Scholarly article on topic 'Investigating the Influence of Polymers on Supersaturated Flufenamic Acid Cocrystal Solutions'

Investigating the Influence of Polymers on Supersaturated Flufenamic Acid Cocrystal Solutions Academic research paper on "Chemical sciences"

0
0
Share paper
Academic journal
Molecular Pharmaceutics
OECD Field of science
Keywords
{""}

Academic research paper on topic "Investigating the Influence of Polymers on Supersaturated Flufenamic Acid Cocrystal Solutions"

molecular

pharmaceutics

flAjtl

vv Penn

Libraries

Subscriber access provided by University of Pennsylvania Libraries

Article

Investigating the Influence of Polymers on Supersaturated Flufenamic Acid Cocrystal Solutions

Minshan Guo, Ke Wang, Noel Anthony Hamill, Keith Lorimer, and Mingzhong Li

Mol. Pharmaceutics, Just Accepted Manuscript • DOI: 10.1021/acs.molpharmaceut.6b00612 • Publication Date (Web): 05 Aug 2016

Downloaded from http://pubs.acs.org on August 7, 2016

Just Accepted

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

Molecular Pharmaceutics 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.

8 Investigating the Influence of Polymers on

12 Supersaturated Flufenamic Acid Cocrystal Solutions

16 112 2 1

17 Minshan Guo , Ke Wang , Noel Hamill, Keith Lorimer and Mingzhong Li *

20 1School of pharmacy, De Montfort University, Leicester, UK

24 Almac Science, Seagoe Industrial Estate, Craigavon, UK

10 11 12

20 21 22

Table of contents graphic

Abstract

The development of enabling formulations is a key stage when demonstrating the effectiveness

understanding of the nucleation and crystal growth is important. In this study, the influence of

10 of pharmaceutical cocrystals to maximize the oral bioavailability for poorly water soluble drugs.

12 Inhibition of drug crystallization from a supersaturated cocrystal solution through a fundamental

17 the three polymers of polyethylene glycol (PEG), polyvinylpyrrolidone (PVP) and a copolymer

19 of N-vinly-2-pyrrodidone (60%) and vinyl acetate (40%) (PVP-VA) on the flufenamic acid

22 (FFA) crystallization from three different supersaturated solutions of the pure FFA and two

24 cocrystals of FFA-NIC CO and FFA-TP CO has been investigated by measuring nucleation

26 induction times and desupersaturation rates in the presence and absence of seed crystals. It was 28

29 found that the competition of intermolecular hydrogen bonding among drug/coformer,

31 drug/polymer and coformer/polymer was a key factor responsible for maintaining

36 solution. The supersaturated cocrystal solutions with predissolved PEG demonstrated more

38 effective stabilization in comparison to the pure FFA in the presence of the same polymer. In

43 better performance than the pure FFA with the same predissolved polymer. The study suggests

45 that the selection of a polymeric excipient in a cocrystal formulation should not be solely

48 dependent on the interplay of the parent drug and polymer without considering the coformer

50 effects.

54 KeyWords: Cocrystal; polymers; Flufenamic Acid; crystal growth; nucleation; supersaturation.

supersaturation through nucleation inhibition and crystal growth modification in a cocrystal

contrast, neither of the two cocrystal solutions, in the presence of PVP or PVP-VA, exhibited a

4 Introduction

5 Development of supersaturating drug delivery systems to enhance oral bioavailability of poorly

8 water soluble drugs has been of interest for many decades 1. In these systems, two essential steps

10 need to be considered: the drug in a high energy form, e.g. amorphous forms, crystalline salts or

12 cocrystals, should dissolve rapidly to generate a high concentration above the saturation

15 solubility and then this supersaturated solution must be maintained for a reasonable period to

17 allow for significant absorption and eventually sufficient bioavailability. This has been referred

20 to as a "spring and parachute" approach . As a supersaturated drug solution is

22 thermodynamically unstable and has the tendency to return to the equilibrium state through drug

24 crystallization, extensive work has been carried out to delay the drug crystallization by inclusion

27 of different excipients as effective crystallization inhibitors in formulations . For example,

29 significant progress has been made in amorphous solid dispersion formations by using polymeric

31 crystallization inhibitors to maintain the solid drug in an amorphous state and also maintain the

33 drug supersaturation after dissolution 4' 5. It has been found that inhibition of the drug

36 crystallization is a result of the polymers interfering in the nucleation and/or crystal growth

38 stages of the more stable phase, through physical or chemical interactions between the drug and

41 polymer excipients, such as; solution viscosity enhancement, non-specific hydrophobic drug-

43 polymer interactions and specific drug-polymer intermolecular interactions through hydrogen

46 bonding 6-12.

48 Compared with amorphous solid forms, the crystalline forms of the drug substances are

50 generally preferred in a formulation because of their thermodynamic stability and purity.

55 to their ability to modulate the physicochemical properties of a drug compound to overcome any

57 solubility limited bioavailability problem 13-16. Similar to the amorphous solid forms, those

Pharmaceutical cocrystals have therefore attracted significant attention over the last decade due

10 11 12

20 21 22

cocrystals with improved solubility and dissolution rates become thermodynamically unstable once dissolved due to supersaturation of the drug . This results in precipitation of stable solid phases of the parent drugs and reduction of the solubility advantage of the cocrystal 17-19. In order

to achieve the full potential of cocrystals, rational strategies are required that identify the

appropriate crystallization inhibitors of polymers and/or surfactants in cocrystal formulations -

. In comparison with the amorphous solid dispersion systems, in which the supersaturated solution behavior is determined by the ternary drug/polymer/solvent interaction, the complexity of a cocrystal supersaturated solution increases considerably due to inclusion of an additional component of a coformer. This can interfere with the drug molecule, polymeric excipients, and/or solvent, resulting in alteration of the inhibition ability of the polymers on the drug. It is not surprising that inclusions of excipients of polymers and surfactants in the indomethacin or carbamazepine cocrystal formulations have not shown effectiveness in capturing the enhanced

solubility advantage ' . Although research has demonstrated that a combination of a cocrystal of celecoxib-nicotinamide or danazol-vanillin with both a polymer and surfactant can provide an enhanced dissolution rate and a high oral bioavailability, there is no mechanistic understanding

of how these additives interact with the drug molecules in solution ' . Therefore, it is of huge importance to investigate the role of polymeric excipients as potential crystallization inhibitors for rational design of cocrystal formulation systems.

In this work, for the first time, a systematic investigation was conducted to explore the impact of different polymeric additives in cocrystal formulations to elucidate the molecular mechanism of polymer/drug/coformer interactions that affect the kinetics of nucleation and growth of the parent drug. In the study, Flufenamic acid form I (FFA I) was selected as a parent model drug along with two coformers of Nicotinamide (NIC) and Theophylline (TP). This was due to their

4 ability to form FFA-NIC cocrystals (FFA-NIC CO) and FFA-TP cocrystals (FFA-TP CO), both

6 of which display different physicochemical properties 26' 27 FFA, a nonsteroidal anti-

8 inflammatory drug (NSAID), has the problem of low bioavailability after oral administration due

10 26 28

11 to its low solubility " . Among its nine reported polymorphs, FFA I (white color) and FFA III

13 (yellow color) have been used in the commercial solid dosage forms . Three chemically diverse

15 polymers including polyethylene glycol (PEG), polyvinylpyrrolidone (PVP) and copolymer of

18 vinyl pyrrolidone/vinyl acetate (PVP-VA) were selected because they have been widely used as

3 30 31

20 crystallization inhibitors in other supersaturating drug delivery systems ' ' . Among these

23 polymers, PEG is the most hydrophilic, containing a high percentage of hydrogen donors . In

25 comparison to PVP, more hydrophobic PVP-VA, containing 40% acetate side chains, was used

27 to investigate the specific intermolecular interaction with the drug and/or coformers. The

32 coformers and polymers. Chemical structures of the model drug, coformers and monomer units

34 of the polymers are shown in Table 1.

37 Equilibrium solubility tests were first carried out to evaluate the potential role of polymers in

39 changing the apparent FFA solubility in solution. A solvent shift method was then used to

41 generate an initial FFA supersaturation condition to study crystallization kinetics of both

solubility parameter was calculated for comparison of the hydrophobicity of the model drug,

' w 33

44 nucleation and growth . Induction time determined by polarized light microscopy was used to

46 quantify the drug nucleation from a supersaturated solution in the absence and presence of

51 by measuring desupersaturation curves in the presence of the seeds of the pure FFA I crystals.

53 The overall impact of polymers on inhibiting FFA crystallization from a supersaturated solution

different pre-dissolved polymers. The impact of different polymers on growth was characterized

was characterized and evaluated by measuring the desupersaturation curves in the absence of the

4 б б 7 В

10 11 12

14 1б 1б 17 1В

1g 20 21 22

24 2б 2б 27 2В

34 Зб

37 ЗВ

44 4б

47 4В 4g б0 б1 б2

64 бб бб б7 бВ

crystal seeds. In order to quantify inhibition ability of polymers to prolong drug supersaturation, supersaturation parameters in different supersaturated solutions were calculated and compared . The solid residues after the solubility and the desupersaturation experiments with and without seeds were examined by differential scanning calorimetry (DSC), X-ray powder diffraction (XRPD), Fourier transform infrared spectroscopy (FTIR) and polarized light microscopy. To further explore the intermolecular interaction mechanisms among a polymer, drug and coformer, solution infrared spectra of the parent drug FFA I and coformers of NIC and TP in combination with different polymers were collected and compared.

Materials and methods Materials

Flufenamic acid form I (FFA I), Nicotinamide (NIC) (>99.5% purity) and Theophylline (TP) (>99.5% purity) were purchased from Sigma-Aldrich (Dorset, UK). Poly (ethylene glycol) 4000 (PEG) was purchased from Sigma-Aldrich (Dorset, UK). Plasdone K-29/32 (PVP) and Plasdone S-630 (PVP-VA) were gifts from Ashland Inc. (Schaffhausen, Switzerland). Methanol (HPLC grade) and ethanol (lab grade) were purchased from Fisher Scientific UK (Loughborough, UK) and used as received. Double distilled water was generated from a Bi-Distiller (WSC044.MH3.7, Fistreem International Limited, Loughborough, UK) and used throughout the study.

Methods

Preparation of FFA-NIC and FFA-TP cocrystals

Flufenamic acid and Nicotinamide cocrystal (FFA-NIC CO) was prepared by a solvent evaporation method. A 1:1 equimolar mixture of FFA I and NIC was dissolved in acetonitrile

4 with stirring at 80°C. The solution was placed in a fume cabinet overnight for solvent

6 evaporation. Flufenamic acid and Theophylline cocrystal (FFA-TP CO) was synthesized by a

8 cooling crystallization method. A 1:1 molar ratio of FFA and TP was dissolved in a cosolvent

11 (7:3 acetonitrile and water) with stirring at 90°C and then the solution was placed into an ice bath

13 for 2 h until the crystals were separated out from the solution. Both FFA-NIC CO and FFA-TP

15 CO were characterized and confirmed by DSC, FTIR and XRPD.

19 Apparent equilibrium solubility determination

22 The apparent equilibrium solubility of FFA I, FFA-NIC CO and FFA-TP CO was determined by

24 suspending an excess amount of crystalline materials in small vials with 20mL of the cosolvent

26 (1:4 ethanol and water) in the absence or presence of 0.2 mg/mL of a pre-dissolved polymer 28

29 (PEG, PVP or PVP-VA). This mixture was kept at 37 ± 0.5°C with shaking (150RPM) for 24 h.

31 The supernatant was separated from excess solids in solution by MSE Micro Centaur at

34 13000RPM for 1 min in a MSB 010.CX2.5 centrifuge (MSE Ltd, London, UK). Subsequently,

36 the supernatant was diluted and the concentrations of FFA and coformers were determined using

38 a high-performance liquid chromatography system (HPLC). The solid residues were retrieved

43 cosolvent of 1:4 ethanol and water was used in this study to increase the apparent FFA

45 equilibrium solubility and thus avoid immediate crystallization of FFA in media through

48 maintaining slower kinetics of nucleation and growth. All experiments were conducted in

50 triplicate and data were reported as an average concentration in solution.

54 Monitoring nucleation induction time using polarized light

55 microscopy

from the tests, dried for 24h at ambient temperature and analyzed by DSC, FTIR and XRPD. The

4 Nucleation induction times were determined from desupersaturation experiments monitored using

6 polarized light microscopy. The FFA stock concentration of pure FFA I, FFA-NIC CO or FFA-TP CO

8 dissolved in ethanol was 1 mg/mL. Different initial supersaturated solutions of 50, 100 and 200 ^g/mL

10 were generated by adding the appropriate amount of the stock solution into a small quartz cell filled with

12 0.5mL of the cosolvent in the absence or presence of 0.2 mg/mL of different polymers. The FFA

14 crystallization behavior from a supersaturated solution was monitored by a LEICA DM 750 polarizing

microscope (Leica Microsystem Ltd, Milton Keynes, UK) with a 200x or 100x objective and recorded

supersaturated solution of pure FFA, FFA-NIC CO or FFA-TP CO in the absence and presence of

19 using a version 4.0 studio capture. Data collection commenced immediately after addition of the drug

21 stock solution to the test medium. The induction time was determined by observing the onset of the FFA

23 crystal formation.

27 Effect of polymers on the supersaturated FFA and cocrystal solutions

29 In order to decouple the nucleation process, the inhibition effect of a polymer on the growth of FFA

31 crystals was assessed from the seeded experiments by measuring the desupersaturation curve of a

36 0.2mg/mL of a pre-dissolved polymer (PEG, PVP or PVP-VA). 50mg of FFA I crystal seeds, which were

38 slightly ground and sieved by a 60 (size) mesh sieve, were added to 50 mL of the cosolvent medium and

40 allowed to equilibrate at 37°C for 24 h. A supersaturated solution was generated by adding 0.3 mL of a 5

42 mg/mL FFA stock solution of pure FFA, FFA-NIC CO or 0.6 mL of 2.5mg/mL FFA-TP CO. The amount

44 of ethanol added to the medium was small and had a negligible impact on the apparent FFA equilibrium

49 5, 15, 30, 60, 120 and 240 min. The supernatant was separated from excess solids by centrifugation at

51 13000rpm for 1 min in a MSE Micro Centaur. The supernatant was diluted to determine the

53 concentrations of FFA and coformer of NIC or TP by HPLC. In order to evaluate the overall inhibition

55 effect of a polymer on FFA crystallization kinetics from a supersaturated solution, unseeded

57 desupersaturation experiments were conducted. A supersaturated solution was generated by adding 20 mL

solubility. 1 mL of each sample was withdrawn from the solution at six predetermined time intervals, i.e.

4 of 500 |g/mL FFA stock solution of pure FFA, FFA-NIC CO or FFA-TP CO to 80 mL of water, resulting

6 in 100 |g/mL FFA in the cosolvent of 1:4 ethanol and water. The solid residues after the

8 desupersaturation experiments were examined by DSC, FTIR and polarized light microscopy.

10 Supersaturation parameters were calculated and compared for both seeded and unseeded experiments to

12 quantify different polymer inhibition abilities .

15 All experiments were conducted in triplicate and data were reported as the average of the experiments.

18 High Performance Liquid Chromatography (HPLC) analysis

21 The sample concentration of FFA, NIC or TP in solution was determined by a Perkin Elmer series 200

23 HPLC system (PerkinElmer Ltd, Beaconsfield, UK). A HAISLL 100 C18 column (5 pm, 250 x 4.6 mm)

25 (Higgins Analytical Inc., Mountain View, CA, USA) was used at ambient temperature. FFA was detected

27 by UV absorbance detection at a wavelength of 286 nm. The mobile phase used consisted of 15% water

29 (including 0.5% formic acid) and 85% methanol and the mobile phase flow rate was maintained at 1.5

34 mobile phase was composed of 55% methanol and 45% water, and the mobile phase flow rate was kept at

36 1 mL/min. The injection volume was 20 pL.

40 Differential Scanning Calorimetry (DSC)

42 The melting point of solids was measured by a PerkinElmer Jade DSC (PerkinElmer Ltd., Beaconsfield,

44 UK) controlled by Pyris Software. The temperature and heat flow of the instrument were calibrated using

46 indium and zinc standards. A test sample (8-10mg) was analyzed in crimped aluminum pan with a pin-

49 hole pierced lid. Measurements were carried out at a heating rate of 20°C/min under a nitrogen flow rate

mL/min. Both NIC and TP were detected by UV absorbance detection at a wavelength of 265 nm, the

of 20mL/min.

54 X-ray powder diffraction (XRPD)

Fourier transform infrared spectroscopy (FTIR)

4 X-ray powder diffraction pattern of solids was recorded from 5o to 50o at a scanning rate of 0.3° (20) min-1

6 by D2 PHASER diffractometer (Bruker UK Limited, Coventry, UK). Cu-Kp radiation was used with a

8 voltage of 30KV and current of 10 mA.

10 11 12

14 FTIR spectra of the solid samples were measured using an ALPHA interferometer (Bruker UK Limited,

16 Coventry, UK) with a horizontal universal attenuated total reflectance (ATR) accessory. Samples were

18 placed on the surface of the diamond ATR plate and the ATR assembly was clamped to ensure good

20 21 22

23 The investigation of the intermolecular interaction among FFA, NIC, TP and polymers (PEG, PVP and

25 PVP-VA) in solution was carried out by FTIR. Solution spectra were collected using the same

27 spectrometer fitted with a transmission accessory and the Bruker 6500S Circular Aperture liquid cell with

29 size of 32x3 mm CaF2 window. The path length was 0.05mm. Methanol was selected for the

31 intermolecular interaction study of FFA, NIC and polymers, in which the concentrations of individual

contact.

components were 50, 21.7 and 20 mg/mL respectively. A cosolvent of 1M HCl and methanol at a ratio of

36 1:6 was selected for the intermolecular interaction study of FFA, TP and polymers, in which the

38 concentrations of individual components were 14.3, 9.14 and 20 mg/mL respectively.

40 In each measurement, 30 scans were collected per spectrum with a resolution of 2 cm-1 in the spectral

42 region of 400 to 4000 cm-1 using OPUS software. All the spectral data were collected at an ambient

44 temperature, between 20 to 23°C.

48 Solubility parameter (SP, supersaturation ratio (SR) and

49 supersaturation parameter (SSP)

52 Solubility parameter (SP) is used to compare the relative hydrophobicity of polymers, FFA and

54 coformers in solution. The SP of an organic compound is estimated by Fedors 34 as

57 sp= ipä= ^ (1)

58 v ^ w

where the Ae; and Avi are the additive atomic and group contributions to the energy of vaporization and

6 molar volume, respectively. AEV is the energy of vaporization at a given temperature and V is the

8 corresponding molar volume which is calculated from the known values of molecular weight and density.

10 The method is based on group additive constants; therefore it requires only knowledge of the structural

12 formula of the compound. Based on the group contributions provided in the literature 34, SP values of the

14 polymers and drug compounds used in the study are shown in Table 1. Details of calculation of SP for

each compound can be found in Table S1 in the supplementary materials. Supersaturation ratio (SR) in this study is defined as

C0 is higher than its equilibrium solubility Ceq. Line C0C0(t) represents an ideal situation where the drug

21 sr = 7L (2) 22 Leq

23 where C is the solute concentration and Ceq is the solute equilibrium solubility.

25 Supersaturation parameter (SSP) is used to evaluate the drug precipitation behavior from a

27 supersaturated system in comparison to a reference system based on the work by Chen et al 8. Fig. 1

29 shows the desupersaturation curves of supersaturated drug systems, in which the initial drug concentration

34 remains in the medium and no crystallization occurs over the time period of t. The curve C0CR(t) is the

36 desupersaturation curve of a reference system. An integration area of Aq^^q^ can be used to indicate

38 the amount of drug precipitated from the solution over time t. For a supersaturated system with the

41 desupersaturation curve of C0Ca(t), the integration area of AQ(jQa^Qo^ is smaller than that of the

43 reference system, indicating less drug precipitation. Compared with the reference system, a supersaturated

45 system with the desupersaturation curve of C0Cb(t) has more precipitated drug solids because of a larger

47 integration area of AQoQb^t^Qo^t-j. To compare the abilities of different systems on maintaining the drug

supersaturation, SSP is defined as

SSp _ ACoCR(t)c0(t)-Ac0c(t)c0(t) x ioo%0 (3)

AC0CR(t)C0(t)

Solid characterization of FFA I, NIC, TP, and FFA cocrystals

4 where C(t) is the desupersaturation curve of an investigated system. SSP is a dimensionless parameter. A

6 system with a positive SSP value has an increased ability to prolong drug supersaturation while a negative

8 SSP value indicates less ability to maintain the supersaturated drug in solution.

11 Results

18 Fig. 2(a) shows the XRPD patterns of individual components of FFA I, NIC, TP, FFA-NIC CO and FFA-

20 TP CO. The significant characteristic diffraction peaks of FFA I are at 20=7.1°, 14.2o, 21.4o and 24.6o.

22 Key characteristic diffraction peaks of NIC are at 20=14.9o and 23.5o. After co-evaporation of FFA and

25 NIC in acetonitrile, the new materials of FFA-NIC CO have been formed, showing the characteristic

26 diffraction peaks at 20=6.7o, 9.6 16.2o, 16.8o and 21.9o ,which are in agreement with those of published 28 26

29 data . The characteristic diffraction peaks of TP are at 20=7.2o, 12.7o and 14.5o. Through the cooling

31 crystallization method described in Section 2, FFA-TP CO was generated, indicated by the characteristic

33 diffraction peaks at 20=5.9o, 11.3o, 15.6o and 26.8o 27.

35 The structures of FFA-NIC CO and FFA-TP CO have been confirmed by the measured IR spectra in

37 Fig. 2(b) 26' 27. FFA-NIC CO is formed through an acid-pyridine heterosynthon involving FFA and NIC

42 C=O stretching frequencies 35. The spectrum of NIC has 2 peaks at 3353 cm-1 and 1592 cm-1,

44 corresponding to N-H and pyridine ring C=N stretching 36. In the spectrum of FFA-NIC CO, the

45 -1 -1

46 frequencies of N-H stretching and C=O stretching of FFA are shifted to 3324 cm-1 and 1660 cm-1 while

48 the peaks of N-H stretching and pyridine ring C=N stretching of NIC shifted to 1608 cm-1 from 1592 cm-1

50 and to 3395 cm-1 from 3353 cm-1. FFA-TP CO is formed through an O-H--O hydrogen bond involving

53 the carboxylic acid of FFA and unsaturated N atom of the imidazole ring of TP . The IR spectrum of TP

54 -1 -1 -1

55 has peaks at 3119 cm , 1660 cm and 1561 cm , corresponding to N-H, C=O and C=N stretching

56 -1 -1 -1

57 frequencies which are shifted to 3068cm-1, 1669 cm-1 and 1558 cm-1 respectively. In the spectrum of FFA-

molecules 26. The IR spectrum of FFA I has peaks at 3318 cm-1 and 1651 cm-1, corresponding to N-H and

Fig.3(a) demonstrates the apparent equilibrium solubility of FFA I, FFA-NIC CO and FFA-TP

TP CO, FFA's N-H stretching and C=O stretching frequencies are shifted to 3280cm-1 and 1647 cm

6 respectively. The summary of IR peak identities of FFA I, NIC, TP and cocrystals are shown in Table S2

8 in the supplementary materials.

10 The DSC curve in Fig. 2(c) shows that the melting point of FFA-NIC CO was 136.4 oC which

12 was higher than both of the melting points of FFA I (133.5 oC) and NIC (128.1 oC). In contrast,

15 the melting point of FFA-TP CO was 185.6 oC which was in the middle of the melting points of

17 FFA I and the TP melting point at 272.8oC.

21 Apparent FFA equilibrium solubility of FFA I, FFA-NIC CO and FFA-TP

24 CO in cosolvent in the absence and presence of different polymers

30 CO in cosolvent media in the absence or presence of predissolved polymers of PVP, PEG and

32 PVP-VA at equilibrium after 24 h. In the absence of a polymer, the apparent FFA equilibrium

35 solubility of FFA-NIC CO (41.9±2.1 pg/mL) was slightly higher than those of FFA I and FFA-

37 TP CO which were comparable (36.0±0.5 pg/mL for FFA-TP CO and 36.8±2.1 pg/mL for the

39 pure FFA I). In the presence of 200pg/mL polymer, PEG, PVP or PVP-VA, the apparent FFA

42 equilibrium solubility of FFA I or FFA cocrystals does not change, indicating that none of the

44 polymers changed the solution properties.

47 The solid residues collected after the solubility tests were analyzed by DSC in Fig.3(b). For

49 pure FFA I, the resultant solid residues were the same as the starting materials after the solubility

51 test in the absence or presence of polymers, indicated by identical DSC thermographs in Fig.

56 polymers, the solid residues formed were yellow FFA III crystals, indicating the cocrystals of

3(b). Following the solubility tests of FFA-NIC CO and FFA-TP CO in presence or absence of

FFA-NIC CO or FFA-TP CO had transformed into FFA III. This was confirmed by DSC

6 thermographs of the solid residues in Fig. 3(b), in which the same thermal events occurred as that

8 of the pure FFA III. Under DSC heating conditions, FFA III melted at 123.1°C and recrystallized

11 to FFA I which then melted at 134.4°C . However, the morphologies of FFA III particles

13 collected from the two cocrystal tests in Fig. 3(c) were significantly different. The FFA III

15 crystals from FFA-NIC CO tests were needle-shaped, whereas those from the FFA-TP CO tests

18 were rod/disc-shaped. FTIR data of the solid residues are shown in Fig. S1 in the supplementary

20 materials.

24 Effect of polymers on the nucleation induction time of FFA

27 crystallization in solution

33 solutions of 50, 100 and 200 pg/mL were corresponding to the SR values of 1.36, 2.72 and 5.44

35 respectively. The nucleation induction times in Table 2 were based on the initial observation

38 times of FFA crystals detectable by polarized light microscopy. Without a predissolved polymer

40 in the cosolvent media, the precipitation of FFA from the pure FFA and two cocrystal solutions

42 occurred rapidly at the low SR of 1.36. The induction times were significantly different in the

45 presence of different polymers, PEG, PVP and PVP-VA. With predissolved PEG in solution, the

47 induction times were increased slightly for all test solutions at the low SR of 1.36. No FFA

52 PVP-VA at a SR 1.36, indicating that PVP or PVP-VA can completely inhibit the crystallization

54 of FFA during the 30 min experiment. In order to differentiate the inhibition abilities of PVP and

57 PVP-VA, the experiments were conducted with a higher initial degree of supersaturation SR

Based on the measured equilibrium solubility of FFA I in Section 3.2, the initial supersaturated

crystals were found for all test solutions in the presence of 200 pg/mL of pre-dissolved PVP or

Fig.4 shows the images of a representative part of the quartz cell, demonstrating the

4 2.72. From the recorded images, it was observed that dense liquid particles appeared

6 immediately in the experiments with the predissolved PVP or PVP-VA and then the formation of

8 the crystal nuclei within the dense liquid clusters followed. It is in excellent agreement with the

11 two-step mechanism of nucleation of crystals in solution . A video clip of FFA crystals

13 nucleation from a supersaturated FFA-NIC CO solution in the presence of predissolved PVP-VA

15 can be found in the supplementary materials. In the presence of pre-dissolved PVP in solution,

18 the order of the induction times was TFFA-TP co < TFFA-NIC co < TFFA. In contrast, PVP-VA can

20 completely inhibit the crystallization of FFA from the three test solutions. Further tests were

23 conducted at the supersaturation level of SK=5.44 with predissolved PVP-VA. It was shown that

25 the induction times were comparable for the two cocrystal solutions, with the longest induction

27 time being 446 s for the pure FFA solution.

32 morphology of the FFA crystals after tests. In cosolvent without a pre-dissolved polymer, the

34 needle shape morphology of FFA crystals from both the FFA-NIC CO and pure FFA solution

37 was similar. In contrast, the FFA crystals from the FFA-TP CO solution were significantly

39 smaller and rod-shaped. In the presence of PEG in solution, the FFA crystals precipitated from

41 the three test samples became smaller. In the presence of PVP or PVP-VA in solution, all

44 crystals precipitated from three test solutions were a similar shape, lacking any distinctive crystal

46 morphology.

56 is represented by the normalized value of Cnorm (t) which is the ratio of the measured FFA concentration

58 via the initial FFA concentration in solution as

Effect of polymers on the FFA crystal growth in solution

Due to the variation of the initial FFA concentrations in the seeded solutions, the desupersaturation curve

the seeded solution without adding the stock solution and Cstock is the FFA concentration of the stock solution.

crystals, with NIC being more effective at 12% of SSP. PEG can slightly reduce the growth rate of the

4 Cnorm (0 = r j-r (4)

^ -o + -stock

5 where C(t) is the measured FFA concentration at sampling time t, CQ is the initial FFA concentration in

12 Fig. 5 shows the desupersaturation curves of the different test samples. The gradient of a FFA

14 desupersaturation curve is directly related to bulk growth rate of FFA crystals in solution. Without a

16 polymer, the growth rate of FFA crystals of the FFA-NIC CO and FFA-TP CO solutions was slower than

18 that of the pure FFA solution, indicating the coformer of NIC or TP can inhibit the growth of FFA

20 21 22

23 FFA crystals in the pure FFA solution with 4 % of SSP. In contrast, PEG reduced the inhibition ability of

25 NIC for the growth of FFA crystals in FFA-NIC CO solution in which SSP was reduced to 1% from 12%

27 shown in Fig. 5(e). Surprisingly, both PVP and PVP-VA were ineffective in inhibiting the growth of FFA

29 crystals and instead accelerated the FFA crystal growth rates, indicating that the FFA concentrations in

31 solution quickly decreased to the equilibrium solubility, shown in Figs. 5(c)-(d). With pre-dissolved PVP,

36 the FFA-TP solution. In the presence of PVP-VA in solution, the crystal growth rates in cocrystal

38 solutions SSP values, -27% in the FFA-NIC CO solution and -24% in FFA-TP solution, were faster than

40 that of the pure FFA solution, SSP of-13%. DSC thermographs and images of the solids isolated from the

42 experiments were exactly the same as that of initial seeds of FFA I, shown in Fig. S2 in the

44 supplementary materials. However, when closely examining the FTIR data of the solids collected in Fig.

47 5(f), it was found that a shift of the carbonyl peak of FFA I at 1651 cm was observed in all the

49 experiments, suggesting that a coformer or polymer was integrated in the solids.

the SSP dropped to -17% in the pure FFA solution, to -28% in the FFA-NIC CO solution and to -12% in

4 The overall polymer inhibition ability on maintaining FFA

7 supersaturation

11 The overall effect of a polymer on inhibition of FFA crystallization from a supersaturated

13 solution was evaluated by unseeded desupersaturation experiments in the absence or presence of

15 200 pg/mL of a pre-dissolved polymer of PEG, PVP or PVP-VA, as described in the previous

18 Section. The initial FFA concentration was 100 pg/mL corresponding to SR=2.72.

20 Fig. 6 shows the desupersaturation curves of the different test samples. It can be seen that the

22 FFA concentrations from different test samples decreased rapidly in the cosolvent media without

25 a pre-dissolved polymer in Fig. 6(a). The FFA-NIC CO and FFA-TP CO solutions show a

27 comparable performance in which the rate of desupersaturation was slower than that of the pure

29 FFA solution. The FFA concentrations in all three test solutions were reduced to the same static

32 level of 42 pg/mL within 2 h, which was slightly higher than its solubility. In the pre-dissolved

34 PEG media, the decreasing rates of the supersaturated FFA concentrations in the FFA-NIC CO

37 and FFA-TP CO solutions are significantly slower in comparison with that of the pure FFA

39 solution, showing an increased SSP of 13.4% for FFA-NIC CO solution, 12.2% for FFA-TP and

41 just 3.2% for the pure FFA solution in Fig. 6(b). Among the three solutions with pre-dissolved

44 PVP, Fig. 6(c) demonstrates that PVP is the effectively inhibitor for the pure FFA solution as

46 seen by a 15.9 % increase in inhibition of FFA. Compared with PEG, PVP has a reduced ability

48 on maintaining FFA in either the FFA-NIC CO or FFA-TP CO solutions. PVP-VA pre-dissolved

51 in solution can significantly reduce the rate of the FFA precipitation from both supersaturated

53 FFA and FFA-NIC CO solutions, showing 27.4% and 26.4% increases of SSP values in Figs.

55 (d)-(e). However, there is no difference between PVP and PVP-VA in maintaining the

58 supersaturated FFA in FFA-TP CO solution in Fig. 6(e).

peak of FFA in methanol was found at 1686cm-1, indicating C=O stretching 35. When a component of

4 The solids precipitated from all of the experiments were yellow needle-shape FFA III crystals

6 confirmed by the DSC results and images in Fig. S3 in the supplementary materials. The FTIR data

8 showed that a shift of the carbonyl peak of FFA III at 1655 cm was observed in all the experiments,

10 suggesting that a coformer or polymer was coprecipitated in the solids in Fig. 6(f).

14 IR spectroscopic investigation of intermolecular interactions among

17 FFA, coformer and polymer in solution

20 Fig. 7 shows the comparison of the solution IR spectra of individual components of FFA, NIC, TP and

22 mixtures of FFA and coformers in the absence and presence of different polymers. In Fig. 7(a) a strong

27 NIC, PVP or PVP-VA was added in the solution, this FFA characteristic peak was shifted to a smaller

29 wavelength number of 1684cm , indicating an intermolecular interaction between them in solution. In

31 contrast, there is no change in the FFA C=O peak in the PEG solution, suggesting no interaction between

33 these two components. NIC can interact with FFA or PVP in solution, demonstrated by a change in the

35 characteristic peak of NIC at 1625cm-1, corresponding to N-H stretching 36, to 1617 cm-1 in the presence

00 of FFA and to 1631 cm-1 in the presence of PVP in Fig. 7(b). Surprisingly there is no interaction between

40 NIC with PVP-VA or PEG in solution, confirmed by no change in the characteristic peak of NIC at

42 1625cm-1. The IR characteristic peak of TP at 925cm-1, corresponding to N-H symmetric stretching 39, has

44 been shifted to a lower wavenumber by inclusion of PVP or PVA-VA and to a higher wavenumber by

46 adding PEG or FFA in solution, indicating TP can interact with any of components, FFA, PEG, PVP or

48 PVP-VA in solution.

Effect of a polymer on the apparent FFA equilibrium solubility of FFA I, FFA-NIC CO and FFA-TP CO in cosolvent

4 Discussion

10 11 12

14 There is widespread acceptance that the crystalline nature of pharmaceutical cocrystals can offer

15 advantages over amorphous materials to formulate drug compounds with limited solubility and

18 bioavailability, due to superior thermodynamic stability and purity. Although significant advances in

20 design and discovery have been made, little work has been conducted to formulate cocrystals into a drug

22 product. Therefore, the behavior of a cocrystal in a formulated product is largely unknown. In order to

24 offer the most desired in vivo performance with the highest bioavailability for many life-saving drugs

26 with poor biopharmaceutical properties, a fundamental understanding of the critical factors that control

29 the dissolution and absorption performance of a cocrystal formulated product is required. In this work, the

41 focus was on understanding the parent drug crystallization kinetics from a supersaturated cocrystal

44 solution in the presence of a polymeric excipient. It aimed to provide the mechanistic understanding of

45 the properties of a polymer as a good inhibitor of crystallization for a given drug cocrystal. Two FFA

47 cocrystals, FFA-NIC CO and FFA-TP CO, were chosen due to significant differences in their

49 physicochemical properties. The low polymer concentration of 200 ^g/mL used in the investigation was

41 based on the rational consideration of a 500 mg tablet containing 250 mg of stabilizing polymer, in which

44 20% of the polymer was released in 250 mL of the GI tract at the beginning stage of dissolution.

46 According to the equilibrium solubility results in Fig. 3(a), the FFA concentrations of FFA I, FFA-NIC

48 CO or FFA-TP CO were constant in solution in the absence and presence of 200 ^g/mL polymer of PEG,

50 PVP or PVP-VA, indicating that the impact of a polymer on FFA crystallization in a supersaturated

52 solution was not caused by a change in the level of supersaturation. Furthermore, due to the low

57 essentially the same as that of the dissolution medium without a predissolved polymer. Therefore, the

molecular weight of the polymer used in the study, the viscosity of the 200 ^g/mL polymer solution was

4 interplay of API-coformer, API-polymer, and coformer-polymer elucidated in this study was not affected

6 by the changes of the solution bulk properties of solubility and mass transport.

9 Effect of intermolecular interactions of drug/coformer, drug/polymer

12 and coformer/polymer on parent drug nucleation and growth kinetics

15 in solution

19 This study has clearly demonstrated that a cocrystal coformer can interfere with a polymer to alter its

21 ability to maintain the parent drug superstation in solution. This property involves both nucleation and

23 growth through competition of the intermolecular interactions of drug/coformer, drug/polymer and

25 coformer/polymer in solution

27 In the solid state, cocrystals are formed through hydrogen bonding between an API and coformer. Once

30 the cocrystals are dissolved in solution, they could be regarded as completely separate individual

32 molecules. For example, the US FDA has elected to classify cocrystals within their framework as

34 dissociable "API-excipient" molecular complexes. However, in this study it was found that the hydrogen

36 bonds between FFA and coformers, NIC or TP, were not broken completely, indicated by the changes in

38 the their characteristic peaks of the solution spectra in Fig. 7(a). This API/coformer interaction in solution

40 certainly affected the formation of nuclei by hindering the reorganization of a cluster of FFA molecules

43 into its ordered structure. Therefore, the coformer of NIC or TP can be regarded as a nucleation inhibitor

45 for FFA crystallization, generating slightly longer induction time (15 s for FFA-NIC CO solution and 24 s

47 for FFA-TP CO solution) compared with the pure FFA supersaturated solution (9 s of induction time) in

49 the absence of a polymer, as shown in Table 2.

51 The FFA molecule, shown in Table 1, has the very strong hydrogen bond donor of O-H combined with

53 a middle strength acceptor of C=O, thus displaying higher hydrophobicity with a low value of SP (18.62

56 MPa1/2). Therefore, FFA self-association should be disrupted by a polymer with strong acceptor groups

58 that can effectively compete with the FFA acceptor group C=O . Indeed, the formation of the hydrogen

4 bonding between the polymer of PVP or PVP-VA with FFA in solution was demonstrated by the IR

6 spectroscopic investigation in Fig. 7(a), as both polymers (N-C=O in PVP and N-C=O and O-C=O in

8 PVP-VA) have strong acceptors. This suggested that both PVP and PVP-VA were able to act as effective

10 nucleation inhibitors, indicated by the significantly increased nucleation induction times at different

12 degrees of supersaturation. A higher level of inhibition effectiveness of PVP-VA in comparison with PVP

14 was due to the presence of carbonyl oxygens C=O on the side chain which contributed to a more

hydrophobic nature and flexibility to interact with FFA molecules in solution. Therefore, evidence for a

19 two-step mechanism of cocrystal nucleation was revealed in the presence of PVP-VA. The precipitated

21 solids in Fig. 4 show a lack of birefringence under polarized light and no distinct particle morphology,

24 indicating the amorphous nature of the particles was due to the integration of PVP or PVP-VA in the FFA

25 crystal structure and/or rapid desupersaturation. The amorphous nature of the precipitated particles could

27 be also related to liquid-liquid phase separation (LLPS) which was observed in amorphous solid

29 41 42

dispersion systems (ASDs) in recent publications ' . The high supersaturation generated by ASDs can

42 lead to a two phase system wherein one phase is an initially nano-dimensioned drug-rich phase and the

44 other is a drug-lean continuous aqueous phase. In those studies the stronger nucleation inhibitors PVP/

46 PVP-VA allowed the system to reach supersaturation levels such that the system underwent LLPS. The

48 excess drug then precipitated forming a dispersed, colloidal amorphous drug-rich phase which resulted in

40 the absence of birefringence in the precipitated particles.

The ineffectiveness of PEG as a nucleation inhibitor was probably due to its structural rigidity in which

45 the hydrogen acceptor, C-O-C, on the main chain had been prevented from interacting with FFA

47 molecules in solution. Thus no change was observed in the characteristic peak of FFA in solution with the

49 predissolved PEG (Fig. 7(a)). The limited inhibition ability of PEG may be due to the steric barrier for the

51 formation of nuclei via the adsorption of the polymer on the surface of pre-nuclear clusters 43. It has to be

53 stressed that although all three polymers of PEG, PVP and PVP-VA interacted with FFA with different

mechanisms, they were all integrated into the FFA crystal lattices, showing as a variation of the FFA III

58 characteristic peak at 1655 cm-1, which corresponds to its C=O stretching frequency in Fig. 6(f). The

4 results are in good agreement with previous studies which have shown that multicomponent molecular

6 complexes in solution lead to a metastable form precipitating preferentially 44' 45. A change in the crystal

8 morphologies, seen in Fig S3 in the supplementary materials, also supported this.

10 It was not surprising that the nucleation induction time was reduced for FFA-NIC CO solution in the

12 presence of PVP compared to the pure FFA solution because the competition between NIC and FFA with

14 PVP weakened the polymer inhibition ability. There was no interaction between NIC with PVP-VA in

solution shown in Fig. 7(b). Therefore, the nucleation induction time from the FFA-NIC CO solution was

19 almost the same as that of the pure FFA solution, in the presence of PVP-VA. As TP can interact with

21 both polymers of PVP and PVP-VA in solution, the nucleation induction time reduced in the FFA-TP CO

23 solution compared to the pure FFA solution in the presence of the polymers, as shown in Table 2. PEG

25 inhibits FFA crystallization using a different mechanism in comparison with PVP or PVP-VA, for reasons

27 outlined above. The nucleation induction time increased in both the FFA-NIC CO and FFA-TP CO

32 polymer on FFA.

34 In order to study the effectiveness of the polymers on inhibiting FFA crystal growth after nucleation,

36 desupersaturation experiments were conducted including the addition of the FFA seeds. A low SR of 1.27

38 (based on 36.6 pg/mL of the solubility of FFA I measured in this study) was used in the growth

40 experiments to avoid secondary nucleation. It was observed that polymer effectiveness at reducing crystal

solutions in the presence of PEG due to the accumulated inhibition effects of both the coformer and

growth rates was not found to have a similar impact on nucleation. In the nucleation induction time study,

45 PVP and PVP-VA were effective nucleation inhibiting agents in the pure FFA solution. In contrast, they

47 were poor at inhibiting growth and actually accelerated the growth of FFA crystal seeds, as seen by the

49 negative values of SSP in Fig. 5(e). It is known that the alteration of crystal growth by additives can be

51 achieved through modifying the step speed or altering the step edge energy, which is classified as step

53 pinning, incorporation, kink block, and step edge adsorption mechanisms 46. To occur, the additives must

58 interactive forces responsible for the adsorption of additive molecules on the solid surface including

be adsorbed on the surface of the crystals to block active crystal growth sites. There are a number of

4 electrostatic, hydrogen bonding and hydrophobic interactions. The electrostatic force was not considered

6 because of the neutral natures of the solution and drug components used in this study.

8 The coformer molecules of NIC or TP in a cocrystal solution were most likely adsorbed on the FFA

10 crystal surface due to hydrogen bonding attraction as the growth rate inhibitor. This led to moderately

12 increased SSP values of 12% for FFA-NIC CO solution and 5% for FFA-TP CP solution, as shown in

14 Fig. 5(e). In the pure FFA solution with the predissolved PEG, hydrogen bonding was not promoted

between PEG and FFA as shown in the IR spectroscopic investigation in Fig. 7(a). Therefore,

19 hydrophobic interaction was the main interactive force to drive PEG molecules to be adsorbed on the

21 surfaces of the FFA crystal seeds. A large difference in their SP values in Table 1 suggests a weak

23 interactive force between FFA and PEG in solution. It was not surprising that PEG was neither an

25 effective FFA nucleation nor growth inhibitor. The decrease in the growth inhibition in the FFA-NIC CO

27 solution in the predissolved PEG can most likely be the reduced NIC, being a more effective inhibitor in

comparison to PEG, when adsorbed on the solid surface due to competition by PEG for the same

32 adsorption sites. In the FFA-TP CO solution in the presence of PEG, there was no noticeable change in

34 the extent of growth inhibition as both TP and PEG were equally effective on an individual basis shown

36 in Fig. 5(e).

38 In the pure FFA solution in the presence of PVP or PVP-VA, acceleration of crystal growth occurred,

40 indicated by the negative SSP values of -17% for PVP and -13% for PVP-VA. Similar phenomena were

47 growth . However, in this study the enhanced crystal growth was not likely to be caused by the reduced

49 interfacial tension between the crystal and solution due to the polymer adsorption. It is known that strong

51 intermolecular hydrogen bonding was occurring between FFA and PVP or PVP-VA in solution. When the

53 polymer molecules were adsorbed on the surface of the FFA seeds, the bound FFA molecules were driven

found in other studies when one or more surfactants were predissolved in the solution, in which it was believed that the adsorbed additives could lead to a decrease in interfacial tension to be favorable to

around the FFA seeds, leading to increase local supersaturation at the surface and contributing to the

5 8 acceleration in crystal growth.

4 In the FFA-NIC CO solution with the predissolved PVP or PVP-VA, acceleration of crystal growth was

6 enhanced in comparison to the pure FFA solution in the presence of the same polymer. In the FFA-TP CO

8 solution with the predissolved PVP, acceleration of crystal growth was reduced in contrast to PVP-VA

10 where growth was promoted. These results demonstrated that the combination of PVP or PVP-VA in the

12 presence of either coformer (NIC or TP) can either enhance or reduce the rate of the crystal growth.

14 Overall, the effect was to accelerate the growth. Thus rational selection of a polymer is required to

enhance the inhibition ability in a cocrystal supersaturation solution.

19 The comparison of the overall desupersaturation profiles of three supersaturation solutions in the

21 absence and presence of a polymer of PEG, PVP, or PVP-VA is given in Fig. 7. In the absence of a

23 polymer, a cocrystal solution showed a better performance to maintain the FFA in solution in comparison

25 to the pure FFA solution due to the enhanced combination effects of the nucleation and growth inhibition

27 abilities of the coformers. In the predissolved PEG, a cocrystal solution showed an increased ability to

maintain supersaturation for extended time periods, which was most likely due to the enhanced

32 combination effects of the individual nucleation and growth inhibition abilities of the coformer and PEG.

34 Clearly the polymer nucleation inhibition effect outweighed its growth acceleration ability for FFA in

36 solution, indicating that the rate of desupersaturation was reduced dramatically in the presence of PVP or

38 PVP-VA. The desupersaturation behavior of FFA cocrystal solutions in the presence of PVP or PVP-VA

40 depends on different interaction mechanisms of the polymer and coformer and on the competition effect

of the polymer and coformer for formation of hydrogen bonding with FFA molecules, in which PVP-VA

45 was a good crystallization inhibitor for FFA-NIC CO solution.

47 It is worth noting that the study has shown the coformers and polymers have been integrated in the

49 solid particles recovered from the seed and unseeded experiments based on the measured IR spectra.

51 However, the IR data cannot quantify the relative proportion of co-former/polymer co-precipitated in the

53 desupersaturated solids, which could be determined by solid state NMR or other techniques. In the

meantime, more fundamental research is required to guide the selection of polymers in co-crystal

58 formulation systems through understanding the parent drug crystallization kinetics.

pharmaceutical cocrystals and is applied to maximize the oral bioavailability for poorly water soluble drugs. Inhibition of the drug crystallization from a supersaturated cocrystal solution through a

3 Conclusions

5 Development of enabling formulations is a key stage when demonstrating the effectiveness of

12 fundamental understanding of the nucleation and crystal growth is important. In this study, the influence

14 of the three polymers PEG, PVP and PVP-VA on the FFA crystallization in three different supersaturated

16 solutions of the pure FFA and two cocrystals of FFA-NIC CO and FFA-TP CO has been investigated by

18 measuring nucleation induction times and desupersaturation rates in the presence and absence of seed

20 crystals. It was found that the competition of intermolecular hydrogen bonding among drug/coformer,

drug/polymer and coformer/polymer was a key factor responsible for maintaining the supersaturation

25 through nucleation inhibition and crystal growth modification in a cocrystal solution. The supersaturated

27 cocrystal solutions with predissolved PEG demonstrated effectiveness at stabilizing supersaturated

29 solution compared to pure FFA in the presence of the same polymer. In contrast, the two cocrystal

31 solutions in the presence of PVP or PVP-VA did not perform as well as pure FFA with the same

33 predissolved polymer. The study suggested that the selection of a polymeric excipient in a cocrystal

formulation should not be solely dependent on the interplay of the parent drug and polymer without

48 considering the coformer effects.

41 Associated Content

44 Supporting Information

46 1) Additional tables of SP values of FFA, NIC, TP, and polymers;

49 2) Summary of IR peak identities of FFA I, NIC, TP, FFA-NIC CO and FFA-TP CO

52 3) Additional figures of the FTIR results of solid residues after the solubility testing of FFA, FFA-NIC

54 CO and FFA-TP CO;

References

4 4) test results of solids collected after the seeded and unseeded desupersaturation experiments including

6 DSC results and images.

9 Author Information

11 Corresponding author

14 School of Pharmacy, De Montfort University, Leicester, LE1 9BH, UK. Tel: +44(0) 1122577132; E-mail address:

16 mli@dmu.ac.uk (M. Li)

19 Acknowledgments

22 The authors would like to thank Dr. Simon Roberts from Ashland Specialty Ingredients for

24 providing materials for this study.

34 1. Brouwers, J.; Brewster, M. E.; Augustijns, P., Supersaturating Drug Delivery Systems: The Answer

35 to Solubility-Limited Oral Bioavailability? Journal of Pharmaceutical Sciences 2009, 98 (8), 2549-2572.

39 2. Guzman, H. R.; Tawa, M.; Zhang, Z.; Ratanabanangkoon, P.; Shaw, P.; Gardner, C. R.; Chen, H.;

40 Moreau, J.-P.; Almarsson, O.; Remenar, J. F., Combined use of crystalline salt forms and precipitation

41 inhibitors to improve oral absorption of celecoxib from solid oral formulations. Journal of Pharmaceutical

42 Sciences 2007, 96 (10), 2686-2702.

46 3. Warren, D. B.; Benameur, H.; Porter, C. J. H.; Pouton, C. W., Using polymeric precipitation

47 inhibitors to improve the absorption of poorly water-soluble drugs: A mechanistic basis for utility. Journal of

48 Drug Targeting 2010, 18 (10), 704-731.

52 4. Crowley, K. J.; Zografi, G., The Effect of Low Concentrations of Molecularly Dispersed

53 Poly(Vinylpyrrolidone) on Indomethacin Crystallization from the Amorphous State. Pharm Res 2003, 20 (9),

54 1417-1422.

6. Alonzo, D. E.; Raina, S.; Zhou, D.; Gao, Y.; Zhang, G. G. Z.; Taylor, L. S., Characterizing the Impact of Hydroxypropylmethyl Cellulose on the Growth and Nucleation Kinetics of Felodipine from Supersaturated Solutions. Crystal Growth & Design 2012, 12 (3), 1538-1547.

7. Trasi, N. S.; Abbou Oucherif, K.; Litster, J. D.; Taylor, L. S., Evaluating the influence of polymers on nucleation and growth in supersaturated solutions of acetaminophen. CrystEngComm 2015, 17 (6), 12421248.

8. Chen, Y.; Liu, C.; Chen, Z.; Su, C.; Hageman, M.; Hussain, M.; Haskell, R.; Stefanski, K.; Qian, F., Drug-Polymer-Water Interaction and Its Implication for the Dissolution Performance of Amorphous Solid Dispersions. Molecular Pharmaceutics 2015, 12 (2), 576-589.

9. Liu, C.; Chen, Z.; Chen, Y.; Lu, J.; Li, Y.; Wang, S.; Wu, G.; Qian, F., Improving Oral Bioavailability of Sorafenib by Optimizing the "Spring" and "Parachute" Based on Molecular Interaction Mechanisms. Molecular Pharmaceutics 2016, 13 (2), 599-608.

3 5. DiNunzio, J. C.; Miller, D. A.; Yang, W.; McGinity, J. W.; Williams, R. O., Amorphous

4 Compositions Using Concentration Enhancing Polymers for Improved Bioavailability of Itraconazole.

5 Molecular Pharmaceutics 2008, 5 (6), 968-980.

10 11 12

20 21 22

34 10. Ilevbare, G. A.; Liu, H.; Edgar, K. J.; Taylor, L. S., Impact of Polymers on Crystal Growth Rate of

35 Structurally Diverse Compounds from Aqueous Solution. Molecular Pharmaceutics 2013, 10 (6), 2381-2393.

39 11. Ilevbare, G. A.; Liu, H.; Edgar, K. J.; Taylor, L. S., Maintaining Supersaturation in Aqueous Drug

40 Solutions: Impact of Different Polymers on Induction Times. Crystal Growth & Design 2013, 13 (2), 740-751.

44 12. Ilevbare, G. A.; Liu, H.; Edgar, K. J.; Taylor, L. S., Understanding Polymer Properties Important

45 for Crystal Growth Inhibition—Impact of Chemically Diverse Polymers on Solution Crystal Growth of

46 Ritonavir. Crystal Growth & Design 2012, 12 (6), 3133-3143.

50 13. Qiao, N.; Li, M.; Schlindwein, W.; Malek, N.; Davies, A.; Trappitt, G., Pharmaceutical cocrystals:

51 An overview. International Journal of Pharmaceutics 2011, 419 (1-2), 1-11.

55 14. Blagden, N.; Berry, D. J.; Parkin, A.; Javed, H.; Ibrahim, A.; Gavan, P. T.; De Matos, L. L.; Seaton,

56 C. C., Current directions in co-crystal growth. New Journal of Chemistry 2008, 32 (10), 1659-1672.

16. Duggirala, N. K.; Perry, M. L.; Almarsson, O.; Zaworotko, M. J., Pharmaceutical cocrystals: along the path to improved medicines. Chemical Communications 2016, 52 (4), 640-655.

17. Qiao, N.; Wang, K.; Schlindwein, W.; Davies, A.; Li, M., In situ monitoring of carbamazepine-nicotinamide cocrystal intrinsic dissolution behaviour. European Journal of Pharmaceutics and Biopharmaceutics 2013, 83 (3), 415-426.

15. Blagden, N.; Coles, S. J.; Berry, D. J., Pharmaceutical co-crystals - are we there yet? CrystEngComm

4 2014, 16 (26), 5753-5761.

10 11 12

20 18. Qiu, S.; Li, M., Effects of coformers on phase transformation and release profiles of carbamazepine cocrystals in hydroxypropyl methylcellulose based matrix tablets. International Journal of Pharmaceutics

22 2015, 479 (1), 118-128.

26 19. Greco, K.; Bogner, R., Solution-Mediated Phase Transformation: Significance During Dissolution

27 and Implications for Bioavailability. Journal of Pharmaceutical Sciences 2012, 101 (9), 2996-3018.

31 20. Li, M.; Qiu, S.; Lu, Y.; Wang, K.; Lai, X.; Rehan, M., Investigation of the Effect of Hydroxypropyl

32 Methylcellulose on the Phase Transformation and Release Profiles of Carbamazepine-Nicotinamide

33 Cocrystal. Pharm Res 2014, 31 (9), 2312-2325.

37 21. Li, M.; Qiao, N.; Wang, K., Influence of Sodium Lauryl Sulfate and Tween 80 on Carbamazepine-

38 Nicotinamide Cocrystal Solubility and Dissolution Behaviour. Pharmaceutics 2013, 5 (4), 508.

42 22. Childs, S. L.; Kandi, P.; Lingireddy, S. R., Formulation of a Danazol Cocrystal with Controlled

43 Supersaturation Plays an Essential Role in Improving Bioavailability. Molecular Pharmaceutics 2013, 10 (8),

44 3112-3127.

48 23. Remenar, J. F.; Peterson, M. L.; Stephens, P. W.; Zhang, Z.; Zimenkov, Y.; Hickey, M. B.,

49 Celecoxib:Nicotinamide Dissociation: Using Excipients To Capture the Cocrystal's Potential. Molecular

50 Pharmaceutics 2007, 4 (3), 386-400.

54 24. Alhalaweh, A.; Ali, H. R. H.; Velaga, S. P., Effects of Polymer and Surfactant on the Dissolution and

55 Transformation Profiles of Cocrystals in Aqueous Media. Crystal Growth & Design 2014, 14 (2), 643-648.

26. Fabian, L.; Hamill, N.; Eccles, K. S.; Moynihan, H. A.; Maguire, A. R.; McCausland, L.; Lawrence, S. E., Cocrystals of Fenamic Acids with Nicotinamide. Crystal Growth & Design 2011, 11 (8), 3522-3528.

27. Aitipamula, S.; Wong, A. B. H.; Chow, P. S.; Tan, R. B. H., Cocrystallization with flufenamic acid: comparison of physicochemical properties of two pharmaceutical cocrystals. CrystEngComm 2014, 16 (26), 5793-5801.

25. Qiu, S.; Lai, J.; Guo, M.; Wang, K.; Lai, X.; Desai, U.; Juma, N.; Li, M., Role of polymers in solution

4 and tablet-based carbamazepine cocrystal formulations. CrystEngComm 2016, 18 (15), 2664-2678.

10 11 12

20 28. Pedersen, S. B., Biopharmaceutical Aspects of Tolfenamic Acid. Pharmacology & Toxicology 1994,

20 75, 22-32.

25 29. Lopez-Mej'as, V.; Kampf, J. W.; Matzger, A. J., Nonamorphism in Flufenamic Acid and a New „6 Record for a Polymorphic Compound with Solved Structures. Journal of the American Chemical Society 2012,

27 134 (24), 9872-9875.

31 30. Vandecruys, R.; Peeters, J.; Verreck, G.; Brewster, M. E., Use of a screening method to determine

32 excipients which optimize the extent and stability of supersaturated drug solutions and application of this

33 system to solid formulation design. International Journal of Pharmaceutics 2007, 342 (1-2), 168-175.

37 31. Tian, F.; Saville, D. J.; Gordon, K. C.; Strachan, C. J.; Zeitler, J. A.; Sandler, N.; Rades, T., The

38 influence of various excipients on the conversion kinetics of carbamazepine polymorphs in aqueous

39 suspension. Journal of Pharmacy and Pharmacology 2007, 59 (2), 193-201.

43 32. Chen, J.; Spear, S. K.; Huddleston, J. G.; Rogers, R. D., Polyethylene glycol and solutions of

44 polyethylene glycol as green reaction media. Green Chemistry 2005, 7 (2), 64-82.

48 33. Yamashita, T.; Ozaki, S.; Kushida, I., Solvent shift method for anti-precipitant screening of poorly

49 soluble drugs using biorelevant medium and dimethyl sulfoxide. International Journal of Pharmaceutics 2011,

50 419 (1-2), 170-174.

54 34. Fedors, R. F., A method for estimating both the solubility parameters and molar volumes of liquids.

55 Polymer Engineering & Science 1974, 14 (2), 147-154.

36. Ramalingam, S.; Periandy, S.; Govindarajan, M.; Mohan, S., FT-IR and FT-Raman vibrational spectra and molecular structure investigation of nicotinamide: A combined experimental and theoretical study. Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 2010, 75 (5), 1552-1558.

37. Alvarez, A. J.; Singh, A.; Myerson, A. S., Polymorph Screening: Comparing a Semi-Automated Approach with a High Throughput Method. Crystal Growth & Design 2009, 9 (9), 4181-4188.

35. Jabeen, S.; Dines, T. J.; Leharne, S. A.; Chowdhry, B. Z., Raman and IR spectroscopic studies of

4 fenamates - Conformational differences in polymorphs of flufenamic acid, mefenamic acid and tolfenamic

5 acid. Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 2012, 96, 972-985.

10 11 12

21 38. Vekilov, P. G., The two-step mechanism of nucleation of crystals in solution. Nanoscale 2010, 2 (11),

21 2346-2357.

26 39. Baranska, M.; Schulz, H., Chapter 4 Determination of Alkaloids through Infrared and Raman

27 Spectroscopy. In The Alkaloids: Chemistry and Biology, Geoffrey, A. C., Ed. Academic Press: 2009; Vol.

27 Volume 67, pp 217-255.

32 40. Van Eerdenbrugh, B.; Taylor, L. S., An ab initiopolymer selection methodology to prevent

33 crystallization in amorphous solid dispersions by application of crystal engineering principles.

34 CrystEngComm 2011, 13 (20), 6171-6178.

38 41. Raina, S. A.; Zhang, G. G. Z.; Alonzo, D. E.; Wu, J.; Zhu, D.; Catron, N. D.; Gao, Y.; Taylor, L. S.,

39 Enhancements and Limits in Drug Membrane Transport Using Supersaturated Solutions of Poorly Water

40 Soluble Drugs. Journal of Pharmaceutical Sciences 2014, 103 (9), 2736-2748.

44 42. Indulkar, A. S.; Gao, Y.; Raina, S. A.; Zhang, G. G. Z.; Taylor, L. S., Exploiting the Phenomenon of

45 Liquid-Liquid Phase Separation for Enhanced and Sustained Membrane Transport of a Poorly Water-

46 Soluble Drug. Molecular Pharmaceutics 2016 13 (6), 2059-2069.

50 43. Hendriksen, B. A.; Grant, D. J. W.; Meenan, P.; Green, D. A., Crystallisation of paracetamol

51 (acetaminophen) in the presence of structurally related substances. Journal of Crystal Growth 1998, 183 (4),

52 629-640.

56 44. Lou, B.; Bostrom, D.; Velaga, S. P., Polymorph Control of Felodipine Form II in an Attempted

57 Cocrystallization. Crystal Growth & Design 2009, 9 (3), 1254-1257.

10 11 12

20 21 22

45. Thomas, L. H.; Wales, C.; Zhao, L.; Wilson, C. C., Paracetamol Form II: An Elusive Polymorph through Facile Multicomponent Crystallization Routes. Crystal Growth & Design 2011, 11 (5), 1450-1452.

46. Yoreo, J. J. D.; Vekilov, P. G., Principles of crystal nucleation and growth. 2003; Vol. 54.

47. Ilevbare, G. A.; Liu, H.; Edgar, K. J.; Taylor, L. S., Effect of Binary Additive Combinations on Solution Crystal Growth of the Poorly Water-Soluble Drug, Ritonavir. Crystal Growth & Design 2012, 12 (12), 6050-6060.

10 11 12

20 21 22

48 /10

Table 1: Structure and SP values of FFA, NIC, TP, and polymers

FFA NIC TP PEG PVP PVP-VA

Molecular structure F °Yoh 'Vtb O^N^n CH3 H0-f-CH2CH20)^-H --CH2-CH-k (V ^y.0 °YCHs

SP (MPa1/2) 18.62 29.39 30.21 21.94 21.24 20.98

Table 2: Nucleation induction time

Cosolvent Cosolvent with predissolved PEG Cosolvent with predissolved PVP Cosolvent with predissolved PVP-VA

SR=1.36 FFA 9± 2(sec) 176 ±37 (sec) No crystal appeared No crystal appeared

FFA-NIC CO 15±7 (sec) 288± 172(sec) No crystal appeared No crystal appeared

FFA-TP CO 24± 10(sec) 218± 161(sec) No crystal appeared No crystal appeared

SR=2.72 FFA N/A N/A 658±47 (sec) No crystal appeared

FFA-NIC CO N/A N/A 555±93 (sec) No crystal appeared

FFA-TP CO N/A N/A 510±166 (sec) No crystal appeared

SR=5.44 FFA N/A N/A N/A 446±73 (sec)

FFA-NIC CO N/A N/A N/A 392± 93(sec)

FFA-TP CO N/A N/A N/A 397± 63(sec)

ACS Paragon Plus Environment

10 11 12

20 21 22

Figure. 1: Illustration of supersaturation parameter

10 11 12

20 21 22

Figure. 2: Characterization of solid samples: (a) XRPD patterns; (b) IR spectra; (c) DSC thermographs

10 11 12

20 21 22

10 11 12

20 21 22

Figure. 3: Solubility test results: (a) apparent equilibrium solubility; (b) DSC results of solid residues; (c) images of solid residues

FFA-NIC CO

FFA-TP CO

A. PVP-VA

/V PVP

/ Cosolvent

A/V PVP-VA

/V pvp

/Vy Cosolvent

D 20 40 60 80 100 120 140 160 180 200

Temperature °C

0 20 40 60 80 100 120 140 160 180 200

Temperature °C

10 11 12

20 21 22

Starting materials

Cosolvent

Cosolvent with

predissolved

Cosolvent with

predissolved

Cossolvent with

predissolved

PVP-VA

FFA-NIC CO

_W dtf

FFA-TP CO

10 11 12

20 21 22

48 /10

Figure. 4: Images of FFA crystals after induction time tests

Cosolvent Cosolvent with predissolved PEG Cosolvent with predissolved PVP Cosolvent with predissolved PVP-VA

S=1. 37 FFA

FFA-NIC CO

FFA-TP CO

S=2. 74 FFA Hi

ACS Paragon Plus Environment

10 11 12

20 21 22

Figure. 5: Seeded desupersaturation curves in the absence or presence of polymers: (a) cosolvent; (b) Cosolvent with predissolved PEG; (c) Cosolvent with predissolved PVP; (d) Cosolvent with predissolved PVP-VA; (e) Comparison of supersaturation parameters; (f) FTIR data of solids

10 11 12

20 21 22

4 б б 7 В

10 11 12

14 1б 1б 17 1В

20 21 22

24 2б 2б 27 2В

34 Зб

37 ЗВ

44 4б

47 4В

50 б1 б2

64 бб бб б7 бВ б9 60

Figure. 6: Unseeded desupersaturation curves in the absence or presence of polymers: (a) cosolvent; (b) Cosolvent with predissolved PEG ; (c) Cosolvent with predissolved PVP; (d) Cosolvent with predissolved PVP-VA; (e) Comparison of supersaturation parameters; (f) FTIR data of solids

■gso

4j PEG

FFAsolibilty

100 150 200

Time (min)

10 11 12

20 21 22

10 11 12

20 21 22

Figure. 7: IR spectroscopic investigation of molecular interaction in solution: (a) FFA interaction with NIC and polymers; (b) NIC interaction with FFA and polymers; (c) TP interaction with FFA and polymers