Evidence for Halogen Bonding in Amorphous Solid Dispersions

Article abstract of DOI:10.1021/acs.cgd.0c00073

Carbon-bound halogen atoms (iodine, bromine, chlorine, and occasionally fluorine) are known to act as electron acceptors and form interactions with different species capable of acting as electron donors. This interaction is termed halogen bonding and shar

Full text of DOI:10.1021/acs.cgd.0c00073

Subscriber access provided by University of Rochester | River Campus & Miner Libraries
Article
Evidence for Halogen Bonding in Amorphous Solid Dispersions
Chailu Que, Qingqing Qi, Dmitry Y. Zemlyanov, Huaping Mo, Alexandru Deac,
Matthias Zeller, Anura S. Indulkar, Yi Gao, Geoff G.Z. Zhang, and Lynne S. Taylor
Cryst. Growth Des., Just Accepted Manuscript • DOI: 10.1021/acs.cgd.0c00073 • Publication Date (Web): 20 Mar 2020
Downloaded from pubs.acs.org on March 22, 2020
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 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 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.
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 37
Crystal Growth & Design
1
2
3
4
5
6
7
8
9
Evidence for Halogen Bonding in Amorphous Solid
Dispersions
Chailu Que,1 Qingqing Qi,2 Dmitry Y. Zemlyanov,3 Huaping Mo,4 Alexandru Deac,1 Matthias
Zeller,2 Anura S. Indulkar,5 Yi Gao,6 Geoff G. Z. Zhang,*5 and Lynne S. Taylor*1
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
1. Department of Industrial and Physical Pharmacy, College of Pharmacy, Purdue University,
West Lafayette, Indiana 47907, United States
2. Department of Chemistry, College of Science, Purdue University, West Lafayette, Indiana
47907, United States
3. Birck Nanotechnology Center, Purdue University, West Lafayette, Indiana 47907, United
States
4. Purdue Interdepartmental NMR Facility, Department of Medicinal Chemistry and Molecular
Pharmacology, Purdue University, West Lafayette, Indiana 47907
5. Drug Product Development, Research and Development, AbbVie Inc., North Chicago, Illinois
60064, United States
6. Science and Technology, Operations, AbbVie Inc., North Chicago, Illinois 60064, United
States
Corresponding authors: *Geoff G. Z. Zhang (Telephone: +1-847-937-4702; Fax: +1-847-937-
7756) and *Lynne S. Taylor (Telephone: +1-765-496-6614; Fax: +1-765-494-6545).
E-mail addresses: Geoff.GZ.Zhang@abbvie.com (G.G.Z. Zhang), lstaylor@purdue.edu (L.S.
Taylor)
1
ACS Paragon Plus Environment

Crystal Growth & Design
Page 2 of 37
1
2
3
4
Abstract
5
6
7
8
Carbon-bound halogen atoms (iodine, bromine, chlorine and occasionally fluorine) are known to
act as electron acceptors and form interactions with different species capable of acting as
electron donors. This interaction is termed halogen bonding, and shares some similarities with
hydrogen bonding. While hydrogen bonding has been extensively studied as an important
intermolecular interaction for many aspects drug delivery and dosage form design, halogen
bonding has not been investigated in these contexts. However, approximately one third of all
drug molecules contain halogen atoms. Using X-ray photoelectron spectroscopy, nuclear
magnetic resonance spectroscopy, and infrared spectroscopy, it was demonstrated herein that
halogen-bearing drugs that lack hydrogen bond donors form a halogen bonding interaction with
the carbonyl group of a common pharmaceutical polymer, copovidone, in a solid amorphous
blend. Halogen bonding thus may be a common, but neglected intermolecular interaction in
drug-polymer amorphous blends, and potentially important for the properties and performance of
these systems which are used to enhance drug solubility and delivery. This is the first time that
the occurrence of halogen bonding has been observed in drug-polymer formulations.
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
Keywords
Amorphous solid dispersions, halogen bonding, X-ray photoelectron spectroscopy, nuclear
magnetic resonance, infrared spectroscopy
2
ACS Paragon Plus Environment

Page 3 of 37
Crystal Growth & Design
1
2
3
4
1. Introduction
5
6
7
8
Drugs are rarely delivered to the body in pure form. Instead, highly engineered dosage forms are
necessary in order for the drug to reach its intended site of action at the right dose and rate. Of
the many challenges in drug delivery, poor aqueous solubility of modern drug candidates is one
of the most prevalent issues.1, 2 Miscible blends of a lipophilic drug with a hydrophilic polymer,
termed amorphous solid dispersions (ASDs), are one of the most popular formulation strategies
for oral delivery of poorly soluble drugs. It is widely recognized that using an ASD formulation
can greatly improve drug bioavailability relative to the crystalline form.3, 4 The intermolecular
interactions between drug and polymer are considered critical to ASD formulation
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
performance.5-8 One of the most frequently studied intermolecular interactions in ASDs is
hydrogen bonding. Several studies have suggested that hydrogen bonding helps delay drug
crystallization in the solid state,9-11 maintain the supersaturation generated by dissolving an
ASD,12 and potentially can improve dissolution performance.6 In contrast, halogen bonding,
another noncovalent interaction, has not been investigated for these drug formulations.
Halogen bonding (XB) is the attractive interaction between the electrophilic region of halogen
(F, Cl, Br, I) substituents and Lewis bases. Even though halogen atoms have the capability to
donate a lone pair of electrons, the electrophilic region on the halogen atoms, termed the σ hole,
and located trans to its σ bond, can attract electron-rich sites, such as oxygen, nitrogen, and
halogen atoms.13 Considering the fact that both halogen bond donors and acceptors are
electronegative, the existence of halogen bonding is somewhat counterintuitive. According to a
survey of ligand-protein interactions from the Protein Data Bank, the most common element
acting as electron donor in halogen bonds are carbonyl oxygen atoms, followed by nitrogen.14, 15
Although it has been only approximately a half century since halogen bonding was first
3
ACS Paragon Plus Environment

Crystal Growth & Design
Page 4 of 37
1
2
3
4
5
6
7
8
9
demonstrated by Hasselt in 1960s,16 this interaction has been explored and exploited in many
fields including crystal engineering,17, 18 material science,19 protein-ligand interactions,20, 21
supermolecular gel formation,22 and catalysis.23, 24 Halogen bonding also has been used to form
pharmaceutical co-crystals to modify the physicochemical properties of active pharmaceutical
ingredients and excipients.25-27 In general, the strength of the halogen bond depends on the size
of the σ-hole, following the trend of I > Br > Cl > F.13 However, the strength can be tuned by
changing the chemical environment of the halogens, e.g. by adding electron donating or
withdrawing substituents in the vicinity of the carbon atom attached to the halogens. Among the
halogens, F is the least prone to be involved in halogen bonding, only forming halogen bonds
when fluorine atoms in the vicinity to strong electron-withdrawing groups.28 Halogen bonds tend
to be directional, and the ideal C—XA contact angle between the σ-hole and the halogen bond
acceptor A is close to 180°.29 For optimized systems, the strength of the halogen bond can be
comparable to that of a hydrogen bond.13 In a study of protein-ligand complexes, the interaction
energies for halogen bonds were estimated to range from -1.97 to -5.4 kcal/mol, where the
stronger halogen bonds have comparable interaction energies as a typical hydrogen bond.30 For
example, the interaction energy of the hydrogen bond between hydroxyl in methanol and
carbonyl in acetone is about -5.3 kcal/mol.31 It has been shown in several crystal engineering
studies that when both intermolecular interactions are available, hydrogen bonding is favored
over halogen bonding, but that both interactions can co-exist.18, 32 Another interesting
characteristic of halogen bonding is that it tends to be less sensitive to, and persists in a polar
environment.29, 33
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 most frequently used experimental technique for studying halogen bonding is X-ray
crystallography, which enables measurement of the interatomic distance between halogen donor
4
ACS Paragon Plus Environment

Page 5 of 37
Crystal Growth & Design
1
2
3
4
5
6
7
8
9
and acceptor.34 However, this technique can only be applied to crystalline materials. Other
techniques, such as solid and solution state nuclear magnetic resonance (NMR),33, 35 infrared (IR)
spectroscopy,36 Raman spectroscopy,37 and X-ray photoelectron spectroscopy (XPS)19 have been
employed to characterize halogen bonding. Density functional theory (DFT) calculations are
often used to confirm experimental observations of halogen bonding.38
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
Investigations into halogen bonding are gaining traction, and the occurrence of such interactions
in drug formulations is of considerable relevance given that more than 30% of the drug
molecules in development contain halogens, and halogen bonds have been observed in a variety
of biomolecular systems.14 Medicinal chemists have taken advantage of halogen bonding in the
design of new therapeutics.20, 39 However, the implications of halogen bond formation in drug
formulation has been ignored. Of particular interest in this context, are interspecies interactions
for halogen-bearing drugs dispersed in polymers. Polymers are widely used in drug formulations,
and play a pivotal role in the formation of ASDs, a solubility enhancing platform for low
aqueous solubility drugs.40 While several types of intermolecular drug-polymer interactions have
been described previously including hydrogen bonding and ionic interactions,10, 41-43 halogen
bonds in these systems have not been reported to date.
In the current study, we chose four model compounds that contain Cl, Br or I, but lack hydrogen
bond donors (shown in Figure 1), where clotrimazole and loratadine are marketed drugs.
Brotrimazole is a derivative of clotrimazole obtained by replacing Cl with Br. Compound 3, 3-
bis(3,5-diiodo-4-methoxyphenyl)isobenzofuran-1(3H)-one (Me-DIBF) is a methylated derivative
of a protein inhibitor.44 The polymers investigated (Figure 1) were copovidone (PVPVA), which
is present in many important commercial ASD formulations,40 and the somewhat simpler, but
closely related polymer, povidone (PVP). To demonstrate and investigate halogen bonding
5
ACS Paragon Plus Environment

Crystal Growth & Design
Page 6 of 37
1
2
3
4
5
6
7
8
9
between model compounds and polymers, X-ray photoelectron spectroscopy was used to monitor
changes in the electronic environment of relevant atoms. Solution nuclear magnetic resonance
spectroscopy was used to measure the binding strength between donor and acceptor. Lastly,
infrared spectroscopy was used to provide confirmation of the acceptor group. For the first time,
the presence of drug-polymer halogen bonding is demonstrated in amorphous formulations.
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
2. Materials
Clotrimazole (Figure 1a) was purchased from Sigma-Aldrich (St. Louis, MO, USA). Loratadine
(Figure 1c) was attained from Attix Pharmaceuticals (Toronto, ON, Canada). Compound 3,3-
bis(3,5-diiodo-4-methoxyphenyl)isobenzofuran-1(3H)-one (Me-DIBF) and 1-((2-
bromophenyl)diphenylmethyl)-1H-imidazole, (brotrimazole), shown in Figure 1b and 1d
respectively, were synthesized in-house, and the synthesis procedures are described in detail in
the methods section. Copovidone (PVPVA) 64 (Figure 1e) and povidone (PVP K29-32) (Figure
1f) were obtained from Ashland (Covington, KY). Deuterated toluene, containing 0.03 % (v/v)
tetramethylsilane (TMS), deuterated cyclohexane, TMS, 1-ethylpyrrolidine (Figure 1g), 1-ethyl-
2-pyrrolidone (Figure 1h) were all purchased from Sigma-Aldrich (St. Louis, MO, USA).
Ethanol 200 proof anhydrous, tetrahydrofuran (THF), and all the chemicals for synthesis were
bought from Fisher Scientific (Hampton, NH, USA).
(a)
(b)
6
ACS Paragon Plus Environment

Page 7 of 37
Crystal Growth & Design
1
2
3
4
5
6
Cl
Br
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
N
N
N
N
(c)
(d)
O
OCH3
O
I
I
N
I
I
OCH3
N
O
O
Cl
(e)
(f)
H2
C
H2
C
H2
C
H
C
HC
N
HC
N
O
O
O
O
C
m
n
n
(g)
(h)
CH3
CH3
N
N
O
7
ACS Paragon Plus Environment

Crystal Growth & Design
Page 8 of 37
1
2
3
4
5
6
Figure 1.
Molecular structure of (a) clotrimazole, (b) brotrimazole, (c) loratadine, (d)
3,3-bis(3,5-diiodo-4-methoxyphenyl)isobenzofuran-1(3H)-one (Me-DIBF), (e)
PVPVA , (f) PVP, (g)1-ethylpyrrolidine, and (h) 1-ethyl-2-pyrrolidone.
7
8
9
3. 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
3.1 Synthesis of Brotrimazole
Br
Br
Br
SOCl2, rt
Ether, 0 oC to rt
MgBr
Ph
OH
Ph
Cl
COOCH3
Ph
Ph
Ph
Br
Cl
Br
N
DCM, 0 oC to rt
N
Et3N
N
HN
Ph
Ph
Ph
Ph
Scheme 1. Synthesis of brotrimazole (1-((2-Bromophenyl)diphenylmethyl)-1H-imidazole).
(2-Bromophenyl)diphenylmethanol The preparation was as described previously45: to a stirred
solution of methyl 2-bromobenzoate (2.6 g, 12 mmoL, 1.0 equiv) in anhydrous Et2O (40 mL) at 0
oC under nitrogen was added PhMgBr (8.8 mL, 3.0 M in Et2O, 2.2 equiv). The reaction mixture
was stirred at room temperature overnight. The reaction was quenched by sat. NH4Cl solution (7
mL). The mixture was extracted by Et2O, the organic layers were combined and the solvent was
removed under vacuum. The residue was purified by silica gel chromatography with hexane as the
eluent to afford the product as white solid (3.5 g, 86%).
((2-Bromophenyl)chloromethylene)dibenzene (2-bromophenyl)diphenylmethanol (3.4 g, 10.0
mmol, 1.0 equiv) was mixed with SOCl2 (10 mL, excess) and several drops of DMF were added.
The mixture was stirred at room temperature overnight. The stirring was stopped upon the
completion of reaction as monitored by 1H NMR of the crude mixture. Then SOCl2 was removed
8
ACS Paragon Plus Environment

Page 9 of 37
Crystal Growth & Design
1
2
3
4
5
6
7
8
9
under vacuum. To the residue was added DCM (5 mL) and the solution was concentrated in
volume again and this was repeated for three times to afford ((2-
bromophenyl)chloromethylene)dibenzene which was used directly in the next step without further
purification.
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
Brotrimazole (1-((2-Bromophenyl)diphenylmethyl)-1H-imidazole) To a precooled solution of
2-bromophenyl)chloromethylene)dibenzene (10.0 mmol, 1.0 equiv) in anhydrous DMF (20 mL)
at 0 °C under argon was added 1H-imidazole (1.0 g, 15.0 mmol, 1.5 equiv) and triethylamine
(4.2 mL, 30.0 mmol, 3.0 equiv) via a syringe. The reaction mixture was allowed to warm to
room temperature and stirred overnight. 50 mL ethyl acetate was added and the solution was
washed with brine for three times, and dried over Na2SO4. After filtration and concentration, the
crude mixture was purified by column chromatography (silica gel, 50% ethyl acetate in hexanes)
to afford the title product (3.0 g, 77.1%). 1H NMR (800 MHz, Chloroform-d) δ 7.66 (dd, J = 7.9,
1.5 Hz, 1H), 7.49 (t, J = 1.2 Hz, 1H), 7.36 – 7.32 (m, 6H), 7.30 – 7.28 (m, 1H), 7.24 – 7.21 (m,
13
5H), 7.07 (d, J = 1.3 Hz, 1H), 6.91 (dd, J = 8.0, 1.7 Hz, 1H), 6.76 (d, J = 1.5 Hz, 1H). C NMR
(201 MHz, CDCl3) δ 142.19, 140.36, 139.24, 135.90, 130.93, 130.52, 129.86, 128.46, 128.17,
127.97, 127.56, 125.90, 121.63. HRMS (m/z, APCI): calcd for C22H18BrN2(M+H)+: 389.0653.
Found: 389.0639. 1H and 13C NMR spectra for the final product are provided in the supporting
information.
3.2 Synthesis of 3,3-bis(3,5-diiodo-4-methoxyphenyl)isobenzofuran-1(3H)-one (Me-DIBF)
HO
O
I
I
I
I
I
I
OH
I
O
I
DMF, rt
K2CO3
MeI
O
O
O
O
9
ACS Paragon Plus Environment

Crystal Growth & Design
Page 10 of 37
1
2
3
4
5
Scheme 2. Synthesis of 3,3-bis(3,5-diiodo-4-methoxyphenyl)isobenzofuran-1(3H)-one (Me-
DIBF).
6
7
8
9
To a solution of 3,3-bis(4-hydroxy-3,5-diiodophenyl)isobenzofuran-1(3H)-one (6.6 g, 8.0 mmol,
1.0 equiv) in anhydrous DMF (50 mL) at RT under argon was added K2CO3 (4.4 g, 32.0 mmol,
4.0 equiv) and MeI (2.5 mL, 40.0 mmol, 5.0 equiv) via a syringe. The reaction mixture was
stirred overnight. The mixture was poured into water and filtered to afford the title product as a
white solid (6.5g, 95.6%). 1H NMR (800 MHz, Chloroform-d) δ 8.00 – 7.98 (m, 1H), 7.83 – 7.81
(m, 1H), 7.69 – 7.65 (m, 5H), 7.58 – 7.56 (m, 1H), 3.88 (s, 6H). 13C NMR (201 MHz, CDCl3) δ
168.44, 159.52, 149.85, 139.44, 138.09, 134.97, 130.31, 126.72, 125.09, 123.78, 91.01, 87.36,
60.73. HRMS (m/z, ESI): calcd for C22H15I2O4(M+H)+: 850.7149. Found: 850.7137. 1H and 13C
NMR spectra for the final product are shown in the supporting 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
3.3 Single Crystal X-ray Structure Determination
Crystallization and Data Collection
Single crystals were grown from a solution of dichloromethane and ethyl acetate using hexane
vapor diffusion. Two crystals were selected for data collection and analyzed to ensure
reproducibility of results. The crystals were mounted on a Mitegen micromesh mount using a
trace of mineral oil and cooled to 150 K. Data were collected on a Bruker D8 Quest Single
Crystal CMOS Diffractometer (Bruker, Nano Inc, Madison, WI) with a sealed tube X-ray source
generating Mo-Kα radiation (λ = 0.71073 Å), a curved triumph monochromator with a 10 cm x
10 cm Photon-100 detector and fixed chi angle, and an Oxford Cryostream 800 plus variable
temperature device (80-500K). Reflections were collected and processed, and the data was
corrected for absorption and scaled by the multi-scan method using the Apex3 software v2018.1-
0 (Bruker Nano Inc, Madison, WI).
Structure Determination and Refinement
10
ACS Paragon Plus Environment

Page 11 of 37
Crystal Growth & Design
1
2
3
4
5
6
7
8
9
The space group was assigned and the structure was solved using XPREP within the SHELXTL
suite of programs46, 47 by direct methods. Structure refinement was performed in SHELXL
version 2018/348, 49 by full matrix least squares on F2 with all reflections. Hydrogen atoms
attached to carbons were geometrically placed in calculated positions and were constrained to
ride on their parent atoms, such that the hydrogen-carbon bond distance for the aromatic C-H
was 0.95 Å. Uiso(H) values for the C-H units were set to be a 1.2 multiple of Ueq(C). The
imidazole and one of the phenyl groups were refined as disordered. The imidazole was refined in
three parts such that two imidazole moieties were placed in similar positions but rotated 180⁰
againstone another, while the position of the third imidazole moiety was swapped with the
position of one of the phenyl rings. Similarly, the phenyl ring was refined in three parts such that
two phenyl moieties were placed in similar positions but slightly shifted, coinciding with the
rotation of the imidazole, while the position of the third phenyl moiety was swapped with the
third imidazole moiety. Uij components of all thermal parameters were restrained to be similar to
each other. The bond distance between the central carbon atom and the carbon atom on the
disordered phenyl ring was restrained to be similar to the bond distance between the central
carbon atom and the carbon atom of one of the non-disordered phenyl rings for all disordered
phenyl moieties. Similarly, the bond distance between the central carbon atom and the nitrogen
for the minor moieties was restrained to be similar to that for the major moiety. Lastly, the
diffraction intensities were corrected for extinction. Complete crystallographic data, in CIF
format, have been deposited with the Cambridge Crystallographic Data Center: CCDC 1964776.
Additional crystal data, data collection and refinement details can be found in the supporting
information. The data can be obtained free of charge from The Cambridge Crystallographic Data
Centre via. www.ccdc.cam.ac.uk/datarequest/cif.
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
11
ACS Paragon Plus Environment

Crystal Growth & Design
Page 12 of 37
1
2
3
4
3.4. Preparation of Amorphous Solid Dispersions (ASDs)
5
6
All amorphous solid dispersions were prepared by solvent evaporation using a rotatory
7
8
9
evaporator (Brinkman Instruments, Westbury, NY). The organic solvent used to dissolve both
drug and polymer was ethanol, except for Me-IDBF, where tetrahydrofuran was used. The
temperature of the water bath was 60 °C and 40 °C for ethanol and tetrahydrofuran, respectively.
Subsequently, the ASDs were placed in a vacuum oven for a day to remove residual organic
solvent. Clotrimazole-PVPVA, clotrimazole-PVP, and loratadine-PVPVA ASDs were prepared
with drug loadings of 10, 20, 25, 30, 40 and 50 wt. %. Brotrimazole-PVPVA and Me-DIBF-
PVPVA ASDs were prepared with drug loading of 10, 20, 30 and 40 wt. %.
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
3.5 X-ray Photoelectron Spectroscopy (XPS)
XPS data were obtained by a Kratos Axis Ultra DLD spectrometer using monochromic Al Kα
radiation (1486.6 eV). The spectra was collect at constant pass energy (PE) at 20 and 160 eV for
high-resolution and survey spectra, respectively. A commercial build-in Kratos charge
neutralizer was used to avoid non-homogeneous electric charge of non-conducting powder and to
achieve better resolution. All the ASD samples for XPS were compressed into tablets prior to
examination. 100 mg ASD powder were compressed at 1500 psi using a hydraulic press (Carver
Inc, Wabash, IN) in a round die with diameter of 8 mm, and the compression pressure was held
for 60s. Then the tablets were ejected manually from the die. The tablet were placed on a
stainless steel sample holder bar using a double-sided sticking Cu tape. The XPS spectra were
collected at 5 different points for each tablet. The size of analysis spots was 0.5 mm in diameter.
XPS data were analyzed with the CasaXPS software (www.casaxps.com). Prior to data analysis,
the C-C component of the C 1s peak was set to a binding energy (BE) of 284.8 eV to correct for
charge at each acquisition spot. Curve-fitting was performed following a Shirley background
12
ACS Paragon Plus Environment

Page 13 of 37
Crystal Growth & Design
1
2
3
4
5
6
7
8
9
subtraction using Gaussian/Lorentzian peak shapes. The atomic concentrations of the elements in
the near-surface region were estimated after a Shirley background subtraction taking into account
the corresponding Scofield atomic sensitivity factors and inelastic mean free path (IMFP) of
photoelectrons using standard procedures in the CasaXPS software assuming homogeneous
mixture of the elements within the information depths (~10 nm).
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
3.6 Nuclear Magnetic Resonance (NMR) Spectroscopy
1H and 13C NMR were performed using Bruker Avance-III-800 spectrometer (Billerica, MA)
equipped with a QCI cryoprobe. Titration was used to obtain binding constants. Deuterated
cyclohexane was chosen as a solvent to ensure sufficient solute solubility and to avoid
interference with halogen bonding. Chemical shifts were reported relative to tetramethylsilane
(TMS). All samples were equilibrated at 25 °C for at least 5 min before data collection. To
evaluate the weak interaction between drug or its analogues and 1-ethylpyrrolidinone (which is a
monomeric analogue of PVP and maintains good solubility in toluene), we monitored model
compound chemical shifts by EP titrations. The general procedure for titrations was as follows: a
deuterated toluene stock solution with the host compound (concentration of 14.5 mM) was
divided amongeight NMR tubes (0.6 mL each). A variable amount of 1-ethyl-2-pyrrolidone (EP)
as guest was added to each tube to provide a range of halogen bonding acceptor concentrations.
The guest concentrations used were 11.0, 87.5, 173.2, 339.9, 654.6, 946.8 and 1218.9 mM for
each titration sample. The chemical shift of the host molecule was monitored and the titration
curves were fitted to a 1:1 binding model, which is a hyperbolic function, using OriginPro 2019
(Originlab Corporation, Northampton, MA) to determine the binding constant KF, which is given
by equation 1:
[퐴퐷]
퐾퐹 = [퐴][퐷]
(1)
13
ACS Paragon Plus Environment

Crystal Growth & Design
Page 14 of 37
1
2
3
4
5
6
where [AD] is the concentration of the halogen-bonded complex, while [A] and [D] are the
concentrations of the free acceptor and the free donor, respectively.
7
8
9
Using NMR spectroscopic titration to attain binding constants is a widely used approach.33, 50-52
Titrations were repeated three times for reproducibility and estimation of errors, one set of
representative NMR spectra and the titration curves for each model compound are presented
here. The binding constants were determined by both 1H NMR and 13C NMR. For control
experiments, 1-ethylpyrrolidine, which lacks the carbonyl group present in 1-ethylpyrrolidinone,
was added to the host solution to examine its impact on the chemical shift.
3.7 Bulk Infrared (IR) Spectroscopy
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
Thin films of pure PVPVA, clotrimazole-PVPVA mixtures, brotrimazole-PVPVA mixtures, and
Me-DIBF-PVPVA mixtures were spin-coated on thallium bromoiodide (KRS-5) windows
(Harrick Scientific Corporation, Ossining, NY) using a spin coater (Chemat Technology Inc.,
Northridge, CA). The general procedure for spin coating was as follows: 50 µL of the PVPVA or
drug-PVPVA ethanol solution with solid content of 100 mg/mL was deposited onto the substrate,
then it was spun for 15 s at 50 rpm followed by 50 s at 2500 rpm. The IR spectra were collected
using a Bruker Vertex 70 FTIR spectrometer (Bruker, Billerica, MA). For each sample, 128
scans were collected for both background and samples at a resolution of 4 cm-1. The data were
analyzed utilizing the OPUS software (version 7.2, Bruker Billerica, MA). IR spectra for
loratadine-PVPVA were not collected due to the overlap of the carbonyl group in loratadine with
PVPVA carbonyl group region.
4. Results
4.1 X-ray Photoelectron Spectroscopy (XPS)
14
ACS Paragon Plus Environment

Page 15 of 37
Crystal Growth & Design
1
2
3
4
5
6
7
8
9
The chemical composition of all the neat model compounds and ASDs was analyzed using XPS.
The binding energies of the Cl 2p electrons in clotrimazole and loratadine, the Br 3d electrons in
brotrimazole, and the I 3d electrons in Me-DIBF were measured respectively. The 2p and 3d
core-level spectra are characterized by spin-orbit splitting, and therefore show two components.
The ratios between the two components are 2:1 and 3:2 for p and d orbitals, respectively. The
spin-orbit splitting for Cl 2p, Br 3d, and I 3d are 1.6, 1.0 and 11.5 eV respectively. The Cl 2p
spectrum of pure clotrimazole exhibits one pair of peaks at 200.4 ± 0.1 (Cl 2p3/2) and 202.0 ± 0.1
eV (Cl 2p1/2) as shown in Figure 2a. However, when the clotrimazole molecules were dispersed
into PVPVA, two new pairs of Cl peaks at 198.1 ± 0.2 (Cl 2p3/2), 199.8 ± 0.2 (Cl 2p1/2) and 196.7
± 0.4 (Cl 2p3/2), 198.3 ± 0.1 eV (Cl 2p1/2) emerged (Figure 2a). Similar results were obtained for
clotrimazole and PVP (Figure 2b). The Cl 2p peaks for loratadine showed a similar pattern,
where two extra pairs of Cl 2p peaks with lower binding energy emerged in ASDs (Figure 2 d).
The Br 3d spectrum from neat brotrimazole also showed an extra pair of photoelectron peaks
with lower binding energy, corresponding to a 25% atomic ratio. The probable explanation is
presence of two energetically similar molecular conformations of brotrimazole (see Figure 5S).
This possibility is supported by crystal structure refinement of pure brotrimazole, which revealed
disorder of the imidazole and phenyl substituents switching positions 18% of time, resulting in
Br interacting with a phenyl ring instead of imidazole. In general, the phenyl ring is a better
electron donating group. Therefore, the portion of Br atoms with a lower binding energy are
likely a result of Br interacting with a phenyl ring instead of imidazole. In contrast, in the ASD
matrix, Br 3d peaks, occurring at lower binding energies of 67.1 ± 0.1, 68.1 ± 0.1 eV, were
dominant (Figure 2c). In Figure 2e, the neat Me-DIBF compound showed only one pair of I 3d
photoelectron peaks with binding energy of 620.7 ± 0.1 (I 3d5/2), 632.2 ± 0.1 eV (I 3d1/2) eV,
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
15
ACS Paragon Plus Environment

Crystal Growth & Design
Page 16 of 37
1
2
3
4
5
6
7
8
9
while an extra pair of peaks at 618.1 ± 0.1 and 629.6 ± 0.1 eV were observed in the ASDs.
Binding energies for all systems are summarized in Table 2. For each of these compounds, which
contained different halogen atoms, the halogen atom showed additional lower binding energy
peaks in the XPS spectrum when the compounds were molecularly mixed with PVPVA or PVP.
A binding energy shift to a lower energy is an indication that the atom has become more
electron-rich. Thus, in the ASD matrix, the halogen atoms likely received electrons from
carbonyl groups present in the polymers, PVPVA or PVP.
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
16
ACS Paragon Plus Environment

Page 17 of 37
Crystal Growth & Design
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
Figure 2.
Representative halogen X-ray photoelectron spectrum for (a)
clotrimazole/PVPVA ASDs, (b) clotrimazole/PVP K29-32 ASDs, (c)
brotrimazole/PVPVA ASDs, (d) loratadine/PVPVA ASDs, and (e) Me-
17
ACS Paragon Plus Environment

Crystal Growth & Design
Page 18 of 37
1
2
3
4
5
DIBF/PVPVA ASDs. The X-ray photoelectron spectrum of neat model
compound were also shown for comparison.
6
7
8
Table 1.
Binding energy for neat materials and for peaks shifted to lower binding
energies when dispersed in polymers.
9
Low Binding Energy Peaks when
Neat Compound (eV)
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
Dispersed in Polymer (eV)
Clotrimazole
(PVPVA)
198.1 ± 0.2, 199.8 ± 0.2
200.4 ± 0.1, 202.0 ± 0.1
196.7 ± 0.4, 198.3 ± 0.4
Clotrimazole
(PVP K29-32)
197.9 ± 0.1, 199.5 ± 0.1
200.4 ± 0.1, 202.0 ± 0.1
196.8 ± 0.1, 198.4 ± 0.1
Brotrimazole
(PVPVA)
70.3 ± 0.1, 71.2 ± 0.1
200.6 ± 0.04, 202.2 ± 0.1
620.7 ± 0.1, 632.2 ± 0.1
67.1 ± 0.1, 68.1 ± 0.1
Loratadine
(PVPVA)
198.0 ± 0.1, 199.6 ± 0.1
196.8± 0.0, 198.4 ± 0.0
Me-DIBF
(PVPVA)
618.1 ± 0.1, 629.6 ± 0.1
By deconvoluting the XPS spectra, the percentage of halogen atoms showing low binding energy
peaks was calculated for different drug loadings, with results summarized in Figure 3 and
numerical values of the percentage of the lower binding energy peaks given in Table S2. We
rationalized that at low drug loading, the drug molecules are further apart, and hence a majority
have a higher tendency to interact with polymer. In contrast, at high drug loading, the drug
molecules are in closer proximity to each other, and a majority have a greater likelihood of self-
interactions. The trend among all ASDs is that the percentage of halogen atoms possessing a
lower binding energy decreased as the drug load increased. This supports the idea that PVPVA is
donating electrons to the halogens.
18
ACS Paragon Plus Environment

Page 19 of 37
Crystal Growth & Design
1
2
3
4
5
6
7
8
9
Cl 2p with Higher Binding Energy
Cl 2p with Lower Binding Energy
(a) Clotrimazole/PVPVA
140.0
120.0
100.0
80.0
60.0
40.0
20.0
0.0
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
10% DL 20% DL 25% DL 30% DL 40% DL 50% DL
140.0
120.0
100.0
80.0
60.0
40.0
20.0
0.0
Cl 2p with Higher Binding Energy
Cl 2p with Lower Binding Energy
(b) Clotrimazole/PVP
10% DL
20% DL
30% DL
40% DL
50% DL
19
ACS Paragon Plus Environment

Crystal Growth & Design
Page 20 of 37
1
2
3
4
5
6
7
8
9
Cl 2p with Higher Binding Energy
Cl 2p with Lower Binding Energy
(c) Loratadine/PVPVA
140.0
120.0
100.0
80.0
60.0
40.0
20.0
0.0
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
10% DL
20% DL
25% DL
30% DL
40% DL
140.0
120.0
100.0
80.0
60.0
40.0
20.0
0.0
Br 3d with Higher Binding Energy
Br 3d with Lower Binding Energy
(d) Brotrimazole/PVPVA
10% DL
20% DL
30% DL
40% DL
20
ACS Paragon Plus Environment

Page 21 of 37
Crystal Growth & Design
1
2
3
4
5
6
7
8
9
140.0
120.0
100.0
80.0
60.0
40.0
20.0
0.0
I 3d with Higher Binding Energy
I 3d with Lower Binding Energy
(e) Me-DIBF/PVPVA
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
10% DL
20% DL
30% DL
40% DL
Figure 3.
The percentage of halogen atoms with high (as found in the neat compound)
and low binding energies (new peaks seen in the ASD) as a function of wt. %
drug loading (DL) in the ASD for ASD system (a) clotrimazole/PVPVA, (b)
clotrimazole/PVP K29-32 (c) Loratadine/PVPVA, (d) brotrimazole/PVPVA,
(e) Me-DIBF/PVPVA.
4.2 Nuclear Magnetic Resonance Spectroscopy (NMR)
Halogen bonding interactions between components were examined, quantitatively, in solution by
NMR spectroscopy by measuring the binding constant between the components. The titration
experiments entailed addition of 1-ethyl-2-pyrrolidone (EP), a structural analogue of the repeat
unit in the polymers, to a solution of the model compound dissolved in deuterated cyclohexane.
Deuterated cyclohexane was chosen as solvent to avoid any potential halogen bonding between π
electrons and halogen atoms. For clotrimazole, the 1H NMR spectra showed that the chemical
shift of the proton at the para position relative to the chlorine atom changed as a function of EP
concentration, as presented in Figure 7S (a). The 13C NMR chemical shift for the carbon atom
covalently bonded to the chlorine showed a similar change with addition of EP, as shown in
21
ACS Paragon Plus Environment

Crystal Growth & Design
Page 22 of 37
1
2
3
4
5
6
7
8
9
Figure 7S (b). The change of chemical shift from the 1H and 13C spectra during EP titration is
depicted in Figure 4. The binding constant (KF) values for the halogen bonding interaction
between clotrimazole and EP are reported in Table 2 for each nucleus. Similar KF values were
obtained for the halogen bonding interaction from the 1H and 13C NMR titration experiments.
Analogous experiments were conducted for brotrimazole and loratadine, and the binding
constants are summarized in Table 2. As a negative control, solution 1H NMR of clotrimazole,
brotrimazole, and loratadine in the presence of 1-ethylpyrrolidine (no carbonyl group) showed
only minor proton shifts (as shown in SI). All the NMR spectra and chemical shift change
titration curves are provided in the supporting information. Binding constants were also
measured in deuterated toluene (Table S3). It has been reported that π electrons can be halogen
bond donors.53 In toluene, the binding constants were much smaller compared to those in
cyclohexane. This is consistent with competition between the toluene π electrons and the
carbonyl groups in EP for the halogen atoms. Similar solvent effects have been reported
previously.54 NMR titration experiments were not performed for Me-DIBF due to the complex
potential stoichiometry between halogen bond donor and acceptor.
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
22
ACS Paragon Plus Environment

Page 23 of 37
Crystal Growth & Design
1
2
3
4
5
6
7
8
9
0.10
0.08
0.06
0.04
0.02
0.00
(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
0
200
400
600
800
1000
1200
1400
Concentration of 1-Ethyl-2-pyrrolidone (mM)
23
ACS Paragon Plus Environment

Crystal Growth & Design
Page 24 of 37
1
2
3
4
5
6
7
8
9
0.40
0.35
0.30
0.25
0.20
0.15
0.10
0.05
0.00
(b)
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
0
200
400
600
800
1000
1200
1400
Concentration of 1-Ethyl-2-pyrrolidone (mM)
Figure 4.
Table 2.
(a) 1H NMR chemical shift change of clotrimazole as a function of the
concentration of 1-ethyl-2-pyrrolidone; (b) 13C NMR chemical shift change of
clotrimazole as a function of the concentration of 1-ethyl-2-pyrrolidone.
KF values for halogen bonding between model compounds and 1-ethyl-2-
pyrrolidone from the 1H and 13C NMR titration experiments at 25 °C in
cyclohexane.
KF based on 1H NMR (M-1)
0.63 ± 0.05
KF based on 13C NMR (M-1)
0.97 ± 0.04
Clotrimazole
Loratadine
Brotrimazole
0.65 ± 0.05
0.56 ± 0.06
1.11 ± 0.07
0.81 ± 0.04
4.3 Infrared (IR) Spectroscopy of ASDs
24
ACS Paragon Plus Environment

Page 25 of 37
Crystal Growth & Design
1
2
3
4
5
6
7
8
9
IR spectra of PVPVA in the carbonyl region were examined in the absence and presence of
molecularly dispersed drug. For pure PVPVA, the peak for the vinylpyrrolidone (VP) carbonyl
was around 1684 cm-1, while the position of the vinyl acetate (VA) carbonyl was at 1736 cm-1.
The IR spectra for clotrimazole-PVPVA and brotrimazole/PVPVA ASD systems, shown in
Figure 5a and 5b respectively, illustrate that the peak maximum corresponding to the carbonyl
group in VP shifted slightly to lower wavenumbers, from 1684 cm-1 to 1682 cm-1, as the drug
loading was increased from 10% to 90%. The spectral shift is well resolved, despite the very
small numerical change of only 2 cm-1. The peak position of the carbonyl group in VA did not
change. This 2 cm-1 shift suggests that the carbonyl group in VP is responsible for the halogen
bonding interaction between the model compounds and PVPVA. For Me-DIBF, a similar trend
was observed with a larger downshift of 6 cm-1, suggesting stronger halogen bonding interaction
in Me-DIBF ASDs.
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)
90%
80%
70%
60%
50%
40%
30%
20%
10%
PVPVA
1770
1720
1670
1620
Wavenumber (cm-1)
25
ACS Paragon Plus Environment

Crystal Growth & Design
Page 26 of 37
1
2
3
4
5
(b)
6
7
8
9
90%
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
80%
70%
60%
50%
40%
30%
20%
10%
PVPVA
1770
1720
1670
1620
Wavenumber (cm-1)
(c)
90%
80%
70%
60%
50%
40%
30%
20%
1
PVPVA
1770
1720
Wavenumber (cm-1)
1670
1620
Figure 5.
Infrared spectra of the carbonyl group region of pure PVPVA and (a)
clotrimazole-PVPVA ASDs, (b) brotrimazole-PVPVA ASDs, and (c) Me-
DIBF-PVPVA ASDs.
26
ACS Paragon Plus Environment

Page 27 of 37
Crystal Growth & Design
1
2
3
4
5. Discussion
5
6
7
8
Halogen bonding (X-bonding) is emerging as an important intermolecular interaction in various
areas of science. In the area of drug research, most studies have focused on X-bonding between
halogen bearing drugs and proteins, and its role in improving ligand-receptor interactions. As
knowledge in this area increases and utilization of halogen groups as a means to enhance drug
efficacy becomes more prominent, it becomes increasingly important to understand the relevance
of this interaction for other aspects of drug research such as pharmaceutical formulation
development. Molecular recognition is of particular importance to the solid state, playing a
critical role in both local and long range ordering of materials. In the context of pharmaceutical
systems, non-covalent interactions impact properties such as solubility, solid state form
landscape, and phase transformations.
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
Herein, using a variety of orthogonal analytical techniques, we found compelling evidence of X-
bonding between drug or drug-like molecules containing Cl, Br, or I but lacking hydrogen bond
donors, and polymers, such as PVPVA and PVP. XPS appears to be of particular utility for
studying changes in the halogen environment, being one of the few techniques that allow a direct
probe of the halogen moiety. Further, photoelectron spectra are sensitive to intermolecular
interactions. For example, ionic and hydrogen bond interactions are well known to lead to
binding energy shifts in the photoelectron spectra.55, 56 Thus, we note that the binding energies of
Cl 2p, Br 3d, and I 3d peaks decreased when the model compounds were molecularly dispersed
in PVPVA. The lowering of the binding energy indicates that these halogen atoms gain electron
density when the polymer is present. This trend is consistent with halogen bonding, where the σ-
hole on the halogen atom is known to attract electrons. Moreover, the 2-4 eV binding energy
shift observed is significant and comparable to the binding energy shift produced by hydrogen
27
ACS Paragon Plus Environment

Crystal Growth & Design
Page 28 of 37
1
2
3
4
5
6
7
8
9
bonding.56 This lends support that the shift is due to a specific intermolecular interaction. Shifts
to a lower binding energy, albeit to a lesser extent, have been reported for systems containing
halogen bonds between iodine atoms and nitrogen,57, 58 providing further support to our
interpretation. Oxygen in carbonyl is known to be one of the most common halogen bond
acceptors, in particular for protein-ligand interactions.14 Further, it is well-known that the amide
carbonyl group of the vinyl pyrrolidone group in PVP and PVPVA is a good hydrogen bond
acceptor.41 Therefore, it is not surprising that this group acts as the halogen bond acceptor.
Results from IR spectroscopy confirm that the vinyl pyrrolidone group is preferred relative to the
vinyl acetate group (Figure 5). Further support for the supposition that this oxygen group donates
electrons to the halogen atom is provided by examination of the O1s peak of PVP in the presence
and absence of clotrimazole. For all other systems, XPS spectra contain signal from multiple
oxygen atoms, however, for the PVP-clotrimazole system the O1s peak arises solely from the
vinylpyrrolidone carbonyl. In the dispersion, the O1s binding energy shifts to a higher value by
0.2 eV relative to neat PVP (Figure 6S), suggesting a reduction in electron density consistent
with the formation of a halogen bond. The minimal change in the 1H NMR spectrum of
clotrimazole, brotrimazole and loratadine in the presence of 1-ethylpyrrolidine (which lacks a
carbonyl group) relative to the large shifts seen in the presence of the carbonyl containing
analogue 1-ethyl-2-pyrrolidone, further supports the role of the VP oxygen as the halogen bond
acceptor. Halogen bonding with a carbonyl group has also been seen not only in protein-ligand
binding,59, 60 but also between two small molecules.23, 24 In summary, based on the evidence
presented herein, we propose that halogen bonding exists between the amide carbonyl , on the
polymer, and halogen atoms, Cl, Br, or I, in ASDs. In addition, the interaction occurs to a greater
extent when the drug loading is low and drug-drug intermolecular interactions are less probable.
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
28
ACS Paragon Plus Environment

Page 29 of 37
Crystal Growth & Design
1
2
3
4
5
6
7
8
9
Halogen bonds share many similar characteristics with the more commonly encountered
hydrogen bonds. They are both directional interactions between an electron-deficient moiety and
a high electron density region. They are similar in the strength of interaction. The binding energy
for hydrogen bonding is 0.2 to 40 kcal/mol, while for halogen bonding it is 1.2 to 43 kcal/mol.61,
62 Therefore, the interaction strengths of halogen and hydrogen bonds can be comparable.
Specifically for the systems we have studied here, the halogen bond strength between the
compounds and PVPVA appears to be similar to weak hydrogen bonding interactions. The
binding constants, which reflects the strength of interaction, between clotrimazole, brotrimazole,
and loratadine with EP (Table 2) are between 0.6-1.1 M-1. They are of similar magnitude as the
0.7 M-1 binding constant corresponding to weak hydrogen bonds, where fluorine is the hydrogen
bond acceptor.51 The IR carbonyl absorption peak only shifted 2-6 cm-1 when halogen bonded to
the model compounds in the ASDs. In contrast, red shifts as large as 20-30 cm-1 have been
observed when PVP forms hydrogen bonds with drugs in ASDs.9, 63, 64 In general, the strength of
halogen bonds correlate with the polarizability of the halogen atoms, following the trend I > Br >
Cl. In our study, however, we have observed similar interaction strength between of -Cl and -Br
analogues, based on the binding constant measured from the NMR titration and the wavenumber
shift in the IR absorption of the carbonyl group. This observation can perhaps be rationalized
based on analysis of cohesive interactions in the crystal. In brotrimazole, there is intermolecular
interaction between Br and either the phenyl group or imidazole, however, this interaction was
not observed in the clotrimazole crystal. Thus, the tendency for brotrimazole to self-interact may
be greater compared to clotrimazole, and this interaction can compete with the brotrimazole-
PVPVA intermolecular interaction. The formation of halogen bonds with phenyl groups is
further suggested by the smaller binding constants between model compounds and EP in toluene
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
29
ACS Paragon Plus Environment