Crystal Engineering of Pharmaceutical Co-crystals: "nMR Crystallography" of Niclosamide Co-crystals

Article abstract of DOI:10.1021/acs.cgd.5b01619

Niclosamide is a Biopharmaceutics Classification System (BCS) class II taeniacide currently reconsidered for new promising applications including treatment of rheumatoid arthritis, prevention of protein degeneration in neurodegenerative diseases, or even multi-targeted therapy of cancer and cancer stem cells. Its efficacy in medical treatments, however, is currently limited by its insufficient solubility or bioavailability. Thus, we have further explored the potential of hydrogen-bond-mediated co-crystal formation of niclosamide with suitable co-formers selected from either the "Generally Regarded as Safe" (GRAS) or United States Food and Drug Administration (U.S. FDA) "Everything Added to Food in the United States" (EAFUS) list, respectively. Solvent-assisted solid grinding and/or slow solvent evaporation yielded four new co-crystals: (i) niclosamide-2-aminothiazole (NCL-AT), (ii) niclosamide-benzamide (NCL-BA), (iii) niclosamide-isoniazide (NCL-IN), and (iv) niclosamide-acetamide I and II (NCL-AA-I/NCL-AA-II). The crystal structures of NCL-AA-I/II, NCL-AT were solved from white microcrystalline powder samples on the basis of the combined application of powder X-ray diffraction (PXRD), solid-state NMR, and Density Functional Theory (DFT) chemical shift computation. In addition, the crystal structure of the monohydrate NCL-H_A was reconsidered for comparison. Finally an improvement of the equilibrium solubility of the 1:1 co-crystal NCL-AT could be determined (2.8 times that of pristine NCL and 1.4 times that of NCL-UREA co-crystal), suggesting NCL-AT as a candidate for future medical treatment.

Full text of DOI:10.1021/acs.cgd.5b01619

Subscriber access provided by SUNY DOWNSTATE
Article
Crystal engineering of pharmaceutical co-crystals:
“NMR- crystallography” of Niclosamide co-crystals
David Luedeker, Rebecca Gossmann, Klaus Langer, and Gunther Brunklaus
Cryst. Growth Des., Just Accepted Manuscript • DOI: 10.1021/acs.cgd.5b01619 • Publication Date (Web): 19 Apr 2016
Downloaded from http://pubs.acs.org on April 26, 2016
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 34
Crystal Growth & Design
1
2
3
4
5
6
Crystal engineering of pharmaceutical co-crystals: “NMR-
crystallography” of Niclosamide co-crystals
7
8
9
David Luedeker,† Rebecca Gossmann,‡ Klaus Langer,‡ and
Gunther Brunklaus*,†
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
†Institut für Physikalische Chemie, Westfälische Wilhelms-Universität Münster, Corrensstr. 28, D-48149
Münster, Germany
Fax: (+49)-251-83-36032; email: gbrunklaus@uni-muenster.de
‡Institut für Pharmazeutische Technologie und Biopharmazie, Westfälische Wilhelms-Universität Münster,
Corrensstr. 48, D-48149 Münster, Germany
Abstract: Niclosamide is a Biopharmaceutics Classification System (BCS) class II taeniacide
currently reconsidered for new promising applications including treatment of rheumatoid
arthritis, prevention of protein degeneration in neurodegenerative diseases or even multi-
targeted therapy of cancer and cancer stem cells. Its efficacy in medical treatments, however,
is currently limited by its insufficient solubility or bioavailability. Thus we have further
explored the potential of hydrogen-bond mediated co-crystal formation of niclosamide with
suitable co-formers selected from either the “Generally Regarded as Safe” (GRAS) or United
States Food and Drug Administration (US FDA) “Everything Added to Food in the United
States” (EAFUS) list, respectively. Solvent-assisted solid grinding and/or slow solvent
evaporation yielded four new co-crystals: (i) niclosamide − 2-aminothiazole (NCL-AT), (ii)
niclosamide − benzamide (NCL-BA), (iii) niclosamide − isoniazide (NCL-IN), and (iv)
niclosamide − acetamide I and II (NCL-AA-I/NCL-AA-II). The crystal structures of NCL-
AA-I/II, NCL-AT were solved from white micro-crystalline powder samples based on the
combined application of powder X-ray diffraction (PXRD), solid-state NMR and Density
Functional Theory (DFT) chemical shift computation. In addition, the crystal structure of the
monohydrate NCL-HA was reconsidered for comparison. Finally an improvement of the
equilibrium solubility of the (1:1) co-crystal NCL-AT could be determined (2.8x that of
pristine NCL and 1.4x that of NCL-UREA co-crystal), suggesting NCL-AT as a candidate for
future medical treatment.
Keywords
1
ACS Paragon Plus Environment

Crystal Growth & Design
Page 2 of 34
1
2
3
4
5
NMR crystallography, niclosamide, pharmaceutical co-crystal, PXRD structure solution,
polymorphism, niclosamide hydrate, niclosamide co-crystal.
6
7
8
1. Introduction
9
Niclosamide (2’, 5-dichloro-4’-nitrosalicylanilide, NCL) is a hydrophobic taeniacide and
restricted-use pesticide that may be applied to cure parasite infestations (including most
tapeworms and cestoda)[1] in both humans and animals or for water treatment against e.g. sea
lampreys (Petromyzon marinus)[2] or apple snails (Pomacea canaliculata).[3] Its primary effect
comprises the inhibition of energy production in mitochondria,[4] though NCL may also bind
to DNA after its reductive activation[5] rendering it possibly toxic to some aquatic organisms
or plants.[6] An early long term toxicology survey of molluscicides indicated that NCL neither
has increased mutagenic, oncogenic or embryotoxic activity nor impact on liver and kidney
functions.[7] Despite that a single oral dose of NCL for adults in cestocidal treatment amounts
to 2 g,[8] NCL exhibits a very low acute toxicity in humans[9] reflecting both its insufficient
oral bioavailability (merely 10% in male Sprague-Dawley rats)[10] and the fact that NCL is
poorly absorbed from the intestinal tract.[11] However, in view of recently identified promising
properties of NCL such as anti-inflammatory effects useful for the treatment of rheumatoid
arthritis,[12] the prevention of protein aggregation in neurodegenerative diseases[13] or even for
multi-targeted therapy of cancer and cancer stem cells[14] (including an application of NCL as
effective radio-sensitizer),[15] research efforts are currently devoted to considerably improve
the solubility (and thus bioavailability) of NCL thereby enhancing the achievable maximal
serum concentrations and its efficacy in medical treatments.
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
In addition to suitable drug carriers such as polymeric particles[16-18] or even
supramolecular hosts[19-21] introduced during formulation, either efficient crystallization
screening (including polymorphs, salts, solvates or co-crystals)[22-24] or tailored amorphization
of the active drug[25-28] or its multi-component system of interest (co-amorphous drug
systems)[29] constitute two major approaches that are available for the manipulation of
physicochemical properties of drugs. Though amorphous drugs may have great potential to
reduce issues related to either poor dissolution rate or solubility-limited absorption in
particular in case of drugs where salt formation cannot be applied, this approach is not readily
accessible for the non-glass former NCL.[30] Therefore, the various attempts to counteract
NCLs hydrophobic nature in order to facilitate better aqueous dispersion or solubility of NCL
for example include salt formation of NCL with either ethanolamine[31] or sodium (as part of a
polymer-based controlled-release formulation)[32] and inclusion of NCL in 4-sulphonatao-
2
ACS Paragon Plus Environment

Page 3 of 34
Crystal Growth & Design
1
2
3
4
5
6
7
8
9
calix[n]arene[33] or cyclodextrin[34] host systems. Since the NCL delivery formulations were
mainly considered in the framework of water treatment such as control of snails in farm fields
neither compatibility nor toxicity of certain additives with respect to the human metabolism
was explicitly evaluated which only recently has changed due to the emerging interest in NCL
as a potential candidate in cancer therapy.[35] Consequently, chemically modified but highly
water soluble derivatives of NCL, (e.g. with an attached phosphate group[36] or alkylamino
tethered derivatives[37]) have been introduced, though at the expense of rather tedious
synthetic protocols.
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 1: The molecular structures of niclosamide and the considered co-formers discussed in this study. All
obtained co-crystals of niclosamide and the monohydrate NCL-HA have a (1:1) stoichiometry. Note that the
[38]
crystal structure of NCL-HA was very recently reported
and is reconsidered for comparison, while 2-
[39]
aminothiazole derivatives were suggested as therapeutic leads for the treatment of prion diseases.
Moreover, the concept of hydrogen-bond mediated formation of pharmaceutical co-
crystals[40-46] was successfully applied to obtain co-compounds of NCL with suitable
molecules taken from either the “Generally Regarded as Safe” (GRAS) or US FDA
“Everything Added to Food in the United States” (EAFUS) list[47] including caffeine (CAF),
urea (URE), p-amino-benzoic acid (PABA), theophylline (THPH), nicotinamide (NA) or
isonicotinamide (IA), respectively.[48] The co-crystals were designed according to the
availability of specific supramolecular synthons[49-51] and are mainly based on the robust
{OH···O} synthon though in case of para-amino-benzoic acid the {OH···NH2} synthon was
identified. Since nicotinamide and its stereoisomer isonicotinamide contain CONH2 and
Naromatic functional groups, the likely resulting co-crystals may exhibit {OH···O}, {OH··
3
ACS Paragon Plus Environment

Crystal Growth & Design
Page 4 of 34
1
2
3
4
5
6
7
8
9
·NH2}, or {OH···Naromatic} synthons where the latter is a particularly versatile supramolecular
unit for crystal engineering applications.[52-54] In addition, salt co-crystals with concomitant
presence of NCL both as neutral component and as salt co-former (derived from salification
of NCL with inorganic salts) as well as mixed solvate/hydrate salts of NCL were very recently
reported.[55] The latter partially included the solvent dimethyl sulfoxide (DMSO), which in
view of the identified brain degeneration due to apoptosis in the central nervous system
induced by DMSO already at low doses of 0.3 mL/kg appears pharmaceutically
questionable.[56] Therefore, the intent of this work was to further explore the reliability of
hydrogen-bond mediated co-crystal formation for the improvement of the solubility of NCL
considering suitable co-former molecules that contain either amide groups and/or aromatic
nitrogen atoms, respectively (Figure 1). From solvent assisted solid grinding and/or slow
solvent evaporation new co-crystals of (i) niclosamide – 2-aminothiazole (NCL-AT), (ii)
niclosamide – benzamide (NCL-BA), (iii) niclosamide – isoniazide (NCL-IN) and the
polymorphic system (iv) niclosamide – acetamide I and II (NCL-AA-I/ NCL-AA-II) were
successfully prepared. Since structure solutions were not yet provided in case of the co-
crystals of NCL-NA (vi), NCL-IA (vii), niclosamide − imidazole (NCL-IMI), respectively,
these co-compounds were reproduced for further structural characterization. However, in the
absence of single crystals suitable for single-crystal X-ray diffraction, the corresponding
crystal structures of both NCL-AA-I/II and NCL-AT were solved from micro-crystalline
powdered samples based on the combined application of powder X-ray diffraction (PXRD),
solid-state NMR and Density Functional Theory (DFT) chemical shift computations, an
approach that we have recently applied for the structure solution of co-crystals of ezetimibe
with imidazole and L-proline.[57] Briefly, the currently evolving concept of “NMR
crystallography”[58-64] comprises at least the two-step strategy of an initial structure model
derived from either powder diffraction techniques[65,66] or where feasible from crystal
structure prediction (CSP)[67-69] and subsequent structure validation against experimental data,
often including solid state NMR chemical shifts or chemical shift tensor data, respectively,
that are computed by DFT methods based on the proposed structure models (or identified
building blocks) where the best match indicates the (apparently) most likely crystal
structure.[70-76] Besides, the structural refinement may be supported by complementary data
input obtained from solid state NMR such as the number of independent molecules in the
asymmetric unit (Z´), molecular connectivity, or in particular characteristic distances
extracted from dipolar couplings.[77-87] Experimental 1H solid state NMR chemical shifts have
even been utilized for immediate structure generation based on a genetic algorithm and
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
ACS Paragon Plus Environment

Page 5 of 34
Crystal Growth & Design
1
2
3
applied as pseudo-forces during structure refinement,[88] while in other cases the
4
5
1
corresponding H solid state NMR chemical shifts of protons involved in reasonably strong
6
7
8
9
hydrogen bonding (depending on the actually considered synthon) could be correlated with
the explicit geometry of the hydrogen bond thus allowing for an estimation of corresponding
distances and/or torsion angles from solid state NMR data.[89-92] In an alternative approach,
13C solid-state NMR data was applied to establish a crystal structure even in the absence of
powder diffraction data and single crystals,[93] though such data more typically has been
considered for structure validation or identification of molecular conformation or
polymorphism present in solid materials.[94-97] Despite the presence of many distinct protocols
(including software packages) the high versatility of the interdisciplinary concept of “NMR
crystallography” to support solving or refinement of crystal structures of powdered solids is
indeed convincingly demonstrated by a continuously growing number of contributions that
report successful structure elucidation of a variety of materials.[98-109]
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. Results and Discussion
Rather well defined crystalline drugs are often preferred for the tailored design of solid
oral dosage forms based on crystal engineering strategies.[110] Niclosamide, however, is prone
to solvate, hydrate or mixed solvate/hydrate formation even when reacted under solvent
assisted solid grinding where only a few drops of suitable solvent are present. Thus several
solvates of NCL with methanol (MeOH);[111] acetone and acetonitrile;[38] dimethyl sulfoxide
(DMSO), N,N’-dimethyl formamide (DMF), diethyl ether (Et2O);[112] tetrahydrofuran (THF)
and tetraethylene glycol (TEG);[113] and the polymorphic monohydrates NCL-HA[38] and NCL-
[114,115]
HB
have been reported, thereby clearly emphasizing the necessity to utilize rather
purified water-free solvents for the co-crystal synthesis of NCL to circumvent any hydrate by-
products. While unintentional changes of the level of hydration or even complete dehydration
of pharmaceutical solids ideally can be avoided during formulation development or
manufacture under controlled conditions, this may be more difficult upon prolonged storage
of the final product which then may adversely affect its physical, chemical or bio-medical
properties.[116] Similarly, charge-assisted synthons formed due to salt formation tend to be
hygroscopic and should be avoided as well rendering co-crystallization attempts based on
neutral synthons such as {OH···O} or {NH···O} rather preferred. Note that in the abundant
presence of different hydrogen-bond donor and acceptor moieties in a given crystal structure,
the analysis of data sets from the Cambridge Structural Database (CSD) indicates that the
5
ACS Paragon Plus Environment

Crystal Growth & Design
Page 6 of 34
1
2
3
4
5
majority of hydrogen bond donors (particularly the strong donors) will be satisfied (and hence
involved in hydrogen bonding) while a larger fraction of acceptors may remain “free”.[117]
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
1
Figure 2: H MAS NMR spectra (acquired at 500.1 MHz and 30 kHz spinning frequency) of (a) NCL-IMI, (b)
NCL-AT, (c) NCL-IA, (d) NCL-NA, (e) NCL-IN, (f) NCL-BA, (g) NCL-AA-II, (h) NCL-AA-I, (i) NCL-HA and
(j) NCL anhydrate. The region for the NCL-IMI spectrum between 10-20 ppm is enlarged on top, while the
signals attributed to small impurities from residual solvents (either ethanol or ethyl acetate) are marked by
asterisks.
Indeed, the molecular structure of NCL exhibits a planar, secondary amide group
comprised of a C=O hydrogen bond acceptor and NH hydrogen bond donor. The latter forms
an intramolecular S(6) hydrogen bonded ring with the oxygen atom of a neighboring hydroxyl
group resulting in an exposed OH unit that may be considered as suitable hydrogen donor for
the successful co-crystal and/or solvate formation of NCL, as documented in the pristine
compound NCL where this hydroxyl group connects via intermolecular Cꢀꢀ(6) chains with the
carbonyl hydrogen bond acceptor of an adjacent NCL moiety (note that NCL crystallizes in
6
ACS Paragon Plus Environment

Page 7 of 34
Crystal Growth & Design
1
2
3
the monoclinic space group
P21/c with one molecule in the asymmetric unit). In order to
4
5
6
7
8
9
enhance the probability of successful co-crystal formation of NCL, all of the selected co-
former molecules offer a competitive hydrogen-bond acceptor such as other C=O units or
aromatic nitrogen atoms (Figure 1) that upon either solvent assisted solid grinding or slow
solvent evaporation typically yielded micro-crystalline powdered samples rather than single
crystals.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Solid state MAS NMR. Successful formation of phases different from the starting materials
was monitored by solid-state NMR which is applicable even in cases where possibly
amorphous products would be obtained (particularly in early attempts where the reaction
conditions are not suitably established). The spatial rearrangement of molecules including the
coordination of the exposed OH (or even NH) group of NCL to a hydrogen bond acceptor of
the offered co-former molecules result in different hydrogen bonding pattern compared to the
precursor compounds, which depending on the strength or bond lengths (or even geometry) of
the considered hydrogen bond is typically reflected by quite characteristic changes of the
1
corresponding H magic-angle spinning (MAS) NMR spectra (Figure 2).[118-120] Notably,
based on the analysis of integrated area ratios extracted from deconvolution of experimental
1H NMR spectra (peak fitting) particularly considering the signal fraction of sufficiently
hydrogen bonded protons (e.g., NCL-OH or NCL-NH) to “ordinary” protons the molecular
composition of the compound can be derived, allowing for a comparison with the intended
1
stoichiometry. An inspection of the H MAS NMR spectra of NCL-AT, NCL-IA, NCL-NA
and NCL-IN (Figure 2(b)-2(e)), respectively, where the peaks of NCL-NH and NCL-OH are
clearly separated, revealed integrated area ratios of 1:1:10 (NCL-AT), 1:1:12 (NCL-IA),
1:1:12 (NCL-NA) and 1:1:13 (NCL-IN), in addition to integrated area ratios of 1:1:13 (NCL-
BA), 1:1:11 (NCL-AA-I, NCL-AA-II) and 1:1:8 (NCL-HA) derived from peak fitting (Figure
2(f)-2(i)), which in all considered cases are in excellent agreement with a 1:1 stoichiometry of
NCL and the corresponding co-former. This even holds for the compound NCL-IMI, based on
the 3:9 integrated area ratio of hydrogen bonded protons (distributed in the region from 12.6
to 18.6 ppm) and aromatic protons. In most cases, the 1H chemical shift of the NCL-NH group
amounts to {11.0
thus indicating that the planar conformation including the intramolecular S(6) hydrogen
bonded ring of the NCL anhydrate structure ( -conformation) remained intact in the obtained
co-crystals. This finding was further verified based on the corresponding 15N chemical shift
− 11.7 ppm}, except for NCL-IMI (12.7 ppm) and NCL-URE (9.8 ppm),
β
of {245.5
−
248.9 ppm}, in agreement with quite negligible changes of local electron density
7
ACS Paragon Plus Environment

Crystal Growth & Design
Page 8 of 34
1
2
3
4
5
6
7
1
distributions (Figure S1). In contrast, the H NMR chemical shift of the exposed O
−
H group
of NCL showed a strong dependence on the actually formed synthon upon co-crystallization,
ranging from {11.0 12.3 ppm} and {13.5 15.3 ppm} for {O−H···O} and {O−H···N}
−
−
8
9
hydrogen bonds, respectively. Note that the co-crystal NCL-AT is stabilized by rather strong
1
{O−H···N} hydrogen bonds as reflected by a high H chemical shift of 15.3 ppm for the
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
NCL-OH group. Based on the observed H chemical shifts of the co-crystals that contain
aromatic nitrogen acceptors (including NCL-IN, NCL-NA and NCL-IA) whose crystal
structures are not yet plausibly solved from the available PXRD pattern, it indeed can be
concluded that the NCL-OH group coordinates to the aromatic nitrogen rather than amide or
hydrazine groups of the co-former thereby yielding comparatively stronger hydrogen bonds.
In the case of linear hydrogen bonds with fixed donor (D) and acceptor (A) atom distances
(e.g., OꢀꢀꢀO, OꢀꢀꢀN or NꢀꢀꢀN) the 1H chemical shift of the proton is explicitly correlated with the
1
position of the proton within the hydrogen bond where quite high H chemical shielding
(hence chemical shift) was attributed to either short D-H distances or “stretched” symmetric
hydrogen bonds in which the proton approached the center of the DꢀꢀꢀA bond,[121] though the
1
occurrence of charges due to proton transfer could also result in increased H chemical
shifts.[122] In principle the formation of a charge assisted synthon {NCLO-ꢀꢀꢀ+HNIMI} is feasible
for NCL-IMI in addition to {NCLNHꢀꢀꢀ-ONCL} thereby shifting the peak of NCL-NH to higher
ppm (12.7 ppm) compared to the neutral synthon {NCLNHꢀꢀꢀOH,NCL} that typically comprises
the S(6) intramolecular ring of NCL. If the IMI-NH unit is also taken into account, it becomes
clear why the integrated area ratio of hydrogen bonded to “ordinary” protons amounts to 3:9.
A deconvolution of the 1H MAS NMR spectrum of NCL-IMI revealed an integrated area ratio
of 0.95:0.05:2 for the corresponding peaks at 18.5, 14.9 and 12.7 ppm, respectively, thus
suggesting that the minor peak at 14.9 ppm represents an “impurity”. In addition, the
experimental data indicates that the resonances of both NCL-NH and IMI-NH incidentally
coincide at 12.7 ppm if the peak at 18.5 ppm is tentatively attributed to IMI-NH+. NCL-NH
and IMI-NH should have different protons in their immediate spatial proximity which may be
revealed based on characteristic correlation peaks in a 2D H-1H dipolar double-quantum
1
(DQ) spectrum where so-called double-quantum coherences due to pairs of dipolar coupled
protons are correlated with single-quantum (SQ) coherences.[123,124] DQ signals therefore
appear at the sum of the chemical shifts of the coupled protons (facilitating the identification
of sites obscured due to overlapping peaks) while the DQ signal intensities are proportional to
-6
rij (the internuclear distance) hence yielding sufficient signal intensities only for those
8
ACS Paragon Plus Environment

Page 9 of 34
Crystal Growth & Design
1
2
3
4
5
6
7
8
9
protons in reasonable spatial proximity (distances of up to 4.0 Å). The DQ MAS NMR
spectrum of NCL-IMI (Figure 3) exhibits two particularly revealing DQ peaks with
significantly different intensities at 19.7 ppm (12.7 + 7.0 ppm, 11% of max. intensity) and
20.2 ppm (12.7 + 7.5 ppm, 4% of max. intensity) that reflect dipolar coupling to unlike
aromatic protons (7.0 vs. 7.5 ppm), in good agreement with two different protons (NCL-NH
and IMI-NH) coinciding at 12.7 ppm. In addition, a rather strong DQ (cross)peak at 31.2 ppm
(12.7 + 18.5 ppm) is indicative of spatial proximity of NCL-NH (12.7 ppm) and IMI-NH+
(18.5 ppm) whereas a DQ (auto)peak at 25.4 ppm (12.7 + 12.7 ppm) suggests molecular
packing that result in similar proximity of NCL-NH (12.7 ppm) with NCL-NH (12.7 ppm) or
IMI-NH (12.7 ppm) with IMI-NH (12.7 ppm), since a dipolar coupling of NCL-NH with IMI-
NH (when considering likely distances) is rather unlikely to yield such a DQ signal.
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
Figure 3: H-1H DQ-MAS-NMR spectrum of NCL-IMI, recorded at 500.1 MHz and 30 kHz MAS. 32 positive
contour levels between 1% and 35% of the maximum peak intensity were plotted. The F2 projection is shown
on top; the most important DQ cross-peaks are highlighted, illustrating that two different proton sites
coincide at 12.7 ppm.
Except for NCL-IMI, the 1H-1H DQMAS NMR spectra (Figures S12-S18) were primarily
considered for structure validation with respect to the hydrogen bonding network present in
the final structure solutions derived from Rietveld refinement, though independent distance
constraints and hence insight into the molecular packing may be derived from DQ build-up
curves (monitoring of DQ signal intensities as a function of DQ excitation times).[125,126] In
1
addition to exploiting the integrated area ratio of the H MAS NMR spectra to determine the
ratio of the molecular constituents of the considered co-compound, the content of the
asymmetric unit (which is an essential information with respect to the structure solution from
9
ACS Paragon Plus Environment

Crystal Growth & Design
Page 10 of 34
1
2
3
4
5
6
7
8
9
PXRD data) was estimated based on 15N{1H} (Figure S1) or 13C{1H} (Figure S2) cross-
polarization MAS NMR spectra of the co-crystals, explicitly utilizing that all inequivalent
atoms (those not generated by symmetry elements)[127] comprising a given structure result in
individual NMR signals provided that sufficient spectral resolution could be achieved.
Nevertheless, in case of hydrogen-bonded nitro groups (often if connected to aromatic rings)
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
the occurrence of rotational dynamics may perturb the H-15N polarization transfer thereby
rendering the corresponding 15N signals eventually “invisible”, irrespective of the applied
cross-polarization contact time.[128,129] Most other nitrogen atoms could be identified including
the unique IMI-NH+ at -213.7 ppm.
Powder X-ray Diffraction. In addition to the solid state NMR analysis, the powdered
compounds derived from solvent-assisted grinding of stoichiometric mixtures of the precursor
materials were investigated by powder X-ray diffraction. Indeed, successful formation of
phases different from the starting materials could be readily identified form the occurrence of
new peaks in the corresponding diffraction pattern and in due consideration of the available
solid state NMR data were attributed to the formation of co-crystals. Notably, we have
obtained the (1:1) co-crystals NCL-AT, NCL-IMI, NCL-AA, NCL-BA, NCL-IN as well as
[38]
NCL-NA and NCL-IA while the monohydrate NCL-HA was reconsidered for comparison.
Despite quite substantial efforts, attempts to grow larger crystals suitable for single-crystal
XRD analysis from slow solvent evaporation were not successful so that structure solutions
were derived from powder XRD data taking complementary structural input from solid state
NMR data of the respective co-phases into account. An inspection of the PXRD pattern of the
starting compound NCL anhydrate (Figure S3(a)) revealed that its most intense reflections are
at 2
of the formation of new phases. For NCL-HA the characteristic reflections are at 2
10.6°, 16.9°, 25.6° and 27.2°}[38] (Figure S4(b)) while major reflections at 2
ϴ
= {13.3°, 13.9°, 26.1°, 26.7° and 27.2°} so that changes of those peaks were indicative
= {9.4°,
= {10.4°,
ϴ
ϴ
13.1°, 22.4°, 26.8° and 27.3°} (Figure S4(c)) are identified in the theoretical XRD pattern
computed from available single crystal structure data of NCL-HB. Note that all three XRD
pattern are significantly different from each other as indeed expected in the case of three
independent phases. Likewise, the most significant new reflections for NCL-IMI at 2
8.1, 16.3, 26.6 and 28.2} (Figure 4), for NCL-AT are at 2 = {7.0°, 9.4°, 15.8°, 26.4° and
27.6°} (Figure 6), for NCL-AA-I at 2 = {8.7°, 14.4°, 18.7°, 25.9°, and 26.6°} (Figure 8) and
for NCL-AA-II at 2 = {6.1°, 8.8°, 17.7°, 19.6° and 26.5°} (Figure 9). Also, the co-crystals
NCL-BA and NCL-IN yielded characteristic reflections at 2 = {5.6°, 8.0°, 12.2°, 16.7°, and
ϴ = {6.1,
ϴ
ϴ
ϴ
ϴ
25.6°} (Figure S6(e)) and at 2
ϴ
= {5.1°, 5.6°, 7.6°, 24.0° and 27.3°} (Figure S6(f)). The two
10
ACS Paragon Plus Environment

Page 11 of 34
Crystal Growth & Design
1
2
3
4
5
co-crystals NCL-IA and NCL-NA have been reproduced for further structural characterization
since no explicit structure solution was provided in the original work.[48] Their most
6
7
significant reflections are at: 2
ϴ = {5.4°, 7.7°, 13.9°, 27.0°, and 27.6°} (NCL-IA; Figure
S3(d)) and 2 = {5.5°, 7.7°, 14.0°, 27.1°, and 27.6°}(NCL-NA; Figure S3(e)). The explicit
ϴ
8
9
structure solution in all considered cases started with profile fitting and indexing of the
recorded PXRD pattern with both N-TREOR and DICVOL04 as implemented in the expo2014
and DASH software packages until reasonable solutions were obtained. The indexing attempts
for the co-crystals NCL-NA, NCL-IA, NCL-BA and NCL-IN yielded rather flat unit cells that
were similar to the unit cells of NCL-HA, NCL-AA-II and NCL-MeOH, respectively.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table 1: Crystallographic parameters obtained for the crystal structures of the (1:1) co-crystals of NCL with
imidazole (IMI), 2-aminothiazole (AT) and acetamide (AA), respectively. In case of NCL-HA that was considered
for comparison with NCL-HB, the assignment of the crystal axes in our case was done purposely to demonstrate
its structural relationship to NCL-HB, preferring the space group 21/a over 21/c, while the lattice parameters are
similar to the reported structure of NCL-HA.[38]
NCL-HA
NCL-IMI
NCL-AT
NCL-AA-I
NCL-AA-II
emp. formula
C13H10Cl2N2O5 C16H12Cl2N4O4 C16H12Cl2N4O4 C15H13Cl2N3O5 C15H13Cl2N3O5
S
formula wt.
crystal system
space group
T/ K
345.13
monoclinic
P21/a
395.20
monoclinic
P21/c
427.26
monoclinic
P21/c
386.19
triclinic
P-1
386.19
orthorhombic
P212121
295
295
295
295
295
a/ Å
b/ Å
c/ Å
α/ °
23.05465(54)
16.15304(33)
3.812027(84)
90
92.8243(23)
90
4.02638(12)
21.80412(41)
19.30093(50)
90
94.074(27)
90
12.89932(33)
18.77890(45)
7.44985(17)
90
99.6491(27)
90
11.21040(39)
11.01955(31)
7.57010(18)
99.5245(25)
103.1851(22)
108.3846(23)
835.026(47)
2
20.00562(00)
21.20496(00)
3.88235(00)
90
90
90
β/°
γ/°
volume/ Å3
1417.884(54)
4
1690.179(75)
4
1779.084(76)
4
1647.964(00)
4
Z
Z’
1
1
1
1
1
radiation type,
λ/ Å
Rexp
Cu Kα,
1.54060
1.132
Cu Kα,
1.54060
1.886
Cu Kα,
1.54060
2.586
Cu Kα,
1.54060
1.771
Cu Kα,
1.54060
1.395
Rwp
1.977
2.579
4.227
6.351
6.235
Rp
1.415
1.826
3.327
4.155
4.431
GOF
1.747
3.052
STOE StadiP
1.412
1.994
STOE StadiP
1.634
2.670
Bruker D8
Discover
3.587
12.867
Bruker D8
Discover
4.470
19.981
STOE StadiP
χ2
diffractometer
Niclosamide monohydrate HA (NCL-HA). The structure of the monohydrate NCL-HA was
independently solved in a monoclinic unit cell and space group suggested from a previous Le-
Bail fit[111] similar to the recently reported crystal structure of NCL-HA[38] except that the cell
11
ACS Paragon Plus Environment

Crystal Growth & Design
Page 12 of 34
1
2
3
4
5
6
7
8
9
axes a and c were exchanged for better comparability with the pseudo-polymorphic hydrate
NCL-HB. The structure solution converged quite quickly with excellent reproducibility within
50 consecutive hybrid big bang big crunch (HBB-BC) runs as implemented in expo2014,
where the final refinement (Figure S8) gave an excellent goodness of fit (GOF = 1.747) and
Rwp of 1.977 %. In our case, NCL-HA crystallizes in the monoclinic space group P21/a (No.
14) with one molecule of NCL and one molecule of water in the asymmetric unit (Z’ = 1).
Briefly, the NCL-OH group is hydrogen-bonded to one molecule of water that itself is
coordinated to two other NCL moieties, while the S(6) ring involving NCL-NH is maintained
(Figure S9).[38] Note that the rather short c-axis of the unit cell evokes π−π stacking of
identical layers (at a stacking distance of 3.812 Å), similar but not isostructural to the known
solvate of NCL with methanol,[111] while the trimeric coordination observed in NCL-HA is
also present in the crystal structure of the thermodynamically stable hydrate NCL-HB. The
comparison of the unit cells of both hydrates (NCL-HA: P21/a, a ≈ 23.054(7) Å,
b ≈ 16.153(0) Å, c ≈ 3.812(0) Å, β ≈ 92.824(3)°; HB: P21/c, a = 11.332(2) Å, b = 16.964(2) Å,
c = 7.346(3) Å, β = 98.281(2)°) indicates that the a-axis is approximately halved while the c-
axis is doubled through the transformation from hydrate HA to HB, resulting in much less
efficient π-π stacking in HB. It therefore appears feasible that the π-π stacking of NCL is
responsible for the crystallization of the kinetically favored hydrate HA.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 4: The final Rietveld fit obtained for NCL-IMI: experimental data points (red), Rietveld refinement fit
(blue), background (black line), difference Iobs - Icalc (black) and phase tick marks (blue). The vertical blue line
marks the place at which the data points are multiplied with factor five for better visibility of the data in the
higher 2ϴ region.
12
ACS Paragon Plus Environment

Page 13 of 34
Crystal Growth & Design
1
2
3
4
5
6
7
8
9
Niclosamide – imidazole (NCL-IMI). The crystal structure of NCL-IMI was solved in the
monoclinic space group P21/c (No. 14) with one molecule of NCL and one molecule of IMI in
the asymmetric unit, in agreement with the solid state NMR data, where the structure solution
converged with good reproducibility within consecutive hybrid big bang big crunch (HBB-
BC) runs. The final Rietveld refinement (Figure 4) gave an excellent fit of the experimental
PXRD pattern (GOF = 1.412) and Rwp of 2.579%. As before, the S(6) ring (1.739 Å, 137.6°)
remained intact upon co-crystal formation. Based on the observation that all hydrogen bonded
protons are shifted to higher ppm in the 1H MAS NMR spectrum of NCL-IMI, the presence of
an imidazolium ion is assumed, connecting two molecules of NCL via a charge-assisted
{N−H+···-O} (1.613 Å, 157.1°) and a neutral {N−H···O} (2.223 Å, 132.7°) synthon, thereby
leading to the formation of ꢁꢂꢂ(10) chains which constitute layers. The obtained unit cell is
relatively flat and stabilized by π−π stacking (4.026 Å), (Figure 5).
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
Figure 5: (top left) The NCL-IMI (1:1) salt comprises a C2 (10) chain formed by the deprotonated hydroxyl
group of NCL and the aromatic NH+ group of IMI as well as both the NH group of IMI and carbonyl oxygen of
NCL, respectively; (bottom left) π−π interactions (at a distance of 4.026 Å) result in AAA packing of layers along
the c-axis; (right) unit cell of the crystal structure viewed in the ab-plane.
Niclosamide – 2-aminothiazole (NCL-AT). The structure of the (1:1) co-crystal NCL-AT
was solved in the monoclinic space group P21/c (No. 14) with one molecule of NCL and one
molecule of 2-aminothiazole comprising the asymmetric unit, in agreement with the 1H MAS
NMR spectrum. As for NCL-HA, the refinement converged quickly with a good
reproducibility in repeated HBB-BC runs. The final Rietveld refinement (Figure 6) provided
an excellent goodness of fit (GOF = 1.634) and Rwp of 4.227 %. It was found that the
intramolecular {N−H···O} hydrogen bond of NCL (1.747 Å, 140.7°) persisted in the obtained
co-crystal, while 2-aminothiazole links two molecules of NCL via Cꢃꢃ(10) chains, including a
rather strong intermolecular {O−H···N} hydrogen bond (1.587 Å, 170.6°) of the hydroxyl
13
ACS Paragon Plus Environment

Crystal Growth & Design
Page 14 of 34
1
2
3
4
5
6
7
8
9
group of NCL to an aromatic nitrogen of 2-aminothiazole and intermolecular {N−H···O}
hydrogen bond (2.027 Å, 174.8°), connecting the amino group of 2-aminothiazole to the
amide oxygen of NCL. The molecules are arranged in ABAB packing fashion along the c-axis
through π−π stacking (3.645 Å) resulting in a γ-motif (Figure 7).
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 6: The final Rietveld fit obtained for NCL-AT: experimental data points (red), Rietveld refinement fit
(blue), background (black line), difference Iobs - Icalc (black) and phase tick marks (blue). The vertical blue line
marks the place at which the data points are multiplied with factor five for better visibility of the data in the
higher 2ϴ region.
2
Figure 7: (left) The NCL-AT (1:1) co-crystal comprises a C2 (10) chain motif formed by the hydroxyl group of
NCL and the aromatic nitrogen of 2-aminothiazole as well as both the amino group of 2-aminothiazole and
carbonyl oxygen of NCL, respectively; (right) π−π interactions (at a distance of 3.645 Å) lead to ABAB packing
of layers along the c-axis forming a γ-motif.
Niclosamide – acetamide. Based on PXRD and solid-state NMR data it was identified that
the co-crystallization of NCL and acetamide in all considered cases yielded polycrystalline
mixtures of at least two polymorphs. Using a few drops of ethanol during solvent assisted
solid grinding resulted in compound NCL-AA-I as dominating phase while in case of acetone
NCL-AA-II was found as the main product. Nevertheless, since no experimental condition
could be established that would result in pure co-crystal phases (e.g., only NCL-AA-I or
14
ACS Paragon Plus Environment

Page 15 of 34
Crystal Growth & Design
1
2
3
4
5
6
7
NCL-AA-II), multiphase Rietveld refinements were performed after independent structure
solutions of NCL-AA-I and NCL-AA-II were achieved with sufficient reproducibility in the
HBB-BC runs.
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 8: The final Rietveld fit of the co-crystal polymorph NCL-AA-I: experimental data points (red), Rietveld
refinement fit (blue), background (black line), difference Iobs - Icalc (black), phase tick marks of NCL-AA-I (blue)
and phase tick marks of NCL-AA-II (black). The vertical blue line marks the place at which the data points are
multiplied with factor ten for better visibility of the data in the higher 2ϴ region.
Figure 9: Final Rietveld fit of the co-crystal polymorph NCL-AA-II: experimental data points (red), Rietveld
refinement fit (blue), background (black line), difference Iobs - Icalc (black), phase tick marks of NCL-AA-II (blue)
and phase tick marks of NCL-AA-I (black). The vertical blue line marks the place at which the data points are
multiplied with factor five for better visibility of the data in the higher 2ϴ region.
15
ACS Paragon Plus Environment

Crystal Growth & Design
Page 16 of 34
1
2
3
4
5
6
7
8
The Rietveld refinement yielded quite acceptable values for both NCL-AA-I (GOF = 3.587,
Rwp = 6.351 %; Figure 8) and NCL-AA-II (GOF = 4.470, Rwp = 6.235 %; Figure 9), though
the overall quality of the refinement for NCL-AA-II is rather moderate. The corresponding
fraction of impurity through the presence of the other polymorph could be quantified. In case
of NCL-AA-I, a fraction of 13.5% NCL-AA-II was found while the co-crystal of NCL-AA-II
contained a fraction of 6.6% NCL-AA-I. Notably, a comparison of the cell parameters of both
NCL-AA-I and II revealed that the polymorphs are comparable to the hydrates NCL-HA and
NCL-HB, respectively, though upon transformation of NCL-AA-II into NCL-AA-I the
symmetry is lowered from an orthorhombic to a triclinic cell. Hence, the cell axes a and b are
nearly halved whereas the c-axis is doubled so that in total the unit cell content is halved
(NCL-AA-I: Z = 2; NCL-AA-II: Z = 4). As previously observed for the pseudo-polymorphic
hydrates NCL-HA and NCL-HB, the π−π stacking in NCL-AA-II (AAA layers) seems to be
more efficient than in NCL-AA-I (ABA layers) while the overall crystal packing is more
efficient in NCL-AA-I.
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
Figure 10: (left top) The NCL-AA-I (1:1) co-crystal comprises a R4 (20) ring motif between carbonyl oxygen of
NCL and the amide group of acetamide connect ABAB layers of niclosamide. (right top) Within each layer of
2
niclosamide molecules C2 (12) chains link acetamide and NCL through the nitro group and carbonyl oxygen of
NCL and the amide group of acetamide. (bottom) Two antiparallel acetamide ribbons form a column in which
NCL molecules are stacked through π−π interactions at a distance of 3.404 Å.
Niclosamide – acetamide I (NCL-AA-I). The (1:1) co-crystal NCL-AA-I crystallizes in the
triclinic space group P-1 (No. 2) with one molecule of NCL and acetamide in the asymmetric
4
unit. Acetamide links two different layers of NCL via a R4 (20) ring (reminiscent of the NCL-
16
ACS Paragon Plus Environment

Page 17 of 34
Crystal Growth & Design
1
2
3
4
5
6
7
8
urea co-crystal) through {O−H···O} hydrogen bonds (1.694 Å, 176.4°) of NCL-OH and
amide oxygen of acetamide as well as {N−H···O} hydrogen bonds (1.933 Å, 161.7°) of
acetamide to the amide oxygen of NCL. Within one layer acetamide links NCL molecules
2
through C2 (12) chains over an {N−H···O} hydrogen bond (2.161 Å, 137.4°) to the nitro
9
group and {N−H···O} hydrogen bond to the amide group of NCL, respectively. NCL
molecules are connected to the next layer by π−π stacking (at a distance of 3.404 Å) along the
c-axis in an ABAB fashion (Figure 10).
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
Niclosamide – acetamide II (NCL-AA-II). The polymorph NCL-AA-II crystallizes in the
orthorhombic space group P212121 (No. 19) with one molecule of both NCL and acetamide in
1
the asymmetric unit (Figure 11). Acetamide forms infinite columns according to C1 (4) chains
through {N−H···O} hydrogen bonds (2.050 Å, 157.0°). The acetamide columns link two
3
crossing layers of NCL along the c-axis via D2 (7) hydrogen bonding motifs at which the
hydroxyl group of NCL coordinates the oxygen of acetamide through {O−H···O} hydrogen
bonds (1.756 Å, 175.5°). In addition the amide group of AA coordinates the nitro-group of
NCL {N−H···O} so that as all donors and acceptors are saturated (1.987 Å, 142.3°). The NCL
molecules build two columns that are placed diagonally to each other along the c-axis; within
the columns the NCL molecules are connected via π−π stacking (3.882 Å) in AAA fashion,
similarly to the case of the pseudo-polymorph NCL-HA (Figure 11).
Figure 11: The unit cell of the crystal structure of polymorph NCL-AA-II (viewed in the ab plane). The structure
3
1
comprises D2 (7) hydrogen bonding motifs between NCL-OH and the oxygen of AA that is part of C1 (4)
columns, while NCL layers are linked via π π stacking (3.882 Å) in AAA fashion.
−
17
ACS Paragon Plus Environment

Crystal Growth & Design
Page 18 of 34
1
2
3
4
5
6
7
8
9
Structure validation & DFT calculations. The evaluation of the correctness of structure
solutions derived from profile fitting and analysis of PXRD pattern may be challenging
without the input of complementary data. Therefore, structure solutions resulting from the
final Rietveld refinement were subjected to DFT chemical shift computations where
representative fragments (“cutouts”) including important hydrogen bonding pattern and
packing features (close contacts) of the considered structure were taken into account. Since
reasonable structural models with well localized heavier atoms (“skeleton”) are typically
derived from PXRD data,[65,66] merely proton positions were optimized prior DFT NMR
chemical shift computations. While this may not be suitable to tackle long-ranged van-der-
Waals interactions among ordered aliphatic chains or cooperative π−π stacking effects in
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
13
extended lattices or full periodic boundary conditions, the achievable accuracy in H and C
chemical shifts of ±1 and ±(5-6) ppm at PBE/6-311G(d,p) level of theory[57,130] reflecting
rather localized effects effectively allows a discrimination of different structural models at
affordable computational costs. Though improved accuracy may in principle be expected
from using higher basis sets and/or polarization functions, benchmark computations on a
variety of molecules have revealed that this is not always the case, rendering the chosen level
of theory a suitable compromise among the conflicting demands of required accuracy and
computational costs.[131,132]
Table 2: Experimental and DFT calculated 1H chemical shifts of characteristic hydrogen bonds that stabilize the
structures of the considered NCL hydrates and co-crystals. The superscript DQ denotes for the estimation of
single quantum coherences from 1H-1H DQ MAS NMR data, provided that they could be unambiguously
determined (if not, a ’?’ is set). Co-crystals for which no structure solution was obtained are marked with ‘n.s.’
(no structure).
H-bond motif 1H, δiso
1H, δiso
Angle/ °
d(X-HꢀꢀꢀY)/
Phase
(DFT)/ ppm (exp)/ ppm
11.1
11.6
Å
1.742
1.717
NCLO-HꢀꢀꢀO
NCLN-
11.0DQ
11.6DQ
168.6
140.5
NCL-HA
HꢀꢀꢀOHNCL
H2OO-HꢀꢀꢀOket
H2OO-HꢀꢀꢀOnitro 3.8
NCLO-HꢀꢀꢀO
NCLN-
5.5
?
1.803
2.063
1.616
1.747
176.6
168.1
171.9
140.1
4.3DQ
11.6[55]
11.6[55]
12.6
11.9
NCL-HB
NCL-AT
HꢀꢀꢀOHNCL
H2OO-HꢀꢀꢀOket
H2OO-HꢀꢀꢀOnitro 4.8
NCLO-HꢀꢀꢀNarom 15.7
NCLN-
4.7
4.6[55]
4.6[55]
15.3
1.891
1.981
1.587
1.747
168.6
168.0
170.6
140.7
11.7
11.2
HꢀꢀꢀOHNCL
ATN-HꢀꢀꢀONCL
ATN-HꢀꢀꢀONCL
7.8
5.7
7.1DQ
5.7DQ
2.027
2.271
174.8
160.4
18
ACS Paragon Plus Environment

Page 19 of 34
Crystal Growth & Design
1
2
NCLO-HꢀꢀꢀO
11.7
12.3
11.0
1.694
1.810
176.4
139.4
3
4
5
6
7
8
9
NCL-AA-
I
NCLN-
11.0
HꢀꢀꢀOHNCL
AAN-HꢀꢀꢀOket
AAN-HꢀꢀꢀOnitro
NCLO-HꢀꢀꢀO
8.4
7.4
11.5
?
?
11.8
1.933
2.161
1.756
161.7
137.4
175.5
NCL-AA-
II
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
NCLN-
11.4
10.6
1.780
140.5
HꢀꢀꢀOHNCL
AAN-HꢀꢀꢀOket
AAN-HꢀꢀꢀOnitro
NCLO-HꢀꢀꢀO
NCLN-
HꢀꢀꢀOHNCL
NCLO-HꢀꢀꢀNarom n.s.
NCLN-
HꢀꢀꢀOHNCL
NCLO-HꢀꢀꢀNarom n.s.
NCLN-
HꢀꢀꢀOHNCL
NCLO-HꢀꢀꢀNarom n.s.
7.8
8.2
n.s.
n.s.
?
?
12.5
11.7
2.050
1.987
n.s.
157.0
142.3
n.s.
NCL-BA
NCL-IN
NCL-NA
NCL-IA
NCL-IMI
n.s.
n.s.
13.5
11.1
n.s.
n.s.
n.s.
n.s.
n.s.
13.7
11.0
n.s.
n.s.
n.s.
n.s.
n.s.
14.4
10.7
n.s.
n.s.
n.s.
n.s.
NCLN-
n.s.
HꢀꢀꢀOHNCL
NCLO-ꢀꢀꢀ+HNarom 17.9
18.5/ 18.2[55] 1.613
12.7
157.1
137.6
NCLN-
13.7
1.739
HꢀꢀꢀOHNCL
IMIN-HꢀꢀꢀOket
NCLO-HꢀꢀꢀOH
11.7
13.1
12.7
12.3
2.223
1.526
132.7
173.2
NCL-
MeOH
NCLN-
11.9
11.7
1.724
139.7
HꢀꢀꢀOHNCL
NCLO-HꢀꢀꢀO
NCLN-
HꢀꢀꢀOHNCL
NCLO-HꢀꢀꢀO
NCLN-
10.8
11.4
10.5
11.0
1.661
1.774
177.6
139.3
NCL-CAF
NCL-URE
12.5
10.5
12.2
9.8
1.599
1.792
177.3
138.5
HꢀꢀꢀOHNCL
Co-crystallization of molecular solids typically entails rearrangement of the hydrogen
bonding network thus rendering the involved protons, the donor/acceptor atoms and their
neighboring atoms sufficiently sensitive solid state NMR probes for the validation of
1
proposed structure models. The experimental MAS NMR and DFT computed H chemical
shifts of the important hydrogen bonds that stabilize the resulting crystal packing within the
considered crystal structures are summarized in Table 2, and in all cases fit well within
±1 ppm, including NCL-AA-I and NCL-AA-II. In some cases, however, even upon inspection
19
ACS Paragon Plus Environment

Crystal Growth & Design
Page 20 of 34
1
2
3
4
5
of the corresponding 1H-1H DQMAS NMR spectra, the experimental 1H chemical shifts could
not be unambiguously resolved, and thus were excluded from validation.
6
7
Table 3: Experimental and calculated 13C chemical shifts of NCL hydrate HA and selected co-crystals.
8
9
Phase
δiso 13C(C-OH)
exp., (DFT)
/ppm
δiso 13C(C=O)
exp., (DFT)
/ppm
δiso 13C(Camide
exp., (DFT)
/ppm
)
δiso 13C(C-N-C)
exp., (DFT)
/ppm
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
154.0
163.0
NCL
154.7, (156.7)
155.4, (159.9)
154.1, (156.4)
156.1
155.4
155.9
156.3
157.6, (158.1)
166.3, (172.3)
162.8, (160.3)
163.3, (165.4)
161.8, (170.5)
164.8
163.9
164.7
162.8
165.9, (165.0)
161.6, (167.6)
NCL-HA
NCL-AA-I
NCL-AA-II
NCL-BA
NCL-IN
NCL-NA
NCL-IA
NCL-AT
NCL-IMI
178.7 (171.6)
179.9 (171.5)
173.5
151.9
169.4
150.2/148.2
148.0/149.0
150.2/150.2
134.9, (134.2)
143.0, (137.0)
166.1
172.3 (172.3)
In addition, the experimental and DFT computed 13C chemical shifts of selected co-
crystals are collected in Table 3. While in most cases, the calculated shifts fit reasonably well
(± 6 ppm) with the experimental values, the deviation is somewhat larger in case of NCL-AA-
I and NCL-AA-II, though the agreement is much better than in case of those structure
1
solutions considered as not validated. Nevertheless, since the agreement is good for the H
NMR shifts which are considered to be more reliable probes for the evaluation of
structures,[133-135] the obtained structure solutions for NCL-HA (in agreement with a recently
reported structure[38]), NCL-AT, NCL-AA-I, NCL-AA-II, and NCL-IMI could be successfully
validated. Note that the 15N NMR chemical shifts are not well reproduced at the applied level
of theory, rather reflecting trends, and thus are not included in the structure validation. Since
1
13
both the H and C DFT computed chemical shifts of structure solutions derived for the co-
compounds NCL-BA, NCL-IN, NCL-NA and NCL-IA were inconsistent with the available
experimental solid state NMR data, these structures could not be validated even though they
exhibited plausible hydrogen bonding pattern as well as reasonably good profile fits and R-
values.
Thermal Stability of Niclosamide Co-crystals. The thermal stability and phase transition of
drugs or active pharmaceutical ingredients are highly important to elucidate physicochemical
and/or pharmacokinetic properties and likely changes thereof upon prolonged storage due to
e.g. de-/rehydration or polymorphism, respectively. The DSC curve obtained for NCL-AA-II
20
ACS Paragon Plus Environment

Page 21 of 34
Crystal Growth & Design
1
2
3
4
5
6
7
8
9
indicated two endotherm peaks at 171.5°C and 157.7°C (which could be due to the presence
of residual NCL-AA-I), followed by the sublimation of AA and melting process of pristine
NCL. In contrast, the polymorph NCL-AA-I showed merely one endotherm peak at 159.2°C
in addition to sublimation of AA and melting of NCL even though the multiphase Rietveld
analysis indicated a larger residual fraction of NCL-AA-II. Note that the dehydration of NCL-
HA was observed at 86.9°C while the remaining new co-crystals exhibited a single endotherm
peak at temperatures ranging between those of the pristine drug and co-formers thereby
suggesting no decomposition before melting (Table 4).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table 4: Melting points and endotherm peaks identified in the corresponding DSC curves of NCL co-crystals
and NCL-HA hydrate. In case of NCL-AA-II, the peak marked with an asterisk could be due to residual NCL-AA-
I; in case of NCL-NA, the melting point is not well-defined.
Phase
NCL
Peak /°C
228.3
NCL-HA
NCL-AA-I
NCL-AA-II
NCL-BA
NCL-IN
NCL-NA
NCL-IA
NCL-AT
NCL-IMI
86.9 (dehydration)
159.2
157.7*; 171.5
189.6
185.5
≈222.7
226.7
187.5
213.5
Table 5: The determined equilibrium solubility of the considered NCL co-crystals. Reference data obtained from
already reported co-crystals[48] is given in brackets for comparison. Note that the discrepancies likely result
from normalization with respect to a fixed amount of NCL per sample.
Phase
Solubility at 37° C (24 h) in 40%
isopropanol-water mixture (mg L-1)
87.15
NCL
NCL-AA-I
NCL-BA
NCL-IN
NCL-NA
NCL-IA
NCL-AT
NCL-URE
NCL-CAF
NCL-PABA
122.29 (x1.4)
109.34 (x1.3)
105.29 (x1.2)
114.45 (x1.3) [x1.2][48]
113.45 (x1.3) [x1.7][48]
245.87 (x2.8)
84.89 (x1.0) [x2.0][48]
114.5 (x1.3) [x1.3][48]
109.1 (x1.1) [x0.8][48]
Solubility of Niclosamide Co-crystals. The equilibrium solubility of NCL, the obtained co-
crystals and some reference compounds was determined based on reported procedures,[136,137]
21
ACS Paragon Plus Environment

Crystal Growth & Design
Page 22 of 34
1
2
3
4
5
6
7
8
9
and could be reproduced in repeated measurements (Table 5). Since co-crystals containing
low molecular weight co-formers may exhibit superior solubility compared to other
formulations, we have normalized the amount of NCL actually present in the tested
compounds to 50 mg per sample, thereby adjusting the initially weighted sample accordingly,
in this way allowing better comparability of the determined solubility values (note that NCL-
AA-II and NCL-IMI were both excluded from the solubility tests). Nevertheless, the
solubility values of NCL-NA, NCL-IA, NCL-URE and NCL-PABA co-crystals derived from
our measurements differ from the reported data, likely resulting from the fact that in reference
[48] the authors did not normalize the weights of solid material used for solubility tests to the
actual amount of NCL. Among the considered compounds, the most significant solubility
improvement could be achieved in case of NCL-AT whose equilibrium solubility was found
to be 2.8 times higher than that of pristine NCL, which (to the best of our knowledge) is the
highest equilibrium solubility of NCL co-crystals currently observed. Rather unexpectedly,
the reasonably high solubility of NCL-URE reported in the literature could not be reproduced,
possibly due to undesired hydrate formation upon prolonged storage (a strong tendency for
hydrate formation has previously been observed for NCL-PABA, NCL-URE and NCL-CAF
after a storage period of 4−6 weeks).[48] In contrast, the novel NCL-AT co-crystal is
demonstrably stable for a period of more than 18 months, most probably due to its strong
hydrogen bonding motifs. Also, a moderately increased equilibrium solubility was observed
for the co-crystals of NCL-BA, NCL-IN and NCL-AA-I, respectively.
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. Conclusion
An enhanced equilibrium solubility of the anthelmintic drug niclosamide could be
achieved via hydrogen bond mediated co-crystal formation based on crystal engineering
principles. Suitable co-formers of NCL were taken from the GRAS and EAFUS lists, thereby
rendering all obtained co-phases not involving salt formation pharmaceutically acceptable and
promising candidates for the current trend of repurposing long known drugs for the treatment
of lifestyle diseases such as cancer.[138] Aspects of “green chemistry” were also considered by
producing the new crystalline phases from either solvent-drop assisted grinding and/or
mechanical treatment in addition to slow evaporation from environmentally friendly solvents
such as ethanol. Indeed, structure solutions for NCL-AT, NCL-IMI, the polymorphs NCL-
[38]
AA-I and NCL-AA-II, and NCL-HA
were derived from powdered samples combining
powder X-ray diffraction and solid state NMR data, evidencing successful co-crystal
22
ACS Paragon Plus Environment

Page 23 of 34
Crystal Growth & Design
1
2
3
4
5
formation based on the rather robust {O−H···O}, {N−H···O} and {O−H···N} synthons,
respectively.
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
4. Experimental Section
NCL was purchased from Cayman Chemical Company (Tallinn, Estonia) while AA, AT,
BA and NA were obtained from Acros Organics (Geel, Belgium); IA and IN were delivered
from Sigma Aldrich (Saint-Quentin Fallavier, France). All compounds were used without
further purification.
Synthesis. Niclosamide hydrate, NCL-HA. 100 mg (0.31 mmol) of niclosamide were refluxed
in water (6 mL) for 2 h and the product was cooled to room temperature over 6 h.
Dehydration 86.9 °C.
NCL-IMI (1:1) Co-crystal. 500 mg (1.53 mmol) of niclosamide and 104.06 mg (1.53 mmol)
of imidazol were ground in mortar-pestle for 15 min after adding 5 drops of dry EtOH. Mp
213.5 °C.
NCL-AT (1:1) Co-crystal. 460 mg (1.41 mmol) of niclosamide and 140.82 mg (1.41 mmol) of
2-aminothiazole were ground in mortar-pestle for 15 min after adding 5 drops of dry EtOH.
Mp 187-189 °C.
NCL-AA-I (1:1) Co-crystal. 510 mg (1.56 mmol) of niclosamide and 92.09 mg (1.56 mmol)
of acetamide were ground in mortar-pestle for 15 min after adding 5 drops of dry EtOH. Mp
158-160 °C.
NCL-AA-II (1:1) Co-crystal. 510 mg (1.56 mmol) of niclosamide and 92.09 mg (1.56 mmol)
of acetamide were ground in mortar-pestle for 15 min after adding 5 drops of Acetone. Mp
157-159 °C.
23
ACS Paragon Plus Environment

Crystal Growth & Design
Page 24 of 34
1
2
3
4
5
6
7
NCL-BA (1:1) Co-crystal. 450 mg (1.38 mmol) of niclosamide and 166.65 mg (1.38 mmol) of
benzamide were ground in mortar-pestle for 15 min after adding 5 drops of dry EtOH. Mp
189-191 °C.
8
9
NCL-IN (1:1) Co-crystal. 430 mg (1.31 mmol) of niclosamide and 180.27 mg (1.31 mmol) of
isoniazide were ground in mortar-pestle for 15 min after adding 5 drops of dry EtOH. Mp
185-187 °C.
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
NCL-NA (1:1) Co-crystal. 450 mg (1.38 mmol) of niclosamide and 167.99 mg (1.38 mmol) of
nicotinamide were ground in mortar-pestle for 15 min after adding 5 drops of dry EtOH. Mp
221-223 °C.
NCL-IA (1:1) Co-crystal. 450 mg (1.38 mmol) of niclosamide and 167.99 mg (1.38 mmol) of
isonicotinamide were ground in mortar-pestle for 15 min after adding 5 drops of dry EtOH.
Mp 226−228 °C.
Powder X-ray Diffraction. Bulk samples for phase identification were analyzed by PXRD
on a Bruker D8 Discover powder diffractometer. Experimental conditions: Cu−Kα1/2 radiation
(λ1 = 1.540596 Å, λ2 = 1.544410 Å; ratio Kα1/ Kα2 = 0.441122); 40 kV; 40 mA; scanning
interval 5−50° 2θ at a step size of 0.01°; time per step 3 s; T = 295 K in Bragg-Brentano
geometry. In addition, all samples for structure solution were analyzed with a STOE StadiP
powder diffractometer. Experimental conditions: Cu−Kα1 radiation (λ1 = 1.540598 Å); 45 kV;
30 mA; scanning interval 0−70° 2θ at a step size of 0.1°; time per step 120 s; T = 293 K in
Debye-Scherrer geometry. Indexing of the powder patterns was performed using N-
TREOR[139] and further confirmed with DICVOL04.[140] The corresponding cell volume was
verified by calculating expected cell volumes from volume increments. [141] For space group
determination Le-Bail refinement was used for whole pattern fitting thereby extracting the
integrated peak intensities and their correlations using the expo2014 software package.
Structures were solved in direct space using the simulated annealing (SA) approach or hybrid
big bang big crunch (HBB-BC) algorithm as implemented in expo2014, allowing flexible
torsion angles as well as six degrees of freedom for rotation and translation, respectively, for
each molecule. After an initial structure model was obtained with all hydrogen atoms
removed during the optimization (50 runs) hydrogen atoms were inserted at typical positions
using Mercury 3.3 and Rietveld refinement was then performed using Bruker TOPAS 4.2.
24
ACS Paragon Plus Environment

Page 25 of 34
Crystal Growth & Design
1
2
3
Solid-State NMR spectroscopy. Solid-state NMR measurements were performed on either
4
5
B
RUKER AVANCE III 300 or AVANCE DSX 500 spectrometers, corresponding to magnetic
6
7
8
9
flux densities of 7.05 T and 11.74 T. The spectrometers were equipped with commercially
available BRUKER 4 mm double and triple resonance probes operating at MAS rotation
frequencies between 3.0 and 15.0 kHz while at the 500 MHz spectrometer, either 2.5 mm
triple or double resonance probes operating at MAS rotation frequencies of up to 30 kHz were
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
applied. 13C{1H} cross-polarization (CP)-MAS NMR spectra were measured with H 90°
pulse lengths of 5 ꢁs (corresponding to a radiofrequency field (ν1) of 50 kHz), a contact time
of 2.5 ms with spinning frequency of 12 kHz and relaxation delays of 10 to 120 s, depending
on the proton T1 relaxation times. Hartmann – Hahn conditions were adjusted on 1-13C-15N-
labelled α-glycine; an efficient polarization transfer was achieved by a ramped-amplitude CP
step with ν1(1H) being swept from 54 kHz to 27 kHz in 64 steps (in the case of a 1H 90° pulse
length of 5.0 ꢁs).[142] All the spectra were obtained with TPPM-15 proton decoupling[143]
during the data acquisition applying decoupling pulses of 6.7 ꢁs to 10.0 ꢁs length (∼10/12 π-
pulse). Chemical shifts are reported relative to the secondary standard 1-13C-15N-labelled
glycine (1-13C signal at 176.5 ppm). The 15N{1H} CPMAS-NMR experiments were
1
performed at 7.05 T using the following acquisition parameters: a H 90° pulse length of
7.0 ꢁs, a contact time of 5 ms, a spinning speed of 10 kHz and a relaxation delay of 10-30 s;
15N chemical shifts are reported with respect to 1-13C-15N-labelled glycine (set to -
1
347.6 ppm). All the H MAS NMR spectra were measured at 11.74 T with 2.5 µs 90° pulse
length, a spinning speed of 30 kHz, and relaxation delays of 120-2000 s. A back-to-back
(BaBa) recoupling sequence was used to excite and reconvert double-quantum (DQ)
coherences at 11.74 T with 2.5 µs 90° pulse length, a spinning speed of 29762 Hz (due to
rotor-synchronized detection in the F1 dimension), a dwell time of 5 µs and an excitation time
of τexc=33.6 µs. For the 2D 1H-1H DQ-MAS-NMR experiments 64-256 t1 increments in steps
of 33.6 µs (corresponding to a rotor period) and 16 transients per increment were acquired.
Spectral deconvolution and line-shape analysis were performed with DMFIT (version
2011).[144]
Thermal Analysis. DSC analysis of the co-phases was performed on a NETZSCH DSC 204
Phoenix®. Samples were placed in crimped but vented aluminum pans. A typical sample size
was 4 − 10 mg, the measured temperature range was {20 − 250°C} at 10 °C/min; all samples
were purged with a dry nitrogen flow at 125 mL/min during the measurement.
25
ACS Paragon Plus Environment

Crystal Growth & Design
Page 26 of 34
1
2
3
4
5
6
7
8
9
Solubility Experiments. The quantification of NCL was validated using a Phenomenex
Gemini-NX HPLC-UV/vis and a C18 column (250 mm x 4.6 mm, 5 mm particle diameter).
The mobile phase consisted of a 0.05 M methanol / (NH4)3PO4 (9:1) buffer, adjusted to a pH
value of 3.6 with H3PO4. The considered concentration range of NCL was in between 0.5-
100 µg/mL. The flow rate and the injection volume was set to 1.0 mL/min and 50 µL,
respectively. The limit of detection was 0.1 µg/mL and the limit of quantification 0.5 µg/mL.
The absorbance was measured at 254 nm. The equilibrium solubility was determined in 40%-
isopropanol − water medium using the shake-flask method.[136,137] Accordingly, each solid
sample containing 50 mg of NCL was stirred for 24 h at 37°C in 2.5 mL of the medium. The
concentration of the saturated solution was calculated after 24 h which is referred to as the
equilibrium solubility of the considered solid form.
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
DFT calculations. For the sake of reasonable accuracy at affordable computational costs, the
DFT chemical shift calculations were performed in Gaussian09[145] considering suitably
chosen cluster models that were extracted from the obtained PXRD structure solutions or
where available from single-crystal data. The cluster models were created such that all
characteristic hydrogen bonds or close proximities due to packing were satisfied. Though the
hydrogen atoms were Rietveld refined in the final step of the structure solution process, all
atoms except for hydrogen were “frozen” at given coordinates (assuming that a reasonable
“skeleton” structure was derived from Rietveld refinement) for H-bond length optimization
1
(opt = ModRedundant) at DFT/PBE1PBE/6-311G(d,p) level of theory. The H and 13C
chemical shielding values were computed from optimized cluster models (considering
important hydrogen bonding, short contacts and molecular packing) at PBE1PBE/6-311G(d,p)
level of theory and “translated” into the corresponding chemical shifts with respect to either
benzene (1H, 13C) or methanol (1H, 13C), respectively, depending on the orbital hybridization:
sp3 (methanol); sp2,sp (benzene)) following the multi-standard approach.[131,132]
1
Supporting Information Available: 13C{1H} & 15N{1H} CPMAS and 2D H−1H DQMAS
NMR spectra (except for NCL-IMI) as well as additional PXRD data of the co-crystals are
available free of charge via internet at http://pubs.acs.org. The following crystal structures
have been submitted to the Cambridge Crystallographic Data Centre (CCDC): NCL-AT
(1437254), NCL-AA-II (1437255), NCL-AA-I (1437256), NCL-HA (1437257), NCL-
IMI_salt (1437258).
26
ACS Paragon Plus Environment

Page 27 of 34
Crystal Growth & Design
1
2
3
Notes: Some parts of the publication required for patent application were submitted to the
4
5
European Patent Office (application number EP15193955.0).
6
7
8
Acknowledgements
9
Funding by the Deutsche Forschungsgemeinschaft (DFG) project EC168/10-2 (“Dynamische
Supramolekulare Assoziationen”) is gratefully acknowledged. G. B. appreciates access to the
High Performance Computing facilities (WWU-PALMA), kindly provided by Prof. A. Heuer
(Univ. Münster). Prof. M. Winter (Univ. Münster, MEET) is acknowledged for his support
and for providing access to Bruker D8 powder X-ray diffractometers. In addition, technical
support and maintenance of the STOE StadiP powder diffractometer by T. Fickenscher (Univ.
Münster) is highly appreciated.
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
5. References
(1) Craig, P.; Akira, I. Curr. Opin. Infect. Dis. 2007, 20, 524.
(2) Scholefield, R. J.; Bergstedt, R. A.; Bills, T. D. J. Great Lakes Res. 2003, 29, 493.
(3) Huang, H. C.; Liao, S. C.; Chang, F. R.; Kuo, Y. H.; Wu, Y. C. J. Agr. Food Chem. 2003, 51, 4916.
(4) Park, S. J.; Shin, J. H.; Kang, H.; Hwang, J. J.; Cho, D. H. BMB Rep. 2011, 44, 517.
(5) Abreu, F. C.; Goulart, M. O. F.; Oliveira, A. B. M. Biosens. Bioelectron. 2002, 17, 913.
(6) Lardans, V.; Dissous, C. Parasitol. Today 1998, 14, 413.
(7) Andrews, P.; Thyssen, J.; Lorke, D. Pharmacol. Therapeut. 1983, 19, 245.
(8) Gill, G.; Beeching, N. Lecture Notes: Tropical Medicine 2009, 6th edition, Blackwell Publishing Ltd., pp. 202
− 207.
(9) World Health Organization: Specification and Evaluation for Public Health Pesticides (Niclosamide).
Geneva: WHO 2002, 1.
(10) Chang, Y. W.; Yeh, T. K.; Lin, K. T.; Chen, W. C.; Yao, H. T.; Lan, S. J.; Wu, Y. S.; Hsieh, H. P.; Chen, C.
M.; Chen, C. T. J. Food Drug Anal. 2006, 14, 329.
(11) Al-Hayida, B. M. Profiles Drug Subst. Excip. Relat. Methodol. 2005, 32, 67.
(12) Liang, L.; Huang, M.; Xiao, Y.; Zen, S.; Lao, M.; Zou, Y.; Shi, M.; Yang, X.; Xu, H. Inflamm. Res. 2015,
64, 225.
(13) Gies, E.; Wilde, I.; Winget, J. M.; Brack, M.; Rotblat, B.; Novoa, C. A.; Balgi, A. D.; Sorensen, P. H.;
Roberge, M.; Mayor, T. PLOS One 2010, 5, e14410.
(14) Li, Y.; Li, P. K.; Roberts, M. J.; Arend, R. C.; Samant, R. S.; Buchsbaum, D. J. Cancer Lett. 2014, 349, 8.
(15) Lee, S.; Son, A. R.; Ahn, J.; Song, Y. J. Biomed. Pharmacother. 2014, 68, 619.
(16) Yokoama, M. J. Drug Target. 2014, 22, 576.
27
ACS Paragon Plus Environment

Crystal Growth & Design
Page 28 of 34
1
2
3
4
(17) Crucho, C. I. C. ChemMedChem 2015, 10, 24.
5
6
7
8
(18) Dong, R.; Zhou, Y.; Huang, X.; Zhu, X.; Lu, Y.; Shen, J. Adv. Mater. 2015, 27, 498.
(19) Ma, X.; Zhao, Y. Chem. Rev. 2015, 115, 7794.
9
(20) Zhang, M.; Xuzhou, H.; Feihe, N.; Gibson, H. W. Acc. Chem. Res. 2014, 47, 1995.
(21) Ghosh, I.; Nau, W. M. Adv. Drug. Delivery Rev. 2012, 64, 764.
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) Morissette, S. L.; Almarsson, Ö.; Petersson, M. L.; Remenar, J. F.; Read, M. J.; Lemmo, A. V.; Ellis, S.;
Cima, M. C.; Gardner, C. R. Adv. Drug Delivery Rev. 2004, 56, 275.
(23) Huang, L. F.; Tong, W. Q. Adv. Drug Delivery Rev. 2004, 56, 321.
(24) Storey, R.; Docherty, R.; Higginson, P.; Dallman, C.; Gilmore, C.; Barr, G.; Dong, W. Crystallography
Reports 2004, 10, 45.
(25) Yu, L. Adv. Drug Delivery Rev. 2001, 48, 27.
(26) Laitinen, R.; Löbmann, K.; Strachan, C. J.; Grohganz, H.; Rades, T. Int. J. Pharm. 2013, 453, 65.
(27) Kaushal, A. M.; Gupta, P.; Bansal, A. K. Crit. Rev. Ther. Drug Carrier Syst. 2004, 21, 133.
(28) Craig, D. Q. M.; Royal, P. G.; Kett, V. L.; Hopton, M. L. Int. J. Pharm. 1999, 179, 179.
(29) Löbmann, K.; Laitinen, R.; Grohganz, H.; Gordon, K. C.; Strachan, C.; Rades, T. Mol. Pharmaceutics 2011,
8, 1919.
(30) Alhalaweh, A.; Alzghoul, A.; Kaialy, W.; Mahlin, D.; Bergström, C. A. S. Mol. Pharmaceutics 2014, 11,
3123.
(31) Dai, J.; Wang, W.; Liang, Y.; Li, H.; Guan, X.; Zhu, Y. Parasitol. Res. 2008, 103, 405.
(32) Kenawy, E.; Rizk, E. Macromol. Biosci. Polymer 2004, 4, 119.
(33) Yang, W.; de Villiers, M. M. AAPS J. 2005, 7, E241.
(34) Devarakonda, B.; Hill, R. A.; Liebenberg, W.; Brits, M.; de Villiers, M. M. Int. J. Pharm. 2005, 304, 193.
(35) Bai, M. Y.; Yang, H. C. Colloids and Surfaces A: Physicochem. Eng. Aspects 2012, 419, 248.
(36) Jin, Y.; Lu, Z.; Ding, K.; Li, J.; Du, X.; Chen, C.; Sun, X.; Wu, Y.; Zhou, J.; Pan, J. Cancer Res. 2010, 70,
2516.
(37) Chen, H.; Yang, Z.; Ding, C.; Chu, L.; Zhang, Y.; Terry, K.; Liu, H.; Shen, Q.; Zhou, J. ACS Med. Chem.
Lett. 2013, 4, 180.
(38) Sovago, I.; Bond, A. D. Acta. Cryst. C 2015, 71, 394.
(39) Gallardo-Godoy, A.; Gever, J.; Fife, K. L.; Silber, B. M.; Prusiner, S. B.; Renslo, A. R. J. Med. Chem. 2011,
54, 1010.
(40) Steed, J. W. Trends Pharmacol. Sci. 2013, 34, 185.
(41) Miroshnyk, I.; Mirza, S.; Sandler, N. Expert Opin. Drug Delivery 2009, 6, 333.
(42) Aakeröy, C. B.; Forbes, S.; Desper, J. CrystEngComm 2014, 16, 5870.
(43) Khan, M.; Enkelmann, V.; Brunklaus, G. J. Am. Chem. Soc. 2010, 132, 5254.
28
ACS Paragon Plus Environment

Page 29 of 34
Crystal Growth & Design
1
2
3
4
5
6
(44) Springuel, G.; Norberg, B.; Robeyns, K.; Woutes, J.; Leyssens, T. Cryst. Growth Des. 2012, 12, 475;
Stanton, M. K.; Tufekcic, S.; Morgan, C.; Bak, A. Cryst. Growth Des. 2008, 9, 1344.
7
8
(45) Khan, M.; Enkelmann, V.; Brunklaus, G. CrystEngComm 2011, 13, 3213.
9
(46) Thakuria, R.; Delori, A.; Jones, W.; Lipert, M. P.; Roy, L.; Rodriguez-Hornedo, N. Int. J. Pharm. 2013,
453, 101.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(47) Trask, A. V. Mol. Pharm. 2007, 4, 301.
(48) Sanphui, P.; Kumar, S. S.; Nangia, A. Cyrst. Growth Des. 2012, 12, 4588.
(49) Merz, K.; Vasylyeva, V. CyrstEngComm 2010, 12, 3989.
(50) Desiraju, G. R. Angew. Chem. Int. Ed. 1995, 34, 2311.
(51) Desiraju, G. R. Chem. Commun. 1997, 16, 1475.
(52) Khan, M.; Enkelmann, V.; Brunklaus, G. J. Org. Chem. 2009, 74, 2261.
(53) Khan, M.; Enkelmann, V.; Brunklaus, G. CrystEngComm 2009, 74, 1001.
(54) Khan, M.; Enkelmann, V.; Brunklaus, G. Cryst. Growth Des. 2009, 9, 2354.
(55) Grifasi, F.; Chierotti, M. R.; Gaglioti, K.; Gobetto, R.; Maini, L.; Braga, D.; Dichiarante, E.; Curzi, M.
Cryst. Growth Des. 2015, 15, 1939.
(56) Hanslick, J. L.; Lau, K.; Noguchi, K. K., Olney, J. W.; Zorumski, C. F.; Mennerick, S.; Farber, N. B.
Neurobiol. Dis. 2009, 34, 1.
(57) Luedeker, D.; Brunklaus, G. Solid State Nucl. Magn. Reson. 2015, 65, 29.
(58) Harris, R. K.; Wasylishen, R. E.; Duer, M. J. NMR Crystallography, Wiley, 2009.
(59) Taulelle, F. Solid State Sci. 2004, 6, 1053.
(60) Senker, J.; Seyfarth, L.; Voll, J. Solid State Sci. 2004, 6, 1029.
(61) Harris, R. K. Solid State Sci. 2004, 6, 1025.
(62) Ashbrook, S. E. Phys. Chem. Chem. Phys. 2009, 11, 6892.
(63) Ashbrook, S. E.; Duer, M. J. Concepts Magn. Reson. A 2006, 28, 183.
(64) Harris, R. K.; Hodgkinson, P.; Pickard, C. J.; Yates, J. R.; Zorin, V. Magn. Reson. Chem. 2007, 45, S174.
(65) Harris, K. D. M. Top. Curr. Chem. 2012, 315, 133.
(66) Powell, H. R. Annu. Rep. Prog. Chem. C 2010, 106, 192.
(67) Price, S. L. Acc. Chem. Res. 2009, 42, 117.
29
ACS Paragon Plus Environment