carboxyl)···O=C(amide)
and
N–H(imidazole)···
O=C(carboxyl) hydrogen bonds (i.e. synthon II, Scheme 2).
The established synthon hierarchy could not be correlated with
the strengths in acidy (viz. pKa value) of the carboxylic acid
CCFs. In addition to synthons I and II, the theoꢀbased solids
In earlier work, we reported a series of caf cocrystals involving
(di)hydroxybenzoic acids and established that: a) the imidazole
unit of caffeine predominantly participates in an O–
H(carboxyl)···N(imidazole) synthon with the carboxyl group of
the carboxylic acid, b) both urea and amide moieties participate
in OꢀH···O synthons with the phenolic moiety of the acid, and
c) the formation of the O–H(hydroxyl)···O(urea) synthon is
favored over the O–H(hydroxyl)···O(amide) synthon.31
Although caf and theo are structurally similar from a molecular
perspective, it was expected that theo would participate in
different supramolecular structures than caf. Specifically, theo
exhibits one additional NꢀH(imidazole) hydrogenꢀbondꢀ
donating group positioned in close proximity to the
C=O(amide) hydrogenꢀbondꢀaccepting moiety. Moreover, the
additional hydrogenꢀbondꢀdonating group was expected to
exhibit
O–H(hydroxyl,CCF)···O=C(urea,theo)
and
O–
H(hydroxyl,CCF)···O=C(amide,theo)
interactions
(i.e.
synthons VII and VIII, Scheme 2), which are also present in the
related caf cocrystals. The hierarchy of the hydroxylꢀgroupꢀ
based synthons in theo is consistent with the one found in the
caf cocrystal series, wherein synthon VIII occurs more
frequently then synthon VII. Finally, we highlight differences
in the theo and caf cocrystals that concern the dimensionality
of the supramolecular assemblies, the propensity for hydrate
formation, as well as the involvement of the CCF orthoꢀ
hydroxyl groups in intermolecular interactions.
ACS Paragon Plus Environment

Page 3 of 13
Crystal Growth & Design
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
Scheme 2. Hydrogenꢀbondꢀaccepting and ꢀdonating sites of theo (grey – imidazole, orange – amide, blue – urea) (upper row, on the
very left). The following structures depict supramolecular synthons being observed in cocrystals composed of theo and
hydroxylated carboxylic acids: I) O–H(carboxyl)···N(imidazole) heterosynthon, II) cyclic array of O–H(carboxyl)···O=C(amide)
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
and N–H(imidazole)···O=C(carboxyl) hydrogen bonds, III) O–H(carboxyl)···O(amide) heterosynthon, IV)
C=O(carboxyl)····N–H(imidazole)···O–H(hydroxyl) hydrogen bonds, V) N–H(imidazole)···O(hydroxyl,
bifurcated
VI) O–
H(hydroxyl)···N(imidazole) heterosynthon, VII) O–H(hydroxyl)···O(amide) heterosynthons, VIII) O–H(hydroxyl)···O(urea)
heterosynthons, IX) cyclic arrays of N–H(imidazole)···O=C(urea) and C–H(amide)···O=C(urea) hydrogen bonds, X) N–
H(imidazole)···O=C(amide) hydrogen bonds, XI) N–H(imidazole)···O=C(urea) hydrogen bonds, XII) cyclic arrays of O–
H(hydroxyl)···O–H(hydroxyl)···O=C(carboxyl) hydrogen bonds and XIII) O–H(hydroxyl)···O=C(carboxyl) hydrogen bonds. The
theo solids described in this study are all based on at least one of the following synthons: I, II, VI, VII, IX and XIII.
2. EXPERIMENTAL SECTION
or dihydroxybenzoic acid, and a low volume of acetonitrile (1ꢀ
4 mL). The obtained suspension was sonicated59 to facilitate the
SMPT process. A phase change was visually evident within
few minutes or, in some instances, after only a few seconds. To
ensure complete conversions of theo and the CCF into the
cocrystal, the slurries were equilibrated for a week at ambient
conditions. The suspensions were filtered, and the residual
solids were investigated via powder Xꢀray diffraction (PXRD).
The crystalline phases were further studied via IR spectroscopy
and single crystal Xꢀray diffraction.
2.1. Materials
Theo (anhydrous, minimum 99%), 2HBA (≥99.0% ), 3HBA
(99.0%) , 4HBA (≥99.0%), 23DBA (≥97.0%), 24DBA
(≥98.0%), 25DBA (≥98.0%), 26DBA (98.0%), 34DBA
(≥97.0%), 35DBA (≥97.0%) and acetonitrile (anhydrous,
99.8%) were purchased from SigmaꢀAldrich (St. Louis, MO,
USA) and were used as received. Millex® syringe filters
(PVDF, 0.2 ꢁm, 13 mm) were purchased from Millipore
(Bedford, MA, USA)
2.4. Single crystal preparation
2.2. Survey of Cambridge Structural Database
Single crystals of (theo)·(2HBA) (1:1 cocrystal),
(theo)·(3HBA) (1:1 cocrystal), (theo)·(4HBA) (1:1 cocrystal),
(theo)·(23DBA) (1:1 cocrystal), (theo)·(24DBA) (1:1
cocrystal), (theo)·(25DBA) (1:1 cocrystal), (theo)+·(26DBA)ꢀ
·(H2O) (1:1:1 salt), (theo)·(34DBA) (1:1 cocrystal) and
(theo)·(35DBA) (1:1 cocrystal) were obtained by slow
evaporation from solution. To obtain single crystals of the
theoꢀbased solids, 20 mg of the corresponding polycrystalline
solids, obtained during the screening procedure, were dissolved
in acetonitrile (4 ml) at 80 ºC, filtered through a Millex®
syringe filters (PVDF, 0.2 ꢁm, 13 mm), and left to evaporate at
ambient conditions. Crystals suitable for single crystal Xꢀray
diffraction of all solids were obtained from the filtrate via slow
evaporation within two weeks.
The CSD (version 5.34, update of November 2012) was
searched for entries involving theo using the following filters:
“3D coordinates determined” and “only organics”. The search
was conducted with the help of the program ConQuest58
(version 1.12). The survey revealed 51 entries related to theoꢀ
containing structures, whereof 25 involve theo and
a
carboxylic acid (i.e. CIZTAH, CODCOO, CSATEO,
DEYREF, DUCROJ, IJIBEJ, IJIBEJ01, KIGKAN, KIGLAO,
KIGLES, KIGLIW, KIGLOC, LUKXUL, LEDVOH,
NEXWOD, NEYCIE, UNITER, UNITIV, VAXSAQ,
WUYROX, WUYRUD, XEJWUF, XEJXAM, XEJXEQ,
XEJXIU). A total of 14 structures involved theo and a
hydroxylated carboxylic acid; namely, CIZTAH, CODCOO,
CSATEO, DEYREF, DUCROJ, IJIBEJ, IJIBEJ01, KIGKAN,
KIGLES, KIGLIW, KIGLOC, NEXWOD, NEYCIE,
WUYRUD).
2.5. Powder Xꢀray diffractometry (PXRD)
The diffractograms were obtained using an Inel G3000
diffractometer equipped with a curved position sensitive
detector and parallel beam optics. The instrument was operated
with a copperꢀanode tube (1.5 kW, fine focus) at 40 kV and
30 mA. An incidentꢀbeam germanium monochromator was
utilized to obtain monochromatic Kα1 irradiation. The
diffractometer was calibrated using the attenuated direct beam
2.3. Cocrystal screening
To investigate cocrystal formation of theo with hydroxyꢀ and
dihydroxybenzoic acids, we utilized a SMPTꢀbased suspension
screening method.29 In a typical screening experiment, theo
(1 mmol) was mixed with an equimolar amount of a hydroxyꢀ
ACS Paragon Plus Environment

Crystal Growth & Design
Page 4 of 13
Table 1. General and crystallographic data for the reported theo solid forms.
1
2
3
4
(theo)·(3HBA)
(theo)·(23DBA)
(theo)·(24DBA)
molecular formula
Mr
(C7H8N4O2)ꢂ(C7H6O3)
318.29
(C7H8N4O2)ꢂ(C7H6O4)
334.29
(C7H8N4O2)ꢂ(C7H6O4)
334.29
monoclinic
P21/c
crystal system
space group
triclinic
Pī
monoclinic
P21/c
5
a, Å
b, Å
c, Å
α, °
7.683(2)
8.191(2)
23.307(6)
98.202(5)
92.589(5)
103.123(4)
1409.2(7)
4
1.500
664
0.117
170(2)
0.40 × 0.30 × 0.05
ꢀ9 → 9
ꢀ10 → 10
ꢀ30 → 30
15917
6377
0.034
5210
435
15.327(7)
6.792(3)
14.041(6)
90.00
106.297(4)
90.00
1403.1(10)
4
1.583
696
0.126
100(2)
0.40 × 0.10 × 0.10
ꢀ19 → 19
ꢀ6 → 8
ꢀ18 → 18
10225
3120
0.055
1688
219
0.096
0.131
0.122
0.163
12.565(4)
7.894(3)
14.528(5)
90.00
98.840(6)
90.00
1423.8(9)
4
1.560
696
0.124
170(2)
0.40 × 0.40 × 0.25
ꢀ16 → 16
ꢀ10 → 10
ꢀ19 → 19
16002
3478
0.063
3244
219
0.043
0.045
0.121
0.123
6
7
8
9
β, °
γ, °
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
V, Å3
Z
Dc/gcm−3
F(000)
ꢁ(Mo Kα)/cm−1
T/K
crystal size/mm
range of indices
collected reflections
unique reflections
Rint
reflections with I > 2σ(I)
no. parameters
R(F), F > 2σ(F)
0.054
0.067
0.130
0.138
ꢀ0.325, 0.391
972041
wR(F2), F > 2σ(F)
R(F), all data
wR(F2), all data
ꢂr (min., max.) eÅ−3
CCDC deposition number
ꢀ0.365, 0.336
972042
ꢀ0.353, 0.282
972043
(Hꢀtheo)+·(26DBA)ꢀ ·(H2O)
(C7H9N4O2)+ꢂ(C7H5O4)ꢀꢂ(H2O)
352.31
monoclinic
P21/c
14.6593(19)
6.5583(8)
15.727(2)
90.00
94.0430(10)
90.00
1508.2(3)
4
1.552
736
(theo)·(34DBA)
(theo)·(35DBA)
molecular formula
Mr
crystal system
space group
a, Å
b, Å
c, Å
α, °
(C7H8N4O2)ꢂ(C7H6O4)
334.29
triclinic
Pī
8.0694(12)
8.5782(13)
11.5578(17)
103.147(2)
104.970(2)
105.358(2)
706.70(18)
2
1.571
348
0.125
100(2)
0.40 × 0.40 × 0.10
ꢀ10 → 10
ꢀ11 → 11
ꢀ15 → 15
8226
(C7H8N4O2)ꢂ(C7H6O4)
334.29
triclinic
Pī
7.2253(17)
7.9998(19)
12.650(3)
81.240(4)
85.071(4)
82.509(4)
714.9(3)
2
1.553
348
0.124
170(2)
0.40 × 0.40 × 0.05
ꢀ9 → 9
ꢀ10 → 10
ꢀ16 → 16
8422
β, °
γ, °
V, Å3
Z
Dc/gcm−3
F(000)
ꢁ(Mo Kα)/cm−1
T/K
0.126
100(2)
crystal size/mm
0.30 × 0.15 × 0.15
ꢀ18 → 19
ꢀ8 → 8
ꢀ20 → 20
16040
3436
0.037
2225
252
0.039
0.081
0.113
0.164
ꢀ0.259, 0.293
972044
range of indices
collected reflections
unique reflections
Rint
reflections with I > 2σ(I)
no. parameters
3287
0.027
2609
231
0.035
0.046
0.103
0.120
3399
0.075
2629
223
0.059
0.072
0.154
0.167
R(F), F > 2σ(F)
wR(F2), F > 2σ(F)
R(F), all data
wR(F2), all data
ꢂr (min., max.) eÅ−3
CCDC deposition number
ꢀ0.288, 0.328
972045
ꢀ0.359, 0.334
972046
ACS Paragon Plus Environment

Page 5 of 13
Crystal Growth & Design
at oneꢀdegree intervals. The calibration was established using a of the carboxylꢀgroup geometries and inspections of Fourier
1
2
3
4
5
6
7
8
Si standard reference material (i.e. NIST 640c). The instrument
was operated using the Symphonix program,60 whereas the data
was analyzed using the Jade software (version 6.5).61 The
samples were loaded onto an aluminum sample holder and
leveled with a glass slide.
difference maps. The CꢀO bond distances in eight solids were
consistent with the formation of a cocrystal (Table 2,
supporting info), while one solid (i.e. 26DBA) contained a
carboxylic acid with a geometry characteristic of a carboxylate
anion. At the same time, the Fourier difference maps of the
eight complexes revealed that the acidic protons to be
positioned in close proximity to the Oꢀatom of the carboxylic
acid group, at distances that correspond to a neutral acid
moiety. In the case of 26DBA, the acidic proton was located in
close proximity of the imidazole moiety of theo.
2.6. Singleꢀcrystal Xꢀray crystallography (SXRD)
The
crystallographic
analysis
of
(theo)·(3HBA),
(theo)·(4HBA), (theo)·(24DBA) and (theo)·(35DBA) were
performed on a Bruker SMART diffractometer equipped with
an APEX1 CCD camera, while the data for (theo)·(2HBA),
(theo)·(23DBA), (theo)·(25DBA), (Hꢀtheo)+·(26DBA)ꢀ·(H2O)
and (theo)·(34DBA) were obtained using a Bruker APEX2
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
Crystallographic data pertaining to all solids are summarized in
Table 1. The carboxylicꢀgroup geometries for the solids are
given in Table 2. The structural features of the solid theo
phases are described as follows.
diffractometer.
The single crystals of (theo)·(3HBA),
(theo)·(4HBA), (theo)·(24DBA), and (theo)·(35DBA) were
mounted on a glass pin using epoxy glue, whereas the single
crystal of (theo)·(2HBA), (theo)·(23DBA), (theo)·(25DBA),
(Hꢀtheo)+·(26DBA)ꢀ·(H2O), and (theo)·(34DBA) were
positioned on a MiYeGen mount with vacuum grease. Both Xꢀ
ray diffraction data sets were collected using graphiteꢀfiltered
MoKα radiation (λ = 0.71069 Å) at temperatures indicated in
Table 1, processed using SAINTꢀPlus,62 and corrected for
Lorentzꢀpolarization effects. The data sets were collected with
omega scans with 0.5° steps. Structure solution and refinement
were accomplished using SHELX.63 All nonꢀhydrogen atoms
were refined anisotropically. All hydrogen atoms bonded to
carbon atoms were located in the Fourierꢀdifference electron
density map, fixed in geometrically constrained riding
positions, and refined isotropically on the basis of
corresponding Cꢀatoms [U(H) = 1.2·Ueq(C)]. In the case of (Hꢀ
theo)+·(26DBA)ꢀ·(H2O), the hydrogen atoms associated with
Oꢀ and Nꢀatoms were included in the located positions (i.e.
coordinates refined) and treated isotropically. The hydrogen
atoms belonging to Nꢀ and Oꢀatoms were fixed in
geometrically constrained riding positions, and refined
isotropically in the cases of (theo)·(2HBA), (theo)·(4HBA),
(theo)·(23DBA) and (theo)·(24DBA). In the structures of
(theo)·(3HBA), (theo)·(25DBA), (theo)·(34DBA) and
(theo)·(35DBA), hydrogen atoms bonded to Nꢀ and Oꢀatoms
were refined either independently (i.e. coordinates and Uiso
refined) or in a constrained fashion (i.e. coordinates constrained
and Uiso refined). Crystallographic data of the investigated
compounds are listed in Table 1. Geometry parameters of
selected hydrogen bonds are given in the Supplementary
Information document.
Table 2. Carbonꢀoxygen bond distances of the CCF carboxyl
groups in the theoꢀbased solids.
solid
d(C=O) / (Å)
1.208(2)
1.211(2)
1.233(4)
1.226(6)
1.233(2)
1.225(2)
1.253(3)
1.235(2)
1.218(2)
d(C−O) / (Å)
1.329(2)
1.327(2)
1.351(4)
1.314(5)
1.325(1)
1.323(2)
1.280(3)
1.318(2)
1.319(2)
ꢀd / (Å) (*)
(theo)·(3HBA)
0.121(3)
0.166(3)
0.118(6)
0.088(8)
0.092(2)
0.098(3)
0.027(4)
0.083(3)
0.101(3)
(theo)·(4HBA)
(theo)·(23DBA)
(theo)·(24DBA)
(theo)·(25DBA)
(theo)+·(26DBA)ꢀ·(H2O)
(theo)·(34DBA)
(theo)·(35DBA)
* ꢀd = d(C−O) ꢀ d(C=O)
3.1. Cocrystal (theo)·(2HBA)
PXRD studies of the solid recovered from the screening
experiment involving theo and 2HBA revealed the formation
of (theo)·(2HBA) 1:1 cocrystal (CSD reference code:
KIGLES, Fig 1). The cocrystal was initially reported by
Madarasz et al. in 2002.65 and structurally characterized by
Childs et. al. in 2007. The latter66 showed that theo and 2HBA
crystallize in the monoclinic P21/n space group with both one
molecule of theo and 2HBA in the asymmetric unit. It was
found that the cocrystal components selfꢀassemble to form 1D
ribbons (Fig. 1a).
Within the ribbons, theo and 2HBA interact via O–
H(carboxyl)···N(imidazole) hydrogen bonds (synthon I), while
both theo and 2HBA form dimers. Whereas theo molecules
form dimers via a cyclic array of N–H(imidazole)···O(amide)
hydrogen bonds (synthon IX, Scheme 2), 2HBA molecules
form dimers via a bifurcated hydrogen bond wherein the orthoꢀ
hydroxyl group participates in both an intraꢀ and intermolecular
O–H(hydroxyl)···O(carboxyl) hydrogen bond (Fig. 1a). The
theo:2HBA ribbons form “wavy” 2D sheets held together by
C–H···O and van der Waals forces. Within the sheets, the
ribbons are inclined by φ = 29.2º (Fig 1b). The 2D sheets form
stacks that are sustained by π(theo)···π(2HBA) interactions
(Fig 1c).
2.7. Hydrogenꢀbondꢀpattern analysis
The CCDC Mercury56 program was utilized to identify
hydrogenꢀbond patterns in the structures of the theo solids by
using default definitions of hydrogen bonds. The following
criterion was added to the default definition: DꢀH···A ≥ 100°
(where D = hydrogenꢀbond donor, A = hydrogenꢀbond
acceptor). The hydrogenꢀbond parameters were calculated
using the CIFTAB63 in WinGX.64
3. RESULTS
Nine solid phases involving theo and (di)hydroxybenzoic acids
were identified using SMPT,29ꢀ30, 32, 35ꢀ36 with six phases being
identified as new using PXRD. SXRD was utilized to
determine whether cocrystal or salt formation occurred.
Specifically, the theo solids were identified as either neutral
molecular complexes (i.e. cocrystal) or salts based on analyses
ACS Paragon Plus Environment

Crystal Growth & Design
3.3. Cocrystal (theo)·(4HBA)
Page 6 of 13
1
2
3
4
5
6
The structure of (theo)·(4HBA) was initially reported by
Childs et al. in 2007 (CSD reference code: KIGLOC).66
(Theo)·(4HBA) crystallizes in a 1:1 ratio in the monoclinic
space group P21/c with both one molecule of theo and 4HBA
in the asymmetric unit. Theo and 4HBA form “zigꢀzag”ꢀ
shaped chains being held together by cyclic arrays of N–
7
8
H(imidazole)···O=C(carboxyl)
O=C(amide) hydrogen bonds (synthon II), and
and
O–H(carboxyl)···
O–
a
H(hydroxyl)···N(imidazole) hydrogen bond (synthon VI) that
is supported by a C–H(urea)···O(hydroxyl) hydrogen bonds
(Fig. 3a). Within the “zigꢀzag” chain, the theo:4HBA pairs are
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
inclined
by φ = 39.9º.
The
chains
interact
via
π(theo)···π(4HBA) interactions to form stacks (Fig. 3b).
Neighboring stack are interdigitated and held together by
π(theo)···π(4HBA) and C–H···O forces (Fig 3c).
Figure 1. Perspective views: (a) a molecular ribbon based on theo
and 2HBA being held together by O–H(carboxyl)···N(imidazole),
N–H(imidazole)···O=C(amide) and bifurcated O–H(hydroxyl)···
O(carboxyl)···O–H(hydroxyl) hydrogen bonds, (b) a 2D sheet
composed of theo:2HBA ribbons, and (c) stacks of theo:2HBA
sheets sustained by π(theo)···π(2HBA) interactions viewed along
the (ꢀ201) and (010) crystallographic planes.
3.2. Cocrystal (theo)·(3HBA)
Cocrystal (caf)·(3HBA) crystallizes in the triclinic space group
Pī with both two theo and 3HBA molecules in the asymmetric
unit. Theo and 3HBA form a “wavy” 1D polymer constituted
of centrosymmetric fourꢀcomponent theo:3HBA aggregates
being sustained via O–H(carboxyl)···N(imidazole) and O–
H(hydroxyl)···O=C(urea) bonds (synthons
I
and VIII,
respectively). The aggregates are interlinked to form the 1D
structures through cyclic arrays of N–H(carboxy)···
O=C(amide) hydrogen bonds (synthon IX) (Fig 2a). The
theo:3HBA aggregates within the polymers are inclined by
φ = 39.5º (Fig. 2a, bottom). The polymers selfꢀassemble to
form 2D sheets being held together by π(theo)···π(3DBA) and
C–H···O interactions (Fig. 2b). The sheets stack in an offset
Figure 3. Perspective views: (a) hydrogenꢀbonded “zigꢀzag”
chains involving theo and 4HBA, (b) stacks of theo:4HBA chains
viewed along the crystallographic planes (210) and (001), (c)
intercalated stacks of theo:4HBA held together by faceꢀtoꢀface πꢀ
forces viewed along the crystallographic planes (210) and (001),
and the crystallographic axis a (inset).
3.4. Cocrystal (theo)·(23DBA)
manner to form
a
3D structure, and interact via
Theo and 23DBA crystallize in the monoclinic space group
P21/c with both one molecule of theo and 23DBA in the
asymmetric unit. In the crystal structure, theo and 23DBA form
π(theo)···π(3HBA) and C–H···O forces (Fig. 2c).
a
helical
array
held
together
by
N–
H(imidazole)···O=C(carboxyl)
and O–H(carboxyl)···
O=C(amide) hydrogen bonds (synthon II), and O–
H(hydroxyl)···N(imidazole) hydrogen bonds (synthon VI), as
well as polar N(imidazole)···O(hydroxyl) forces (N···O
distance: 3.05Å). Present are helices of both handedness being
related by a center of inversion. The pitch of the helix is 6.8 Å,
with two molecules of theo and 23DBA per turn. The helix
exhibits a diameter of approximately 16 Å (Fig. 4a).
Helices of opposite handedness are intercalated and held
together via π(theo)···π(23DBA) forces to form a 2D sheet
that propagates along the crystallographic cꢀaxis (Fig. 4b). The
2D sheets selfꢀassemble into a 3D structure being held
together by C–H···N and C–H···O interactions (Fig. 4c).
Figure 2. Perspective views: (a) fourꢀcomponent assemblies of
theo and 3HBA forming a “wavy” 1D polymer, (b) molecular
sheet composed 1D polymers held together by π(theo)···π(3HBA)
and C–H···O interactions (viewed along the crystallographic
planes (100) and (001)), (c) 3D structure composed of the stacked
2D sheets held together by π(theo)···π(3HBA) interactions and
C–H···O hydrogen bonds (viewed along the crystallographic
planes (1ꢀ10) and (001)).
ACS Paragon Plus Environment

Page 7 of 13
Crystal Growth & Design
along the crystallographic c axis (b), and the crystallographic
1
planes (110) and (001) (c).
2
3
3.6. Cocrystal (theo)·(25DBA)
The crystal structure of (theo)·(35DBA) was reported by
Aitipamula et al. in 2009 (CSD reference code: DUCROJ).67
Theo, and 25DBA cocrystallize in the triclinic space group Pī.
The asymmetric unit contains one molecule of each
component. Theo and 25DBA selfꢀassemble into molecular
sheets sustained, and are held together by O–
H(carboxyl)···N(imidazole) and O–H(hydroxyl)···O=C(urea)
hydrogen bonds (synthons I and VIII). The theo molecules
form dimers via cyclic arrays of N–H(imidazole)···
O=C(carboxyl) and O–H(carboxyl)···O=C(amide) hydrogen
bonds (synthon IX), while the 25DBA molecules dimerize via
bifurcated hydrogen bonds. Specifically, the orthoꢀhydroxyl
moiety of 25DBA acts as bifurcated donor, and is involved in
both an intraꢀ and intermolecular O–H(hydroxyl)···
O=C(carboxyl) hydrogen bonds (as seen in the case of the
(theo)·(2HBA) cocrystal), Fig. 6a. The 2D structures stack
offset along the (21ꢀ1) crystallographic plane and are held
together by π(theo)···π(25DBA) forces (Fig. 6b).
4
5
6
7
8
9
10
11
12
13
Figure 4. Perspective views: (a) helical theo:23DBA assemblies,
(b) interdigitated helices sustained via π(theo)···π(4HBA)
interactions viewed along the crystallographic a axis (the red and
blue helices are of opposite handedness), c) sheets of
theo:23DHBA helices interlinked via C–H···N and C–H···O
interactions viewed along the crystallographic planes (101) and
(2ꢀ11) (the green and yellow molecular aggregates represent
neighboring sheets of interdigitated helicies).
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
3.5. Cocrystal (theo)·(24DBA)
The cocrystal of (theo)·(24DBA) crystallizes in the monoclinic
space group P21/c with both one molecule of theo and 24DBA
in the asymmetric unit. The crystal structure reveals that theo
and 23DBA form 1D polymers held together by cyclic arrays
of N–H(imidazole)···O(carboxyl) and O–H(carboxyl)···
O=C(amide)
hydrogen
bonds
(synthon
II),
O−H(hydroxyl)···O=C(urea) hydrogen bonds (synthon VIII),
as well as intramolecular O−H(hydroxyl)···O=C(carboxyl)
hydrogen bonds. The 1D polymers form bilayers that resemble
2D grids (Fig. 5a). Within each of the two layers, the 1D
polymers are aligned in parallel being separated by d ~ 6 Å.
The 1D polymers are hydrogenꢀbonded to a set of parallel
polymers from the second layer via O–H(hydroxyl)···
N(imidazole) hydrogen bonds (Fig. 5b). Within the bilayer, the
two sets of parallel polymers are inclined by φ = 70.9º (Fig.
5a.). The bilayers form a 3D structure wherein neighboring
Figure 6. Perspective views: (a) 2D theo:25DBA network
involving theo and 25DBA sustained by hydrogen bonds, and (b)
stacked theo:24DBA networks held together by π···π interactions.
bilayers
are
intercalated
interlinked
through
π(theo)···π(24DBA) and C–H···O interactions.
3.7. Salt (Hꢀtheo)+·(26DBA)ꢀ·(H2O)
Theo and 26DBA crystallize to form a monohydrate salt from
acetonitrile in the monoclinic space group P21/c. A single (Hꢀ
theo)+ cation, (26DBA)ꢀ anion, and water molecule are present
in the asymmetric unit. (Hꢀtheo)+ and (26DBA)ꢀ form a
centrosymmentric fourꢀcomponent assembly wherein (Hꢀtheo)+
and (26DBA)ꢀ interact via a N+–H(imidazole)···Oꢀ(carboxyl)
hydrogen bond, while (26DBA)ꢀ anions interact with each other
via bifurcated O–H(hydroxyl)···O=C(carboxyl) hydrogen bond
(Fig. 7a).
The fourꢀcomponent assemblies are interlinked through water
molecules to form molecular ribbons sustained by
N–H(imidazole)···O(water) and O–H(water)···O=C(amide)
hydrogen bonds (Fig. 7b, upper inset). The ribbons are stacked
offset to form a layered 2D structure (Fig. 7b). The layers are
held together by O–H(water)···Oꢀ(carboxyl) hydrogen bonds
(Fig 9b, lower inset). The 2D structures further selfꢀassemble in
to form a 3D structure, and are sustained by C–H···O forces
(Fig. 7c).
Figure 5. Perspective views of a molecular bilayer composed of
hydrogenꢀbonded 1D theo:24DBA polymers viewed along the
crystallographic c axis (the inset depicts a closeꢀup view a
structural fragment of the bilayer highlighting the hydrogenꢀbond
patterns occurring within and among 1D polymers in the bilayer)
(a), intercalated bilayer held together by π···π forces, viewed
ACS Paragon Plus Environment

Crystal Growth & Design
Page 8 of 13
fashion viewed along the crystallographic planes (01ꢀ1) and
1
(ꢀ221).
2
3
3.9. Cocrystal (theo)·(35DBA)
Theo and 35DBA crystallize in a 1:1 ration in the triclinic
space group Pī with one molecule of each component in the
asymmetric unit. The components form a planar 2D structure
comprised of both theo and 35DBA that form dimers (Fig. 9a).
Within the 2D structure, the theo dimers are held together by
cyclic arrays of N–H(imidazole)···O=C(carboxyl) and
O–H(carboxyl)···O(amide) hydrogen bonds (synthon IX),
while the 35DBA molecules form dimers through O–
H(hydroxyl)···O=C(carboxyl) hydrogen bonds (Fig. 9a). The
theo and 35DBA dimers are connected through O–
H(hydroxyl)···N(imidazole) and O–H(hydroxyl)···O=C(urea)
hydrogen bonds (synthons VI and VIII). The theo:35DBA
sheets are stacked in an offset manner being held together by
both π(theo)···π(35DBA) and weak C–H···O interactions (Fig.
9b)
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
Figure 7. Perspective views: (a) fourꢀcomponent (Hꢀ
theo)+:(26DBA)ꢀ assemblies, (b) the 2D structure of
(Hꢀtheo)+:(26DBA)ꢀ:(H2O) assemblies viewed along the
crystallographic a axis, and (d) the 3D structure of the (Hꢀ
theo)+·(26DBA)ꢀ·(H2O) salt viewed along the crystallographic b
axis.
3.8. Cocrystal (theo)·(34DBA)
Theo and 34DBA selfꢀassemble to form a 1:1 complex that
crystallizes in the triclinic space group Pī with both one
molecule of theo and 34DBA in the asymmetric unit. The
cocrystal components form a planar sheet. Within the sheet,
theo and 34DBA are hydrogen bonded via O–
H(hydroxyl)···N(imidazole) and O–H(hydroxyl)··· O=C(urea)
interactions (synthons VI and VIII, respectively). At the same
time, both theo and 34DBA form dimers within the molecular
sheet. Specifically, theo dimer are sustained by
N–H(imidazole)···O(amide) hydrogen bonds (synthon IX),
while the 34DBA dimers are being held together via O–
H(hydroxyl)···O=C(carboxyl) hydrogen bonds (Fig. 8a). The
occurrence of carboxylicꢀacid homosynthons in the presence of
an imidazole moiety is uncommon,32 and thus surprising. The
planar sheets are stacked in an offset and “headꢀtoꢀtail” manner
being sustained by π(theo)···π(34DBA) forces, as well as weak
C–H···O, and C–H···π interactions (Fig. 8b).
Figure 9. Perspective views: (a) the 2D structure of theo and
35DBA dimers viewed along the crystallographic planes (116)
and (1ꢀ20) (inset: spaceꢀfill view of the theo:35DBA 2D sheet)
(b) stacked sheets sustained by π(theo)———π(35DBA) and C–H———O
forces viewed along the crystallographic planes (20ꢀ1) and (010).
4. DISCUSSION
4.1. Synthon hierarchies: synthon I vs. synthon II
A survey of the CSD reveals that theo and carboxylic acids
predominantly selfꢀassemble through synthon I (Scheme 2).
More specifically, out of 25 CSD entries involving theo and a
carboxylic acid, 15 structures exhibit a theo:acid interaction via
synthon I (i.e. CIZTAH, CODCOO, CSATEO, DUCROJ,
KIGLES, KIGLIW, LUKXUL, NEXWOD, NEYCIE,
UNITER, UNITIV, XEJWUF, XEJXAM, XEJXEQ, XEJXIU),
whilst seven theo:carboxylic acid structures exhibited synthon
II (i.e. DEYREF, IJIBEJ, IJIBEJ01, KIGLAO, KIGLOC,
LUKXUL, VAXSAQ). One CSD entry exhibited both
synthons I and II (i.e. LUKXUL), while the remaining four
entries (LEDVOH, KIGKAN, WUYROX, WUYRUD) do not
exhibit either synthon I and II. The persistence of synthon I in
theo:carboxylicꢀacid solids is in line with the reported
cocrystals based on of caf and carboxylicꢀacids (Scheme 3).31
Our study shows that five out of nine theo:(di)hydroxybenzoic
acid cocrystals exhibit synthon I (i.e. cocrystals involving
2HBA, 3HBA, 25DBA, 26DBA, and 35DBA), whereas three
cocrystals exhibited synthon II (i.e. cocrystals based on 4HBA,
23DBA and 24DBA).
Figure 8. Perspective views: (a) molecular sheets composed of
theo and 34DBA molecules viewed along the crystallographic
planes (110) and (001), and (b) sheets stacked in a “headꢀtoꢀtail”
ACS Paragon Plus Environment

Page 9 of 13
Crystal Growth & Design
Table 3. Overview of structural features in the cocrystals of theo and (di)hydroxybenzoic acids.
1
2
3
4
5
6
7
8
structure of theo:CCF
assembly
hydrogenꢀbond acceptor group of theo
compound
(theo)·(2HBA)
theo:CCF ratio
imidazole
urea
amide
total
1D infinite
1D infinite
1D infinite
1D infinite
2D infinite
2D infinite
discretea
2
3
2
2
3
3
2
3
3
1:1
1:1
1:1
1:1
1:1
1:1
1:1
1:1
1:1
√
√
√
√
√
√
√
√
√
×
√
√
√
√
√
√
√
(theo)·(3HBA)
(theo)·(4HBA)
×
×
√
√
×
(theo)·(23DBA)
(theo)·(24DBA)
(theo)·(25DBA)
(theo)+·(26DBA)ꢀ ·(H2O)
(theo)·(34DBA)
(theo)·(35DBA)
b
√
9
2D infinite
2D infinite
√
√
√
√
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
a 1D infinite based on included H2O; b hydrogenꢀbond acceptor interacts with H2O, rather than a hydroxyl group of the CCF.
Table 4. Overview of structural features in the cocrystals of caf and (di)hydroxybenzoic acids.
hydrogenꢀbond acceptor group of caf
structure of caf:CCF
assembly
compound
caf:CCF ratio
imidazole
urea
amide
total
discrete
discrete
2D infinite
discrete
1
2
3
2
2
3
2
3
(caf)·(2HBA)
(caf)·(3HBA)
(caf)·2(4HBA)
2(caf)·(4HBA)
(caf)·(23DBA)·(H2O)
(caf)·(24DBA)·(H2O)
(caf)·(25DBA)
1:1
2:2
1:2
2:1
1:1
1:1
1:1
2:2
√
√
√
√
√
√
√
×
√
√
×
×
√
√
×
discretea
discreteb
1D infinite
discretec
d
√
√
×
d
d
√
×
√
d
(caf)·(35DBA)·(H2O)
√
√
√
a
b
c
d
1D infinite based on included H2O; 2D infinite based on included H2O; 3D infinite based on included H2O, hydrogenꢀbond
acceptor interacts with H2O, rather than a hydroxyl group of the CCF.
The components of (theo)·(34DBA) do not exhibit either
synthon I or II. Specifically, (theo)·(34DBA) exhibits both
theo and 34DBA hydrogenꢀbonded dimers sustained via the N–
H(imidazole)··· O=C(amide) synthon (synthon VII) and O–
H(carboxyl)··· O=C(carboxyl) homosynthon (Fig. 8). The theo
Table 5. Modes of binding of theo:(di)hydroxybenzoic acid
solids correlated with the CCF’s calculated pKa values.a
CCF
26DBA
23DBA II
2HBA
25DBA
24DBA II
synthon pKa
CCF
synthon pKa
I
1.30±0.1 35DBA
2.96±0.1 3HBA
I
I
3.96±0.1
4.08±0.1
4.45±0.1
4.57±0.1
and
34DBA
dimers
are
linked
through
O–
I
I
3.01±0.1 34DBA N/A
3.01±0.1 4HBA
H(hydroxyl)···N(imidazole) and O–H(hydroxyl)···O=C(urea)
hydrogen bonds (synthon VI and VII, respectively). The
formation of a carboxylic acid homosynthon in the presence of
an N(aromatic) functional groups is rare in the crystalline
state.32 However, a recent computational study has indicated
that carboxylic acid homodimers can form in the presence of
N(aromatic) moieties, while achieving a global minimum in the
crystal energy landscape.68 Such carboxylic acid homodimers
were not observed in the caf cocrystals.31
II
3.32±0.1
a
Values calculated using Advanced Chemistry Development
software.69
The cocrystals involving theo provide the first O–
H(hydroxyl)···N(imidazole) hydrogen bonds (synthon VI)
involving the phenolic groups of 4HBA, 23DBA, 24DBA, and
34DBA and the imidazole moiety (Figure 3a, 4a, 5a, 8a). The
presence of one O−H(hydroxyl)···N(imidazole) hydrogen bond
(i.e. synthon VI) in the theo solids can be attributed to the
presence of a second structural moiety in theo interacting with
the CCF via forces of comparable strengths; namely, a
Our structural analyses revealed that the interactions of theo
with CCFs through synthons I and II cannot be correlated with
the strengths in acidities (viz., pKa value) of the carboxylic
acidꢀbased CCFs. It was likely that a CCF with pKa < 3 would
participate in synthon I, while CCFs with higher pKa values
would form synthon II. The hypothesis was based on reported
theo cocrystal structures involving (di)hydroxybenzoic acids,
namely 2HBA, 4HBA and 25DBA (KIGLES, KIGLOC and
DUCROJ, respectively). While 2HBA and 25DBA exhibit a
pKa ≈ 3 and synthon I in respective theo cocrystals, 4HBA
exhibits a pKa > 4.5 and interacts with theo through synthon II
(Table 5). Our study, however, shows that 23DBA with a pKa <
3 participates in the formation of synthon II, whereas 3HBA
and 35DBA with pKa ≈ 4 form synthon I in their respective
cocrystals (Table 5). These observations suggest that the pKa of
the carboxylic acids CCF does not significantly affect the
outcome of the selfꢀassembly process involving theo and
(di)hydroxybenzoic acids.
O−H(carboxyl)···O=C(amide) and/or
N−H(imidazole)···
O=C(carboxyl) hydrogen bond.70 The close proximity of the
hydrogenꢀbond donating N–H(imidazole) and accepting
C=O(amide) likely facilitates the formation of synthon II, thus,
leaving the hydrogenꢀbond accepting N(imidazole) group
available to form synthon VI. It is, therefore, not surprising that
all theo solids that exhibit synthon II also display synthon VI.
The presence of synthon VI was expectedly not observed in
related caf cocrystals, as caf lacks a structural moiety to enable
the formation of synthon II. The absence of synthon II in
caf:(di)hydroxybenzoicꢀacid cocrystals is in agreement with a
recent study that established that O−H(carboxyl)···N(pyridine)
hydrogen bonds persist in cocrystals that also contain a
hydroxyl functional group (Scheme 3).40
ACS Paragon Plus Environment

Crystal Growth & Design
Page 10 of 13
Overall, a lesser persistence of synthon I in the theo cocrystals 34DBA and 35DBA form infinite 2D structures. A 0D, or
1
2
3
4
5
6
7
8
compared to analogous caf cocrystals attests to a difficulty to
predict hydrogenꢀbond patterns in those cocrystals sustained by
multiple homoꢀ and heterosynthons.
discrete, assembly based on theo was identified in the case of
26DBA. Specifically, theo and 26DBA form a discrete fourꢀ
component assembly. Within the assemblies, the carboxylicꢀ
acid protons undergo proton transfer to the imidazoles, thus,
forming a salt (Fig. 7a). The proton transfer is not surprising
given that theo and 26DBA exhibit a negative ꢃpKa value
4.2. Synthon hierarchies: synthon VII vs. synthon VIII
Further discrepancies between the theo and caf cocrystal series
are associated with the involvement of the C=O(amide) and
C=O(urea) moieties in hydrogen bonding. The theo cocrystals
showed that the C=O(amide) moiety participates in hydrogen
bonding in all nine solids, while the C=O(urea) unit was
involved in hydrogen bonding in only five solids (Table 3).
That the C=O(amide) group participates in hydrogen bonding
to a larger extent than the C=(urea) group is not unexpected
considering that the C=O(amide) moiety recurrently
participates in interactions with either the CCF carboxyl group
or another theo molecule via synthons II and IX. The
C=O(amide) moiety, however, does not participate in O–
H(hydroxyl)···O=C(amide) hydrogen bonding (i.e. synthon
VIII) in any of the studied solids (Scheme 3). The C=O(urea)
group, on the other hand, participates in O–
H(hydroxyl)···O=C(urea) hydrogen bonding (i.e. synthon VII)
in theo solids involving 3HBA, 24DBA, 25DBA, 34DBA and
35DBA.
[ꢃpKa ∼ 0.34;
pKa(theo) = 1.64 ± 0.7,
pKa(26DBA)
= 1.30 ± 0.1]. A recent computational study shows that two
components exhibiting a ꢃpKa value [ꢃpKa = pKa(conjugated
acid of the base) ꢀ pKa(acid)] greater than 4 will most likely
undergo proton transfer to form a salt, whereas a ꢃpKa< ꢀ1
suggests formation of a neutral species (i.e. cocrystal).71
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
4.4. Hydrate formation
All theo cocrystals were anhydrous, whereas the 26DBA salt
crystallized as a monohydrate. The analogous series of caf
cocrystals, prepared under similar crystallization conditions,
exhibited three cocrystal hydrates. Given that water molecules
commonly become incorporated into crystal lattices to establish
a balance of hydrogenꢀdonors and hydrogen acceptors,72 we
attribute the exclusive formation of anhydrous theo cocrystals
to an equal (or very similar) number of donors and acceptors in
both theo and CCF (hydrogenꢀbond donor:acceptor (D/H)
ratios: D/H=3:3 for hydroxybenzoic acid CCFs, D/H=4:3 for
dihydroxybenzoic acid CCFs). The formation of a hydrate in
the case of the (Hꢀtheo)+·(26DBA)ꢀ·(H2O) salt is ascribed to
steric effects, rather than a mismatch in the number of
hydrogenꢀbond donating and accepting functional groups. In
particular, water molecules position themselves between two
theo molecules in the 1D ribbon (Fig. 7b) to minimize
The overall participation of the C=O(urea) and C=O(amide)
moieties in hydrogen bonding in the theo cocrystal series is
lower than seen in the related caf solids (Tables 3 and 4).31 The
occurrence of synthons VII and VIII in the theo cocrystals is,
however, consistent with the one found in the caf solids.
Specifically, four caf cocrystals exhibit the O–
H(hydroxyl)···O(urea) heterosynthon (i.e. synthon VIII), two
solids reveal the presence of an O–H(hydroxyl)···O=C(amide)
heterosynthon (i.e. synthon VII), while one cocrystal involves
both synthon VII and VIII.
repulsive
contacts
between
benzene(26DBA)
and
CH3(amide,theo) moieties that would likely emerge upon
formation of theo dimers synthon IX.
4.4. CCF orthoꢀhydroxyl groups form intermolecular hydrogen
bonds
The orthoꢀhydroxyl group of 2HBA, 24DBA, 25DBA and
26DBA acted as bifurcated hydrogenꢀbond donors in the theo
solids. Each orthoꢀhydroxyl group participates in both an
intramolecular O–H(hydroxyl)···O=C(carboxyl) hydrogen
bond and an intermolecular O–H(hydroxyl)···O=C(carboxyl)
or O–H(hydroxyl)···N(imidazole) hydrogen bond. Similarly,
the orthoꢀhydroxyl group of 23DBA participates in an
intramolecular O–H(hydroxyl)···O(carboxyl) hydrogen bond
and
a
polar intermolecular N(imidazole)···O(hydroxyl)
interaction. In contrast, the orthoꢀhydroxyl groups of the CCFs
did not form intermolecular hydrogen bonds in those cocrystals
involving caf. Origins of the different behaviors of the orthoꢀ
hydroxyl groups in the theo and caf solids are, at this point, not
understood and currently under investigation.
5. CONCLUSION
We have prepared and structurally characterized nine theo
solids based on structural isomers of (di)hydroxybenzoic acids.
The solids were discovered using SMPT as a screening
method. Structural analyses revealed the formation of
crystalline architectures based on intermolecular homoꢀ and
heterosynthons, as well as intramolecular heterosynthons. The
most prevalent intermolecular interactions were found to be:
the O−H(carboxyl)···N(imidazole) and O−H(hydroxyl)···
O=C(urea) heterosynthons, and the homosynthon comprised of
Scheme 3.
4.3. Dimensionalities of supramolecular assemblies
All theo cocrystals are based on 1:1 theoꢀtoꢀCCF ratios that
predominantly form infinite hydrogenꢀbonded assemblies.
Cocrystals involving 2HBA, 3HBA, 4HBA and 23DBA
generate infinite 1D structures, whereas 24DBA, 25DBA,
ACS Paragon Plus Environment

Page 11 of 13
Crystal Growth & Design
6. Jones, W.; Motherwell, W. D. S.; Trask, A. V., MRS Bulletin
a cyclic array of N–H(imidazole)···O=C(amide) hydrogen
bonds (all synthons occur in five out of nine theo solids). The
intermolecular O−H(hydroxyl)···N(imidazole) and
2006, 31, 875ꢀ879.
7. Trask, A. V.; Motherwell, W. D. S.; Jones, W., Int. J. Pharm.
2006, 320, 114ꢀ123.
1
2
3
4
5
6
7
8
9
O−H(hydroxyl)···O=C(carboxyl) heterosynthons were found in
four out of nine solids, whereas the heterosynton comprised of
a cyclic array of N–H(imidazole)···O=C(carboxyl) and O–
H(carboxyl)···O=C(amide) hydrogen bonds occurs three times.
The intermolecular O−H(carboxyl)···O=C(carboxyl) homoꢀ
synthon formed in only one solid. The pKa values of the CCFs
were shown to have no impact on the selfꢀassembly process.
Our study demonstrates the difficulty in predicting outcomes of
selfꢀassembly processes involving two molecules whereby both
consist of multiple hydrogenꢀbonding functional groups. We
have also shown that structurally similar PAs (i.e. theo and
caf) exhibit distinct supramolecular behavior. Altogether, the
reported findings impose the need for systematic structural
studies of cocrystal based on multiple synthons to establish
synthons hierarchies in complex supramolecular systems. The
outcomes of these studies are expected to provide further
insights into selfꢀassembly processes wherein multiple
hydrogenꢀbonding groups compete for synthon formation.
8. Pharmaceutical Salts: Properties, Selection, and Use. 2nd ed.; P.
Heinrich Stahl; Wermuth, C. G., Eds. Wiley VCH: Weinheim,
2008.
9. Berstein, J., Polymorphism in Molecular Crystals Oxford
University Press: New York, 2002; Vol. 14.
10. Gavezzotti, A., J. Pharm. Sci. 2007, 96, 2232ꢀ2241.
11. Day, G. M.; Cooper, T. G.; CruzꢀCabeza, A. J.; Hejczyk, K. E.;
Ammon, H. L.; Boerrigter, S. X. M.; Tan, J. S.; Valle, R. G. D.;
Venuti, E.; Jose, J.; Gadre, S. R.; Desiraju, G. R.; Thakur, T. S.;
Eijck, B. P. v.; Facelli, J. C.; Bazterra, V. E.; Ferraro, M. B.;
Hofmann, D. W. M.; Neumann, M. A.; Leusen, F. J. J.; Kendrick,
J.; Price, S. L.; Misquitta, A. J.; Karamertzanis, P. G.; Welch, G.
W. A.; Scheraga, H. A.; Arnautova, Y. A.; Schmidt, M. U.;
Streek, J. v. d.; Wolf, A. K.; Schweizerr, B., Acta Crystallogr. B
2009, 65, 107ꢀ125.
12. Dunitz, J. D.; Bernstein, J., Acc. Chem. Res. 1995, 28, 193ꢀ200.
13. Bauer, J.; Spanton, S.; Henry, R.; Quick, J.; Dziki, W.; Porter,
W.; Morris, J., Pharm Res. 2001, 18, 859ꢀ866.
14. Chemburkar, S. R.; Bauer, J.; Deming, K.; Spiwek, H.; Patel, K.;
Morris, J.; Henry, R.; Spanton, S.; Dziki, W.; Porter, W.; Quick,
J.; Bauer, P.; Donaubauer, J.; Narayanan, B. A.; Soldani, M.;
Riley, D.; McFarland, K., Org. Process Res. Dev. 2000, 4, 413ꢀ
417.
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: crystallographic data (CCDC 972041ꢀ
972046) and hydrogenꢀbond parameters. This material is available
free of charge via the Internet at http://pubs.acs.org.
15. Di Martino, P.; GuyotꢀHermann, A. M.; Conflant, P.; Drache, M.;
Guyot, J. C., Int. J. Pharm. 1996, 128, 1ꢀ8.
16. Desiraju, G. R., Angew. Chem. Int. Ed. 1995, 34, 2311ꢀ2327.
17. Desiraju, G. R., Angew. Chem. Int. Ed. 2007, 46, 8342ꢀ8356.
18. Trask, A. V., Mol. Pharm. 2007, 4, 301ꢀ309.
19. Nehm, S. J.; RodríguezꢀSpong, B.; RodríguezꢀHornedo, N.,
Cryst. Growth Des. 2006, 6, 592ꢀ600.
20. Remenar, J. F.; Morissette, S. L.; Peterson, M. L.; Moulton, B.;
MacPhee, J. M.; Guzman, H. R.; Almarsson, O., J. Am. Chem.
Soc. 2003, 125, 8456ꢀ8457.
21. Childs, S. L.; Chyall, L. J.; Dunlap, J. T.; Smolenskaya, V. N.;
Stahly, B. C.; Stahly, G. P., J. Am. Chem. Soc. 2004, 126, 13335ꢀ
13342.
22. 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, 1888ꢀ1897.
23. Trask, A. V.; Motherwell, W. D. S.; Jones, W., Cryst. Growth
Des. 2005, 5, 1013ꢀ1021.
24. Sun, C. C.; Hou, H., Cryst. Growth Des. 2008.
25. Karki, S.; Friščić, T.; László, F.; Peter, R. L.; Graeme, M. D.;
William, J., Adv. Mater. 2009, 21, 3905ꢀ3909.
26. Chattoraj, S.; Shi, L.; Sun, C. C., CrystEngComm 2010, 12, 2466ꢀ
2472.
27. Shan, N.; Zaworotko, M. J., Drug Discov. Today 2008, 13, 440ꢀ
446.
ACKNOWLEDGEMENT
We thank AbbVie Inc. for funding this study. DKB
acknowledges the U.S. Pharmacopeia for a Fellowship
(2008ꢀ2009).
AUTHOR INFORMATION
Corresponding Author
Geoff.GZ.Zhang@abbvie.com and lenꢀmacgillivray@uiowa.edu
Present Addresses
† Department of Chemistry, University College London, London
WC1H 0AJ, UK
§This study was funded by AbbVie Inc. AbbVie participated in the
study design, research, data collection, analysis and interpretation
of data, as well as writing, reviewing, and approving the
publication. DejanꢀKrešimir Bučar is an employee at Department
of Chemistry, University College London, London, United
Kingdom, and has no conflicts of interest to report. Leonard R.
MacGillivray is an employee at Department of Chemistry,
University of Iowa, Iowa, USA, and has no conflicts of interest to
report. Rodger F. Henry and Geoff G. Z. Zhang are AbbVie
employees and may own AbbVie stock/options.
28. Schultheiss, N.; Newman, A., Cryst. Growth Des. 2009, 9, 2950ꢀ
2967.
29. Zhang, G. G. Z.; Henry, R. F.; Borchardt, T. B.; Lou, X., J.
Pharm. Sci. 2007, 96, 990ꢀ995.
30. Bučar, D.ꢀK.; Henry, R. F.; Lou, X.; Duerst, R. W.; Borchardt, T.
B.; Zhang, G. G. Z., Chem. Commun. 2007, 525ꢀ527.
31. Bučar, D.ꢀK.; Henry, R. F.; Lou, X.; Duerst, R. W.; MacGillivray,
L. R.; Zhang, G. G. Z., Cryst. Growth Des. 2009, 9, 1932ꢀ1943.
32. Bučar, D.ꢀK.; Henry, R. F.; Lou, X.; Duerst, R. W.; Borchardt, T.
B.; MacGillivray, L. R.; Zhang, G. G. Z., Mol. Pharm. 2007, 4,
339ꢀ346.
33. Bučar, D.ꢀK.; Henry, R.; Duerst, R.; Lou, X.; MacGillivray, L.;
Zhang, G., J. Chem. Crystallogr. 2010, 40, 933ꢀ939.
34. Sander, J. R. G.; Bučar, D.ꢀK.; Henry, R. F.; Baltrusaitis, J.;
Zhang, G. G. Z.; MacGillivray, L. R., J. Pharm. Sci. 2010, 99,
3676ꢀ3683.
REFERENCES
1. Qiu, Y.; Chen, Y.; Zhang, G. G. Z.; Liu, L.; Porter, W.,
Developing Solid Oral Dosage Forms: Pharmaceutical Theory &
Practice. 1 ed.; Elsevier: New York, 2009.
2. Bastin, R. J.; Bowker, M. J.; Slater, B. J., Org. Process Res. Dev.
2000, 4, 427ꢀ435.
3. Morissette, S. L.; Almarsson, Ö.; Peterson, M. L.; Remenar, J. F.;
Read, M. J.; Lemmo, A. V.; Ellis, S.; Cima, M. J.; Gardner, C. R.,
Adv. Drug Del. Rev. 2004, 56, 275ꢀ300.
4. Serajuddin, A. T. M., Adv. Drug Del. Rev. 2007, 89, 603ꢀ616.
5. Vishweshwar, P.; McMahon, J. A.; Bis, J. A.; Zaworotko, M. J.,
J. Pharm. Sci. 2006, 95, 499ꢀ516.
35. Sander, J. R. G.; Bučar, D.ꢀK.; Henry, R. F.; Giangiorgi, B. N.;
Zhang, G. G. Z.; MacGillivray, L. R., CrystEngComm 2013, 15,
4816ꢀ4822.
ACS Paragon Plus Environment

Crystal Growth & Design
Page 12 of 13
36. Elacqua, E.; Bučar, D.ꢀK.; Henry, R. F.; Zhang, G. G. Z.;
MacGillivray, L. R., Cryst. Growth Des. 2012, 13, 393ꢀ403.
37. Bučar, D.ꢀK.; Day, G. M.; Halasz, I.; Zhang, G. G. Z.; Sander, J.
R. G.; Reid, D. G.; MacGillivray, L. R.; Duer, M. J.; Jones, W.,
Chem. Sci. 2013, 4, 4417ꢀ4425.
38. Bis, J. A.; Vishweshwar, P.; Weyna, D.; Zaworotko, M. J., Mol.
Pharm. 2007, 4, 401ꢀ416.
1
2
3
4
5
39. Kavuru, P.; Aboarayes, D.; Arora, K. K.; Clarke, H. D.; Kennedy,
A.; Marshall, L.; Ong, T. T.; Perman, J.; Pujari, T.; Wojtas, L.;
Zaworotko, M. J., Cryst. Growth Des. 2010, 10, 3568ꢀ3584.
40. Shattock, T. R.; Arora, K. K.; Vishweshwar, P.; Zaworotko, M.
J., Cryst. Growth Des. 2008, 8, 4533ꢀ4545.
41. Aakeröy, C. B.; Epa, K. N.; Forbes, S.; Desper, J.,
CrystEngComm 2013, 15, 5946ꢀ5949.
42. MoraguesꢀBartolome, A. M.; Jones, W.; CruzꢀCabeza, A. J.,
CrystEngComm 2012, 14, 2552ꢀ2559.
43. B. Aakeröy, C.; Schultheiss, N.; Desper, J.; Moore, C., Cryst.
Growth Des. 2007, 7, 2324ꢀ2331.
44. Aakeröy, C. B.; Chopade, P. D.; Desper, J., Cryst. Growth Des.
2013, 13, 4145ꢀ4150.
45. Aakeröy, C. B.; Desper, J.; Elisabeth, E.; Helfrich, B. A.; Levin,
B.; Urbina, J. F., Z. Kristallogr. 2005, 220, 325ꢀ332.
46. Aakeröy, C. B.; Beatty, A. M.; Helfrich, B. A., Angew. Chem. Int.
Ed. 2001, 40, 3240ꢀ3242.
47. Bhogala, B. R.; Nangia, A., New. J. Chem. 2008, 32, 800ꢀ807.
48. Lee, H. G.; Zhang, G. G. Z.; Flanagan, D. R., J. Pharm. Sci.
2011, 100, 1736–1744.
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
49. Grobelny, P.; Mukherjee, A.; Desiraju, G. R., CrystEngComm
2011, 13, 4358ꢀ4364.
50. Évora, A. O. L.; Castro, R. A. E.; Maria, T. M. R.; Rosado, M. T.
S.; Ramos Silva, M.; Matos Beja, A.; Canotilho, J.; Eusébio, M.
E. S., Cryst. Growth Des. 2011, 11, 4780ꢀ4788.
51. Cherukuvada, S.; Nangia, A., CrystEngComm 2012, 14, 2579ꢀ
2588.
52. Cheney, M. L.; Weyna, D. R.; Shan, N.; Hanna, M.; Wojtas, L.;
Zaworotko, M. J., J. Pharm. Sci. 2011, 100, 2172–2181.
53. Bhatt, P. M.; Azim, Y.; Thakur, T. S.; Desiraju, G. R., Cryst.
Growth Des. 2009, 9, 951–957.
54. Aitipamula, S.; Chow, P. S.; Tan, R. B. H., CrystEngComm 2009,
11, 1823ꢀ1827.
55. Žegarac, M.; Lekšić, E.; Šket, P.; Plavec, J.; Devčić Bogdanović,
M.; Bučar, D.ꢀK.; Dumić, M.; Meštrović, E., CrystEngComm
2014, 16, 32ꢀ35.
56. Scheindlin, S., Mol. Interv. 2007, 3, 236ꢀ242.
57. Allen, F., Acta Crystallogr. B 58, 380ꢀ388.
58. Bruno, I. J.; Cole, J. C.; Edgington, P. R.; Kessler, M.; Macrae, C.
F.; McCabe, P.; Pearson, J.; Taylor, R., Acta Crystallogr. B 2002,
58, 389ꢀ397.
59. Sonicator bath: Branson 2510RꢀDTM (frequency: 42kHz±6% at
100W).
60. Symphonix, Inel Corp.: Artenay, France.
61. Jade 6.5; Materials Data, Inc.: Livermore, CA, USA.
62. SaintPlus, Bruker AXS Inc.: Madision, WI, USA, 2007.
63. Sheldrick, G. M., Acta Cryst. 2008, A64, 112ꢀ122.
64. L. J. Farrugia, J. Appl. Cryst. 1999, 32, 837ꢀ838.
65. Madarasz, J.; Bombicz, P.; Jarmi, K.; Ban, M.; Pokol, G.; Gal, S.,
J. Therm. Anal. Calorim. 2002, 69.
66. Childs, S. L.; Stahly, G. P.; Park, A., Mol. Pharm. 2007, 4, 323ꢀ
338.
67. Aitipamula, S.; Chow, P. S.; Tan, R. B. H., Acta Crystallogr. E
2009, 65, o2126ꢀo2127.
68. Habgood, M.; Price, S. L., Cryst. Growth Des. 2010, 10, 3263ꢀ
3272.
69. IꢀLab, 11.02; Advanced Chemistry Development, Inc.
(ACD/Labs): 1994ꢀ2011.
70. Gilli, G.; Gilli, P., The nature of the hydrogen bond ꢀ Outline of a
comprehensive hydrogen bond theory. Oxford University Press:
Oxford, 2009.
71. CruzꢀCabeza, A. J., CrystEngComm 2012, 14, 6362ꢀ6365.
72. Desiraju, G. R., J. Chem. Soc., Chem. Commun. 1991, 426ꢀ428.
ACS Paragon Plus Environment

Page 13 of 13
Crystal Growth & Design
1
2
Graphical Abstract
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
A systematic study on the solidꢀstate structures of coꢀcrystals of theophylline and a variety of
hydroxybenzoic acids is reported. Structural analyses revealed architectures based on intermolecular
homoꢀ and heteroꢀ supramolecular synthons, as well as intramolecular heterosynthons. Synthon
hierachies are discussed and compared to caffeineꢀbased cocrystals
ACS Paragon Plus Environment