Molecular Salts of the Antidepressant Venlafaxine - ACS Publications

Jun 16, 2017 - ABSTRACT: The possibility of decreasing the solubility of the antidepressant drug venlafaxine hydrochloride by for- mation of molecular...
0 downloads 0 Views 1MB Size
Subscriber access provided by NEW YORK UNIV

Article

Molecular salts of the antidepressant venlafaxine: an effective route to solubility properties modifications Floriana Spinelli, Elena Dichiarante, Marco Curzi, Stefano L. Giaffreda, Michele R. Chierotti, Roberto Gobetto, Federica Rossi, Laura Chelazzi, Dario Braga, and Fabrizia Grepioni Cryst. Growth Des., Just Accepted Manuscript • DOI: 10.1021/acs.cgd.7b00606 • Publication Date (Web): 16 Jun 2017 Downloaded from http://pubs.acs.org on June 20, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Crystal Growth & Design is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Crystal Growth & Design

Molecular salts of the antidepressant venlafaxine: an effective route to solubility properties modifications Floriana Spinelli,a Elena Dichiarante,b Marco Curzi,b Stefano L. Giaffreda,b Michele R. Chierotti,*,c Roberto Gobetto,c Federica Rossi,c Laura Chelazzi,d Dario Braga,a Fabrizia Grepioni*,a a

Dipartimento di Chimica Giacomo Ciamician, Università di Bologna, Via F. Selmi 2, 40126

Bologna, Italy b

PolyCrystalLine s.p.a., Via S.R. Fabri, 127/1, 40059, Medicina (Bologna) Italy

c

Dipartimento di Chimica and NIS Centre, Università di Torino, Via P. Giuria 7, 10125 Torino,

Italy d

Centro di Servizi di Cristallografia Strutturale, Università degli studi di Firenze, Via della

Lastruccia n°3, 50019, Sesto Fiorentino (Firenze), Italy

ACS Paragon Plus Environment

1

Crystal Growth & Design

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 37

ABSTRACT

The possibility of decreasing the solubility of the antidepressant drug venlafaxine hydrochloride by formation of molecular salts with organic acids accepted by the Pharmacopeia has been successfully investigated. Reacting venlafaxine with coumaric, ferulic, oxalic, salicylic, fumaric and citric acids results in the protonation of the amino group and formation of the corresponding 1:1 molecular salts. All compounds have been characterized by a combination of solid state techniques, i.e. single crystal and powder X-ray diffraction, thermogravimetric analysis, differential scanning calorimetry, FT-IR and solid-state NMR (1H MAS, 1

13

C and

15

N CPMAS,

H DQ MAS, 2D 13C-1H HETCOR and 2D 14N-1H J-HMQC) spectroscopy. Intrinsic dissolution

tests were performed on the pure salts, and suitable candidates were selected for the preparation of solid formulations with excipients; dissolution profiles for the solid formulations were measured in water and sodium chloride solution, and compared with that of the commercial form of venlafaxine, showing that the coumarate salt might represent an improvement for extended release administrations.

INTRODUCTION The search for multicomponent crystals, especially molecular and ionic co-crystals, and salts, is at the forefront of crystal engineering studies.1-6 The goal is that of being able to devise new “collective” properties by binding together within the same crystal structure an active molecule, e.g. an active pharmaceutical ingredient (API), and one or more partner molecules, whether active or “innocent”, in order to obtain new and improved crystal properties with respect to those of the original solid.7-9 Understanding the differences in properties of multicomponent crystals as

ACS Paragon Plus Environment

2

Page 3 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Crystal Growth & Design

a function of their structural composition is critical for establishing uniqueness in a pharmaceutical patent, as well as for optimizing conditions for tablet compression and powder flow required for pharmaceutical compounding.10-11 Venlafaxine is the common name for the compound (±)1-[2-(dimethylamino)-1-(4methoxyphenyl)ethyl]cyclohexan-1-ol, a serotonin-norepinephrine reuptake inhibitor, which is used for treating depression and thymoanaleptic and anxiolytic disorders.12 The molecular structure is shown in Scheme 1.

Scheme 1. The molecular structure of venlafaxine, methoxyphenyl)ethyl]cyclohexan-1-ol (ven in the following).

1-[2-(dimethylamino)-1-(4-

Venlafaxine is extremely insoluble (267 mg L-1 in water at 25°C),13 and for this reason it is administered orally as its chloride salt. Venlafaxine was marketed in 1993 as an immediaterelease form “EFFEXOR®” (IR) and since 1997 as an extended-release formulation “EFFEXOR® XR” (ER), approved by the US Food and Drug Administration. It has been listed among the 10 most-prescribed antidepressant medications in US in 2017. The extended-release formulation has been developed as an alternative to the immediaterelease that – due to the high solubility of venlafaxine hydrochloride (572 mg mL-1 in water at 25°C), 13 had to be administered two or three times per day, causing side effects in patients.14-15

ACS Paragon Plus Environment

3

Crystal Growth & Design

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 37

The extended-release formulation possesses the same properties as the immediate-release one, with the advantage for the patient that it needs to be administered only once per day. 16 In the extended-release formulation the API is contained in spheroids with microcrystalline cellulose and hydroxypropyl methylcellulose (HPMC); the spheroids are coated with a mixture of ethyl cellulose and HPMC; in this way the right in vitro dissolution profile can be obtained: the amount of sample dissolved must be within the parameters dictated by TEST 2 for bioequivalence.17 There is a strong interest in the crystal engineering approach applied to the pharmaceutical field for the preparation of new formulations of active pharmaceutical ingredients (APIs), in particular co-crystals and salts, that often show physical and chemical properties (solubility, intrinsic dissolution rate, melting point, color, etc.) different from those of the pure components.18 During the last decade venlafaxine solid state properties have been intensely investigated.19 Venlafaxine is a highly polymorphic system. Not less than six polymorphs have been reported.20-21 Research on venlafaxine is mainly aimed to the discovery of new crystalline forms, salts and co-crystals, which might show a decrease in their dissolution rate with respect to venlafaxine current commercial form, i.e. the hydrochloride salt; less water-soluble crystalline forms are of interest because they can be more easily formulated into extended-release formulations.22 A careful investigation of the composition in terms of polymorph quantification in commercially available raw materials has been recently published.23 The presence of an amino group in the venlafaxine molecule makes it a good candidate for the design of hydrogen bonded adducts, i.e. co-crystals or molecular salts of protic organic acids.24

ACS Paragon Plus Environment

4

Page 5 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Crystal Growth & Design

In the present work we report the results of the co-assembly processes of venlafaxine with coformers accepted by the pharmacopeia, i.e. coumaric ((E)-3-(4-hydroxyphenyl)-2-propenoic acid), ferulic ((E)-3-(4-hydroxy-3-methoxy-phenyl)prop-2-enoic acid), oxalic, salicylic, fumaric and citric acid, obtained by solution crystallization with different solvents (water, ethanol, ethyl acetate, transcutol, propylenic glycol and dimethyl sulfoxide). The products were characterized by differential scanning calorimetry (DSC) and thermogravimetric analysis (TGA), solid-state nuclear magnetic resonance (SS NMR) and infrared spectroscopy (FTIR); the molecular salts were also structurally characterized via single crystal or powder X-ray diffraction, with the exception of the [venH][H2citrate] salt, whose structure was studied by solid state NMR. For four of these compounds, namely [venH][coumarate], [venH][ferulate], [venH][Hoxalate] and [venH][salicylate], it was possible to obtain single crystals suitable for X-ray diffraction, while in the case of [venH][Hfumarate] the structure was solved from X-ray powder diffraction data. Dissolution profiles in water were measured for all compounds, and the results were compared with those obtained on venlafaxine hydrochloride. Selected salts were also formulated as tablets with excipients, and their intrinsic dissolution rate was measured, in accordance to the United States Pharmacopeia (USP).17 The synthetic procedures were also optimized to allow for the preparation of larger quantities of pure salts for tablets formulation.

EXPERIMENTAL SECTION Materials. All reactants were purchased from Sigma-Aldrich and used without further purification. For dissolution tests distilled water was used. All salts were synthesized starting from a 1:1 stoichiometric ratio of venlafaxine base (50 mg, 0.18 mmol) and carboxylic acid (coumaric, ferulic, oxalic, salicylic, fumaric and citric acids).

ACS Paragon Plus Environment

5

Crystal Growth & Design

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 37

Solution synthesis. A mixture of venlafaxine base (50 mg, 0.18 mmol) and organic acid was added to 2 mL of solvent and heated up to 50°C, to enhance solubility. Solvents accepted by the Pharmacopeia were used (ethanol and ethyl-acetate). DMSO was also employed in the case of coumaric acid, with the only intent of obtaining single crystals suitable for X-ray diffraction analysis. Slurry. Synthesis of the [Hven][ferulate] salt was also conducted via slurry in ethyl acetate. A 1 mL suspension, containing venlafaxine (50 mg, 0.18 mmol) and ferulic acid (35.1 mg, 0.18mmol) in 1:1 ratio, was kept under stirring for 3 days in a closed vessel. The solid product was recovered after solvent evaporation at ambient conditions. Thermogravimetric analysis. Thermogravimetric analyses were performed using a simultaneus Mettler Toledo TGA/DSC equipped with a thermocouple for the direct measurement of DSC signal and a thermobalance for the measurement of the TGA signal. The samples (5-15 mg) were placed in aluminium pierced pans, and the heating was carried out at 10°C min–1 in N2 atmosphere (see Supporting Information). Differential scanning calorimetry. Calorimetric measurements were performed using a DSC 200 F3 Maia® differential scanning calorimeter equipped with an intra-cooler. The samples (2–4 mg) were placed in aluminium pierced pans, and the heating was carried out at 10°C min–1 in N2 atmosphere. (See Supporting Information) Single crystal X-ray diffraction. Single-crystal data were collected on an Oxford X’Calibur CCD diffractometer equipped with a graphite monochromator (MoKα radiation, λ = 0.71073) and operated at room temperature. Structures were solved by direct methods and refined by fullmatrix least-squares on F2 with the SHELX9725 program package. All non-hydrogen atoms were refined anisotropically. HCH atoms for all salts were added in calculated positions and refined

ACS Paragon Plus Environment

6

Page 7 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Crystal Growth & Design

riding on their respective carbon atoms; HNH atoms were either directly located or added in calculated positions. Crystal data and details of measurement for all compounds are reported in Table SI-1. The program Mercury was used for molecular graphics.26 X-Ray Powder diffraction (XRPD). Diffraction patterns in the 2θ range 3-40° (step size, 0.02°; time/step, 20 s; 0.04 rad soller) were collected, using a glass sample holder, on a PANalytical X’Pert PRO automated diffractometer equipped with an X'Celerator detector and in Bragg-Brentano geometry, using Cu Kα radiation without a monochromator. The tube voltage and amperage were set to 40 kV and 40 mA, respectively. The program Mercury26 was used for simulation of X-ray powder patterns on the basis of single crystal data. Chemical and structural identity between bulk materials and single crystals was always verified, when possible, by comparing experimental and simulated powder diffraction patterns. Structure solution and refinement from powder data. Powder diffraction data for [Hven][Hfumarate] were collected in the 2θ range 5–60° (step size, 0.003°; time/step, 99s; 0.02 rad soller; V x A 40 × 40) on a Panalytical X’Pert PRO automated diffractometer operated in transmission mode (capillary spinner) and equipped with a Pixel detector. The data were analyzed with the software Highscore plus and unit cell parameters were found thanks to the algoritm DICVOL.27 Simulated and annealing runs with structure fragments were performed with EXPO201028, all options were left as default if not specifically stated and the best solutions were chosen for Rietveld refinements. Salt vs. co-crystal attribution was first made on the basis of pH values difference between venlafaxine and fumaric acid, and was confirmed by SSNMR spectroscopy (see below).18 The structure was refined by the Rietveld method with the software TOPAS429, treating the single molecules as rigid bodies.30-32 and using a spherical harmonics

ACS Paragon Plus Environment

7

Crystal Growth & Design

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 37

model to describe preferred orientation. Figure SI-16 shows the final Rietveld refinement plot. Refinements converged with χ2= 2,074 and Rwp= 5,167%. FT-IR. All FT-IR measurements were performed using a Nicolet FT-IR 6700 Thermo Fischer equipped with an ATR device. The analyses were performed on all samples and for each measurement 16 scans were recorded. (See Supporting Information) SSNMR measurements. 1D solid-state NMR measurements were performed on a Bruker Avance II 400 instrument operating at 400.2, 100.6 and 40.5 MHz for 1H,

13

C and

15

N nuclei,

respectively. Cylindrical 4 mm o.d. zirconia rotors with a sample volume of 80 µL were employed and spun at 12 (13C) and 9 (15N) kHz. All experiments employed the RAMP-CP pulse sequence (1H 90° pulse=3.05 µs) with the TPPM 1H decoupling with an rf field of 75 kHz during the acquisition period. Detailed acquisition parameters may be found in the Supporting Information (Table SI-2).

13

C and

15

N chemical shifts were referenced with the resonance of

hexamethylbenzene (13C methyl signal at 17.4 ppm) and (NH4)2SO4 (15N signal at -355.8 ppm with respect to CH3NO2). For [venH][Hfumarate] and [venH][H2citrate], 1H MAS, 1H DQ MAS, 2D

13

C-1H HETCOR

(indirect detection), and 2D 14N-1H J-HMQC (indirect detection) spectra were collected on a Jeol ECZR 600 instrument, operating at a frequency of 600.1, 150.9, and 43.4 MHz for 1H, 13C, and 14

N, respectively. Samples were packed in 1 mm (o.d.) cylindrical zirconia rotors (sample

volume 0.8 µl). All measurements were acquired at probe temperature with a spinning speed of 70 kHz. 1H MAS spectra were performed with an echo pulse sequence (90°-τ-180°-τ) to remove the probe background (1H 90° pulse=0.77 µs; 3 transients for all sample). The 2D 1H DQ MAS experiments were performed with the back-to-back (BABA) recoupling pulse sequence with excitation time durations of one rotor period (1H 90° =0.47 µs; 12 scans; t1 increments = 64;

ACS Paragon Plus Environment

8

Page 9 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Crystal Growth & Design

relaxation delay = 3.6 and 5.1 s for [venH][Hfumarate] and [venH][H2citrate], respectively. The 13

C-1H HETCOR were measured according to the method described in detail in earlier studies.33

In short, the experiment is composed by four parts: 1) a 1H-13C RAMP-CP (1H 90° = 0.77 µs; contact time 1: 1.8 and 1.5 ms, respectively); 2) a t1 period during which

13

C magnetization

evolves in the presence of heteronuclear 1H decoupling (TPPM; rf field= 32.5 kHz); 3) the

13

C

magnetization is stored along the z-axis (13C 90°= 0.86 µs), while the 1H magnetization remaining after the first CP transfer is canceled by reintroducing the 1H-1H dipolar interactions with a 1H irradiation following the homonuclear Rotary Resonance Recoupling (R3 with n=½: HORROR) condition34 with two x and y phases;35 4) the 1H magnetization is detected following the final

13

C-1H RAMP-CP transfer (contact time 2: 1.0 or 0.05 and 1.55 or 0.45 ms,

respectively), this time under heteronuclear 13C WALTS decoupling (rf field 10.5 kHz). In the JHMQC experiments, 1H and 14N 90° pulses were set to 0.77 and 6.5 µs, respectively. Relaxation delays of 3.6 s ([venH][Hfumarate]) and 5.1 s ([venH][H2citrate]) with rotor-synchronization of the t1 increment, ∆t1 =1/νr (14.3 µs) were employed. 200 transients were averaged for 50 t1 experiments. Optimum values for the magnetization transfer periods (texc=trec) were set to 1.25 ms for [venH][Hfumarate] and 0.65 ms for [venH][H2citrate]. 1

H, 13C, 14N and

15

N chemical shift scales were referenced with the resonance of adamantane

(1H signal at 1.87 ppm), glycine (13C methylene signal at 43.5 ppm), (NH4)2SO4(14N signal at 0 ppm and 15N signal at 24.6 ppm with respect to NH3), respectively as external standards. Dissolution measurements on pure salts. A calibration curve was built by plotting the absorbance against concentrations for five standard solutions of the sample (a calibration curve was built for each salt) in a sodium chloride solution 0.9% in standard concentrations. Analyses were performed using a dissolution tester. Methods are described in the “European

ACS Paragon Plus Environment

9

Crystal Growth & Design

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 37

Pharmacopeia 5.0”.36 The detector used was a UV-Vis Cary 50 Varian, equipped with an optic fiber. The program used was “Concentration” (Cary 50 WinUV Software V.3) and the measurement for each standard was recorded five times at a fixed wavelenght (λ=280nm) (see Supporting Information). Intrinsic dissolution rates were measured with a Hanson’s Vision Classic 6 dissolution tester coupled with a Varian Cary 50 UV-Vis Spectrophotometer. The program used was “Kinetic” (Cary 50 WinUV Software V.3), which recorded continuously the absorbance, at λ =280nm, of a sodium chloride solution 0.9% (80mL) under stirring (100 rpm) at 37°C, in which was added a tablet of sample. The values obtained were converted from Abs/sec into mg/sec and for each sample two measurements were recorded. Tablets (13 mm diameter) were obtained by compressing the powder with a tablet press CRUSHIR™ 15 TON HYDRAULIC PRESS. Dissolution measurements on formulations with excipients. The standard solutions of venlafaxine hydrochloride at various concentrations, for the calibration curve, were injected directly into the HPLC system. The procedure was repeated three times for each concentration. The areas of the analyte peaks (mAU s) have been plotted as a function of analyte concentration (expressed in g/mL), thus obtaining calibration curves by means of least squares method. The analytical method for the determination of venlafaxine hydrochloride has been validated following the guidelines Q2 (R1) "Validation of analytical procedures: Text and Methodology" of the International Conference of Harmonisation (ICH). The dissolution tests were performed by using a dissolution apparatus Agilent 708-DS Agilent Technologies equipped with autosampler 8000 Dissolution Sampling Station (see Supporting Information) with the following parameters: (a) Medium: water, 900mL; (b) Apparatus 1 (basket): 100rpm; (c) Times: 2, 4, 8, 12, 20 h; (d) Temperature: 37°C.

ACS Paragon Plus Environment

10

Page 11 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Crystal Growth & Design

The withdrawals were analyzed by using the HPLC method reported in the Venlafaxine Hydrochloride Monography17 and the data were quantified using the calibration curves built in the first part of the experiment (see Supporting Information). The chromatographic system used in this work consists of a chromatographic pump model Thermo Finnigan Surveyor MS Pump (Waltham, MA, USA), an autosampler Thermo Finnigan Surveyor Autosampler Plus and a spectrophotometric detector Thermo Finnigan Surveyor PDA Plus. The separation was achieved through the use of a chromatographic column Zorbax SB-C18 (250x4.6 mm i.d., 5 mM) of the company Agilent (Santa Clara, CA, USA) maintained at a temperature of 30 °C. The mobile phase used is a mixture of 10 mM of phosphate buffer, pH 3.0, containing 0.5% (v / v) of triethylamine, and acetonitrile in proportions of 75/25 (v / v). The phase eluent was then filtered using filters in PES 0.45 uM of VWR company (Milan, Italy). The analyzes were carried out in an isocratic way, by adopting a rate of flow of 1.0 mL/min and injecting through a loop from 20 L. The UV absorption was detected at a wavelength of 225 nm. The data processing was done through the use of “ChromeQuest 5.0” software (ThermoScientific). Tablets (13 mm diameter) were obtained by compressing the powder with a tablet press CRUSHIR™ 15 TON HYDRAULIC PRESS.

RESULTS AND DISCUSSION Venlafaxine salts [venH][coumarate], [venH][ferulate], [venH][Hoxalate], [venH][salicylate], [venH][Hfumarate] and [venH][H2citrate] are summarized in Scheme 2. In all cases protonation of the venlafaxine amino group and formation of the [Hven]+ cation could be detected, via either X-ray diffraction or solid state NMR (see SI for crystal data and NMR spectra). Hence all solids

ACS Paragon Plus Environment

11

Crystal Growth & Design

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 37

ought to be described as molecular salts, whereby proton transfer from the organic acid conformer to the API has taken place.

Scheme 2 – Acids employed for reactions with venlafaxine free base, and the molecular salts obtained.

Crystalline [venH][coumarate] and [venH][ferulate] are isostructural: the arrangements of protonated venlafaxine, [venH]+, and of the anions are similar (see Figure 1) and both solids are characterized by a zig-zag hydrogen bonded chain formed by the coumarate and ferulate anions, respectively. Venlafaxine cations are placed on both sides of the chain and are linked to it via

ACS Paragon Plus Environment

12

Page 13 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Crystal Growth & Design

N(H)+··· OCOO- and O(H)··· OCOO- hydrogen bonds [N(H)+··· OCOO- = 2.628 – 2.649 (1) Å, O(H)··· OCOO- = 2.726 (2) Å, OCOOH(H)··· OCOO- = 2.607 (5) - 2.636 (4) Å ] (see Figure 1).

Figure 1. Hydrogen bonding pattern in [venH][coumarate] (top) and [venH][ferulate] (bottom).

ACS Paragon Plus Environment

13

Crystal Growth & Design

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 37

The structure of the hydrogen oxalate salt is also characterized by the presence of an anionic chain (see Figure 2), with venlafaxine cations linked to carboxylate groups on both sides of the chain; the carboxylate groups also interact with the –OH groups on the cations [N(H)+··· OCOO- = 2.781 (2) Å, O(H)··· OCOO- = 2.793 (1) Å, OCOOH(H)··· OCOO- = 2.625 (1) Å].

Figure 2. Hydrogen bonding pattern in [venH][Hoxalate]. Note the hydrogen bonded chain formed by the hydrogen oxalate anions. The salicylate salt presents a different packing pattern: due to the strong intramolecular hydrogen bond, which is formed upon deprotonation of the carboxylic group in the salicylic acid, the chain motif is lost, and intermolecular hydrogen bonds are observed between the carboxylate group on the anion and (i) the protonated amino group and (ii) the –OH group on the cation (see Figure 3.) [N(H)+··· OCOO- = 2.632 (3) Å, O(H)··· OCOO- = 2.909 (3) Å, OCOOH(H)··· OCOO- = 2.460 (3) Å]

ACS Paragon Plus Environment

14

Page 15 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Crystal Growth & Design

Figure 3. Hydrogen bonding pattern in crystalline [venH][salicylate]. A similar inter-anionic hydrogen bonded chain (as in the case of the structure of [venH][Hoxalate]) is observed in the structure of the hydrogen fumarate salt (see Figure 4) [OCOOH(H)··· OCOO-

2.416(4) Å]. The venlafaxine cations are linked to the chain via

N+(H)···OCOO- bonds [N···O distance 2.838(3) Å].

Figure 4. Hydrogen bonding pattern in crystalline [venH][Hfumarate]. Incidentally, these five molecular salts provide further interesting examples of how the hydrogen bonds are “used” by the crystal packing to minimize repulsions – as those acting

ACS Paragon Plus Environment

15

Crystal Growth & Design

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 37

between charged fragments of the same sign – and maximize attractions at in the case of fragments having opposite formal charges.37-38 Solid-state NMR measurements. The aim of the SSNMR measurements was threefold: a) to investigate the ionic or neutral nature of all samples, i.e. whether they can be regarded as salts or co-crystals. Indeed, it is well known that the COOH 13C chemical shift as well as the 15N chemical shift of amine moieties are extremely sensitive to the position of the hydrogen atom along the O⋅⋅⋅N contacts.30-40 In the

13

C SSNMR spectra, COO- signals are shifted at higher

frequencies (up to ca 7 ppm) with respect to COOH peaks while

15

N signals, upon salification,

shift up to 50-100 ppm at high or low frequencies depending on the type of nitrogen atom (whether aliphatic or aromatic, respectively);41-43 b) To assist and support the structure solution of [venH][Hfumarate] from XRPD data as well as to check the reliability of the solved structure thank to the well known “NMR crystallography” approach based on the chemical shift and dipolar analysis;44-46 c) to provide any information able to describe the structure of [venH][H2citrate] for which the X-ray structure is not available thanks to the NMR ability of providing 1H-1H, 1H-13C and 1H-14/15N proximities.47-51 13C and were acquired for all samples. The carboxylic region of the

13

15

N CPMAS SSNMR spectra

C CPMAS SSNMR spectra with

assignments are reported in Figure 5, for the whole spectra with all chemical shifts see Figure SI22 in the supporting information. 15N CPMAS spectra with assignments are reported in Figure 6. The analysis of the

13

C and

15

N shifts experienced by the signals assigned to the COOH and

N(CH3)2 moieties upon co-crystallization (see Table 1) shows that all samples can be regarded as salts i.e. the hydrogen is transferred from the acid to the base. Indeed, the COOH peaks shift at higher frequencies according to the formation of COO- groups. In a similar way also the N(CH3)2

ACS Paragon Plus Environment

16

Page 17 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Crystal Growth & Design

resonances undergo a high-frequency shift of about 10 ppm upon crystallization, indicating the formation of N+H(CH3)2 groups.

Figure 5. 13C CPMAS NMR spectra (100.6 MHz; spinning speed = 12 kHz) with assignments of a) ven, b) [venH][Hoxalate], c) [venH][salicylate], d) [venH][ferulate], e) [venH][Hfumarate], f) [venH][coumarate] and g) [venH][H2citrate]. Red lines indicate the COOH signal position of the free acids.

ACS Paragon Plus Environment

17

Crystal Growth & Design

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 37

Figure 6. 15N CPMAS NMR spectra (40.5 MHz; spinning speed = 9 kHz) with assignments of a) ven, b) [venH][Hoxalate], c) [venH][salicylate], d) [venH][ferulate], e) [venH][Hfumarate], f) [venH][coumarate] and g) [venH][H2citrate]. The red dotted line indicates the N(CH3)2 signal position of venlafaxine free base. Table

1.

13

C

and

15

N

chemical

shifts

(ppm)

(carboxylic/carboxylate

and

dimethylamino/dimethylammonium signals) in free acids, venlafaxine and in the salts under investigations. 13

[venH][Hoxalate] [venH][Hsalicylate]

free acid COOH 160.6 176.3

15

C Salts COO- COOH 164.5 173.1

venlafaxine free N 3.45 3.45

ACS Paragon Plus Environment

N Salts N+-H 14.3 11.3

18

Page 19 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Crystal Growth & Design

173.5 172.8 172.8 179.7

[venH][Hferulate] [venH][Hfumarate] [venH][Hcoumarate] [venH][H2citrate]

175.4 173.5 174.5 181.7

3.45 3.45 3.45 3.45

17.6 18.9 16.3 15.3

The SSNMR data useful for assisting the XRPD structure solution of [venH][Hfumarate] were obtained from

13

C CPMAS and

neutral character and from

13

14

N-1H J-HMQC which provide information on Z’ and ionic or

C-1H HETCOR (indirect detection) and 1H DQ MAS for probing

packing and conformation. The 13C CPMAS spectrum confirms the presence of one venlafaxine and one fumaric acid in the asymmetric unit. In particular, the two signals at 173.5 and 169.3 ppm attributed to COO- and COOH groups respectively, indicate a partial salification of the acid. Rotor-synchronised 1H-14N J-HMQC experiment (Figure 8a) provides direct evidence of the ionic character, with N+-H⋅⋅⋅O formation. Indeed, this experiment achieves indirect detection of 14

N lineshapes through a combination of J-coupling and residual dipolar splitting (RDS).52,§

Therefore, the presence of a correlation peak between the 1H signal at 11.6 ppm and the 14N peak at -243.4 ppm¥ in the 2D spectrum implies a 1J-coupling and thus a covalent bond. Two

13

C-1H HETCOR (indirect detection) experiments were performed: (i) a short-range

version (see Figure SI-23 in the Supporting Information) highlighting correlations between directly bonded C and H, provided a complete assignment of both 13C and 1H signals (Table SI-3 in Supporting Information); (ii) a long-range version (Figure 7a) enabling the identification of 1

H-13C through-space long range (up to ∼3Å) proximities, gave packing information. The long-

range version presents several inter- and intra-molecular proximities, such as C20-H17, C21H1/H2, NH-C1/C2/C3/C4 that are consistent with both crystal packing and N+-H formation found in the structure solved from XRPD.

ACS Paragon Plus Environment

19

Crystal Growth & Design

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 7. (a)

13

Page 20 of 37

C-1H HETCOR NMR spectrum (150.9 and 600.1 MHz; spinning speed = 70

kHz) of [venH][Hfumarate] acquired with CT1=1.8 ms and CT2=1.0 ms providing long range proximities. Red and blue horizontal lines highlight intra- and intermolecular through space 13C1

H proximities, respectively. (b) [venH][Hfumarate] chemical structure highlighting intra- (red)

and intermolecular (blue) proximities found in the 13C-1H HETCOR spectrum. The 1H DQ MAS NMR spectrum (Figure 8b) is characterized by correlations due to pairs of through-space dipolar coupled protons. The main correlations, as highlighted in Figure 8c, are: N+-H with H1/H2/H3 and OH21 with H17/H19. In particular, the two intermolecular proximities (i.e. OH21-H17/H19) agree with the structure solved from XRPD. The former accounts for the

ACS Paragon Plus Environment

20

Page 21 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Crystal Growth & Design

proximity between the 4-methoxyphenyl venlafaxine moiety and the fumarate, whereas the latter is due to the formation of the inter-anionic hydrogen-bonded chain (see above X-ray part).

Figure 8. (a) 2D 14N-1H J-HMQC (43.4 and 600.1 MHz, spinning speed = 70 kHz) spectrum of [venH][Hfumarate] recorded using texc=trec=1.24 ms. (b) 2D 1H DQ MAS (600.1 MHz, spinning speed = 70 kHz) spectrum of [venH][Hfumarate] recorded using one period of BABA recoupling. Red and blue horizontal lines highlight intra- and intermolecular through space 1H-

ACS Paragon Plus Environment

21

Crystal Growth & Design

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

1

Page 22 of 37

H proximities, respectively. Both spectra were recorded with 1H 90° pulse = 0.77 µs. (c)

[venH][Hfumarate] chemical structure highlighting intra- (red) and intermolecular (blue) proximities found in the 1H DQ MAS spectrum. Concerning [venH][H2citrate], whose X-ray structure was not available,

13

C CPMAS (Figure

5), 13C-1H HETCOR (useful for a complete assignment of both 13C and 1H signals, Figure SI-24. and Table SI-4 in the Supporting Information)) and performed. The

13

14

N-1H J-HMQC (Figure 9) spectra were

C CPMAS spectrum was instrumental to identify one venlafaxine and one

citric acid in the asymmetric unit. The presence of one COO- resonance (181.7 ppm) and two overlapped COOH peaks (177.8 and 177.4 ppm) indicate a partial salification of the acid. Moreover, a correlation between the 1H signal at 8.9 ppm and the 14N peak at -279.8 ppm ¥ in the 2D

14

N-1H J-HMQC spectrum (Figure 9), is clearly indicating the presence of a N+-H group

pointing toward a molecular salt formation.

Figure 9. 2D

14

N-1H J-HMQC spectrum of [venH][H2citrate] (600 MHz, spinning speed = 70

kHz) recorded using texc=trec=0.65 ms.

ACS Paragon Plus Environment

22

Page 23 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Crystal Growth & Design

Intrinsic dissolution rate of pure molecular salts. In view of the high solubility of venlafaxine hydrochloride, the easy and unexpensive preparation of derivatives that, albeit biologically acceptable, would allow a controlled decrease of the solubility and intrinsic dissolution rate, is an attractive task. For this reason the six new compounds described above have been subjected to dissolution tests (see Supporting Information), and the dissolution profiles have been compared with the one obtained for venlafaxine hydrochloride. All compounds show a decrease in solubility with respect to the commercial form of venlafaxine (venlafaxine hydrochloride); this is a positive result for a drug that needs to be slowly released in the body. The best result has been obtained with the coumarate salt, which shows the lowest intrinsic dissolution rate (see Figure SI-25 in the Supporting Information). Table 2 reports melting points and dissolution rate of the new salts; data for venlafaxine hydrochloride are also added for comparison.

Table 2.

Melting points (onset temperature, DSC) and intrinsic dissolution rate for the

venlafaxine molecular salts. Salt [venH][Cl]

m.p. (°C) 211.1

Intrinsic dissolution rate (10-4 mg s-1) 30

[venH][Hoxalate]

162.4

14

[venH][salycilate]

115.5

6

[venH][Hfumarate]

119.0

6

[venH][H2citrate]

164.7

6

[venH][ferulate]

147.5

1

[venH][coumarate]

177.6

0.4

ACS Paragon Plus Environment

23

Crystal Growth & Design

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 24 of 37

On the basis of these data we can divide the new salts in three groups characterized by different

intrinsic

dissolution

rates:

(i)

high

([venH][Hoxalate]),

(ii)

medium

([venH][Hfumarate], [venH][H2citrate] and [venH][salycilate]) and (iii) low ([venH][coumarate] and [venH][ferulate]). For each group one salt was selected; the resulting three salts (coumarate, salicylate and hydrogen oxalate) were in turn used for the preparation of formulated tablets with excipients (see below).

Dissolution profiles of molecular salts in formulations. The three selected salts (coumarate, salicylate and hydrogen oxalate) were used for the preparation of formulated tablets with excipients.14 Dissolution tests were then performed in vitro in accordance to the performance Test 2 reported in the USP Monographs of Venlafaxine Hydrochloride Extended-Release Capsules,17 and the results compared with those obtained for venlafaxine hydrochloride. To this end three different formulations were prepared for each of the selected salts and for venlafaxine hydrochloride, in combination with pharmaceutically acceptable eccipients: tablets compositions, shown in Tables 3 and 4, were prepared both for immediate-release (IR) (one composition, see Table 3) and for extended-release (ER) (two compositions, differing for the amount of the HPMC excipient; see Table 4), as reported in the United States patent for low water-soluble venlafaxine salts.14

ACS Paragon Plus Environment

24

Page 25 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Crystal Growth & Design

Table 3. Tablet formulation for immediate-release (IR).

Ingredients Venlafaxine hydrochloride Venlafaxine coumarate Venlafaxine hydrogen oxalate Venlafaxine salicylate Microcrystalline cellulose Lactose monohydrate Magnesium stearate

mg/tablet 42.40 59.61 49.60 56.10 29.88 16.00 1.00

Table 4. Tablet formulations for extended-release (ER). Two different salt:HPMC weight ratios were used. Ingredients Venlafaxine hydrochloridea Venlafaxine coumarateb Venlafaxine hydrogen oxalatec Venlafaxine salicylated HPMC Microcrystalline cellulose Dibasic calcium phosphate anhydrous Magnesium stearate

mg/tablet 42.40 59.61 49.60 56.10 salt:HPMC=1:1 salt:HPMC=1:1.48 12.00 5.00 1.25

The dissolution profiles are reported in Figure 10 for the IR formulation and in Figures 11 and 12 for the two ER formulations. Formulations for immediate release (IR) show the same dissolution profiles as those containing venlafaxine hydrochloride: the API is released within the first two hours. The difference in the dissolution profiles becomes significant in the extended release formulations; for example, the salycilate shows an improvement in the dissolution profile (about 50% of API dissolved in two hours). The best result is clearly obtained with the coumarate salt in the formulation ER (1:1) (see Figure 12) for which dissolution data fall within the parameters dictated by the TEST 2 for

ACS Paragon Plus Environment

25

Crystal Growth & Design

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 26 of 37

bioequivalence (see Table SI-13)17.

Figure 10. Dissolution profiles for immediate-release (IR) formulations containing: a) [venH][Cl], b) [venH][coumarate], c) [venH][Hoxalate], d) [venH][salicylate].

Figure 11. Dissolution profiles for the extended-release formulations (ER) containing

(a)

[venH][Cl], (b) [venH][coumarate], (c) [venH][Hoxalate] and (d) [venH][salicylate]. In this case the salt:HPMC stoichiometric ratio is 1:1.

ACS Paragon Plus Environment

26

Page 27 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Crystal Growth & Design

Figure 12. Dissolution profiles for the extended release formulations (ER) containing

(a)

[venH][Cl], (b) [venH][coumarate], (c) [venH][Hoxalate] and (d) [venH][salicylate]. In this case the salt:HPMC stoichiometric ratio is 1:1.48.

CONCLUSIONS In this paper we have reported the preparation of a series of novel derivatives of the antidepressant drug venlafaxine. The aim was that of discovering biologically acceptable systems with a more favorable (i.e. less soluble) solubility profile with respect to the available commercial form of venlafaxine hydrochloride. The new compounds are all characterized by the transfer of one proton from the acid to the nitrogen atom of the amino group on the structure of venlafaxine as spectroscopically confirmed by SSNMR measurements (13C and 15N CPMAS and 2D

14

N-1H J-HMQC spectra). More specifically the crystal structures of the coumarate and

ferulate salts display a “chevron-like” arrangement of the organic acid moieties linked on the vertices to the venlafaxine cations. The hydrogen oxalate and the hydrogen fumarate salts also shows formation of an inter-anion hydrogen bonded chain interaction with the venlafaxine cations whereas the salicylate systems shows direct cation-anion interactions and no interanionic chains. Multinuclear SSNMR experiments were fundamental in providing information on [venH][Hfumarate] for which the structure has been solved from XRPD data and on [venH][H2citrate] whose X-ray structure was not available. For both samples,

13

C CPMAS

spectra provided the number of independent molecules in the unit cell (Z’). The molecular salt formation was definitively confirmed by 14N-1H J-HMQC which provides direct evidence of the presence of N+-H group. For [venH][Hfumarate], other advanced 2D SSNMR experiments, such

ACS Paragon Plus Environment

27

Crystal Growth & Design

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 28 of 37

as 13C-1H HETCOR and 1H DQ MAS, provided useful insights on crystal packing and hydrogen bond network, thus supporting and confirming the structure solved by XRPD. In terms of solubility properties the coumarate derivative shows, with respect to both the hydrochloride and the other salts, a significant decrease in intrinsic dissolution profile. In order to put the analysis on a pharmaceutically practical basis of formulation, both immediate-release and extended-release formulations of the tablets have been tested, confirming the favorable behavior of the coumarate derivative in water and sodium cloride solutions. These are preliminary results: further work would of course be necessary to test the dissolution properties in buffered systems, and to assess the exented release performance of the tablets depending on physical properties as porosity, particle size, etc.

ASSOCIATED CONTENT Supporting Information. Crystallographic information files (cif) for the structure described (CCDC 1543887-1543891). The Supporting Information is available free of charge on the ACS Publications website at DOI: xxxxxxx. Crystal data, X-Ray powder patterns, TGA and DSC measurements, solid state NMR spectra, calibration lines for dissolution tests (PDF). AUTHOR INFORMATION Corresponding Authors *Dipartimento di Chimica and NIS Centre, Università di Torino, Via P. Giuria 7, 10125 Torino, Italy. E-mail: [email protected]

ACS Paragon Plus Environment

28

Page 29 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Crystal Growth & Design

*Dipartimento di Chimica Giacomo Ciamician, Università di Bologna, Via F. Selmi 2, 40126 Bologna, Italy. E-mail: [email protected]. The authors declare no competing financial interest. Notes § In case of N⋅⋅⋅H-O interaction, often the RDS alone without J-coupling is not sufficient to provide 14N-1H correlations. ¥ the 14N chemical shifts differ from the 15N chemical shifts due to the quadrupole-induced shift which depends on the quadrupolar coupling constant.

ACKNOWLEDGMENTS M.R.C. and R.G. are indebted with Jeol Company for helpful technical assistance and cooperation. F.R. thanks Jeol (Italia) S.p.A. for a Ph.D. scholarship. The University of Bologna is acknowledged (F.S., D.B. and F.G.) for financial support. The EU COST Action CM1402 “Crystallize” is also acknowledged (M.R.C., R.G., F.S., D.B. and F.G.). REFERENCES (1) Braga, D.; Grepioni, F. Making Crystals by Design – methods, techniques and application. Wyley-VCH, Weinheim, DE. 2007. (2) Aitipamula, S.; Banerjee, R.; Bansal, A. K.; Biradha, K.; Cheney, M. L.; Choudhury, A. R.; Desiraju, G. R.; Dikundwar, A. G.; Dubey, R.; Duggirala, N.; Ghogale, P. P.; Ghosh, S.; Goswami, P. K; Goud, N. R.; Jetti, R. R. K. R.; Karpinski, P.; Kaushik, P.; Kumar, D.; Kumar, V.; Moulton, B.; Mukherjee, A.; Mukherjee, G.; A. S.; Puri, V.; Ramanan,

ACS Paragon Plus Environment

29

Crystal Growth & Design

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 30 of 37

A.; Rajamannar, T.; Reddy, C. M.; Rodriguez-Hornedo, N.; Rogers, R. D.; Guru Row, T. N.; Sanphui, P.; Shan, N.; Shete, G.; Singh, A.; Sun, C. C.; Swift, J. A.; Thaimattam, R.; Thakur, T. S.; Thaper, R. K.; Thomas, S. P.; Tothadi, S.; Vangala, V. R.; Variankaval, N.; Vishweshwar, P.; Weyna, D. R.; Zaworotko, M. J. Polymorphs, Salts, and Cocrystals: What’s in a Name? Cryst. Growth Des. 2012, 12, 2147-2152. (3) Steed, J. W. The role of co-crystals in pharmaceutical design. Trends Pharmacol. Sci. 2013, 34, 3. (4) Braga, D.; Grepioni, F.; Lampronti, G. I.; Maini, L.; Turrina, A. Ionic Co-crystals of Organic Molecules with Metal Halides: A New Prospect in the Solid Formulation of Active Pharmaceutical Ingredients. Cryst. Growth Des. 2011, 11, 5621-5627. (5) Braga, D.; Grepioni, F.; Maini, L.; Prosperi, S.; Gobetto, R.; Chierotti, M. From unexpected reactions to a new family of ionic co-crystals: the case of barbituric acid with alkali bromides and caesium iodide. Chem. Commun. 2010, 46, 7715-7717. (6) Braga, D.; Grepioni, F.; Maini, L.; Capucci, D.; Nanna, S.; Wouters, J.; Aerts, L.; Quéré, L. Combining piracetam and lithium salts: ionic co-crystals and co-drugs? Chem. Commun. 2012, 48, 8219-8221. (7) Samie, A.; Desiraju, G. R.; Banik, M. Salts and Cocrystals of the Antidiabetic Drugs Gliclazide, Tolbutamide, and Glipizide: Solubility Enhancements through Drug– Coformer Interactions. Cryst. Growth Des., DOI: 10.1021/acs.cgd.6b01804. (8) Duggirala, N.K.; Perry, M.L.; Almarsson, O.; Zaworotko, M.J. Pharmaceutical Cocrystals: Along the Path to Improved Medicines. Chem. Commun. 2016, 52, 640-655.

ACS Paragon Plus Environment

30

Page 31 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Crystal Growth & Design

(9) Schultheiss, N.; Newman, A. Pharmaceutical Cocrystals and Their Physicochemical Properties. Cryst. Growth Des. 2009, 9, 2950−2967. (10) Friscić, T.; Jones, W. Benefits of cocrystallisation in pharmaceutical materials science: an update. J. Pharm. Pharmacol. 2010, 62, 1547−1559. (11) Babu, N. J.; Nangia, A. Solubility Advantage of Amorphous Drugs and Pharmaceutical Cocrystals. Cryst. Growth Des. 2011, 11, 2662−2679. (12) Gutierrez, M. A.; Stimmel G. L.; Aiso J. Y. Venlafaxine: a 2003 update. Clin. Ther. 2003, 25, 2138-2154. (13) https://pubchem.ncbi.nlm.nih.gov/compound/venlafaxine#section=ExperimentalProperties. (14) Oosterbaan, M. J. M.; Keltjens, R. Low water-soluble venlafaxine salts. US, 6, 696, 496, B2. 2004. (15) Oosterbaan, M. J. M.; Keltjens, R. Low water-soluble venlafaxine salts. WO 03/082805 A1. 2003. (16) Wright, C. W.; Aikman, M. S.; Werts, E.; Seabolt, J.; Haeusler, J. M. C. Bioequivalence of Single and Multiple Doses of Venlafaxine Extended-Release Tablets and Capsules in the Fasted and Fed States: Four Open-Label, Randomized Crossover Trials in Healthy Volunteers. Clin. Ther. 2009, 31, 2722-2734. (17) Venlafaxine Hydrochloride/ Official Monographs. United States Pharmacopeia 2014, 37-NF 30.

ACS Paragon Plus Environment

31

Crystal Growth & Design

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 32 of 37

(18) Wouters, J.; Quéré, L. Pharmaceutical salts and co-crystals. RSC 2011. (19) van Eupen, J.Th.H.; Westheim, R.; Deij, M.A.; Meekes, H.; Bennema, P.; Vlieg, E. Crystal polymorphism of pharmaceuticals. Int. J. Pharm. 2009, 368, 146–153. (20) Roy, S.; Bhatt, P. M.; Nangia, A.; Kruger, G. J. Stable Polymorph of Venlafaxine Hydrochloride by Solid-to-Solid Phase Transition at High Temperature. Cryst. Growth Des. 2007, 7, 476–480. (21) Roy, S.; Aitipamula, S.; Nangia, A. Thermochemical Analysis of Venlafaxine Hydrochloride Polymorphs 1-5. Cryst. Growth Des. 2005, 5, 2268-2276. (22) Shan, N.; Zaworotko, M. J. The role of cocrystals in pharmaceutical science. Drug Discov. Today 2008, 13, 440−446. (23) Bernardi, L. S.; Ferreira, F. F.; Cuffini, S. L.; Campos, C. E. M.; Monti, G. A.; Kuminek, G.; Oliveira, P. R.; Cardoso, S. G.; Solid-state evaluation and polymorphic quantification of venlafaxine hydrochloride raw materials using the Rietveld method. Talanta 2013, 117,189–195. (24) Rodríguez-Hornedo, N. Cocrystals: Molecular Design of Pharmaceutical Materials. Mol. Pharm. 2007, 4, 299−300. (25) Sheldrick, G. M. SHELX97. Program for Crystal Structure Determination; University of Göttingen: Göttingen, Germany 1997. (26) Macrae, C. F.; Bruno I. J; Chisholm, J. A.; Edgington, P. R.; McCabe, P.; Pidcock, E.; Rodriguez-Monge, L.; Taylor, R.; van de Streek, J.; Wood, P. A. Mercury CSD 2.0 -

ACS Paragon Plus Environment

32

Page 33 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Crystal Growth & Design

New Features for the Visualization and Investigation of Crystal Structures. J. Appl. Crystallogr. 2008, 41, 466. (27) Boultif, A.; Louer, D. Powder pattern indexing with the dichotomy method. J. Appl. Crystallogr. 2004, 37, 724–731. (28) Altomare, A.; Camalli, M.; Cuocci C.; Rizzi R. EXPO2009: structure solution by powder data in direct and reciprocal space J. Appl. Crystallogr. 2009, 42, 1197. (29) Cohelo, A. TOPAS-Academic, Coelho Software, Brisbane, Australia 2007. (30) Young, R. A.; Ed. The Rietveld Method. International Union of Crystallography and Oxford University Press: Oxford 1993. (31) Poojary, D. M.; Clearfield, A. Application of X-ray powder diffraction techniques to the solution of unknown crystal structures. Acc. Chem. Res. 1997, 30, 414-422. (32) Harris, K. D. M.; Tremayne, M.; Kariuki, B. M. Contemporary advances in the use of powder X-ray diffraction for structure determination. Angew. Chem., Int. Ed. Engl. 2001, 40, 1626- 1651. (33) Althaus, S. M.; Mao, K.; Stringer, J. A.; Kobayashi T.; Pruski, M. Indirectly detected heteronuclear correlation solid-state NMR spectroscopy of naturally abundant 15N nuclei. Solid State Nucl. Magn. Reson. 2014, 57–58, 17–21. (34) Nielsen, N. C.; Bildso/e, H.; Jakobsen, H. J.; Lewitt, M.H. Double‐quantum homonuclear rotary resonance: Efficient dipolar recovery in magic‐angle spinning nuclear magnetic resonance. J. Chem. Phys. 1994, 101, 1805–1812.

ACS Paragon Plus Environment

33

Crystal Growth & Design

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 34 of 37

(35) Wiench, J. W.; Bronnimann, C. E.; Y Lin, V. S.; Pruski, M. Chemical Shift Correlation NMR Spectroscopy with Indirect Detection in Fast Rotating Solids:  Studies of Organically Functionalized Mesoporous Silicas. J. Am. Chem. Soc. 2007, 129, 12076– 12077. (36) Dissolution test for solid dosage forms. European Pharmacopeia 5.0 2005, 228-230. (37) Braga, D.; Grepioni, F.; Novoa, J. J. Inter-anion O–H–···O– hydrogen bond like interactions: the breakdown of the strength–length analogy. Chem. Commun. 1998, 19591960. (38) Dunitz, J. D.; Gavezzotti, A.; Rizzato, S. “Coulombic Compression”, a Pervasive Force in Ionic Solids. A Study of Anion Stacking in Croconate Salts. Cryst. Growth Des, 2014, 14, 357–366. (39) Gobetto, R.; Nervi, C.; Valfrè, E.; Chierotti, M.R.; Braga, D.; Maini, L.; Grepioni, F.; Harris, R.K.; Ghi, P.Y. 1H MAS, 15N CPMAS, and DFT Investigation of HydrogenBonded Supramolecular Adducts between the Diamine 1,4-Diazabicyclo-[2.2.2]octane and Dicarboxylic Acids of Variable Chain Length. Chem. Mater. 2005, 17,1457–1466. (40) Grifasi, F.; Chierotti, M.R.; Gaglioti, K.; Gobetto, R.; Maini, L.; Braga, D.; Dichiarante, E.; Curzi, M. Using Salt Cocrystals to Improve the Solubility of Niclosamide. Cryst. Grow. Des. 2015, 15, 1939-1948. (41) Chierotti, M. R.; Gobetto, R. Solid-state NMR studies of weak interactions in supramolecular systems. Chem. Comm. 2008, 1621-1634;

ACS Paragon Plus Environment

34

Page 35 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Crystal Growth & Design

(42) Duthaler, R. O.; Roberts, J. D. Nitrogen-15 nuclear magnetic resonance spectroscopy. Solvent effects on the15N chemical shifts of saturated amines and their hydrochlorides. J. Magn. Reson. 1979, 34, 129. (43) Botto, R.E.; Roberts, J. D. Nitrogen-15 magnetic resonance spectroscopy. Naturalabundance nitrogen-15 spectra of some 2-amino-2-deoxy-D-hexose derivatives. J. Org. Chem. 1977, 42, 2247. (44) Chierotti, M.R.; Gobetto, R. NMR crystallography: the use of dipolar interactions in polymorph and co-crystal investigation. Cryst.Eng.Comm. 2013, 15, 8599-8612. (45) Harris, R. K. NMR crystallography : the use of chemical shifts. Solid State Sci. 2004, 6, 1025-1037. (46) Brown, S.P. Applications of high-resolution 1H solid-state NMR. Solid State Nucl. Magn. Reson. 2012, 41, 1-27. (47) Bacchi, A.; Cantoni, G.; Chierotti, M.R.; Girlando, A.; Gobetto, R.; Lapadula, G.; Pelagatti, P.; Sironi, A.; Zecchini, M. Water vapour uptake and extrusion by a crystalline metallorganic solid based on half-sandwich Ru(II) building-blocks. Cryst. Eng. Comm. 2011, 13, 4365-4375. (48) Braga, D.; Chelazzi, L.; Grepioni, F.; Dichiarante, E.; Chierotti, M.R..; Gobetto, R. Crys. Grow. Des. 2013, 13, 2564-2572. (49) Gumbert, S.D.; Korbitzer, M.; Alig, E.; Schmidt, M.U.; Chierotti, M.R.; Gobetto, R.; Li, X.Z.; van de Streek, J. Dyes Pigm. 2016, 131, 364-372.

ACS Paragon Plus Environment

35

Crystal Growth & Design

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 36 of 37

(50) Reddy, G.N.M.; Cook, D.S.; Iuga, D.; Walton, R.I.; Marsh, A.; Brown, S.P. An NMR crystallography study of the hemihydrate of 2′, 3′-O-isopropylidineguanosine. Solid State Nucl. Magn. Reson. 2015, 65, 41-48. (51) Nishiyama, Y.; Endo, Y.; Nemoto, T.; Utsumi, H.; Yamauchi, K.; Hioka, K.; Asakura, T. Very fast magic angle spinning 1H-14N 2D solid-state NMR: Sub-micro-liter sample data collection in a few minutes. J. Magn. Reson. 2011, 208, 44-48. (52) Cavadini , S.; Antonijevic, S.; Lupulescu, A.; Bodenhausen, G. J. Magn. Reson. 2006, 182, 168–172.

ACS Paragon Plus Environment

36

Page 37 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Crystal Growth & Design

For Table of Contents Use Only

Molecular salts of the antidepressant venlafaxine: an effective route to solubility properties modifications Floriana Spinelli, Elena Dichiarante, Marco Curzi, Stefano L. Giaffreda, Michele R. Chierotti,* Roberto Gobetto, Federica Rossi, Laura Chelazzi, Dario Braga, Fabrizia Grepioni*

Reacting venlafaxine with coumaric, ferulic, oxalic, salicylic, fumaric and citric acids results in the protonation of the amino group and formation of the corresponding 1:1 molecular salts. All compounds were characterized by a combination of diffraction, thermal and spectroscopic solid state techniques. Solid formulations with excipients were tested for dissolution rate, showing that the coumarate salt might represent an improvement for extended release administrations.

ACS Paragon Plus Environment

37