Crystal Morphology Modification by the Addition of Tailor-Made

Article pubs.acs.org/molecularpharmaceutics

Crystal Morphology Modification by the Addition of Tailor-Made Stereocontrolled Poly(N-isopropyl acrylamide) Tommy Munk,† Stefania Baldursdottir,*,† Sami Hietala,‡ Thomas Rades,†,§ Sebastian Kapp,§ Markus Nuopponen,‡ Katriina Kalliomak̈ i,‡ Heikki Tenhu,‡ and Jukka Rantanen† †

Department of Pharmacy, Faculty of Health and Medical Sciences, University of Copenhagen, Denmark Laboratory of Polymer Chemistry, Department of Chemistry, University of Helsinki, Finland § School of Pharmacy, University of Otago, New Zealand ‡

ABSTRACT: The use of additives in crystallization of pharmaceuticals is known to influence the particulate properties critically affecting downstream processing and the final product performance. Desired functionality can be build into these materials, e.g. via optimized synthesis of a polymeric additive. One such additive is the thermosensitive polymer poly(N-isopropyl acrylamide) (PNIPAM). The use of PNIPAM as a crystallization additive provides a possibility to affect viscosity at separation temperatures and nucleation and growth rates at higher temperatures. In this study, novel PNIPAM derivatives consisting of both isotactic-rich and atactic blocks were used as additives in evaporative crystallization of a model compound, nitrofurantoin (NF). Special attention was paid to possible interactions between NF and PNIPAM and the aggregation state of PNIPAM as a function of temperature and solvent composition. Optical light microscopy and Raman and FTIR spectroscopy were used to investigate the structure of the NF crystals and possible interaction with PNIPAM. A drastic change in the growth mechanism of nitrofurantoin crystals as monohydrate form II (NFMH-II) was observed in the presence of PNIPAM; the morphology of crystals changed from needle to dendritic shape. Additionally, the amphiphilic nature of PNIPAM increased the solubility of nitrofurantoin in water. PNIPAMs with varying molecular weights and stereoregularities resulted in similar changes in the crystal habit of the drug regardless of whether the polymer was aggregated or not. However, with increased additive concentration slower nucleation and growth rates of the crystals were observed. Heating of the crystallization medium resulted in phase separation of the PNIPAM. The phase separation had an influence on the achieved crystal morphology resulting in fewer, visually larger and more irregular dendritic crystals. No proof of hydrogen bond formation between PNIPAM and NF was observed, and the suggested mechanism for the observed dendritic morphology is related to the steric hindrance phenomenon. PNIPAM can be used as a crystallization additive with an obvious effect on the growth of NF crystals. KEYWORDS: crystallization, crystal morphology, polymer, thermosensitivity, drug−additive interaction, poly(N-isopropyl acrylamide)

■

INTRODUCTION Crystallization of an active pharmaceutical ingredient (API) is a critical manufacturing step determining the downstream processability of any pharmaceutical solid dosage form. The solid state form of the API and the crystal morphology can often be controlled by the applied crystallization conditions,1 e.g. by optimizing solvent composition, temperature, pH and ionic strength and by strategic selection of additives. Modification of crystal properties in a controlled manner is known as crystal engineering and is used intensively in design of particles with desired properties, for example with respect to specific particle shape and size distributions. These particulate © 2012 American Chemical Society

properties have great influence on drug product performance and related critical quality attributes, such as dissolution rate and downstream processability (e.g., flowability, compressibility and mixing ability with excipients).2 Many new APIs are poorly water-soluble but permeate well through the intestinal membrane and are classified as BCS class 2 compounds.3 The dissolution rate and thus potentially the bioavailability of Received: Revised: Accepted: Published: 1932

December 14, 2011 May 7, 2012 May 16, 2012 May 16, 2012 dx.doi.org/10.1021/mp200643c | Mol. Pharmaceutics 2012, 9, 1932−1941

Molecular Pharmaceutics

Article

nitrogen containing groups) enabling various interaction modes between nitrofurantoin and the additive. In the presence of water, NF will crystallize as nitrofurantoin monohydrate (NFMH), with the possibility of forming two different monohydrate crystal structures known as nitrofurantoin monohydrate I (NFMH-I, CSD refcode = HAXBUD01) and nitrofurantoin monohydrate form II (NFMH-II, CSD refcode = HAXBUD) (Figures 1B and 1C, respectively).26,28,29 Usually NFMH crystallizes as needle shaped NFMH-II crystals (figure 2A), or with change of water activity as a mixture of crystals with needle and plate habit (NFMH-II and NFMH-I, respectively).29 NFMH-I crystallizes in a planar arrangement, whereas the nitrofurantoin molecules in NFMH-II are arranged in a herringbone motif. Even though the solubility difference between NFMH-I and NFMH-II is small (131 ± 12 mg/L and 110 ± 4 mg/L,29 respectively), preparation of pure NFMH-I is difficult to achieve. Use of thermosensitive polymers as additives has only been sparingly utilized in the field of crystallization. The use of polymer solutions undergoing reversible phase transitions upon heating will change the nucleation and growth rates at higher temperatures due to increased viscosity and consequently lower diffusion rates. At the same time, optimal separation of the crystals at low temperatures due to lower viscosity is possible in such a system.30 We have previously shown that addition of thermosensitive PNIPAM to the crystallization medium can change the morphology of NFMH-II crystals.30 In a similar manner a thermoreversible gel system based on poloxamer 407 has recently been applied to control the crystallization of lactose and to achieve easy recovery of the crystals.31 Poly(Nisopropyl acrylamide) (PNIPAM, Figure 1D) is one polymer possessing thermosensitive properties in aqueous solutions or mixtures of water and organic solvents.32−34 PNIPAM can interfere with the crystallization process by hydrogen bond formation via the amide group in the side chain or by specific adsorption onto a preferred crystal face. In the present case two novel “tailor-made” forms of PNIPAM were investigated which consisted of both isotactic rich and atactic blocks. Blocks with different stereostructures show different solution properties (such as solubility and cloud point), suggesting that the hydration of the polymer depends on the stereoregularity.35 This gives a unique possibility for investigating the influence of the polymer stereostructure, and thus hydrophobic nature of the crystallization additive, as well as the state of aggregation of the additive on crystallization of the model compound. The specific aim of this study is to investigate the use of PNIPAM as a crystallization additive, and its effect on the crystallization of the model drug nitrofurantoin. Both atactic PNIPAM and PNIPAMs with isotactic rich blocks were investigated with a special focus on the physical state of the polymer as a function of temperature and solvent composition.

these compounds can be improved by minimizing the particle (crystal) size, thus increasing the surface area to mass ratio. In many practical situations, the classical approach for reducing particle size is milling. However, the milling process transfers energy to the material increasing the risk of converting a stable crystalline form into a metastable or unstable form followed by possible recrystallization in an uncontrollable manner. Milling can for instance affect the polymorphic composition,4,5 reduce crystallinity,5,6 cause particle size changes during storage7,8 and decrease stability.9 These risks of solid form changes could be avoided by controlled crystallization as secondary manufacturing technology, optimizing performance and potentially making a milling step dispensable.10−12 Crystallization in the presence of additives13−15 (usually polymers and surfactants) is a promising way to obtain material with desired properties, and therefore the use of additives has gained more attention in the field of crystallization of APIs. Polymers and surfactants have been used to change crystal habit by interference with specific crystal faces directing the growth into specific directions. Crystallization of mebendazole in the presence of polyvinyl pyrrolidone or sodium lauryl sulfate affected the particle morphology and the dissolution profile of the crystals,16 and similar results has been achieved for siramesine hydrochloride.17 Different polymeric additives have been used for particle size control in direct precipitation of micrometer-sized salbutamol sulfate particles for inhalation.18 The nucleation rate of aspirin and paracetamol can be controlled by presence of cross-linked polyethylene glycol diacrylate particles with specific microstructure.19 A new polymorphic form of celecoxib was prepared using polysorbate 80 and hydroxypropyl methylcellulose as additives during crystallization.20 Additives have been used to selectively grow the metastable α form of L-glutamic acid.21 Nucleation and growth of crystals on self-assembled monolayers (SAMs) of excipients has been introduced as a new approach for controlling crystal morphology.22,23 Furthermore additives have been used to control crystal growth in melt crystallization.24 It has also been found that Ostwald’s law of stages25 does not always fully apply when additives are used in crystallization. In this study nitrofurantoin (NF) (Figure 1A) has been chosen as a low-soluble model compound. It has the ability to exist in different solid state forms,26,27 and it has chemical groups capable of forming hydrogen bonds (oxygen and

■

EXPERIMENTAL SECTION Materials. Nitrofurantoin anhydrate (β form, CSD refcode = LABJON02) was purchased from Unikem, Denmark. Atactic poly(N-isopropyl acrylamide) (atactic PNIPAM) was purchased from Sigma Aldrich, Germany. HPLC grade acetone and Milli-Q water were used as solvents. Triblock PNIPAMs were synthesized as described previously.35 It has been shown that the isotactic rich blocks are more hydrophobic than the atactic block leading to polymer aggregation and changed phase separation behavior in aqueous solution.30,35 Two stereoblock PNIPAM polymers were

Figure 1. Chemical structure and molecular packing of the investigated compounds. (A) chemical structure of nitrofurantoin, (B) molecular packing of nitrofurantoin monohydrate form I (HAXBUD01), (C) molecular packing of nitrofurantoin monohydrate form II (HAXBUD), (D) chemical structure of PNIPAM. 1933

dx.doi.org/10.1021/mp200643c | Mol. Pharmaceutics 2012, 9, 1932−1941


Article

Figure 2. Nitrofurantoin monohydrate crystallized from acetone:water mixtures in the presence of (A) no additive (needle growth of NFMH-II), (B) atactic PNIPAM (controlled dendritic growth of NFMH-II), (C) atactic PNIPAM (uncontrolled growth of NFMH-I).

concentration of 2.0 mg/mL and PNIPAM concentrations of 0.01, 0.033, 0.1, 0.33, 1, 3.3, 10, 25 or 50 mg/mL covering the molar ratio of N-isopropylacrylamide monomers to NF in the range of ∼0.01 to 53. Forty microliters of this solution was placed on a glass slide, and the solvent was allowed to evaporate to create supersaturation for the crystallization to occur. The concentration 3.3 mg/mL of PNIPAM was used in the results shown except where another concentration is explicitly mentioned, giving a molar ratio of N-isopropyl acrylamide monomers to nitrofurantoin molecules of approximately 3.5:1. The crystallization experiments were conducted at four different temperatures, 7, 20, 27 and 35 °C. The temperature during crystallization was controlled using a cold room or a Krüss G12 hot stage (Krüss GmbH, Hamburg, Germany). The temperature of the glass slide was equilibrated and checked with an IR thermometer before the solution was added. Light Microscope. A Zeiss Axiolab microscope (Carl Zeiss) was used to observe the morphology of the crystals using a 10× magnification objective and with the option of using polarized light. The microscope was equipped with a DeltaPix digital camera (DeltaPix Aps, Maaloev, Denmark) with a resolution of 1280 × 1024 pixels and a pixel size of 0.83 × 0.83 μm. DeltaPix software version 1.6 was used to acquire pictures of the grown crystals. Raman Spectroscopy. Raman spectra were recorded on a Bruker Senterra dispersive Raman microscope using Opus version 6.5 software (Bruker Optics, Ettlingen, Germany) to identify the solid state form of the crystals. Raman shifts in the spectral range of 441 to 1800 cm−1 with a resolution of 3−5 cm−1 were recorded. A 785 nm laser with an intensity of 100 mW was used to illuminate the sample placed on an aluminum slide. 10×, 20× and 50× magnification objectives were used, and the aperture was set to 25 × 1000 μm. The integration time was 2 s, and at least 8 spectra were coadded for the final spectrum. ATR-FTIR Spectroscopy. Attenuated total reflectance Fourier transformed infrared (ATR-FTIR) spectra were recorded on a Nicolet 380 FTIR spectrometer (Thermo Fisher Scientific Inc., Waltham, MA, USA) equipped with SMART iTR accessories with a diamond window. The spectral range of 400−4000 cm−1 was recorded with a resolution of 4 cm−1 and taking the average of 32 scans using OMNIC software version 8.1.11. A spectrum of the clean dry diamond was used as background. Either the samples were scraped off the glass slide and placed on the diamond or the glass slide with sample was placed directly on the diamond. Before measurement pressure

investigated and compared with atactic PNIPAM (Table 1). The structure of the polymers can be identified from their Table 1. Overview of Investigated PNIPAM Polymers polymera atactic PNIPAM a12i10a12 i5a70i5

structureb

Mn(theor) (g/mol)

Mn, SECc (g/mol)

Mw/ Mn

atactic

20,000−25,000

atactic−isotactic− atactic isotactic−atactic− isotactic

34,300

34,500

1.37

80,200

64,700

1.31

a

Numbers given are the block size in kg/mol. bIsotactic content of isotactic block ∼70−80%.35 cDetermined by size exclusion chromatography against poly(methylmethacrylate) standards.35

abbreviation (e.g.: a12i10a12), with a/i referring to the location of the isotactic (i) and atactic (a) part in the backbone and the number referring to the block sizes in (kg/mol). The isotactic content of an isotactic block is approximately 70−80%.35 An atactic polymer has a backbone consisting of monomer with random stereoconfiguration; in an isotactic polymer the backbone consists of repeating monomers with the same stereoconfiguration (either R or S configuration).36 Solubility and Phase Separation. The solubility of NF was measured with UV spectrometry using a Thermo Scientific evolution 300 UV/vis spectrophotometer (Thermo Fisher Scientific Inc., Waltham, MA, USA) at 361 nm. The solubility was calculated based on a prepared standard curve. The solubility of NF in the presence and absence of 10 mg/mL atactic PNIPAM was determined at three temperatures, 7 °C, 24 °C, 35 °C, and in three different solvent mixtures, 0:100, 40:60, 67:33 (vol%) acetone:water. The different solvent mixtures and temperatures were chosen to cover the different physical states that PNIPAM can have in aqueous solution: nonaggregated, aggregated (due to the hydrophobic interaction) or phase separated (due to cononsolvency, see below, or elevated temperature). Solubility of NF in the presence of 10 mg/mL a12i10a12 and i5a70i5 was measured at 7 and 35 °C and in 0:100 and 40:60 (vol%) acetone:water. NF was dissolved in the medium for 3 days before measurement. The phase separation behavior of PNIPAM solutions was investigated using a previously described turbidity method.30 Crystallization Experiments. PNIPAM was dissolved in water and nitrofurantoin anhydrate in acetone. The aqueous phase and the acetone phase were mixed to achieve a 2:1 acetone to water ratio with a nitrofurantoin anhydrate 1934



Article

Figure 3. Polarized light micrographs of dendritic shaped nitrofurantoin monohydrate crystals (controlled growth of NFMH-II) crystallized at 7 °C in the presence of 3.3 mg/mL of (A) atactic PNIPAM, (B) a12i10a12, (C) i5a70i5.

2A), whereas NFMH-II grown in the presence of the atactic PNIPAM polymer resulted in crystals with a dendritic shaped morphology (Figure 2B). Each dendritic NFMH-II crystal grows as a well-separated individual crystalline cluster characterized by branching and resulting in crystals with a “tree-like” structure. The building blocks of these branched crystals are significantly shorter and thinner compared to the needles achieved from crystallization without additive. In some parts of the droplets placed on the glass slide, uncontrolled growth of the monohydrate form I (NFMH-I) was observed (Figure 2C). The uncontrolled growth has intense curvature in a flat structure with less clear branching. This phenomenon was most likely related to the fast evaporation of the solvent, high supersaturation of nitrofurantoin in the solution and, finally, the nucleation of the less stable NFMH-I. This form was further stabilized by the presence of polymeric excipient increasing the viscosity of the solution, reducing diffusion and decreasing the transformation rate to the stable monohydrate form II (NFMH-II). The crystallization experiments were performed in the presence of 0.01 to 50 mg/mL of atactic PNIPAM, causing a large variation in the viscosity of the crystallization medium. However, no significant changes in the crystal morphology were observed except for the lowest concentration (0.01 mg/ mL PNIPAM) where a decrease in the branching tendency was observed, indicating a minimum amount of PNIPAM needed to gain the dendritic morphology. At high PNIPAM concentrations, especially at 25 and 50 mg/mL, less nucleation was observed compared to the lower concentration levels. This can be explained with the solubility enhancing effect of PNIPAM shown later in this paper. The viscosity of the crystallization medium often affects the habit of the resulting crystals. The viscosity is affected by polymer type, concentration and solvent composition.30 The viscosity of the aqueous and acetone−water solutions of atactic PNIPAM containing 10 mg/mL polymer was in the range 4− 12 mPa·s below the phase separation temperature and was almost independent of temperature and solvent composition. For the experiments performed with 1 and 3.3 mg/mL PNIPAM an even lower viscosity increase can be expected. In contrast, the viscosity of the polymers with isotactic blocks (10 mg/mL) depends on both temperature and acetone−water mixing ratios. The viscosity varies in the range 140−2000 mPa·s for a12i10a12 and 8−32 mPa·s for i5a70i5.30 It was decided to evaluate the effect of viscosity on crystal morphology by comparing three polymer types at 7 °C. Using this temperature, the influence of phase separation (Figure 8) was avoided as the

was applied to ensure close contact between diamond and sample. X-ray Powder Diffraction (XRPD). XRPD diffractograms were measured using a PANalytical X’Pert PRO X-ray diffractometer (PANalytical B.V., Almelo, The Netherlands) equipped with a PIXcel detector using Cu Kα1 radiation (λ = 1.5406 Å). The Kβ radiation was eliminated by a nickel filter. The voltage and current settings were 45 kV and 40 mA, respectively. Samples were placed on silicon disks and measured in reflection mode in the range of 5−40° 2θ with a step size of 0.0394° 2θ and a scan speed of 0.0540° 2θ/s. Data was collected using X’Pert Data Collector software version 2.2 and analyzed with X’Pert highscore plus version 2.2.4. Pulsed Field Gradient Spin Echo NMR. The self-diffusion coefficients (Ds) of NF and PNIPAM in acetone(D6):D2O mixtures were measured using pulsed field gradient spin echo nuclear magnetic resonance (PFG-NMR) spectroscopy at 22 °C. Nonoverlapping proton peaks from NF at ∼8.0 ppm and PNIPAM at ∼3.9 ppm were used for the calculations. The measurements were performed on a Varian UNITYINOVA spectrometer operating at 300 MHz for protons. In the stimulated echo pulse sequence employed the gradient pulse time (σ) and the dwell time (Δ) were 2 and 120 ms respectively between the 90° radio frequency pulses. The attenuation of the signal was measured as a function of the applied field gradient strength, and the Ds was calculated according to ⎛I⎞ ⎛ σ⎞ −kDs = ln⎜ ⎟ = −(ΥMGσ )2 ⎜Δ − ⎟Ds ⎝ 3⎠ ⎝ I0 ⎠

(1)

where I and I0 are the signal intensities with and without the gradients, γM is the gyromagnetic ratio of protons, G is the applied gradient strength, σ is the length of the field gradient pulse, and Δ is the dwell time between pulses. Plotting the natural logarithm of the measured intensities against the gradient strength provides the Ds as the slope of the linear fit.

■

RESULTS AND DISCUSSION Crystallization of NF in the Presence of Atactic PNIPAM. It has previously been shown that the presence of the polymeric additive HPMC can result in dendritic crystal growth of NFMH-II with intense branching.37 A similar behavior was observed when NFMH-II was grown in the presence of the thermosensitive atactic PNIPAM polymer.30 NFMH-II crystals grown from an acetone−water mixture without additive obtain a needle shaped morphology (Figure 1935



Article

three PNIPAMs remained fully dissolved at all acetone−water ratios. The viscosity difference between atactic and isotactic containing PNIPAM solutions only had a minor influence on the crystal size and shape, as shown in Figure 3. Solid state identification (Figure 4) of the dendritic crystals (Figure 2B) using Raman microscopy confirmed that these

Figure 5. Solubility of nitrofurantoin as a function of temperature and acetone content in the presence and absence of 10 mg/mL atactic PNIPAM (n = 3). Closed symbols: without additive. Open symbols: 10 mg/mL atactic PNIPAM below the phase separation temperature. Open symbols with “+” inside: 10 mg/mL PNIPAM above the phase separation temperature.

with decreased supersaturation,39 likewise should the decreased supersaturation result in less uncontrolled growth of NFMH-I. However, an increase both in the dendritic growth tendency and in uncontrolled growth of NFMH-I were observed, indicating that the presence of PNIPAM results in a complex interference with the crystallization process. Crystallization of NFMH in the Presence of PNIPAM with Isotactic Blocks. The solution properties of PNIPAM change by varying the stereochemistry of the backbone. Increased isotacticity decreases the water solubility of PNIPAM, and, if the isotactic sequences are polymerized in a block fashion, the polymers form aggregates below phase transition temperature.35,40 The advantage of such stereochemical modifications is that the chemical structure otherwise remains unchanged and the interaction with NFMH can be studied without considering changes due to an otherwise altered chemical structure. Previous investigations have shown that in dilute solutions a−i−a (a12i10a12) and i−a−i PNIPAM (i2a28i2 and i2a40i2) form aggregates consisting of 15−26 polymer chains at room temperature, i.e. below the phase separation temperature.40 A rather uniform aggregate size with hydrodynamic radii of 21−29 nm and a polydispersity index (Mw/Mn) of 1.29−1.37 was observed in the concentration range of 0.1 to 5.0 mg/mL at 20 °C. Further it has been shown that a−i−a PNIPAM forms branched star-shaped micelles in the studied concentration range while the i−a−i PNIPAM forms flowerlike micelles.40 Both micelles are thought to have a hydrophobic core consisting of isotactic blocks and a hydrophilic shell with the atactic blocks. The i5a70i5 is assumed to also form flowerlike micelles as it has isotactic end blocks. Two different PNIPAMs with isotactic blocks (Table 1) were evaluated as possible crystallization additives to identify if the change in structure influences the morphology of the NFMH-II crystals. It was found that the morphology of the dendritic NFMH-II crystals remained the same regardless of which of the two PNIPAM modifications was used as an additive. Thus the different micelle types of PNIPAM did not have an influence on the morphology of the achieved crystals, which remained the

Figure 4. Raman spectroscopic identification of nitrofurantoin monohydrate (NFMH) crystallized in presence of atactic PNIPAM. (A) NFMH-II with dendritic structure crystallized in presence of PNIPAM (see Figure 2B), (B) reference spectra of NFMH-II,38 (C) uncontrolled growth of NFMH-I in the presence of PNIPAM (see Figure 2C), (D) reference spectra of NFMH-I.38

crystals were NFMH-II, whereas the uncontrolled growth crystals in Figure 2C were confirmed as NFMH-I. Several spots on each dendritic crystal were measured to ensure that the branching was not caused by changes in the solid state composition. Solid state identification using ATR-FTIR spectroscopy and XRPD revealed a mixture of NFMH-I and NFMH-II after crystallization in the presence of PNIPAM (data not shown), because these techniques gives an average of all the crystals in the sample and not the identity of a single crystal. Despite intensive work on optimizing the temperature, solvent selection and compound concentrations, it was not possible to grow one specific NFMH form selectively by using atactic PNIPAM as additive. The solubility of NF in the absence and presence of PNIPAM was measured to reveal a possible solubility enhancing effect of the polymer. It was observed that the presence of atactic PNIPAM in all investigated combinations of temperature and solvent ratios resulted in increased solubility of NF (Figure 5). The relative solubility increase was particularly large in water, where a solubility increase of 58− 74% was observed. In the presence of acetone, the solubility increase was 9−24% in 40 vol % acetone and 3−7% in 67 vol % acetone. It is interesting to note that atactic PNIPAM increased the solubility of NF even at temperatures above phase separation (24 °C/40:60, 35 °C/0:100 and 35 °C/40:60 vol % acetone:water). Due to the solubility increasing effect of PNIPAM on NF and the expected decrease in solvent evaporation rate due to the viscosity increase, a lower achievable supersaturation level is expected for the experiments conducted in the presence of PNIPAM. The tendency of dendritic growth of crystals is related to high supersaturation level during crystallization and should therefore be less likely 1936



Article

Figure 6. Polarized light micrographs of dendritic shaped nitrofurantoin monohydrate crystals (controlled growth of NFMH-II) crystallized at 20 °C in the presence of 3.3 mg/mL of (A) atactic PNIPAM, (B) a12i10a12, (C) i5a70i5.

°C the addition of any of the three different PNIPAMs results in approximately the same solubility increase of NF to the medium, which was in the range 57−70%. The reason for this similar increase is that at 35 °C all the PNIPAMs are phase separated and the atactic block is dehydrated, making the polymers in general more equally hydrophobic, allowing increased solubility of NF. The addition of acetone (40 vol%) increased the solubility of NF to the medium in all cases although this mixture is a poorer solvent for PNIPAM due to the cononsolvency. At 7 °C all samples showed similar NF solubilizing capability, apart from the i5a70i5, which resulted in an increased NF solubility. Again the increase in temperature from 7 to 35 °C increased the NF solubility. At 35 °C the presence of atactic PNIPAM resulted in a slightly higher NF solubility compared to a12i10a12, i5a70i5 and pure water. The variation of NF solubility in 40 vol % acetone is thought to be caused by complex interactions between water, acetone, PNIPAM and NF. Further studies are needed in order to gain a detailed understanding of all the various factors. The results show that increasing the temperature and the addition of acetone significantly increases the NF solubility. Addition of PNIPAM significantly increases the NF solubility in pure aqueous media both in the dissolved and in the phase separated state, whereas in 40 vol % acetone the solubilization effect of adding PNIPAM is smaller and more complex. Effect of Temperature and Phase Separation Behavior of PNIPAM on Crystallization. PNIPAM shows a thermosensitive character both in aqueous solution and in acetone−water mixtures in certain concentrations. The polymer is soluble in the medium at low temperatures but upon heating loses its solubility at a certain temperature (cloud point), phase separates and, depending on concentration, forms either a suspension of polymeric aggregates (mesoglobules) or precipitates. This behavior originates from the break-up of hydrogen bonds between polymer and solvent molecules and formation of internal hydrogen bonds.41 The phase separation temperature of PNIPAM depends on the solvent composition and the order and length of the isotactic blocks as shown in previous studies.30,40 Further, both acetone and water are good solvents for atactic PNIPAM, but in certain mixtures of these solvents the phase separation temperature of PNIPAM is decreased. This phenomenon, a mixture of two good solvents becoming a poorer solvent for dissolving PNIPAM, is termed cononsolvency.30,33,34,42 The phase separation behavior of the different PNIPAMs is presented in Figure 8.

same as with atactic PNIPAM (Figure 6). Since the changed structure of PNIPAM does not affect the crystal morphology, it is assumed that the nucleation takes place outside the micellar cores. If the nucleation were to take place in the core of the micelles, different growth behavior would be expected when using the polymers with isotactic blocks compared to the atactic PNIPAMs. Both monohydrate forms of NF were found using PNIPAM with isotactic blocks as additives. Hence, similar to the results with atactic PNIPAM, the isotactic block containing PNIPAMs could not be used to selectively grow one specific form of NFMH. The solubility of NF in the presence of PNIPAM with isotactic blocks was measured in order to compare it with atactic PNIPAM (Figure 7). Two temperatures (7 and 35 °C) and two solvents (0:100 and 40:60 vol % acetone:water) were used.

Figure 7. Comparison of the solubility enhancing effect of (10 mg/ mL) atactic, a12i10a12 and i5a70i5 PNIPAM at two temperatures and two solvent mixtures (n = 3). Significant differences between columns are marked with an asterisk (*). A column marked with * and long vertical line indicates significant difference from the columns it is compared to, which are marked with short vertical line.

In water, at 7 °C, the addition of atactic PNIPAM, a12i10a12, i5a70i5 resulted in an increase in NF solubility of 56%, 147% and 180% compared to water, respectively. The increased solubility of NF in the presence of PNIPAM bearing isotactic blocks compared to atactic PNIPAM shows that the increased hydrophobicity promotes the dissolution of NF. In water at 35 1937



Article

Figure 8. Upper left: Diagram of the phase separation temperature for the three different PNIPAM polymers. The dashed horizontal lines indicate the solvent changes in the crystallization medium during evaporation at the 4 temperatures, moving from right to left. Right: Polarized light micrographs of NFMH-II crystallized in the presence of 3.3 mg/mL atactic PNIPAM at different temperatures: (A) 7 °C, (B) 20 °C, (C) 27 °C, (D) 35 °C.

solubility and thereby an interaction between PNIPAM and NFMH is to be expected at all temperatures. This is also clearly the case in the crystallization experiments, where the dendritic growth persisted above the phase separation temperature. Interaction between NF and PNIPAM. Dendritic crystals are frequently observed in general crystallization without additives (e.g., minerals and snow crystals), where the dendritic growth is a result of instable interfacial morphology caused by temperature and concentration gradients.43,44 With the presence of large molecules, for instance polymeric additives close to or adsorbed onto the crystals, an inhomogeneous concentration gradient around the crystal is to be expected and is likely to be one reason for the dendritic growth. Adsorption of foreign molecules at the surface of the crystal will also interrupt the crystal growth and possibly force the crystal to grow in other directions resulting in the changed morphology. A third possible interaction mechanism behind the dendritic growth could be imperfect fusion of newly formed nuclei with existing crystals facilitating growth resulting in branched morphology. The interaction mechanism between NFMH and PNIPAM was investigated using ATR-FTIR, PFG-NMR and Raman spectroscopy to gain more insight into the dendritic growth mechanism. The FTIR technique is well suited for probing polar bonds including hydrogen bonds. If hydrogen bonds between NFMH and PNIPAM are formed to a large extend or if PNIPAM is incorporated in or adsorbed onto the crystal, changes in the infrared (IR) spectra such as peak shifts or formation of new peaks can be expected. However, the IR spectra only contained peaks similar to a physical mixture of

The influence of the polymer phase separation on the crystallization of NFMH was investigated. The solvent composition changes continuously during the evaporative crystallization, which may result in phase separation during the crystallization process depending on the temperature. NFMH was crystallized at four temperatures, to investigate the influence of different aggregation states of the PNIPAMs on the resulting nitrofurantoin crystal structure (Figure 8). The crystal morphology was affected by the temperature at which the crystallization was performed. Panels A, B, C and D in Figure 8 show typical examples of the achieved crystals when the crystallization was done at 7, 20, 27 and 35 °C, respectively. At 7 and 20 °C PNIPAM stayed dissolved during the whole or most of the evaporation process resulting in crystals with uniform size and shape (Figures 8A and 8B). At 27 and 35 °C, PNIPAM is phase separated during the whole or at least major parts of the crystallization process, giving fewer but slightly larger and more irregular crystals (Figures 8C and 8D). Independent of temperature all crystals retained the dendritic growth pattern. At increased temperatures, typically lower nucleation rates and increased growth rates can be achieved due to a change in solubility and diffusion,12 explaining the presence of larger crystals. The presence of PNIPAM did not change this trend. It is likely that phase separated PNIPAM mesoglobules and NFMH-II crystals can aggregate, assisting the formation of the irregular crystals at 27 and 35 °C, which was not observed at 7 and 20 °C. Using PNIPAM with isotactic blocks resulted in similar crystal morphology (micrographs not shown) independent of temperature. The solubility data from the previous sections showed that the hydrophobic modification changes the 1938



Article

Figure 9. Raman spectra of NFMH-II crystallized in the presence of different PNIPAMs at 7, 20 and 27 °C. Blue shifts and peak shape changes were observed for especially the −CN− (1620 cm−1) and the furan group (1025 cm−1) in the NFMH-II.

the crystals due to further growth around the additive capturing the polymer in the crystal, or the PNIPAM could also desorb again before getting trapped in the crystal. In both cases the presence of PNIPAM will leave an imperfect crystal with a changed crystal lattice. The blue shift and peak deformation were investigated at different temperatures to see if the interaction changes (Figure 9). Upon increasing the crystallization temperature from 7 to 27 °C the peak shifts and the shape deformation diminished. The Raman spectrum of NFMH-II crystallized in the presence of PNIPAM becomes more similar to the Raman spectrum of NFMH-II crystallized in the absence of PNIPAM. This indicates that the interaction between PNIPAM and NFMH is sensitive to temperature, which is likely to be due to increased molecular mobility and decreased stability of the interaction between NF and PNIPAM. No proof of hydrogen bond formation between NF and PNIPAM was found. The dendritic morphology of NFMH-II in the presence of PNIPAM is most likely related to locally varying supersaturation effects around the NF crystals in the crystallization medium caused by steric hindrance phenomenon and adsorption.

NFMH-I/NFMH-II and PNIPAM, showing that PNIPAM is not largely incorporated in or adsorbed onto the crystal, and that any bond formation between NFMH and PNIPAM is below the detection limit. The dendritic growth of NFMH-II could be a result of a very short-lived interaction between NFMH-II and PNIPAM, disrupting the incorporation of a single or a few NF molecules into the crystal without incorporation of the polymer, resulting in the modified morphology. The branching in the dendritic crystals is clearly visible under the optical microscope, which means that numerous NF molecules are incorporated together with crystal water into the growing NFMH-II crystal between each branch supporting the hypotheses of a short-lived and temporary interaction. PFG-NMR was used to measure the self-diffusion coefficient of NF molecules in the absence and presence of PNIPAM. Due to the size of the polymers or their aggregates their selfdiffusion coefficient is roughly 10-fold lower than that of NF (data not shown). The ratio of amide groups in the PNIPAM polymer to the number of NF molecules in the experiments was roughly 40:1 to ensure an excess of possible interaction moieties from the polymer. If hydrogen bonds are formed between NF and PNIPAM or other types of interaction take place, a decrease in the self-diffusion coefficient of NF molecules is expected. However, no significant decrease in the self-diffusion rate of NF molecules was observed taking into account the viscosity increasing effect of PNIPAM. Based on the IR and PFG-NMR data the interaction between NF and PNIPAM is limited. With Raman spectroscopy spectral changes in the form of two peak shifts and shape deformation were observed when NFMH-II was crystallized in the presence of PNIPAM compared to crystallization without additives. The two peaks blue-shifted and were identified45 as signals from the furan ring (∼1020 cm−1) and the −CN− group (∼1620 cm−1) (Figure 9). The blue shift shows that there is an increased stress in these particular moieties which could be due to stretching or deformation of bonds. It is possible that PNIPAM temporarily adsorbs onto the crystal surface and then is incorporated into

■

CONCLUSION The crystallization of NF into dendritic shaped NFMH-II crystals in the presence of PNIPAM was an inherent property of the current system regardless of the molecular weight or stereoconfiguration of the polymer, solvent composition, viscosity or concentration. The presence of PNIPAM in aqueous medium resulted in a large solubility increase of NF, whereas the solubility enhancing effect of PNIPAM only partly remained above the phase separation temperature and with addition of acetone to the solvent. The solubility increasing effect was affected by the stereoregularity of PNIPAM, where the presence of isotactic blocks led to an enhanced solubility effect. No direct evidence for specific interactions between PNIPAM and NF was found using spectroscopic techniques apart from blue-shifting in Raman spectra, indicating differences in the crystal lattice. Thus it appears that the dendritic growth is mainly induced by steric factors and to some extent through the 1939



Article

(15) Weissbuch, I.; Popovitzbiro, R.; Lahav, M.; Leiserowitz, L. Understanding and control of nucleation, growth, habit, dissolution and structure of 2-dimensional and 3-dimensional crystals using tailormade auxiliaries. Acta Crystallogr., Sect. B 1995, 51, 115−148. (16) Kumar, S.; Chawla, G.; Bansal, A. K. Role of additives like polymers and surfactants in the crystallization of mebendazole. Yakugaku Zasshi 2008, 128 (2), 281−289. (17) Zimmermann, A.; Millqvist-Fureby, A.; Elema, M. R.; Hansen, T.; Mullertz, A.; Hovgaard, L. Adsorption of pharmaceutical excipients onto microcrystals of siramesine hydrochloride: Effects on physicochemical properties. Eur. J. Pharm. Biopharm. 2009, 71 (1), 109− 116. (18) Tan, R. B. H.; Xie, S. Y.; Poornachary, S. K.; Chow, P. S. Direct precipitation of micron-size salbutamol sulfate: New insights into the action of surfactants and polymeric additives. Cryst. Growth Des. 2010, 10 (8), 3363−3371. (19) Diao, Y.; Helgeson, M. E.; Myerson, A. S.; Hatton, T. A.; Doyle, P. S.; Trout, B. L. Controlled nucleation from solution using polymer microgels. J. Am. Chem. Soc. 2011, 133 (11), 3756−3759. (20) Lu, G. W.; Hawley, M.; Smith, M.; Geiger, B. M.; Pfund, W. Characterization of a novel polymorphic form of celecoxib. J. Pharm. Sci. 2006, 95 (2), 305−317. (21) Davey, R. J.; Blagden, N.; Potts, G. D.; Docherty, R. Polymorphism in molecular crystals: Stabilization of a metastable form by conformational mimicry. J. Am. Chem. Soc. 1997, 119 (7), 1767−1772. (22) Myerson, A. S.; Lee, A. Y.; Ulman, A. Crystallization of amino acids on self-assembled monolayers of rigid thiols on gold. Langmuir 2002, 18 (15), 5886−5898. (23) Myerson, A. S.; Singh, A.; Lee, I. S.; Kim, K. Crystal growth on self-assembled monolayers. CrystEngComm 2011, 13 (1), 24−32. (24) Taylor, L. S.; Kestur, U. S.; Lee, H.; Santiago, D.; Rinaldi, C.; Won, Y. Y. Effects of the molecular weight and concentration of polymer additives, and temperature on the melt crystallization kinetics of a small drug molecule. Cryst. Growth Des. 2010, 10 (8), 3585−3595. (25) Ostwald, W. Studien uber die Bildung und Umwandlung fester Korper. Z. Phys. Chem., Stoechiom. Verwandtschaftsl. 1897, 22, 289− 330. (26) Pienaar, E. W.; Caira, M. R.; Lötter, A. P. Polymorphs of nitrofurantoin. I. Preparation and X-ray crystal-structures of 2 monohydrated forms of nitrofurantoin. J. Crystallogr. Spectrosc. Res. 1993, 23 (9), 739−744. (27) Pienaar, E. W.; Caira, M. R.; Lötter, A. P. Polymorphs of nitrofurantoin. 2. Preparation and X-ray crystal structures of 2 anhydrous forms of nitrofurantoin. J. Crystallogr. Spectrosc. Res. 1993, 23 (10), 785−790. (28) Caira, M. R.; Pienaar, E. W.; Lotter, A. P. Polymorphism and pseudopolymorphism of the antibacterial nitrofurantoin. Mol. Cryst. Liq. Cryst. 1996, 278, A241−A264. (29) Tian, F.; Qu, H.; Louhi-Kultanen, M.; Rantanen, J. Crystallization of a polymorphic hydrate system. J. Pharm. Sci. 2010, 99 (2), 753−763. (30) Munk, T.; Hietala, S.; Kalliomäki, K.; Nuopponen, M.; Tenhu, H.; Tian, F.; Rantanen, J.; Baldursdottir, S. Behaviour of stereoblock poly(N-isopropyl acrylamide) in acetone-water mixtures. Polym. Bull. 2011, 67 (4), 677−692. (31) Cespi, M.; Bonacucina, G.; Casettari, L.; Mencarelli, G.; Palmieri, G. F. Poloxamer thermogel systems as medium for crystallization. Pharm. Res. 2012, 29 (3), 818−826. (32) Ganji, F.; Vasheghani-Farahani, E. Hydrogels in controlled drug delivery systems. Iran. Polym. J. 2009, 18 (1), 63−88. (33) Heskins, M.; Guillet, J. Solution properties of poly(Nisopropylacrylamide). J. Macromol. Sci., Part A 1968, 2 (8), 1441− 1455. (34) Winnik, F. M.; Ringsdorf, H.; Venzmer, J. Methanol water as a co-nonsolvent system for poly(N-Isopropylacrylamide). Macromolecules 1990, 23 (8), 2415−2416.

hydrophobic interactions between PNIPAM and NF, shown by the increased solubility of NF in the presence of PNIPAM. Nevertheless, the finding that minute amounts of PNIPAM affect NFMH-II morphology indicates the potential of using polymeric additives in crystallization as a particle engineering approach.

■

AUTHOR INFORMATION

Corresponding Author

*Faculty of Pharmaceutical Sciences, University of Copenhagen, Universitetsparken 2, 2100 Copenhagen, Denmark. Phone: +45 353 36105. Fax: +45 353 36001. E-mail: sba@ farma.ku.dk. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS Keith Gordon from the Department of Chemistry, University of Otago, New Zealand, and Haiyan Qu from the University of Southern Denmark are gratefully acknowledged for providing access to two Bruker Senterra Raman Microscope systems.

■

REFERENCES

(1) Mullin, J. Crystallization, 4th ed.; Elsevier ButterworthHeinemann: Burlington, 2001. (2) Chen, J.; Sarma, B.; Evans, J. M. B.; Myerson, A. S. Pharmaceutical crystallization. Cryst. Growth Des. 2011, 11 (4), 887−895. (3) Amidon, G. L.; Lennernas, H.; Shah, V. P.; Crison, J. R. A theoretical basis for a biopharmaceutic drug classification - the correlation of in-vitro drug product dissolution and in-vivo bioavailability. Pharm. Res. 1995, 12 (3), 413−420. (4) Forbes, R. T.; Chikhalia, V.; Storey, R. A.; Ticehurst, M. The effect of crystal morphology and mill type on milling induced crystal disorder. Eur. J. Pharm. Sci. 2006, 27 (1), 19−26. (5) Otsuka, M.; Matsuda, Y. Effect of environmental humidity on the transformation pathway of nitrofurantoin modifications during grinding and the physicochemical properties of ground products. J. Pharm. Pharmacol. 1993, 45 (5), 406−413. (6) Wildfong, P. L. D.; Hancock, B. C.; Moore, M. D.; Morris, K. R. Towards an understanding of the structurally based potential for mechanically activated disordering of small molecule organic crystals. J. Pharm. Sci. 2006, 95 (12), 2645−2656. (7) Ng, W. K.; Kwek, J. W.; Tan, R. B. H. Anomalous particle size shift during post-milling storage. Pharm. Res. 2008, 25 (5), 1175− 1185. (8) Ward, G. H.; Schultz, R. K. Process-induced crystallinity changes in albuterol sulfate and its effect on powder physical stability. Pharm. Res. 1995, 12 (5), 773−779. (9) Kitamura, S.; Miyamae, A.; Koda, S.; Morimoto, Y. Effect of grinding on the solid-state stability of cefixime trihydrate. Int. J. Pharm. 1989, 56 (2), 125−134. (10) Byrn, S. R.; Pfeiffer, R. R.; Stowell, J. G. Solid-state chemistry of drugs, 2nd ed.; SSCI, Inc: West Lafayette, 1999. (11) Hilfiker, R.; Blatter, F.; von Raumer, M. Relevance of solid-state properties for pharmaceutical products. In Polymorphism in the pharmaceutical industry, 1st ed.; Hilfiker, R., Ed.; Wiley-VCH Verlag GmbH & Co.: Weinheim, 2006. (12) Myerson, A.; Jacovs, J.; Jacovs, V. Handbook of industrial crystallization, 2nd ed.; Butterworth-Heinemann: New York, 1992. (13) Rodríguez-Hornedo, N. Crystallization and the properties of crystals. In Encyclopedia of pharmaceutical technology, 1st ed.; Swarbrick, J., Boylan, J., Eds.; Marcel Dekker, Inc.: New York, 1990; Vol. 3, pp 399−434. (14) Weissbuch, I.; Addadi, L.; Lahav, M.; Leiserowitz, L. Molecular recognition at crystal interfaces. Science 1991, 253 (5020), 637−645. 1940



Article

(35) Nuopponen, M.; Kalliomäki, K.; Laukkanen, A.; Hietala, S.; Tenhu, H. A-B-A stereoblock copolymers of N-isopropylacrylamide. J. Polym. Sci., Part A: Polym. Chem. 2008, 46 (1), 38−46. (36) Carey, F. Organic Chemistry, 5th ed.; McGraw Hill: New York, 2003. (37) Tian, F.; Baldursdottir, S.; Rantanen, J. Effects of polymer additives on the crystallization of hydrates: A molecular-level modulation. Mol. Pharmaceutics 2009, 6 (1), 202−210. (38) Tian, F.; Qu, H.; Louhi-Kultanen, M.; Rantanen, J. Mechanistic insight into the evaporative crystallization of two polymorphs of nitrofurantoin monohydrate. J. Cryst. Growth 2009, 311 (8), 2508− 2589. (39) Bernstein, J. Polymorphism and structure-property relations. In Polymorphism in Molecular Crystals, 1st ed.; Clarendon Press: Oxford, 2002; p 208. (40) Nuopponen, M.; Kalliomäki, K.; Aseyev, V.; Tenhu, H. Spontaneous and thermally induced self-organization of A-B-A stereoblock polymers of N-isopropylacrylamide in aqueous solutions. Macromolecules 2008, 41 (13), 4881−4886. (41) Schmaljohann, D. Thermo- and pH-responsive polymers in drug delivery. Adv. Drug Delivery Rev. 2006, 58 (15), 1655−1670. (42) Winnik, F. M.; Ottaviani, M. F.; Bossmann, S. H.; Garciagaribay, M.; Turro, N. J. Cononsolvency of poly(N-Isopropylacrylamide) in mixed water-methanol solutions - a look at spin-labeled polymers. Macromolecules 1992, 25 (22), 6007−6017. (43) Mullins, W. W.; Sekerka, R. F. Morphological stability of a particle growing by diffusion or heat flow. J. Appl. Phys. 1963, 34 (2), 323−329. (44) Sunagawa, I. Crystal growth. In Crystals: Growth, morphology, and perfection, 1st ed.; Cambridge University Press: Cambridge, 2005; pp 20−59. (45) Aaltonen, J.; Strachan, C. J.; Pöllänen, K.; Yliruusi, J.; Rantanen, J. Hyphenated spectroscopy as a polymorph screening tool. J. Pharm. Biomed. Anal. 2007, 44 (2), 477−483.

1941


Crystal Morphology Modification by the Addition of Tailor-Made

Recommend Documents