New Solid Forms of the Antiviral Drug Arbidol: Crystal Structures

Article pubs.acs.org/molecularpharmaceutics

New Solid Forms of the Antiviral Drug Arbidol: Crystal Structures, Thermodynamic Stability, and Solubility Artem O. Surov,† Alex N. Manin,† Andrei V. Churakov,‡ and German L. Perlovich*,† †

G.A. Krestov Institute of Solution Chemistry, Russian Academy of Sciences, 153045, Ivanovo, Russia Institute of General and Inorganic Chemistry, Russian Academy of Sciences, Leninskii Prospekt 31, 119991 Moscow, Russia

‡

S Supporting Information *

ABSTRACT: Salts of the antiviral drug Arbidol (umifenovir) with pharmaceutically relevant benzoate and salicylate anions were obtained, and their crystal structures were described. For Arbidol salicylate, an unstable solvate with acetonitrile was also found and characterized. Analysis of the conformational preferences of the Arbidol molecule in the crystal structures showed that it adopts two types of conformations, namely “open” and “closed”, both of which correspond to local conformational energy minima of the isolated molecule. Thermal stability of the Arbidol salicylate solvates with chloroform and acetonitrile was analyzed by means of differential scanning calorimetry and thermogravimetric analysis. The standard thermodynamic functions of the salt formation were determined. The Gibbs energy change of the process was found to be negative, indicating that the formation of the salts from individual components is a spontaneous process. The dissolution study of the Arbidol salts performed in aqueous buffer solutions with pH 1.2 and 6.8 showed that both salts dissolve incongruently to form an Arbidol hydrochloride monohydrate at pH 1.2 and an Arbidol base at pH 6.8, respectively. KEYWORDS: Arbidol, umifenovir, pharmaceutical salts, solvates, X-ray diffraction, conformation analysis, thermodynamics, dissolution

1. INTRODUCTION Arbidol (umifenovir) (1-methyl-2-phenylthiomethyl-3-carbethoxy-4-dimethylaminomethyl-5-hydroxy-6-bromoindole) is a Russian-made indole-derivative molecule with antiviral activity1,2and immunomodulatory effect3,4 (Figure 1). The antiviral efficacy of Arbidol is comparable to that of some other wellknown antiviral drugs such as rimantadin (Roflual), oseltamivir (Tamiflu), and ribavirin.5−7 Arbidol is widely used in Russia

and China for treating acute respiratory infections, such as influenza A and B, and others.8−10 It has been recently found that Arbidol could be used to treat hepatitis B and C and chikungunya virus.11,12 Besides this, Arbidol can also be administered for the complex therapy of patients with chronic bronchitis, pneumonia, or recidivist HSV infection as well as for postoperative infection prevention.13 It has also been confirmed that Arbidol has antioxidant activity.14 Furthermore, one of the advantages of Arbidol compared to other antiviral drugs is that Arbidol belongs to the category of the least toxic drugs (LD50 > 4 g/kg) which do not have adverse effects on the human body when taken orally at the recommended dosage.12,15 However, it should be noted that, to date, no studies have addressed the toxicity of Arbidol in long-term use, such as treating chronic diseases.4 Arbidol (Arb) as a free base is practically insoluble in water and, as a consequence, has poor bioavailability. Therefore, Arbidol is used as the hydrochloride monohydrate [Arb+HCl +H2O] under the brand name Arbidol. But even [Arb+HCl +H2O] has low bioavailability, as approximately 40% of Arb is excreted with feces.16 Efforts have been made to improve Arbidol water solubility through chemical grafting of polymers Received: Revised: Accepted: Published:

Figure 1. Molecular structures of Arbidol, benzoic, and salicylic acids used in this work. The flexible torsion angles in the Arbidol molecule are numbered and indicated by τ1, τ2, and τ3. © 2015 American Chemical Society

4154

August 15, 2015 October 10, 2015 October 13, 2015 October 13, 2015 DOI: 10.1021/acs.molpharmaceut.5b00629 Mol. Pharmaceutics 2015, 12, 4154−4165

Article

Molecular Pharmaceutics to the Arbidol molecule17,18 and by mixing it with an immunomodulator drug (IMOD).12 Despite the good solubility levels (in particular, those of Arb in a water-soluble polymer complex with arabinogalactan18), these polymer complexes have not been widely used. Besides polymer preparation of complexes, a number of approaches are used to improve the solubility of active pharmaceutical ingredients (API), such as preparation of solid dispersions,19 micronization,20 polymorphs,21 and nanocrystals.22 One of the best approaches to overcoming the solubility challenge without modification of the pharmacophore structure of an API is developing new crystalline forms such as salts or cocrystals. In fact, salt formation is the most common method of improving solubility, and today more than 50% of APIs are marketed as salts.23 The ΔpKa rule or the “rule of three” was formulated to predict the possibility of salt/cocrystal formation. It was generally accepted that the reaction of acid with base will form salt when ΔpKa > 3.23−25 However, later the ΔpKa interval was extended. It has been found that components with ΔpKa < −1 mostly form cocrystals, while systems with ΔpKa> 4 tend to form salts.26 The pKa value of Arb has been reported to be 6.0.27 So, for Arbidol and carboxylic acids, the pKa rule strongly suggests proton transfer and salt formation. Indeed, Arbidol was confirmed to form salts with glutaric, gentisic, and salicylic acid and chloroform solvate of Arbidol salicylate.28 In the present work, we focus on crystal structures, thermal analysis, and aqueous solubility of the Arbidol salts with benzoic (BA) and salicylic (SA) acids and two solvates of Arbidol salicylate with acetonitrile (ACN) and chloroform (CHCl3). It is important to note that benzoic and salicylic acids are well-known for their analgesic and anti-inflammatory properties.29,30 Inflammation usually accompanies viral illnesses, such as coronavirus infection and flu diseases, for the treatment of which Arb is used. Thus, this research is an attempt to design a drug−drug cocrystal with double functional action. This concept of improving bioavailability has been gaining popularity in recent years.31 Additionally, the Arbidol conformational analysis is performed with the aim to rationalize the diversity of molecular conformations of the drug in various crystalline forms. There are a number of works analyzing cocrystal formation enthalpy. One of the approaches relies on crystal structure prediction (CSP) with anisotropic potential32,33 and quick methods of energy estimation based on molecular electrostatic potential surfaces.34 Abramov et al.35 applied COSMO-RS fluid-phase thermodynamics computations describing miscibility of cocrystal formers in a supercooled liquid (melt) phase to virtual coformer screening. Moreover, Hansen solubility parameters were recently used to describe miscibility of API and coformers to predict cocrystal formation in order to guide cocrystal screening.36 An essential fault of all these approaches is that they analyze the enthalpic characteristics of cocrystal formation, while the entropic terms are not taken into account. The existence of such approaches can be attributed to the fact that there is almost no experimental data about the Gibbs energies of cocrystal formation. As a rule, researchers focus only on cocrystal dissolution kinetics or on studying thermodynamic solubility at one temperature. Other studies center on studying thermodynamics of cocrystal component complex formation at one temperature.37,38 However, such investigations cannot analyze the entropic terms and, as a rule, are unable to give a full overview of the thermodynamic process. A number of

studies have been made to analyze only enthalpic terms of cocrystal formation by using solubility calorimetry39,40 or DSC.41 A fundamental fault of this approach is the lack of information about the Gibbs energy and, as in the previous case, its inability to describe cocrystal formation thermodynamics. If more attention were paid to studying the problem of dissolution of cocrystals and individual compounds (as part of cocrystals) at different temperatures, we could make better progress in understanding thermodynamic aspects of cocrystal formation. It should be mentioned that only several works deal with this problem. For example, Oliver et al.42 studied the thermodynamics of [carbamazepine+saccharin] cocrystal formation by using two independent techniques: solubility calorimetry (to analyze the enthalpic term) and solubility measured by the saturation method (to analyze the Gibbs energy). Lin et al.,43 Zhang et al.,44,45 Zhao et al.,46 Yu et al.,47 and Evora et al.48 applied only one technique (measuring solubility by the saturation method−temperature dependences) to obtain all thermodynamic characteristics. However, it is worth mentioning that, in this scarce number of works, some of the thermodynamic data are of doubtful reliability. For instance, in the study made by Lin et al.43 the Gibbs energy and enthalpy of [dipfluzine+benzoic acid] cocrystal formation in aqueous ethanol solutions are equal to −55.3 and −59.9 kJ·mol−1, respectively. All this proves again that progress in this area can only be made through researchers’ significant efforts taken to obtain representative samples of thermodynamic data. A significant distinction of the present study from the already published works is that we tried to obtain experimentally all thermodynamic characteristics (Gibbs energy, enthalpy, and entropy) of cocrystal/salt formation from the temperature dependencies. Investigations with similar data are practically absent in the literature. In addition, in this work we propose an approach for estimation of desolvation enthalpies of crystallosolvates in comparison with vaporization enthalpies of the considered solvents. The presented approach is quite novel in the field as well.

2. MATERIALS AND METHODS 2.1. Compounds and Solvents. Arbidol hydrochloride monohydrate (C22H28BrClN2O4S, 98%) was purchased from Sichuan Baili Pharmaceutical Co., Ltd. (China). Arbidol base was prepared according to a procedure described by Orola et al.28 Benzoic acid (C7H6O2, 99.5%) and salicylic acid (C7H6O3, 99.5%) were purchased from Sigma-Aldrich. Acetonitrile (ACN) and chloroform (CHCl3) were obtained from RCI Labscan. All the solvents were of analytical grade and used as received without further purification. 2.2. Crystallization Procedure. 2.2.1. [Arb+BA] (1:1). Arbidol base (50 mg, 0.10 mM) was dissolved with benzoic acid in a 1:1 molar ratio in 10 mL of ethyl acetate and stirred at room temperature until a clear solution was obtained. The solution was filtered and allowed to evaporate slowly in a fume hood at room temperature. Diffraction quality crystals were grown over a week. A bulk sample of the [Arb+BA] salt was obtained by slurring equimolar amounts of Arbidol base and benzoic acid in ethyl acetate for 3 h at room temperature. 2.2.2. [Arb+SA] (1:1). Arbidol base (50 mg, 0.10 mM) was dissolved with salicylic acid in the 1:1 molar ratio in 10 mL of acetonitrile and stirred at room temperature until a clear solution was obtained. The solution was filtered and allowed to evaporate slowly in a fume hood at room temperature. 4155

DOI: 10.1021/acs.molpharmaceut.5b00629 Mol. Pharmaceutics 2015, 12, 4154−4165

Article

Molecular Pharmaceutics

4.6 mm i.d., 5 μm particle size, and 100 Å pore size). Elution was achieved by mobile phase consisting of water−0.1% trifluoroacetic acid (A) and acetonitrile (B). The flow rate of the mobile phase was 1 mL·min−1. To analyze the samples from the hydrochloric buffer, gradient elution was performed by changing the mobile phase from 20% to 50% B for the first 5 min, then it was maintained for the next 6 min, and then the percentage of the mobile phase component B was decreased from 50% to 20% for the next 3 min. The Arbidol retention time was found to be 9 min. To analyze the samples from the phosphate buffer, gradient elution was performed by changing the mobile phase from 20% to 40% B for the first 5 min, then it was maintained for the next 5 min, and then the percentage of the mobile phase component B was decreased from 50% to 20% for the next 4 min. The retention time of Arbidol was found to be ∼11.5 min. The detection of Arbidol was carried out at a wavelength of 314 nm for both buffers: pH 1.2 and pH 6.8. The concentrations were calculated according to an established calibration curve. 2.7. Thermodynamics of the Salt Formation. In case of a 1:1 stoichiometry, the reaction of two-component compound formation from a pure API (A) and a pure coformer (B) may be described as

Diffraction quality crystals were grown over a week. A bulk sample of the [Arb+SA] salt was obtained by slurring equimolar amounts of Arbidol base and salicylic acid in ethanol for 3 h at room temperature. 2.2.3. [Arb+SA+ACN] (1:1:2). Arbidol base (60 mg, 0.13 mM) was dissolved with salicylic acid in a 1:1 molar ratio in 10 mL of acetonitrile and stirred at 60 °C until a clear solution was obtained. The hot solution was placed in a freezer and stored at 5 °C. Diffraction quality crystals of the solvate were grown over a period of 1−2 days. A solvate of [Arb+SA] with chloroform ([Arb+SA+CHCl3]) (1:1:1) was prepared according to procedure described by Orola et al.28 2.3. X-ray Diffraction Experiments. Single-crystal X-ray diffraction data were collected on a Bruker SMART APEX II diffractometer using graphite-monochromated Mo Kα radiation (λ = 0.71073 Å). The structures were solved by direct methods and refined by full matrix least-squares on F2 with anisotropic thermal parameters for all non-hydrogen atoms.49 Absorption corrections based on measurements of equivalent reflections were applied.50 In the [Arb+SA] structure, all hydrogen atoms were found from the difference Fourier map and refined isotropically. As for [Arb+SA+ACN] and [Arb+BA], only amino atoms H2 were located from the difference map and refined with isotropic thermal parameters. All other hydrogen atoms (carbon) were placed in calculated positions and refined using a riding model. In [Arb+SA+ACN], the solvent acetonitrile molecule is disordered over two sites with occupancy ratio 0.71/0.29. X-ray powder diffraction (XRPD) data were recorded under ambient conditions in Bragg−Brentano geometry with a Bruker D8 Advance diffractometer with Cu Kα1 radiation (λ = 1.5406 Å). 2.4. DSC Experiments. Thermal analysis was carried out using a PerkinElmer DSC 4000 differential scanning calorimeter with a refrigerated cooling system (USA). The sample was heated in sealed aluminum sample holders at the rate of 10 °C·min−1 in a nitrogen atmosphere. The unit was calibrated with indium and zinc standards. The accuracy of the weighing procedure was ±0.01 mg. 2.5. Thermogravimetric Analysis (TGA). TGA was performed on a TG 209 F1 Iris thermomicrobalance (Netzsch, Germany). Approximately 10 mg of the sample was added to a platinum crucible. The samples were heated at a constant heating rate of 10 °C·min−1. The samples were purged with a stream of flowing dry Ar at 30 mL·min−1 throughout the experiment. 2.6. Aqueous Dissolution Experiments. Dissolution measurements were carried out by the shake-flask method in the hydrochloric buffer with pH 1.2 and the phosphate buffer with pH 6.8 at 25 ± 0.1 °C. The medium at pH 1.2 was prepared with 0.1 N aqueous hydrochloric acid solution and potassium chloride. For the phosphate buffer, 0.05 M solution of Na2HPO4 was adjusted to pH 6.8 with sodium hydroxide. An excess amount of each sample was suspended in the respective buffer solution in Pyrex glass tubes. The amount of drug dissolved was measured by taking aliquots of the respective media. The solid phase was removed by isothermal filtration (Rotilabo syringe filter, PTFE, 0.2 μm).The concentration was determined by HPLC. The results are stated as the average of at least three replicated experiments. HPLC was performed on Shimadzu Prominence model LC-20AD equipped with a PDA detector and a C-18 column (150 mm ×

A solid + Bsolid → ABsolid

(1)

It has been established in the literature that the standard freeenergy change, ΔG°f, for the reaction given above may be expressed through the solubility data of each of the materials.51 In the case of salt dissolution, however, the dissociation constants of A (pKa,A) and B (pKa,B) have to be taken into account. If A is a weak acid and B is a weak base, then the Gibbs energy of 1:1 salt formation may be expressed by the following equation:52 ⎛ 10 pKa,B− pKa,A Sp ·S p ⎞ A B ⎟⎟ ΔG°f = −RT ·ln⎜⎜ SAsalt ·SBsalt ⎝ ⎠ ⎛ 10 pKa,B− pKa,A Sp ·S p ⎞ A B⎟ = −RT ·ln⎜⎜ ⎟ K ⎝ ⎠ sp

(2)

where SpA and SpB are the solubility values of pure A and B in a solvent, while Ssalt and Ssalt are the solubility of the salt A B components in a solution, when in equilibrium with the pure salt. For the sake of simplicity, the activities of the components are approximated by molar concentrations. The product of Ssalt A and Ssalt B is generally known as the Ksp of a salt. Comparison of eq 2 with the expression of the standard freeenergy change for the isothermal reaction ΔG°f = −RT ·ln K f

(3)

shows that the ratio appearing in the logarithmic term of eq 2 has the character of an equilibrium constant at the given temperature. Therefore, this quantity can be defined as Kf. If the Kf values are known at different temperatures, the van’t Hoff relation may be used to derive the formation enthalpy, ΔH°f, of a multicomponent compound: d ln K f ΔH °f =− d(1/T ) R

(4)

Finally, the formation entropy change can be estimated from the general equation relating different thermodynamic functions: 4156


Article


Table 1. Crystallographic data for Arbidol salts with benzoic and salicylic acids and Arbidol salicylate solvate with acetonitrile chem formula cryst syst a/Å b/Å c/Å α/deg β/deg γ/deg unit cell vol/Å3 temp/K space group no. of formula units per unit cell, Z no. of reflns measd no. of indep reflns Rint final R1 values (I > 2σ(I)) final wR(F2) values (I > 2σ(I)) final R1 values (all data) final wR(F2) values (all data) goodness of fit on F2 largest diff peak and hole, e·Å−3 CCDC no.

[Arb+BA] (1:1)

[Arb+SA] (1:1)

[Arb+SA+ACN] (1:1:2)

C22H26BrN2O3S·C7H5O2 monoclinic 9.9760(10) 19.3572(19) 14.2997(14) 90.00 95.874(2) 90.00 2746.9(5) 180(2) P21/c 4 28231 6633 0.0405 0.0335 0.0743 0.0501 0.0798 1.035 0.402/-0.429 1418368

C22H26BrN2O3S·C7H5O3 triclinic 11.3617(15) 11.5300(15) 12.7182(17) 65.296(2) 66.413(2) 83.009(2) 1385.1(3) 183(2) Pl ̈

C22H26BrN2O3S·C7H5O3·2(C2H3N) triclinic 10.7143(12) 13.0489(14) 13.8206(15) 63.734(2) 79.129(2) 77.862(2) 1683.9(3) 180(2) Pl ̈

2 14395 6673 0.0156 0.0344 0.0885 0.0411 0.0918 1.043 0.792/-0.232 1418366

2 17593 8115 0.0241 0.0375 0.0922 0.0542 0.0988 1.035 0.594/-0.487 1418367

Figure 2. (a) Hydrogen bonded molecular unit in the [Arb+BA] crystal. (b) Molecular packing projections for [Arb+BA]. The Arbidol molecules are colored in green. The benzoic acid molecules are colored in blue. The H atoms are omitted.

ΔG°f = ΔH °f − T ·ΔS°f

performed using the GAUSSIAN03 program at the PBE/6311++G(d,p) level of theory53 and the scan step of 5°.

(5)

The solubility of the salts and their constituents was measured at 20.0, 25.0, 30.0, and 35.0 ± 0.1 °C in ethyl acetate for [Arb+BA] and in ethanol for [Arb+SA]. An excess of the solid was placed in an Eppendorf tube, and 2 mL of solvent was added. After 24 h, the suspension was filtered through a Rotilabo syringe filter (PTFE, 0.2 μm), and the concentration in the supernatant was determined by UV−vis spectroscopy (Varian Cary 50) for [Arb+BA] and by HPLC for [Arb+SA], as described above. The solubility of pure benzoic and salicylic acids in organic solvents was measured by the gravimetric method. The results are stated as the average of at least three replicated experiments. 2.8. Computational Procedure. Geometric optimization and calculations of the torsion profile for the Arbidol ion were

3. RESULTS AND DISCUSSION 3.1. Crystal Structures. Crystallographic data are summarized in Table 1; the asymmetric units and the packing arrangements of the salts and their solvates are shown in Figures 2, 3, 4, and 5 (the asymmetric units of the Arb salts with displacement ellipsoids are presented in Figure S1). The [Arb+BA] salt crystallizes in the monoclinic P21/c space group with one Arbidol cation and one benzoate anion in the asymmetric unit. The molecules are connected by two different hydrogen bonds, i.e., the charge assisted N2+−H2···O5− and conventional O1−H1···O6 H-bonds, to form a dimeric unit with R22(10) graph set notation (Figure 2a).54,55 In the crystals of the known salts of Arbidol, the flat indole fragments of Arbidol molecules tend to be packed in parallel layers, which 4157


Article


Figure 3. (a) Hydrogen bonded molecular unit in the [Arb+SA] crystal. (b) Molecular packing projections for [Arb+SA]. The Arbidol molecules are colored in green. The salicylic acid molecules are colored in blue. The H atoms are omitted.

Figure 4. (a) Hydrogen bonded molecular unit in the [Arb+SA+ACN] (1:1:2) crystal. (b) Molecular packing projections for [Arb+SA+ACN]. The Arbidol molecules are colored in green. The salicylic acid molecules are colored in blue. The solvent molecules and H atoms are omitted.

are held via π−π interactions. The crystal structure of [Arb +BA] is not an exception. It consists of alternating layers containing the π-stacks of the indole moieties (3.529 Å) and phenyl rings of Arbidol along the a-axis (Figure 2b).The benzoate ions occupy the space inside the layer formed by the phenyl fragments and stabilize the structure mainly on account of weak van der Waals interactions. Similar to [Arb+BA], the asymmetric unit of the [Arb+SA] salt also contains a hydrogen bonded dimer of Arbidol and salicylate ions (R22(10)) linked by N2+−H2···O5− and O1− H1···O6 H-bonds (Figure 3a). The packing arrangement of [Arb+SA], however, is considerably different from that of Arbidol benzoate. As Figure 3b shows, the crystal consists of discrete units formed by a pair of centrosymmetric Arbidol molecules. The neighboring units interact with each other via van der Waals and π−π interactions (3.414 Å).The salicylate ions are located between the adjacent Arb units to form a layer in the (010) plane. The solvate formation of the [Arb+SA] salt with acetonitrile leads to a significant change in the molecular conformation of Arbidol compared to that of the nonsolvated structure (Figure

4a). It is evident that the main conformational difference between the Arb molecules is related to the orientation of the phenyl fragment in relation to the indole moiety. Arb conformation in the [Arb+SA+ACN] solvate is found to be similar to that of the [Arb+BA] salt. Moreover, the solvate crystal structure has more common features with [Arb+BA] than with its parent form, i.e., the [Arb+SA] salt. Figure 4b shows that, as for Arbidol benzoate, the packing arrangement of [Arb+SA+ACN] can be described as alternating layers which contain either π-stacks of the indole moieties (3.432 Å) or phenyl rings of the Arbidol molecules, with the salicylate ions mainly occupying the space between the phenyl fragments. The solvent molecules are, in turn, deposited in the interlayer voids along the (001) plane (not shown in Figure 4b). In the crystal of the previously reported [Arb+SA] solvate with chloroform ([Arb+SA+CHCl3]),28 no conventional hydrogen bonded dimer is formed between the Arb and SA molecules. Instead, each salicylate ion accepts hydrogen bonds from two neighboring Arb molecules to form a closed-ring tetrameric unit across a crystallographic inversion center that can be described in graph set notation as R24(16) (Figure 5a). 4158


Article


Figure 5. (a) Hydrogen bonded tetrameric molecular unit in the [Arb+SA+CHCl3] (1:1:1) crystal. (b) Molecular packing projections for [Arb+SA +CHCl3]. The Arbidol molecules are colored in green. The salicylic acid molecules are colored in blue. The solvent molecules and H atoms are omitted.28

forms are collected in Table 2. The overlay of conformations of the Arbidol ions is shown in Figure 6.

Despite the difference in hydrogen bond organization, the overall packing arrangement of the solvate is similar to that of the nonsolvated structure. As Figure 5b shows, the Arb molecules are packed in distinct pairs, so that the molecule’s phenyl rings point toward each other, while the indole fragments form π-stacks (3.470 Å). In contrast to [Arb+SA], however, salicylate ions are generally located inside the cavity formed by a pair of Arb molecules. The size of those cavities is large enough to accept the chloroform molecules as well (not shown in Figure 5b). It is evident that the crystal structures of the salts and their solvates are considerably different, despite the common packing features such as π-stacks of the indole fragments and hydrogen bond network. The main reason for the packing diversity of the salts is most probably related to the high conformational flexibility of the Arbidol molecule. This, in turn, reflects the relatively low energy barrier separating one stable conformer from another, which allows the Arb molecule to adopt a suitable orientation for different supramolecular surroundings. The phenomenon is typical of the so-called conformational polymorphism.56 In case of a multicomponent crystal, however, an excess of conformational energy penalty (conformational strain) caused by the deviation of an API molecule from its optimal geometry may be compensated by different intermolecular interactions with the molecules of the second component and/or solvent. Thus, a crystal, which itself corresponds to a minimum of the packing energy, may contain relatively higher energy molecular conformations. Therefore, as the next step, we performed conformational analysis of Arbidol in order to identify the most energy preferred molecule conformers and to estimate the conformational energies associated with the molecular conformations in various crystal forms of the API. 3.2. Conformational Analysis. The Arb molecule conformation can be defined in terms of at least three torsion angles, namely, τ1 (∠C1−C2−S1−C3), τ2 (∠C2−S1−C3− C4), and τ3 (∠C5−C6−C7−N2) (see Figure 1). The values of the selected torsion angles for the Arbidol ions in all known salt

Table 2. Selected Torsion Angles τ1, τ2, τ3 for Arbidol Ion in Crystals of the Known Salts and Calculated Relative Conformational Energies for the τ1 (Eτ1) Torsion Angle (See Text) τ1,a deg (∠C1−C2− S1−C3) [Arb+BA] (1:1) [Arb+SA] (1:1) [Arb+SA +ACN] (1:1:2) [Arb+SA +CHCl3] (1:1:1) [Arb+Maleic acid] (1:1) [Arb+Gentisic acid] (1:1) [Arb+Glutaric acid] (1:1) [Arb+HCl +H2O] (1:1)

τ2, deg (∠C2−S1− C3−C4)

τ3, deg (∠C5− C6−C7−N2)

Eτ1, kJ·mol−1

177.2

−157.2

104.6

2.9

−81.8

−117.5

108.3

1.7

167.7

−146.5

110.0

3.3

−58.9

−105.7

97.9

1.9

−58.9

−98.5

110.3

2.0

−45.6

−62.8

101.3

3.4

−170.2

−34.6

81.4

−108.9

141.4

78.6

The numbers written in bold correspond to “closed” conformation. The numbers written in italic correspond to “open” conformation.

a

Table 2 and Figure 6 indicate that the torsion angle τ1 (∠C1−C2−S1−C3), which determines the orientation of the phenyl ring in relation to the indole moiety, is the most widely varying molecule fragment. Moreover, it is the τ1 angle that determines the overall geometry of the Arb molecule to a large extent. If τ1 < 90°, the molecule has a “closed” conformation. This type of conformation is observed in the crystals of [Arb +SA], [Arb+SA+CHCl3], [Arb+Maleic acid], and [Arb +Gentisic acid]. On the other hand, the molecular conformations with τ1 > 90° may be defined as “open”. The 4159


Article


< 90°) to form the N2+−H2···O3 intramolecular contacts16 which, in turn, decrease the molecule total energy. Thus, a comparison of the conformational energy of these molecules with that of the first set of molecules (where τ3> 90°) would not be correct due to essentially different factors that affect the energy value. On this account, the molecules with τ3 < 90° will not be considered in further discussion. It is evident that the torsion angle τ1 is found to be the most informative parameter to analyze in terms of the conformational energy, because it is mainly responsible for the overall shape of the Arb molecule in the crystal (“closed” or “open”). Figure 7 shows the τ1 torsion angle relaxed energy profile calculated for the protonated Arb molecule at the PBE/6-311+ +G(d,p) level of theory and the scan step of 5°. The conformational profile has two distinct asymmetric energy minima at ≈80° (min 1) and ≈−70° (min 2), that are separated by the energy barrier with the height of ca. 13 kJ· mol−1. In addition, two relatively higher energy states are observed at −180 and 180°. Figure 7 indicates that all the molecules with “closed” conformation in the crystal correspond to the min 2 energy minimum centered at −70°. Relatively large deviations of the τ1 angle from the equilibrium position are seen in the [Arb +Gentisic acid] structure. It is accompanied by the conformational energy increase by ca. 3 kJ·mol−1, which is in general comparable to the energy level for the “open” molecular conformations (τ1 ≈ 180°). It should be stressed that none of the crystals described contains an Arb conformation with the τ1 angle located in the min 1 region, despite the fact that it corresponds to the lowest energy state. It seems likely that the min 1 conformation is less preferred compared to that of min 2 in terms of crystal packing forces. The packing energy gained in the case of min 2 conformation must be significantly greater than the marginal energy difference (less than 2 kJ·mol−1) between the two conformers.

Figure 6. Overlay of Arbidol ionic conformations in the known salt forms of the compound: [Arb+BA], red; [Arb+SA], blue; [Arb+SA +ACN], green; [Arb+SA+CHCl3], cyan; [Arb+Maleic acid], purple; [Arb+Gentisic acid], violet; [Arb+Glutaric acid], orange; [Arb+HCl +H2O], magenta. H atoms are omitted.

structures that contain “open” conformation of Arb are [Arb +BA], [Arb+SA+ACN], [Arb+Glutaric acid], and [Arb+HCl +H2O]. A considerable variation is also found for the torsion angle τ2 (∠C2−S1−C3−C4), which is responsible for the rotation of the phenyl ring around the S1−C3 bond in general. The τ2 values change from −35 to −160° (an exception is observed for [Arb+HCl+H2O]). The rotation of the phenyl fragment is not expected to be associated with a large amount of the conformational energy, while its orientation in the crystal seems to be determined by the energy balance between the intra- and intermolecular interactions. The third flexible torsion angle τ3 (∠C5−C6−C7−N2) in the Arb molecule refers to the conformation of the dimethylamino group. In most cases, the protonated quaternary nitrogen atom points toward the −O1H1 hydroxy group, so that τ3 > 90° ([Arb+BA], [Arb+SA], [Arb+SA+CHCl3], [Arb +SA+ACN], [Arb+Maleic acid], [Arb+Gentisic acid]). However, in the [Arb+HCl+H2O] and [Arb+Glutaric acid] crystals, the dimethylamino group approaches the ethyl ester group (τ3

Figure 7. τ1 torsion angle relaxed energy profile for the Arbidol ion performed at the PBE/6-311++G(d,p) level of theory. The experimental values of τ1 in the Arbidol salts are marked as red dots. 4160


Article


desolvation process is accompanied by a broad peak on the DSC curve over the temperature range of ∼30−80 °C. Such a large difference between the solvates’ thermal stability suggests that the interaction energies between the solvent molecules and the host structure of the salts are different. The binding strength of the solvent in [Arb+SA+CHCl3] and [Arb +SA+ACN] can be estimated by calculating the vaporization enthalpy (ΔHS) of the salt-bound solvent using the following relationship:57

It can be concluded that, despite the conformational diversity of Arb in the crystals, all the molecules are located not far from the local conformational energy minima of the isolated molecule. It was also found that the crystal structures of [Arb +SA], [Arb+SA+CHCl3], [Arb+Maleic acid], and [Arb +Gentisic acid] contain a relatively low energy (“closed”) molecular conformation. The [Arb+BA] and [Arb+SA+ACN] salts consist of Arb ions with “open” conformation which is found to be a relatively higher energy state. 3.3. Thermal Analysis. The DSC traces for Arbidol base, [Arb+BA], and [Arb+SA] and its solvates are shown in Figure 8, and the thermal data are represented in Table 3.

T ΔHS = (ΔHdesolv ·100/ΔmS) ·MS

(6)

ΔHTdesolv

where is the enthalpy of desolvation derived from the DSC data, ΔmS is the percent mass loss measured in the TG experiment, and MS is the molecular weight of the solvent. The resulting ΔHS values of the solvates are shown in Table 3. It is evident that, in [Arb+SA+CHCl3], the solvent molecules are more tightly bound with the host structure than those in the [Arb+SA+ACN] solvate. Moreover, the vaporization enthalpy of pure chloroform (31.4 kJ·mol−1) is found to be ca. 5 kJ· mol−1 lower than the ΔHS value in the solvate crystal. For [Arb +SA+ACN], however, the ΔHS values indicate 1.5 times weaker interactions of acetonitrile molecules with the crystal environment than in the pure liquid (33.2 kJ·mol−1). Apparently, the binding strength of solvent molecules in solvates depends on various factors such as the packing arrangement of the host structure, the solvent accommodation in the crystal, etc. In the case of [Arb+SA+CHCl3] and [Arb+SA+ACN], the large difference in the ΔHS values is well correlated with the crystal structure features of the solvates. Indeed, in the [Arb+SA +CHCl3] crystal, the solvent is embedded into the cavity formed by a pair of Arb molecules, while in [Arb+SA+ACN], acetonitrile molecules occupy the interlayer voids. It has been reported that the thermal stability of a solvate can also be estimated by the difference between the desolvation onset temperature and the boiling point of the pure solvent.58,59 The Tdesolv − Tboil values equal +44 °C for the [Arb+SA+CHCl3] solvate (Tboil = 61.2 °C for CHCl3) and approximately −40 °C for [Arb+SA+ACN] (Tboil = 82.0 °C for ACN). These results qualitatively agree with the relationships between the ΔHS and ΔHvap(solvent) parameters. Thermal analysis of the [Arb+SA+CHCl3] solvate reveals that a solvent release from the host crystal structure yields a metastable desolvated material. It undergoes phase transition to form the [Arb+SA] salt, and this process is accompanied by a noticeable exo-event (≈−6 kJ·mol−1) seen in the DSC curve at about 130 °C. (Figure 8). We were able to isolate a sample of the desolvated form ([Arb+SA]desolv) by heating the [Arb+SA +CHCl3] solvate to the desolvation temperature (100 °C), which is followed by a rapid cooling to room temperature (≈25 °C) in an inert nitrogen atmosphere. The sample obtained by the procedure described above was immediately subjected to DSC and XRPD analyses (see Figure S3). The results of the

Figure 8. DSC curves for Arbidol base, salts, and [Arb+SA] solvates recorded at 10 °C·min−1 heating rate.

For Arb and the salts, DSC curves show one endotherm which corresponds to the melting process, whereas other phase transitions are not observed. The melting temperature of the [Arb+BA] salt is found to be ca. 2 °C lower than that of the Arbidol base and close to the Tfus value of the pure benzoic acid (121.9 °C). Similar behavior is observed for the [Arb+SA] salt. It melts at a temperature which is significantly closer to the melting point of salicylic acid (158.9 °C) rather than the Arbidol base. However, the salt fusion enthalpy values are comparable. The thermophysical data of the desolvation processes of the [Arb+SA] solvates with chloroform and acetonitrile are shown in Table 3. It is found that the number of solvent molecules observed in the crystal structure analysis is in agreement with the results of TG analyses within experimental uncertainty (see Figure S2). For [Arb+SA+CHCl3], the desolvation process is registered by DSC as a sharp endotherm with the desolvation onset temperature equaling 105.4 ± 1.6 °C. For [Arb+SA +ACN], the solvent already starts releasing at ca. 30 °C, and the

Table 3. Thermophysical Data for Arbidol Salts and [Arb+SA] Solvates Tdesolv, °C (onset) Arbidol base [Arb+BA] [Arb+SA] [Arb+SA+CHCl3] (1:1:1) [Arb+SA+ACN] (1:1:2) a

105.4 ± 1.6 ≈40

ΔHTdesolv, J·g−1

48.9 ± 3.2 ≈64

ΔmS, %

15.8 (16.2)a 11.2 (11.8)a

Tfus, °C (onset) 124.8 122.6 148.6 148.4 149.7

± ± ± ± ±

1.0 0.7 1.3 1.5 1.3

ΔHTfus, kJ·mol−1

ΔHS, kJ·mol−1

± ± ± ± ±

36.9 23.6

44.6 62.9 65.4 62.4 60.4

2.8 1.5 2.4 3.1 2.4

The values in parentheses indicate the calculated TG mass loss. 4161


Article

Molecular Pharmaceutics Scheme 1. Pathways for Synthesis and Phase Transformations of [Arb+SA] Solvatesa

a Blue arrows indicate solution transformations in chloroform (CHCl3). Green arrows indicate solution transformations in acetonitrile (ACN). Red arrows show transformations occurring as a result of heating a solid sample.

The thermodynamic parameters of the salt formation are shown in Table 4. The Gibbs energies of the salt formation

XRPD study clearly show that [Arb+SA]desolv is not amorphous, but has a long-range order with a molecular arrangement that differs from that of the [Arb+SA] salt as well as its parent form [Arb+SA+CHCl3] (Figure S3a). Hence, the exo-event of the phase transition [Arb+SA]desolv → [Arb+SA] detected by DSC (Figure S3b) indicates energy difference between the two solid forms of the salt. In contrast, the thermal desolvation of the [Arb+SA+ACN] solvate directly resulted in the [Arb+SA] salt formation, which was confirmed by the XRPD analysis. Pathways for phase transformations of the [Arb+SA] solvates described above are illustrated in Scheme 1. Scheme 1 also helps to elucidate the thermodynamic relationships between the [Arb+SA] salt and its solvates. Apparently, the most thermodynamically stable form of the salt in chloroform at room temperature (Troom) is the [Arb+SA +CHCl3] solvate. In acetonitrile, however, the [Arb+SA+ACN] solvate is thermodynamically less stable than the pure salt at Troom, since the slurry and slow evaporation experiments only resulted in the unsolvated form of Arbidol salicylate. 3.4. Thermodynamics of the Salt Formation. As the next step, the standard thermodynamic functions of the formation of the Arb salts with benzoic and salicylic acid were estimated using the procedure described in Materials and Methods. Since the thermodynamic parameters of multicomponent compound formation are not a solvent function, the solubility of the [Arb+SA] salt and its constituents was measured in ethanol, while for [Arb+BA] and its constituents, ethyl acetate was used as the solvent. The congruent solubility of the salts was observed at each temperature, and the solid phase, recovered after the experiment, was identified by the XRPD as the starting material. In case of the [Arb+SA] salt, we did not use acetonitrile as the solvent for the solubility experiments in order to avoid contamination of the bottom phase by the [Arb+SA+ACN] solvate. On the other hand, the [Arb+BA] salt was found to dissolve incongruently in ethanol (and methanol), probably, due to a large difference in the solubilities of the pure components in this solvent. Therefore, two different solvents were used to obtain the solubilities and to derive the formation thermodynamics of the salts. The experimental solubility values of the [Arb+BA], [Arb+SA] salts and their constituents in the respective solvents are shown in Table S1.

Table 4. Standard Thermodynamic Functions of Formation for [Arb+BA] and [Arb+SA] Salts in Ethyl Acetate and Ethanol, Respectively a

[Arb+BA] [Arb+SA]b

ΔG°f, kJ·mol−1

ΔH°f, kJ·mol−1

ΔS°f, J·mol−1·K−1

−18.6 ± 0.3 −25.2 ± 0.2

−18.7 ± 0.9 −12.8 ± 0.6

0±4 41 ± 3

The pKa value for benzoic acid equals 4.20; ΔpKa = 1.8. bThe pKa value for salicylic acid equals 2.97; ΔpKa = 3.0. a

(ΔG°f) are negative, indicating a spontaneous process. In addition, the absolute value of the driving force of the formation process of [Arb+SA] is more than by a quarter larger than that of [Arb+BA]. It suggests a greater affinity between Arbidol and salicylic acid compared to that of the API and benzoic acid. These results should be treated with some caution, however, since calculation of the Gibbs energy change of a salt requires the dissociation constants of components (see eq 2), which are solvent-specific and generally known only for water solutions. In an organic solvent, the acid−base properties of a compound may alter significantly, so that application of the water pKa values in such conditions is at best a rough approximation. The formation enthalpies (ΔH°f) derived from eq 4 are pKa independent and, therefore, more reliable in this case. In contrast to the ΔG°f order, the [Arb+BA] formation enthalpy is found to be ca. 6 kJ·mol−1 larger than that of [Arb +SA]. This indicates that formation of the [Arb+BA] salts is accompanied by a greater packing energy gain as compared to that of the [Arb+SA] salt. The entropy change of the process for [Arb+BA] is close to zero, while for [Arb+SA], ΔS°f has a positive value. It should be noted that the entropy change estimated from the eq 5 is also implicitly depend on the pKa difference of the components. Therefore, as noted previously, adjustment of the dissociation constants to a particular medium is needed to provide reliable results. 3.5. Aqueous Dissolution of the Arbidol Salts. As it was mentioned above, Arb is a relatively strong base with pKa value near 6.0.27 Therefore, the solubility of the drug as well as its salts in aqueous solutions is expected to be pH-dependent. In order to compare dissolution behavior of the Arb salts in different media, the experiments (at 25 °C) were performed in 4162


Article


acidic media compared to [Arb+BA]. Table 5 shows that the final concentration level for all the compounds is found to be in agreement with the solubility value of [Arb+HCl+H2O]. In the pH 6.8 buffer solution, the Arb solubility is low, reaching ca. 0.004 ± 0.001 mg·mL−1, which is approximately 6 times lower than that at pH 1.2 (Figure 9b). As in acidic conditions, in the solution with pH 6.8 all the salts dissolve incongruently and undergo a solution-mediated transformation to form abridol base, which was confirmed by XRPD analysis of the solid phase recovered after the experiment (Figure S5). Nevertheless, both salts demonstrate elevated concentration level during the first 30 min of dissolution. The maximal concentration of Arb released from the [Arb+BA] and [Arb +SA] salts is found to be approximately twice higher than the Arb solubility value under the current conditions. For the next 4 h, however, the Arb concentration decreases down to the plateau level, corresponding to solubility of the Arb base.

pharmaceutically relevant hydrochloric buffer with pH 1.2 and the phosphate buffer with pH 6.8. The results of the dissolution studies are shown in Table 5 and Figure 9. Table 5. Results of Dissolution Studies of Arb Salts and Arb Base in pH 1.2 and pH 6.8 Media at 25 ± 0.1 °C Cmax,a mg·mL−1 Arb base [Arb +HCl +H2O] [Arb+BA] [Arb+SA]

0.30 ± 0.01 0.13 ± 0.01

Arb base [Arb +HCl +H2O] [Arb+BA] [Arb+SA]

0.37 ± 0.02 0.36 ± 0.02

solubility,b mg·mL−1 pH 1.2 0.15 ± 0.01 0.13 ± 0.01

solid phase recovered after solubility exptc [Arb+HCl+H2O] [Arb+HCl+H2O] [Arb+HCl+H2O] [Arb+HCl+H2O]

0.005 ± 0.001 0.042 ± 0.005

0.16 ± 0.01 0.17 ± 0.01 pH 6.8 0.004 ± 0.001 0.005 ± 0.001

0.017 ± 0.001 0.013 ± 0.001

0.004 ± 0.001 0.004 ± 0.001

Arb Arb

Arb Arb

4. CONCLUSIONS Salts of antiviral drug Arbidol with pharmaceutically relevant benzoate and salicylate anions have been obtained, and their crystal structures have been determined. For Arbidol salicylate, an unstable solvate with acetonitrile was also found and characterized. The conformational analysis of Arbidol in different crystal forms has revealed that the most widely varying torsion angle in the molecule is ∠C1−C2−S1−C3 (τ1), which determines the orientation of the phenyl ring in relation to the indole moiety. With respect to the τ1 values, all the Arb conformations can be conventionally divided into two groups: “open” and “closed” types. DFT calculations have shown that both types of conformations correspond to local conformational energy minima of the isolated molecule. The “closed” conformation, however, was found to be a relatively low energy state compared to the “open” one. Thermal analysis of the solvates of Arbidol salicylate has revealed that in [Arb+SA +CHCl3] the solvent molecules are more tightly bound with the host structure than those in the [Arb+SA+ACN] solvate. It was also found that the solvent release from the [Arb+SA +CHCl3] crystal structure leads to formation of a metastable desolvated material. On the contrary, the thermal desolvation of the [Arb+SA+ACN] solvate directly resulted in [Arb+SA] salt formation. The standard thermodynamic functions of the salt

a

Maximum concentration of Arb in solution. bConcentration after 6 h of the experiment. cThe residual materials were identified by XPRD analysis (see Supporting Information).

During the first hour of [Arb+BA] and [Arb+SA] dissolution in pH 1.2 medium, the concentration of Arbidol increases by at least 3 times compared to its commercial form, i.e., [Arb+HCl +H2O] (Figure 9a). This is followed by a relatively fast decline of the drug concentration, which can be attributed to a solution-mediated transformation of the bottom phase during the experiment. XRPD analyses of the solid phases recovered after the experiment revealed transformation of the [Arb+BA] and [Arb+SA] to a hydrochloride salt of Arb, which is apparently the most thermodynamically stable form of the drug under the current conditions. Similar behavior is also observed for the arbridol base. It was found that the [Arb+BA] salt totally converts into [Arb+HCl+H2O] during the experiment time (Figure S4d). In the case of [Arb+SA], however, a minimal amount of the Arbidol salicylate was identified by XRPD along with [Arb+HCl+H2O] as a main component of the bottom phase. This fact suggests a greater stability of [Arb+SA] in the

Figure 9. Dissolution profiles for the salts and pure Arb in (a) pH 1.2 and (b) pH 6.8 at 25 °C. 4163


Article


of Viferon and Arbidol in Adult Influenza. Vopr. Virusol. 2008, 53, 31− 33 (Russian). (7) Leneva, I. A.; Shuster, A. M. Antiviral Etiotropic Chemicals: Efficacy Against Influenza A Viruses A Subtype H5N1. Vopr. Virusol. 2006, 51, 4−7 (Russian). (8) Brooks, M. J.; Sasadeusz, J. J.; Tannock, G. A. Antiviral Chemotherapeutic Agents Against Respiratory Viruses: Where Are We Now and What’s in the Pipeline? Curr. Opin. Pulm. Med. 2004, 10, 197−203. (9) Berendsen, B. J.; Wegh, R. S.; Essers, M. L.; Stolker, A. A.; Weigel, S. Quantitative Trace Analysis of a Broad Range of Antiviral Drugs in Poultry Muscle Using Column-Switch Liquid Chromatography Coupled to Tandem Mass Spectrometry. Anal. Bioanal. Chem. 2012, 402, 1611−1623. (10) Delogu, I.; Pastorino, B.; Baronti, C.; Noudairède, A.; Bonnet, E.; de Lamballerie, X. In Vitro Antiviral Activity of Arbidol Against Chikungunya Virus and Characteristics of a Selected Resistant Mutant. Antiviral Res. 2011, 90, 99−107. (11) Pécheur, E. I.; Lavillette, D.; Alcaras, F.; Molle, J.; Boriskin, Y. S.; Roberts, M.; Cosset, F. L.; Polyak, S. J. Biochemical Mechanism of Hepatitis C Virus Inhibition by the Broad-Spectrum Antiviral Arbidol. Biochemistry 2007, 46, 6050−6059. (12) Arastoo, M.; Khorram Khorshid, H. R.; Radmanesh, R.; Gharibdoust, F. Combination of IMOD and Arbidol to Increase Their Immunomodulatory Effects as a Novel Medicine to Prevent and Cure Influenza and Some Other Infectious Diseases. J. Med. Hypotheses Ideas 2014, 8, 53−56. (13) Glushkov, R. G.; Gus’kova, T. A. Arbidol: a New Domestic Immunomodulant (a Review). Pharm. Chem. J. 1999, 33 (3), 115− 122. (14) Glushkov, R. G.; Gus’kova, T. A.; Golikov, P. P.; Davydov, B. V.; Klychnikova, E. V. Study of the Antioxidant Properties of Arbidol. Pharm. Chem. J. 1996, 30 (1), 1−3. (15) Glushkov, R. G. Arbidol Antiviral, Immunostimulant, Interferon Inducer. Drugs Future 1992, 17, 1079−1081. (16) Chernyshev, V. V.; Davlyatshin, D. I.; Shpanchenko, R. V.; Nosyrev, P. V. Structural Characterization of Arbidol®. Z. Kristallogr. 2011, 226, 832−836. (17) Eropkin, M. Y.; Solovskii, M. V.; Smirnova, M. Y.; Bryazzhikova, T. S.; Gudkova, T. M.; Konovalova, N. I. Synthesis and Biological Activity of Water-Soluble Polymer Complexes of Arbidol. Pharm. Chem. J. 2009, 43, 563−567. (18) Babkin, V. A.; Kiselev, O. I. Method for preparing antiviral water-soluble polymer complexes of Arbidol. RU 2475255C1, 2013. (19) Leuner, C.; Dressman, J. Improving Drug Solubility for Oral Delivery Using Solid Dispersions. Eur. J. Pharm. Biopharm. 2000, 50, 47−60. (20) Khadka, P.; Ro, J.; Kim, H.; Kim, I.; Kim, J. T.; Kim, H.; Cho, J. M.; Yun, G.; Lee, J. Pharmaceutical Particle Technologies: An Approach to Improve Drug Solubility, Dissolution and Bioavailability. Asian J. Pharm. Sci. 2014, 9, 304−316. (21) Singhal, D.; Curatolo, W. Drug Polymorphism and Dosage Form Design: a Practical Perspective. Adv. Drug Delivery Rev. 2004, 56, 335−347. (22) Hecq, J.; Deleers, M.; Fanara, D.; Vranckx, H.; Amighi, K. Preparation and Characterization of Nanocrystals for Solubility and Dissolution Rate Enhancement of Nifedipine. Int. J. Pharm. 2005, 299, 167−177. (23) Pudipeddi, M.; Serajuddin, A. T. M.; Grant, D. J. W.; Stahl, P. H. In Handbook of pharmaceutical salts, properties, selection and use; Stahl, P. H., Wermuth, C. G., Eds.; Wiley-VCH: Weinheim, Germany, 2002; pp 19−40. (24) Childs, S. L.; Stahly, G. P.; Park, A. The Salt-Cocrystal Continuum: the Influence of Crystal Structure on Ionization State. Mol. Pharmaceutics 2007, 4, 323−338. (25) Mohamed, S.; Tocher, D. A.; Vickers, M.; Karamertzanis, P. G.; Price, S. L. Salt of Cocrystal? A New Series of Crystal Structures Formed From Simple Pyridines and Carboxylic Acids. Cryst. Growth Des. 2009, 9, 2881−2889.

formation were estimated from the solubility data of the salts and their pure constituents in organic solvents at different temperatures. The Gibbs energy change of the formation process was found to be negative in both cases, which indicates that the formation of the salts from individual components is a spontaneous process. The most significant contribution to the driving force is provided by the formation enthalpy. The dissolution study of the Arbidol salts performed in aqueous buffer solutions with pHs 1.2 and 6.8 has shown that both salts dissolve incongruently to form an Arbidol hydrochloride monohydrate at pH 1.2 and an Arbidol base at pH 6.8, respectively. In the acidic media, the amount of Arb reached the concentration level, which was found to be ca. 3 times higher than the solubility of the commercially available form ([Arb +HCl+H2O]). In the pH 6.8 solution, the salts demonstrated a moderate solubility enhancement due to low stability under the current conditions.

■

ASSOCIATED CONTENT

* Supporting Information S

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.molpharmaceut.5b00629. Asymmetric units of the Arb salts, TG, DSC, and XRPD analyses, XRPD patterns, and solubility data (PDF)

■

AUTHOR INFORMATION

Corresponding Author

*Tel: +7-4932-533784. Fax: +7-4932- 336237. E-mail [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was supported by the Russian Scientific Foundation (No. 14-33-00017). We thank “the Upper Volga Region Centre of Physicochemical Research” for technical assistance with Xray powder diffraction experiments.

■

REFERENCES

(1) Archakov, A. I.; Guseva, M. K.; Uchaykin, V. F.; Ipatova, O. M.; Dochshitsin, Y. F.; Medvedeva, N. V.; et al. Pharmaceutical composition containing Arbidol in the form of phospholipid nanoparticles, United States Patent Application 2012/0052099 A1. (2) Brooks, M. J.; Burtseva, E. J.; Ellery, P. J.; Marsh, G. A.; Lew, A. M.; Slepushkin, A. N.; Crowe, S. M.; Tannock, G. A. Antiviral Activity of Arbidol, a Broad-Spectrum Drug for Use Against Respiratory Viruses, Varies According to Test Conditions. J. Med. Virol. 2012, 84, 170−181. (3) Semenenko, T. A.; Sel’kova, E. P.; Gotvianskaia, T. P.; Gaidarenko, A. D.; Polezhaeva, N. A.; Evseeva, L. F.; Nikolaeva, O. G. Characteristics of the Immune Status in Specific and Nonspecific Prophylaxis of Influenza in Elderly Persons. Zh. Mikrobiol. Epidemiol. Immunobiol. 2005, 6, 24−28 (Russian). (4) Blaising, J.; Polyak, S. J.; Pécheur, E.-I. Arbidol as a BroadSpectrum Antiviral: An Update. Antiviral Res. 2014, 107, 84−94. (5) Gatich, R. Z.; Kolobukhina, L. V.; Vasil’ev, A. N.; Isaeva, E. I.; Burtseva, E. I.; Orlova, T. G.; Voronina, F. V.; Kol’tsov, V. D.; Malinovskaia, V. V. Viferon Suppositories in the Treatment of Influenza in Adult. Antibiot. Khimioter. 2008, 53, 13−17 (Russian). (6) Kolobukhina, L. V.; Malinovskaia, V. V.; Gatich, R. Z.; Merkulova, L. N.; Burtseva, E. I.; Isaeva, E. I.; Parshina, O. V.; Guseva, T. S.; Orlova, T. G.; Voronina, F. V. Evaluation of the Efficacy 4164


Article

Molecular Pharmaceutics (26) Cruz-Cabeza, A. J. Acid-Base Crystalline Complexes and the pKa Rule. CrystEngComm 2012, 14, 6362−6365. (27) Socrates, G. Infrared and Raman Characteristic Group Frequencies: Tables and Charts; John Wiley & Sons: Chichester, 2004; 366 pp. (28) Orola, L.; Sarcevica, I.; Kons, A.; Actins, A.; Veidis, M. V. Conformation of the Umifenovir Cation in the Molecular and Crystal Structures of Four Carboxylic Acid Salts. J. Mol. Struct. 2014, 1056− 1057, 63−69. (29) Mehler, E. L.; Gerhards, J. Electronic Determinants of the AntiInflammatory Action of Benzoic and Salicylic Acids. Mol. Pharmacol. 1987, 31 (3), 284−293. (30) Madan, R. K.; Levitt, J. A Review of Toxicity From Topical Salicylic Acid Preparation. J. Am. Acad. Dermatol. 2014, 70 (4), 788− 792. (31) Grobelny, P.; Mukherjee, A.; Desiraju, G. Drug-Drug CoCrystals: Temperature-Dependent Proton Mobility in the Molecular Complex of Isoniazid with 4-Aminosalicylic Acid. CrystEngComm 2011, 13, 4358−4364. (32) Issa, N.; Karamertzanis, P. G.; Welch, G. W. A.; Price, S. L. Can the Formation of Pharmaceutical Cocrystals Be Computationally Predicted? I. Comparison of Lattice Energies. Cryst. Growth Des. 2009, 9 (1), 442−453. (33) Karamertzanis, P. G.; Kazantsev, A. V.; Issa, N.; Welch, G. W. A.; Adjiman, C. S.; Pantelides, C. C.; Price, S. L. Can the Formation of Pharmaceutical Cocrystals Be Computationally Predicted? 2. Crystal Structure Prediction. J. Chem. Theory Comput. 2009, 5 (5), 1432− 1448. (34) Musumeci, D.; Hunter, C. A.; Prohens, R.; Scuderi, S.; McCabe, J. F. Virtual Cocrystal Screening. Chem. Sci. 2011, 2 (5), 883−890. (35) Abramov, Yu. A.; Loschen, C.; Klamt, A. Rational Coformer or Solvent Selection for Pharmaceutical Cocrystallization or Desolvation. J. Pharm. Sci. 2012, 101 (10), 3687−3697. (36) Mohammad, M. A.; Alhalaweh, A.; Velaga, S. P. Hansen Solubility Parameter as a Tool to Predict Cocrystal Formation. Int. J. Pharm. 2011, 407, 63−71. (37) Alhalaweh, A.; Roy, L.; Rodríguez-Hornedo, N.; Velaga, S. P. pH-Dependent Solubility of Indomethacin−Saccharin and Carbamazepine−Saccharin Cocrystals in Aqueous Media. Mol. Pharmaceutics 2012, 9, 2605−2612. (38) Good, D. J.; Rodríguez-Hornedo, N. Solubility Advantage of Pharmaceutical Cocrystals. Cryst. Growth Des. 2009, 9 (5), 2252− 2264. (39) Surov, A. O.; Simagina, A. A.; Manin, N. G.; Kuzmina, L. G.; Churakov, A. V.; Perlovich, G. L. Fenamate Cocrystals with 4,4′Bipyridine: Structural and Thermodynamic Aspects. Cryst. Growth Des. 2015, 15 (7), 228−238. (40) Surov, A. O.; Voronin, A. P.; Manin, A. N.; Manin, N. G.; Kuzmina, L. G.; Churakov, A. V.; Perlovich, G. L. Pharmaceutical Cocrystals of Diflunisal and Diclofenac with Theophylline. Mol. Pharmaceutics 2014, 11 (10), 3707−3715. (41) Zhang, S.-W.; Harasimowicz, M. T.; de Villiers, M. M.; Yu, L. Cocrystals of Nicotinamide and (R)-Mandelic Acid in Many Ratios with Anomalous Formation Properties. J. Am. Chem. Soc. 2013, 135, 18981−18989. (42) Oliveira, M. A.; Peterson, M. L.; Davey, R. J. Relative Enthalpy of Formation for Co-Crystals of Small Organic Molecules. Cryst. Growth Des. 2011, 11, 449−457. (43) Lin, Y.; Yang, H.; Yang, C.; Wang, J. Preparation, Characterization, and Evaluation of Dipfluzine-Benzoic Acid Co-Crystals with Improved Physicochemical Properties. Pharm. Res. 2014, 31, 566−578. (44) Zhang, S.; Rasmuson, Å.C. Thermodynamics and Crystallization of the Theophylline−Glutaric Acid Cocrystal. Cryst. Growth Des. 2013, 13, 1153−1161. (45) Zhang, S.; Rasmuson, Å.C. The Theophylline−Oxalic Acid CoCrystal System: Solid Phases, Thermodynamics and Crystallisation. CrystEngComm 2012, 14, 4644−4655. (46) Zhao, L.; Raval, V.; Briggs, N. E. B.; Bhardwaj, R. M.; McGlone, T.; Oswald, I. D. H.; Florence, A. J. From Discovery to Scale-Up: α-

Lipoic Acid: Nicotinamide Co-Crystals in a Continuous Oscillatory Baffled Crystalliser. CrystEngComm 2014, 16, 5769−5780. (47) Yu, Z. Q.; Chow, P. S.; Tan, R. B. H. Operating Regions in Cooling Cocrystallization of Caffeine and Glutaric Acid in Acetonitrile. Cryst. Growth Des. 2010, 10, 2382−2387. (48) Évora, A. O. L.; Castro, R. A. E.; Maria, T. M. R.; Silva, M. R.; ter Horst, J. H.; Canotilho, J.; Eusébio, M. E. S. A Thermodynamic Based Approach on the Investigation of a Diflunisal Pharmaceutical Co-Crystal with Improved Intrinsic Dissolution Rate. Int. J. Pharm. 2014, 466, 68−75. (49) Sheldrick, G. M. Acta Crystallogr., Sect. A: Found. Crystallogr. 2008, 64, 112−122. (50) Sheldrick, G. M. SADABS, Program for scaling and correction of area detector data; University of Göttingen: Germany, 1997. (51) Schartman, R. R. On the Thermodynamics of Cocrystal Formation. Int. J. Pharm. 2009, 365, 77−80. (52) Rager, T.; Hilfiker, R. Stability Domains of Multi-Component Crystals in Ternary Phase Diagrams. Z. Phys. Chem. 2009, 223, 793− 813. (53) Frisch, M. J.; et al. GAUSSIAN03, Revision B.03; Gaussian Inc.: Pittsburgh, PA,2003. (54) Etter, M. C. Encoding and Decoding Hydrogen-Bond Patterns of Organic Compounds. Acc. Chem. Res. 1990, 23, 120−126. (55) Bernstein, J.; Davis, R. E.; Shimoni, L.; Chang, N.-L. Patterns in Hydrogen Bonding: Functionality and Graph Set Analysis in Crystals. Angew. Chem., Int. Ed. Engl. 1995, 34, 1555−1573. (56) Cruz-Cabeza, A. J.; Bernstein, J. Conformational Polymorphism. Chem. Rev. 2014, 114, 2170−2191. (57) Caira, M. R.; Bettinetti, G.; Sorrenti, M. Structural Relationships, Thermal Properties, and Physicochemical Characterization of Anhydrous and Solvated Crystalline Forms of Tetroxoprim. J. Pharm. Sci. 2002, 91, 467−481. (58) Caira, M. R.; Nassimbeni, L. R. Phase transformations in inclusion compounds, kinetics and thermodynamics of enclathration. In Comprehensive Supramolecular Chemistry; Atwood, J. L., Davies, J. E. D., Macnicol, D. D., Vögtle, F., Eds.; Pergamon: New York, 1996; Vol. 6, Chapter 25. (59) Lemmer, H.; Stieger, N.; Liebenberg, W.; Caira, M. R. Solvatomorphism of the Antibacterial Dapsone: X-ray Structures and Thermal Desolvation Kinetics. Cryst. Growth Des. 2012, 12, 1683− 1692.

4165


New Solid Forms of the Antiviral Drug Arbidol: Crystal Structures

Recommend Documents