Two anhydrous and a trihydrate form of tilorone dihydrochloride: Hydrogen-bonding patterns and reversible hydration/dehydration solid-state transformation

Article abstract of DOI:10.1021/cg3012475

During the polymorph screening of an active pharmaceutical ingredient tilorone dihydrochloride (chemical name 2,7-bis[2-(diethylamino)ethoxy]-9- fluorenone dihydrochloride) two new polymorphic modifications - III and IV - were obtained. The crystal structures of both polymorphs were established from X-ray powder diffraction. An interesting phenomenon has been observed at ambient conditions for III, which transforms into a novel hydrated form IIIh during 2-3 h when the humidity in the storage room increases to 70% or more. As soon as the relative humidity falls to 30% or less, IIIh transforms into the parent anhydrous form III within an hour. Form IIIh has been identified as a trihydrate of tilorone dihydrochloride, and its crystal structure has been established from X-ray powder diffraction. On the basis of the crystal structures and hydrogen-bonding patterns, a mechanistic model of reversible hydration/dehydration solid-state transformation III ? IIIh depending on the relative humidity is proposed.

Full text of DOI:10.1021/cg3012475

Article
pubs.acs.org/crystal
Two Anhydrous and a Trihydrate Form of Tilorone Dihydrochloride:
Hydrogen-Bonding Patterns and Reversible Hydration/Dehydration
Solid-State Transformation
Published as part of the Crystal Growth & Design virtual special issue In Honor of Prof. G. R. Desiraju
,
†,‡
†
§
§
Vladimir V. Chernyshev,* Alexandr V. Yatsenko, Sergey V. Pirogov, Tamara F. Nikulenkova,
§
‡
†
†
Ekaterina V. Tumanova, Ivan S. Lonin, Ksenia A. Paseshnichenko, Andrei V. Mironov,
and Yurii A. Velikodny†
†Department of Chemistry, M. V. Lomonosov Moscow State University, 1-3 Leninskie Gory, Moscow 119991, Russian Federation
‡A. N. Frumkin Institute of Physical Chemistry and Electrochemistry RAS, 31 Leninsky Prospect, Moscow 119071, Russian
Federation
§
The Medical-Biological Research-Industrial Complex CYTOMED, 2 Muchnoy pereulok, St. Petersburg 191023, Russian Federation
*
S Supporting Information
ABSTRACT: During the polymorph screening of an active
pharmaceutical ingredient tilorone dihydrochloride (chemical
name 2,7-bis[2-(diethylamino)ethoxy]-9-fluorenone dihydro-
chloride) two new polymorphic modifications  III and IV

were obtained. The crystal structures of both polymorphs
were established from X-ray powder diffraction. An interesting
phenomenon has been observed at ambient conditions for III,
which transforms into a novel hydrated form IIIh during 2−3
h when the humidity in the storage room increases to 70% or
more. As soon as the relative humidity falls to 30% or less, IIIh
transforms into the parent anhydrous form III within an hour. Form IIIh has been identified as a trihydrate of tilorone
dihydrochloride, and its crystal structure has been established from X-ray powder diffraction. On the basis of the crystal structures
and hydrogen-bonding patterns, a mechanistic model of reversible hydration/dehydration solid-state transformation III ↔ IIIh
depending on the relative humidity is proposed.
INTRODUCTION
ethoxy]-9-fluorenone, is an orally active antiviral agent with a
■
1
−6
broad spectrum of action. In 1996 tilorone was approved by
Pharmacological Committee of Russian Federation for viral
infection therapeutics and prophylaxis, and since 2005 it has
been widely used in Ukraine.
Tilorone dihydrochloride (trade name Amixin IC), the orange
water-soluble dihydrochloride salt of 2,7-bis[2-(diethylamino)-
Scheme 1. Tilorone dihydrochloride
Pharmaceutical solids can exist in multiple crystalline forms,
7
−12
and this phenomenon is known as polymorphism.
A recent
study estimated that 80−90% organic compounds are able to
13
exist in polymorphic forms. Physicochemical properties of
polymorphs, e.g., melting point, density, morphology, solubility
and color, can differ significantly. In turn, this may have an
impact on the stability, bioavailability and processability during
14
manufacturing and/or at the stage of final product. It should
be noted that transformations between various crystalline forms
(
e.g., hydration, dehydration and polymorphic transformations)
are often induced by environmental changes and may occur
during drug manufacturing processes or during storage of a
15
drug substance or dosage form. The multitude and diversity
Figure 1. Microscopy photographs of the studied samples.
Received: August 28, 2012
Revised: October 24, 2012
©
XXXX American Chemical Society
A
dx.doi.org/10.1021/cg3012475 | Cryst. Growth Des. XXXX, XXX, XXX−XXX

Crystal Growth & Design
Article
Table 1. Crystal Data for III, IIIh and IV
III
IIIh
IV
2
2+
−
C H N O 2+·2Cl−
483.46
empirical formula
formula weight
particle morphology, color
wavelength, Å
crystal system
space group
a, Å
C H N O +·2Cl−
483.46
C H N O ·2Cl ·3H O
25 36 2 3 2
537.51
2
5
36
2
3
25 36
2
3
no specific habit, orange
1.54059
no specific habit, dark orange
1.54059
needle, orange
1.78897
monoclinic
monoclinic
monoclinic
P2 /c
1
P2 /c
1
P2 /n
1
15.7674(18)
17.7102(19)
18.7680(17)
19.6955(19)
6.9450(11)
95.096(18)
2557.0(5)
45
b, Å
13.3128(15)
13.3601(17)
c, Å
β, o
12.2800(14)
12.1598(16)
95.71(2)
96.39(2)
volume, Å3
2564.9(5)
2859.2(6)
a
M20
25
30
b
F30
51 (0.008, 62)
72 (0.006, 69)
117 (0.004, 58)
4
Z
4
4
−
3
D , g cm
1.252
1.249
1.256
x
2
θmin − 2θmax, increment, o
3.00 − 80.0, 0.01
3.00 − 80.0, 0.01
5.00 − 80.0, 0.01
172/110
no. params/restraints
172/110
184/102
c
Rp/Rwp/Rexp [R /Rwp, Pawley fit]
0.028/0.036/0.020 [0.024/0.032]
1.695
0.025/0.032/0.022 [0.021/0.029]
1.437
0.024/0.032/0.022 [0.022/0.030]
1.352
p
goodness-of-fit
a
b
c
M20 is defined according to ref 50. F is defined according to ref 51. R , R and Rexp are defined according to ref 52.
30
p
wp
of solid forms require a thorough understanding of solid-state
phenomena that may occur in pharmaceutical materials. This
can be achieved by applying various analytical methods.
exposure to humid air upon handling and storage. Hydrates
often have unique spectra, diffraction patterns, or thermograms
and thus can be easily distinguished from other hydrate species
or anhydrous states using the respective methods, and many of
these methods were used in recent investigations of hydration
1
6,17
So far two polymorphic modifications of anhydrous tilorone
dihydrochloride  monoclinic I (a = 8.32 Å, b = 43.32 Å, c =
17,20
7.08 Å, β = 93.9°) and orthorhombic II (a = 8.55 Å, b = 39.81
processes in pharmaceuticals.
For the crystalline pharma-
Å, c = 7.38 Å)  along with the characteristic powder patterns
ceutical solids, the known crystal structures of anhydrous and
hydrated forms offer valuable information, which helps to
understand the specificity of hydration/dehydration processes
and apply ab initio calculations in their quantifications. The
most powerful technique used for crystal structure determi-
nation is single-crystal X-ray diffraction. However, it is often
difficult to grow suitable-quality single crystals of biologically
active organic compounds and therefore, pharmaceuticals: the
growing procedure results only in polycrystalline powders. Even
if single crystals of good quality are successfully grown, they
may lose their integrity during hydration/dehydration trans-
formations, producing a polycrystalline phase. Fortunately,
there are methods for structure determination of organic
1
8
were reported. In the framework of the polymorph screening
carried out by our group, two new polymorphic modifications

III and IV  were obtained and characterized by various
methods including DSC, TGA, IR and NMR spectroscopy, and
X-ray powder diffraction. The powder patterns of III and IV
were initially measured in February and successfully indexed in
3
monoclinic unit cells with the volumes of 2565 and 2557 Å ,
respectively. Three months later (in May) new X-ray
measurements of the same samples revealed that IV was stable,
while III transformed into a new form (IIIh) with a new
powder pattern indexed in the monoclinic unit cell with the
3
volume of 2850 Å . All measurements were carried out at one
diffractometer and at the same temperature 25 °C. The only
difference was found outside the experimental room, and it was
the difference in the outdoor temperature, which, however,
resulted in different relative humidity in experimental room 
around 30% in February and around 70% in May. That is why
the hydration seems to be responsible for the solid-state
transformation III → IIIh. The volume of the asymmetric part
21−29
molecular materials from powder X-ray diffraction data.
These methods were successfully applied in structural studies of
polymorphic modifications and hydrated forms of various
30
31
pharmaceuticals, such as acrinol, finasteride and sulfathia-
zole among many others.
32
Inspired by the efficiency of powder diffraction methods in
crystal structure determination, we have applied them to three
aforementioned forms of tilorone dihydrochloride, namely, III,
IIIh and IV (see Figure 1), in an attempt to find: (i) the
reasons for different physicochemical properties of III and IV,
and (ii) a model of reversible hydration/dehydration solid-state
transformation III ↔ IIIh at ambient conditions. Herewith we
present the results of our study.
3
of IIIh is 75 Å (Z = 4) larger than that of III, assuming that
three or four water molecules per asymmetric unit could be
absorbed. Visually, samples III and IIIh can be distinguished
only by a slight color difference (III is orange and IIIh dark
orange), while the morphology of both compounds is identical,
because their crystallites have no specific habit (Figure 1). The
particle morphology of IV is easily identified as needle.
The importance of hydration/dehydration processes in
pharmaceuticals has been well-known for a long time,
because pharmaceutical solids as raw materials or as dosage
forms often come in contact with water in the course of
processing and storage. This may occur during crystallization,
lyophilization, wet granulation, or spray drying, and because of
19
EXPERIMENTAL SECTION
■
Materials. All initial reagents and solvents were commercially
available and used as received. Tilorone base, i.e., 2,7-bis[2-
(diethylamino)ethoxy]-9-fluorenone, was prepared following the
3
known method.
B
dx.doi.org/10.1021/cg3012475 | Cryst. Growth Des. XXXX, XXX, XXX−XXX

Crystal Growth & Design
Article
Figure 2. Model used in a search for structure solution. Torsion angles
varied in simulated annealing runs are shown by arrows. The N···Cl
distances (dashed lines) were fixed to 3.1 Å.
IR-Spectra and Elemental Analysis. IR absorption spectra of
compounds III and IV were recorded in the wave range 3000−400
−1
−1
cm with a resolution of 2 cm using a Shimadzu FTIR-8400S
spectrometer with KBr pellets. C, H and N analysis was carried out
using a LECO CHNS-932 elemental analyzer.
Differential Scanning Calorimetry (DSC) and Thermogravi-
metric Analysis (TGA). Thermal studies were carried out using a
Netzsch STA 449 F3 differential scanning calorimeter with Netzsch
Proteus 5.2.1 software. The instrument was calibrated for temperature
with pure benzophenone (mp 48.0 °C) and indium (mp 156.6 °C),
and the energy calibration was performed with pure indium (purity
−
1
9
9.999%, heat of fusion 28.45 J g ). 10−15 mg of each sample was
poured into a capped Al O pan and heated to 280 °C with a ramp
2
3
−1
rate of 10 K min under dry argon flow (protective purge 20 mL
−1
−1
min , sample purge 30 mL min ).
NMR Measurements. NMR spectra were measured at 305 K
using a Bruker Advance III (600 MHz) spectrometer for D O solution.
2
1
13
H and C NMR spectra were recorded with TMS as the external
standard. The 15N scale is given relative to 100% liquid NH (δN =
3
0
.0). All experiments were based on the Bruker standard techniques.
1
15
H− N gHMBC spectrum was recorded with 125 ms delay for
evolution of long-range coupling.
Synthesis of III. Tilorone base (31.5 g, 0.077 mol) was mixed with
0 mL of acetone with stirring and then 13.7 mL of 35% HCl was
4
added to pH 3−4 (Congo). The reaction mixture was incubated for 2
h at room temperature. The precipitate was filtered, washed with
acetone, dried in air and then in vacuo to constant mass at 40−45 °C.
A 32.95 (0.068 mol) mass of III was recovered as orange crystalline
powder. Yield: 88%. Elemental analysis and IR data for III are
presented in Supporting Information.
III can also be obtained from IIIh under ambient conditions when
the relative humidity is decreased to 30% or less.
Synthesis of IIIh. Tilorone dihydrochloride (25.0 g, 0.0517 mol)
was dissolved in 50 mL of water at 65−70 °C. Solution was filtered,
and 300 mL of acetone was added to the filtrate. After being stirred for
15 min the mixture was cooled to −20 °C. The upper acetone layer
was removed by decantation after 2 h. The lower layer containing the
main part of tilorone dihydrochloride was exposed to air at room
temperature to complete evaporation of solvent. The solid residue was
triturated. A 22.0 g (0.041 mol) mass of IIIh was recovered as dark
orange crystalline powder. Yield: 79%.
The DTA curve (see Figure S1 in Supporting Information) shows a
.79% mass loss for IIIh on heating from ambient temperature to 205
9
°
C, attributed to hydration water release. Such mass loss corresponds
to 2.9 H O molecules per one formula unit of tilorone
2
dihydrochloride. On the basis of these data and the difference in
unit cell volume of IIIh and III, it was proposed that the IIIh form is a
trihydrate. This conclusion agrees with the elemental analysis data
presented in Supporting Information.
IIIh can also be obtained from III under ambient conditions when
the relative humidity is increased to 70% or more.
Figure 3. The Rietveld plots for III (top), IIIh (middle), and IV
(bottom) showing the experimental (black dots), calculated (blue)
and difference (red) curves. The vertical bars denote calculated
positions of the diffraction peaks.
Synthesis of IV. Tilorone base (31.0 g, 0.076 mol) was mixed with
00 mL of dry iso-propanol with stirring and then 13.4 mL of 35%
1
HCl were added to pH 3−4 (Congo). The reaction mixture was
heated to 70−75 °C, cooled to room temperature (a seed can be used)
and then to 7−10 °C. After incubation for 4 h the precipitate was
filtered, washed with cold dry isopropanol, dried in air and then in
vacuo to constant mass at 40−45 °C. A 28.25 g (0.058 mol) mass of IV
was recovered as orange crystalline powder. Yield: 76%. Elemental
analysis, IR and NMR data for IV are presented in Supporting
Information.
C
dx.doi.org/10.1021/cg3012475 | Cryst. Growth Des. XXXX, XXX, XXX−XXX

Crystal Growth & Design
Article
Figure 4. Asymmetric part of III (top), IIIh (middle) and IV (bottom) viewed in two approximately orthogonal projections. Displacement spheres
are drawn at the 50% probability level; H atoms are omitted for clarity.
Powder X-ray Diffraction. Powder patterns were measured using
three powder diffractometers, namely, Huber G670 Guinier camera
for III and IV was 18  3 translational, 3 orientational and 12
torsional, shown by arrows in Figure 2.
(
(
Cu Kα radiation, λ = 1.54059 Å, transmission mode), Stoe STADI-P
CoKα1 radiation, λ = 1.788965 Å, transmission mode) and
The use of initial model with three separate fragments  one cation
and two anions, and six more DOFs gave the same solutions for III
and IV. Nine more DOFs were used in a solution search for IIIh,
where each water oxygen required three translational parameters. The
1
EMPYREAN (Cu Kα radiation, Ni β-filter, Bragg−Brentano mode).
The latter diffractometer equipped with a vacuum camera has been
used for quick data collection in vacuo. The powder patterns for
structure solution and subsequent refinement were collected at Huber
G670 (for III and IIIh) and Stoe STADI-P (for IV); see Table 1 for
data collection details.
Structure Determination. For all compounds, the monoclinic cell
dimensions were determined using three indexing programs:
TREOR90, ITO, and AUTOX.
extinctions, the space group for III and IIIh was determined to be
P2 /c, and for IV − P2 /n. The unit-cell parameters and space groups
were further tested using a Pawley fit and confirmed by crystal
structure solution. The crystal data, data collection and refinement
parameters for three compounds are given in Table 1. The crystal
structures have been solved with the use of a simulated annealing
4
5
solution found was fitted with the program MRIA in the bond-
restrained Rietveld refinement using a split-type pseudo-Voigt peak
46
profile function. A comparison of powder patterns measured in
transmission and reflection modes for all compounds showed weak
texture effects in the patterns of III and IIIh with [001] as a direction
of preferred orientation, and moderate texture in IV, where the
crystallites are needle-shaped (Figure 1). Therefore, two formalisms of
33
34
35,36
On the basis of systematic
47
preferred orientation correction  March-Dollase (for III and IIIh)
1
1
48,49
37
and spherical harmonics expansion up to the sixth order
(for IV)

were used in the refinement. Restraints were applied to the
intramolecular bond lengths and contacts (<2.8 Å); the strength of the
restraints was a function of interatomic separation and, for
intramolecular bond lengths, corresponded to r.m.s. deviation 0.02
Å. Additional restraints were applied to the planarity of ring systems
with the attached atoms, with the maximal allowed deviation from the
mean plane of 0.03 Å. All non-H atoms were refined isotropically.
During the refinement of IIIh, the isotropic displacement parameter
38
technique. The Cambridge Structural Database (CSD, Version
5
ethoxy]-9-fluorenone fragment. Therefore, the initial molecular model
of tilorone dihydrochloride (Figure 1) has been obtained in the course
of density functional theory (DFT) calculations in vacuo using the
39
.33) contains no structures with the 2,7-bis[2-(diethylamino)-
−2
40−43
Uiso of Cl1 reached a high value of 0.3 Å and more, while Uiso of one
water oxygen (O31) was negative, indicating that sorts of these two
atoms have to be permutated. After such permutation, the Uiso
parameters of Cl1 and O31 in new positions were refined to normal
values with the decreased values of profile R-factors. In all structures,
H atoms were positioned geometrically (C−H 0.93−0.97 Å; N−H
0.91 Å; O−H 0.85 Å) and not refined. For all compounds, the
quantum-chemical code Priroda
gradient approximation (GGA) and PBE exchange correlation
employing the generalized-
44
functional. During the subsequent direct space search for structure
solution, the tetrahedral environment of N atoms in the obtained
molecular model was considered as a rigid unit with the N···Cl
distance fixed to 3.1 Å; H atoms were omitted. In the simulated
annealing runs, the total number of varied degrees of freedom (DOF)
D
dx.doi.org/10.1021/cg3012475 | Cryst. Growth Des. XXXX, XXX, XXX−XXX

Crystal Growth & Design
Article
diffraction profiles after the final bond-restrained Rietveld refinements
are shown in Figure 3 and molecular structures  in Figure 4.
RESULTS AND DISCUSSION
For all compounds, the crystal packing exhibits multiple
hydrogen bonds, both strong and weak,
summarized in Table 2.
■
5
3
which are
Table 2. Hydrogen-Bonding Geometry (Å, °) in III, IIIh, and
a
IV
D−H···A
D−H H···A
D···A
D−H···A
III
N3−H3···Cl1
N21−H21···Cl2
C22−H22A···Cl2
C23−H23A···O26
C27−H27A···Cl1
C29−H29B···O26
0.91
0.91
0.97
0.96
0.97
0.96
2.11
2.12
2.66
2.50
2.59
2.53
3.011(10)
3.021(9)
169
172
148
164
152
144
i
3.521(11)
3.438(13)
3.483(12)
3.358(13)
ii
iii
iv
IIIh
N21−H21···O31
N3−H3···Cl1
O31−H31A···Cl1
O32−H32A···Cl1
O32−H32B···O31
O33−H33A···Cl1
O33−H33B···O32
0.91
0.91
0.85
0.85
0.85
0.85
0.85
0.97
0.93
0.97
0.96
0.97
2.11
1.89
2.50
2.47
1.87
2.41
1.99
2.50
2.57
2.50
2.48
2.76
3.021(13)
2.759(16)
3.343(12)
3.263(11)
2.712(15)
3.150(11)
2.828(15)
3.420(19)
3.480(16)
3.334(19)
3.40(2)
175
159
172
156
171
146
167
159
167
144
160
141
Figure 5. The layers in III (top) and IIIh (bottom) viewed along the c
axis direction. Dashed lines denote strong hydrogen bonds. The C-
bound H atoms are omitted for clarity.
v
vi
vii
C28−H28B···O26
located in the channels along the b axis direction, whereas the
rest (Cl2) are held between the cationic stacks of the lower and
upper layers.
viii
C15−H15···Cl2
C4−H4B···O33
i
v
C30−H30C···O32
C22−H22B···Cl2
iii
Because of the nonequivalence of anionic positions illustrated
in Figure 8, Cl1 forms only one Cl···H−C contact shorter than
the sum of van der Waals radii (Cl1···H27A), whereas Cl2 has
five short contacts.
In structure IIIh, the Cl2 anion retains a position inside the
stack, whereas the Cl1 anion is displaced to the interstack space
by one of water molecules, although is still within the layer
3.568(15)
IV
N3−H3···Cl1
N21−H21···Cl2
C4−H4B···O26
0.91
0.91
0.97
0.97
0.93
2.13
2.12
2.40
2.79
2.77
3.043(6)
3.028(6)
3.305(10)
3.511(8)
3.613(8)
178
173
155
132
152
ix
x
C19−H19B···Cl1
xi
C9−H9···Cl2
(
Figure 6, bottom). This is not an unexpected result: statistical
a
Symmetry codes: (i) x, 3/2 − y, 1/2 + z; (ii) −x, 1/2 + y, 3/2 − z;
analysis of the structures of hydrates of hydrochloric salts of
(
(
iii) x, 3/2 − y, z − 1/2; (iv) −x, 1 − y, 1 − z; (v) 1 − x, 2 − y, 1 − z;
vi) 1 − x, 1/2 + y, 1/2 − z; (vii) −x, 1 − y, −z; (viii) −x, 2 − y, 1 −
aliphatic tert-amines containing two or more water molecules
21
per each chloride anion (CSD, Version 5.33) demonstrates
that hydrogen bonds N−H···Cl and N−H···O are formed with
approximately equal probability  58 examples of the first type
z; (ix) 2 − x, −y, 2 − z; (x) x − 1/2, 1/2 − y; z − 3/2; (xi) 3/2 − x,
1
/2 + y, −3/2 − z.
(
N−H···Cl) against 50 examples of the second type (N−
The crystal structures of III and IIIh are closely related to
H···O). In IIIh, one water molecule (O31) forms two hydrogen
bonds: one bond O···H−N with the diethylammonium group,
and the other of the O−H···Cl type  with the Cl1 anion
(Figure 6, bottom). Two remaining water molecules (O32 and
O33) are joined by hydrogen bonds with Cl1 (Table 2)
forming the nets parallel to (011) (Figure 9). These nets
separate the neighboring stacks of cations; thus the stacks
within each layer are pushed apart, as presented in Figure 7
(bottom). As a result, parameter a in IIIh becomes 1.94 Å
longer than in III, whereas changes in b and c parameters are
small.
each other. In both of them, cations and anions form layers
parallel to (101) (Figure 5). In III, there are only two short
contacts between these layers: O26···H23A and O26···H29B,
both involving methyl H atoms (Table 2). In IIIh, the layers
are joined by O−H···O hydrogen bonds between water
molecules.
The layers in III and IIIh are presented in Figure 6. In both
structures, the cations form stacks along the c axis direction.
Within the stack, only the utmost benzene rings of adjacent
cations overlap with centroid−centroid separations of 3.70(1)
Å both in III and IIIh.
+
In IV, the conformation of (CH ) NH (C H ) side groups
2
2
2
5 2
In III, chloride anions are located at the sides of stacks
between diethylammonium groups, and each Cl− forms a
hydrogen bond Cl···H−N with one of these groups (Table 2).
The layers in III are superimposed so that the stacks of cations
in adjacent layers are slightly displaced relative to each other
along the a axis direction (Figure 7, top). As a result, chloride
anions have different environments: half of them (Cl1) are
in the cation is distinct from that in III and IIIh (Figure 4).
Packing of cations and anions in IV (Figure 10) also differs
from those observed in III and IIIh. In IV, cations related by
the c translation form stacks. Two adjacent stacks give a pillar,
and chloride anions are located between the pillars. Each anion
forms hydrogen bond with one diethylammonium group
(Table 2). There are no layers in IV.
E
dx.doi.org/10.1021/cg3012475 | Cryst. Growth Des. XXXX, XXX, XXX−XXX

Crystal Growth & Design
Article
Figure 7. The crystal packing of III (top) and IIIh (bottom) viewed
along the b axis direction and showing the overlapping of layers and
nonequivalence of chloride anions. Dashed lines denote strong
hydrogen bonds. The C-bound H atoms are omitted for clarity.
Figure 6. The layers in III (top) and IIIh (bottom) viewed along the b
axis direction and showing cationic stacks. Dashed lines denote strong
hydrogen bonds. The C-bound H atoms are omitted for clarity.
Structural data offer an explanation of easy reversible
hydration of III: water molecules can penetrate into a crystal
via interlayer boundaries and then enter the layers without
disruption of the cation packing motif. In the same way, the
absorbed water molecules can easily leave the crystals of IIIh
keeping the cation packing motif without disturbance. Ease of
dehydration of IIIh is supported by fast IIIh → III
transformation in vacuo. Two powder patterns of IIIh were
quickly (3 min each) measured using a diffractometer equipped
with a vacuum pump. One pattern was measured at ambient
conditions (red curve in Figure 11), and another one in vacuo
(
blue curve in Figure 11).
One can see that full solid-state transformation IIIh (red) →
Figure 8. Schematic illustration of the nonequivalence of chloride
anions in III. Anions Cl2 are held between the cationic stacks
represented by rectangles.
III (blue) in vacuo occurred within 1 min (or even faster). As
follows from Figure 11, powder diffraction is a quick and
reliable method for estimation of the phase content of a sample
F
dx.doi.org/10.1021/cg3012475 | Cryst. Growth Des. XXXX, XXX, XXX−XXX

Crystal Growth & Design
Article
is why hydration of this polymorph should be accompanied by
complete rearrangement of crystal structure.
This study allowed us to clarify some aspects of reversible III
↔
IIIh transformation. However, several questions still remain
unsolved. For example, in powder patterns measured for III
and IIIh at various conditions we observed no additional peaks;
all peaks were attributed either to III or to IIIh. Why were the
hydrated crystalline forms of tilorone dihydrochloride with low
water content  mono- or dihydrate  not observed?
Another question concerns the minimum at 226.3 °C on the
DSC curve for IV (Figure S1, top)  whether it corresponds
to the phase transition or not. We hope to find answers to these
new questions in the future.
CONCLUSIONS
■
In summary, three new forms of tilorone dihydrochloride 
two anhydrous forms, III and IV, and a trihydrate form IIIh 
were obtained, and their crystal structures were established
from X-ray powder diffraction. Different crystal packings and
hydrogen-bonding patterns of III and IV elucidate why one
form is unstable to hydration at ambient conditions, while
another is stable. In III, there are separate layers of hydrogen-
bonded cations and anions, and atmospheric water can easily
penetrate into a crystal via interlayer boundaries thus leading to
solid-state formation of IIIh. The solid-state hydration/
dehydration transformation III ↔ IIIh is reversible at ambient
conditions and depends on the relative humidity only. In IV,
there are no layers and thus no ways suitable for penetration of
water molecules into a crystal, so hydration of this polymorph
requires a complete rearrangement of crystal structure; that is
why IV is much more stable than III during the long-term
storage.
Figure 9. Hydrogen-bonding net in IIIh formed by anions and
crystalline water molecules.
ASSOCIATED CONTENT
■
Figure 10. The crystal packing of IV viewed along the c axis direction.
Dashed lines denote N---H···Cl hydrogen bonds. The C-bound H
atoms are omitted for clarity.
*
S
Supporting Information
Crystallographic information (.cif files), elemental analysis and
IR data, experimental and calculated powder patterns for all
compounds and NMR spectrum for IV. This material is
available free of charge via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
■
Corresponding Author
E-mail: vladimir@struct.chem.msu.ru. Tel.: +7-8-(495)
393654. Fax: +7-8-(495)9390898.
*
9
Author Contributions
The manuscript was written through contributions of all
authors. All authors have given approval to the final version of
the manuscript.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
DSC and TGA studies were carried out in the X-ray diffraction
Centre of St. Petersburg State University.
Figure 11. Two patterns of IIIh measured on EMPYREAN at T = 25
°
−
C: red −3 min measurement at ambient conditions, P = 1 Kbar; blue
REFERENCES
■
(
(
3 min measurement started 1 min after the vacuum pump was
−6
(1) Krueger, R. F.; Mayer, G. D. Science 1970, 169, 1213−1214.
switched on, P = 10 mbar.
2) Mayer, G. D.; Krueger, R. F. Science 1970, 169, 1214−1215.
3) Andrews, E. R.; Fleming, R. W.; Grisar, J. M.; Kihm, J. C.;
of III during the storage. In IV, the ways suitable for
penetration of water molecules into a crystal are absent; that
Wenstrup, D. L.; Mayer, G. D. J. Med. Chem. 1974, 17, 882−886.
(4) Chandra, P.; Wright, G. J. Top. Curr. Chem. 1977, 72, 125−148.
G
dx.doi.org/10.1021/cg3012475 | Cryst. Growth Des. XXXX, XXX, XXX−XXX

Crystal Growth & Design
Article
(
1
(
3
5) Stringfellow, D. A.; Glasgow, L. A. Antimicrob. Agents Chemother.
972, 2, 73−78.
6) Fleming, R. W.; Wenstrup, D. L.; Andrews, E. R. U.S. Patent
592819, 1971.
(45) Zlokazov, V. B.; Chernyshev, V. V. J. Appl. Crystallogr. 1992, 25,
447−451.
(46) Toraya, H. J. Appl. Crystallogr. 1986, 19, 440−447.
(47) Dollase, W. A. J. Appl. Crystallogr. 1986, 19, 267−272.
(
(
7) Dunitz, J. D. Pure Appl. Chem. 1991, 63, 177−185.
̈
(48) Ahtee, M.; Nurmela, M.; Suortti, P.; Jarvinen, M. J. Appl.
Crystallogr. 1989, 22, 261−268.
8) Bernstein, J. Polymorphism in Molecular Crystals; Oxford
(
(
(
(
̈
49) Jarvinen, M. J. Appl. Crystallogr. 1993, 26, 525−531.
University Press: Oxford, 2002.
9) Rodriguez-Spong, B.; Price, C. P.; Jayasankar, A.; Matzger, A. J.;
Rodriguez-Hornedo, N. Adv. Drug Delivery Rev. 2004, 56, 241−274.
10) Hilfiker, R. Polymorphism in the Pharmaceutical Industry; Wiley-
VCH Verlag GmbH & Co. KGaA: Weinheim, 2006.
50) de Wolff, P. M. J. Appl. Crystallogr. 1968, 1, 108−113.
51) Smith, G. S.; Snyder, R. L. J. Appl. Crystallogr. 1979, 12, 60−65.
52) Young, R. A.; Wiles, D. B. J. Appl. Crystallogr. 1982, 15, 430−
(
(
4
38.
(
53) Desiraju, G. R.; Steiner, T. The Weak Hydrogen Bond in
Structural Chemistry and Biology; Oxford University Press: Oxford,
999.
(11) Desiraju, G. R. Cryst. Growth Des. 2008, 8, 3−5.
(12) Brittain, H. G. Polymorphism in Pharmaceutical Solids, 2nd ed.;
1
Informa Healthcare: New York, 2009; Vol. 192.
(
13) Stahly, G. Cryst. Growth Des. 2007, 7, 1007−1026.
(
14) Byrn, S. R.; Pfeiffer, R. R.; Stowell, J. G. Solid-State Chemistry of
Drugs, 2nd ed.; SSCI Inc.: West Lafayette, IN, 1999.
15) Morris, K. R.; Griesser, U. J.; Eckhardt, C. J.; Stowell, J. G. Adv.
Drug Delivery Rev. 2001, 48, 91−114.
16) Stephenson, G. A.; Forbes, R. A.; Reutzel-Edens, S. M. Adv.
Drug Delivery Rev. 2001, 48, 67−90.
17) Chieng, N.; Rades, T.; Aaltonen, J. J. Pharm. Biomed. Anal. 2011,
5, 618−644.
18) Ignatenko, O. A.; Kontsevoi, I. A.; Gabitov, A. F.; Pirogov, S. V.
Patent application RU 2011152830 (15.12.2011).
19) Byrn, S. R. Solid-State Chemistry of Drugs; Academic Press: New
York, 1982.
(
(
(
5
(
(
(
(
20) Brittain, H. G. J. Pharm. Sci. 2012, 101, 464−484.
21) Harris, K. D. M.; Tremayne, M.; Lightfoot, P.; Bruce, P. G. J.
Am. Chem. Soc. 1994, 116, 3543−3547.
(
22) Harris, K. D. M.; Tremayne, M.; Kariuki, B. M. Angew. Chem.,
Int. Ed. 2001, 40, 1626−1651.
(
(
23) Chernyshev, V. V. Russ. Chem. Bull. 2001, 50, 2273−2291.
24) David, W. I. F.; Shankland, K.; McCusker, L. B.; Baerlocher, C.,
Eds. Structure Determination from Powder Diffraction Data, OUP/IUCr,
002.
25) Cheung, E. Y.; Kitchin, S. J.; Harris, K. D. M.; Imai, Y.; Tajima,
2
(
N.; Kuroda, R. J. Am. Chem. Soc. 2003, 125, 14658−14659.
(26) Tremayne, M. Phil. Trans. R. Soc. 2004, 362, 2691−2707.
(27) Favre-Nicolin, V.; Cerny, R. Z. Kristallogr. 2004, 219, 847−856.
(28) Tsue, H.; Horiguchi, M.; Tamura, R.; Fujii, K.; Uekusa, H. J.
Synth. Org. Chem. Jpn. 2007, 65, 1203−1212.
29) David, W. I. F.; Shankland, K. Acta Crystallogr., Sect. A 2008, 64,
2−64.
30) Fujii, K.; Uekusa, H.; Itoda, N.; Hasegawa, G.; Yonemochi, E.;
Terada, K.; Pan, Z.; Harris, K. D. M. J. Phys. Chem. C 2010, 114, 580−
86.
31) Byard, S.; Abraham, A.; Boulton, P. J. T.; Harris, R. K.;
Hodgkinson, P. J. Pharm. Sci. 2012, 101, 176−186.
32) Munroe, A.; Rasmuson, A. C.; Hodnett, B. K.; Croker, M. D.
Cryst. Growth Des. 2012, 12, 2825−2835.
33) Werner, P.-E.; Eriksson, L.; Westdahl, M. J. Appl. Crystallogr.
(
5
(
5
(
(
(
1
(
985, 18, 367−370.
34) Visser, J. W. J. Appl. Crystallogr. 1969, 2, 89−95.
(35) Zlokazov, V. B. J. Appl. Crystallogr. 1992, 25, 69−72.
(36) Zlokazov, V. B. Comput. Phys. Commun. 1995, 85, 415−422.
(37) Pawley, G. S. J. Appl. Crystallogr. 1981, 14, 357−361.
(38) Zhukov, S. G.; Chernyshev, V. V.; Babaev, E. V.; Sonneveld, E.
J.; Schenk, H. Z. Kristallogr. 2001, 216, 5−9.
(39) Allen, F. H. Acta Crystallogr., Sect. B 2002, B58, 380−388.
(40) Laikov, D. N. Chem. Phys. Lett. 1997, 281, 151−154.
(41) Laikov, D. N. Priroda Code, version 5; Moscow State University:
Moscow, 2004.
42) Laikov, D. N.; Ustynyuk, Y. A. Izv. Akad. Nauk, Ser. Khim. 2005,
(
8
(
(
04−810; (In Russian) Russ. Chem. Bull. 2005, 54, 820−826.
43) Laikov, D. N. Chem. Phys. Lett. 2005, 416, 116−120.
44) Perdew, J. P.; S. Burke, S.; Ernzerhof, M. Phys. Rev. Lett. 1996,
77, 3865−3868.
H
dx.doi.org/10.1021/cg3012475 | Cryst. Growth Des. XXXX, XXX, XXX−XXX