Polymers' ultrafine particles for drug delivery systems precipitated by supercritical carbon dioxide + organic solvent mixtures

Powder Technology 292 (2016) 140–148

Contents lists available at ScienceDirect

Powder Technology journal homepage: www.elsevier.com/locate/powtec

Polymers' ultrafine particles for drug delivery systems precipitated by supercritical carbon dioxide + organic solvent mixtures Valentina Prosapio, Ernesto Reverchon, Iolanda De Marco ⁎ Department of Industrial Engineering, University of Salerno, Via Giovanni Paolo II, 132, 84084, Fisciano (SA), Italy

a r t i c l e

i n f o

Article history: Received 12 November 2015 Received in revised form 20 January 2016 Accepted 23 January 2016 Available online 27 January 2016 Keywords: Expanded liquid antisolvent Sodium alginate Polyvinyl alcohol Supercritical fluids Drug delivery systems

a b s t r a c t The applicability of supercritical antisolvent precipitation (SAS) is restricted to hydrophobic substances because of the very limited solubility of water in CO2 at ordinary SAS operating conditions (40–60 °C, 10–25 MPa). To overcome this limitation, a technique has been developed, named expanded liquid antisolvent (ELAS), in which mixtures of supercritical carbon dioxide (scCO2) and organic solvents, at expanded liquid conditions, are used as the antisolvent: water solubility is widely enhanced. In this work, sodium alginate and polyvinyl alcohol (PVA), two water-soluble polymers, used as carrier for drug delivery systems, were successfully micronized by ELAS. Different antisolvent mixtures were used: scCO2 + ethanol, scCO2 + acetone and scCO2 + isopropyl alcohol. Operating at 15 MPa and 40 °C, varying the organic co-antisolvent, the co-antisolvent mole fraction and the concentration of the polymer in the aqueous solution, nanoparticles (with a mean diameter of about 200 nm), microparticles with smooth surface (with a mean diameter in the range of 0.9–12.5 μm for sodium alginate and 2–9 μm for PVA) and nanostructured microparticles (with a mean diameter of about 11 μm) were produced. XRD analyses on the processed powders revealed that no modifications in the polymer structure were induced by ELAS processing. Solvent residue analyses revealed that the co-antisolvent residue ranged between 50 and 300 ppm depending on the organic solvent used. © 2016 Elsevier B.V. All rights reserved.

1. Introduction A controlled release system is designed to deliver the drug at a predetermined rate. The drug can be surrounded by a polymeric shell or can be uniformly dispersed in it [1]. Morphology, mean size, particle size distribution and polymer molecular weight can influence the ways in which the drug release occurs [2]. The polymer used as carrier has to satisfy some requirements, such as biocompatibility and biodegradability [3]. Among the natural and synthetic polymers, sodium alginate and polyvinyl alcohol (PVA) have been widely proposed for drug delivery systems [4–7]. Sodium alginate is a random, linear and anionic polysaccharide, formed by linear copolymers of α-L-guluronate (G) and β-Dmannuronate (M) residues. The composition and sequence of G and M can influence polymer properties and depend from the source from which alginate is obtained. Sodium alginate is generally regarded as safe (GRAS) by the FDA (Food and Drug Administration) and is considered as one of the most versatile polysaccharides for a large number of industrial applications. Indeed, it is stable, biodegradable, biocompatible, non-toxic, mucoadhesive, non-immunogenic and low in cost [8]. Microparticles of sodium alginate are largely used in biomedical field for drug delivery systems [9–11]. Drug release from alginate matrix

⁎ Corresponding author. E-mail address: [email protected] (I. De Marco).

http://dx.doi.org/10.1016/j.powtec.2016.01.033 0032-5910/© 2016 Elsevier B.V. All rights reserved.

occurs through two mechanisms: drug diffusion and degradation of the polymer [9]. PVA is a water-soluble synthetic polymer, used in pharmaceutical and biomedical applications for its remarkable properties such as mechanical resistance, biocompatibility, biodegradability and non-toxicity [12]. These properties are influenced by polymer molecular weight, cross-linking density and crystallinity [13]. Furthermore, because of PVA rapid hydrolysis, a long time retention of the carrier in the human body is avoided [14]. Traditional techniques used for polymer micronization include solvent/emulsion evaporation [15], jet milling [16] and spray drying [17]. However, they show several limitations such as the use of large quantities of organic solvent difficult to recover, high operating temperatures, lack of control over particle size and particle size distribution [18]. Generally, sodium alginate microparticles are obtained by gelation methods [19–21]; however, these techniques led the production of irregular particles and suffer from some operational limits, such as needle blockage and cleaning if the solution is extruded through a needle to form alginate droplets [22]. Nan et al. [23] produced microparticles of alginate using an emulsification + extrusion process, but using this technique long time processing was required; moreover, harmful solvents, such as petroleum ether, were used and no residual solvent analyses were performed on the final product. Ting et al. [13] proposed the micronization of PVA using spraydrying and spray-desolvation techniques; but, they obtained collapsed

V. Prosapio et al. / Powder Technology 292 (2016) 140–148

large particles with a mean diameter in the range of 17–55 μm and with yield lower than 50%; moreover, the characterization of the powder, to verify the structure of the material after processing, was missing. In order to overcome the limitations of the conventional techniques, supercritical fluid (SCF) based processes have been extensively proposed in the literature [24–35]. In particular, they have been used for the micronization of several kinds of materials [25,26,28–31,33,34], impregnation of drugs in aerogels [27,35] membrane formation [24] and scaffold production [32]. Indeed, a SCF shows liquid-like properties (such as high density and solvent power) and gas-like properties (such as low viscosity, high diffusivity and no surface tension), that can be easily tuned, changing pressure and temperature. Among the different SCF based techniques, microparticles and nanoparticles of different polymers have been frequently obtained using the supercritical antisolvent (SAS) process [36–44]. For example, PVA, that is soluble in some organic solvents, was precipitated from dimethylsulfoxide (DMSO) using SAS technique [36]. Well-defined microparticles were obtained, but the residual solvent amount was not measured. In order to properly apply SAS, the following are required: (a) an organic solvent that, at the process conditions, is completely miscible with supercritical carbon dioxide (scCO2), (b) a solute to be micronized that is soluble in the organic solvent and not soluble in the scCO2 + organic solvent mixture. The requirement of miscibility between solvent and scCO2 restricts the applicability of the SAS technique to hydrophobic compounds, considering that, at ordinary SAS conditions (temperature lower than 60 °C and pressure lower than 25 MPa), the solubility of water in scCO2 is very low [45]. But, in many cases, there is also a need to produce water soluble compounds, such as proteins, enzymes and hydrosoluble polymers in micron sizes; therefore, various attempts at modifying the SAS process have been performed [46–49]. Indeed, the addition of a given quantity of an organic solvent to scCO2 leads to the formation of a supercritical or expanded liquid mixture, in which water solubility can considerably increase. Some papers on the micronization of water-soluble compounds using this SAS-modified technique have been published [46–49]. Among these, protein particles were obtained by adding ethanol to scCO2 but they were irregular and aggregated [48, 49]. Bouchard et al. [46,47] added methanol, ethanol, acetone or 2propanol to scCO2, to process different solutes, and obtained large and irregular particles; moreover, the measure of the solvent residue was not reported. De Marco and Reverchon [50], using bovine serum albumin (BSA) as a model compound, proposed a new technique in which water solutions were precipitated; the antisolvent was formed by scCO2 + ethanol at expanded liquid conditions; therefore, they named ELAS (expanded liquid antisolvent) the developed process. In a subsequent paper, Prosapio et al. [51] observed that nanoparticles, microparticles and expanded microparticles of BSA could be obtained by ELAS varying not only the operating conditions, but also the co-antisolvent. Fourier transform infrared analysis on BSA powders revealed that no modifications of the protein secondary structure were induced by ELAS processing, especially when acetone was used as co-antisolvent. In a recent paper [52], lysozyme microparticles with a mean diameter ranging between 2.8 and 13.8 μm were obtained by ELAS, using scCO2 + acetone and scCO2 + isopropyl alcohol as antisolvents. However, until now ELAS has been applied only to BSA and lysozyme; therefore, polymers have never been processed using this technique. Since controlled delivery is mainly based on the proper dimensions of polymer particles, the possibility to micronize two water-soluble polymers, relevant in the pharmaceutical field, like sodium alginate and PVA, should be regarded as an improvement of the ELAS applications. Moreover, these two polymers have different chemical formulas and can be considered as model compounds: indeed, sodium alginate (C6H7NaO6)n is representative of natural polymers, whereas PVA (C2H4O)n is a synthetic polymer. Therefore, the scope of this work is to extend ELAS application to the micronization of these polymers, using ethanol, acetone and isopropyl alcohol as coantisolvents.

141

2. Materials, apparatus and methods 2.1. Materials Sodium alginate (SA, MW ≈ 240.000 Da, β-D-mannuronic/α-Lguluronic acid ratio 1:2), Polyvinyl alcohol (PVA, MW: 30.000–70.000, degree of alcoholysis: 99.8%), distilled water (H2O), ethanol (EtOH, purity 99.5%), acetone (AC, purity 99.5%) and isopropyl alcohol (iPrOH, purity 99.5%) were supplied by Sigma-Aldrich (Italy). CO2 (purity 99%) was purchased from SON (Italy). All materials were used as received. Solubility tests performed at room temperature showed that sodium alginate solubility in water is about 20 mg/mL, whereas PVA solubility is about 50 mg/mL. 2.2. ELAS apparatus A detailed representation of the equipment used for ELAS experiments has been previously published [50] and a scheme is reported in Fig. 1. It mainly consists of an HPLC pump (Gilson, model 805) used to deliver the co-antisolvent (EtOH, AC or iPrOH) and two diaphragm high-pressure pumps (Milton Roy, model Milroyal B and Milroyal D) used to deliver carbon dioxide and aqueous solution. To ensure a large contact surface between co-antisolvent and CO2, a high-pressure vessel with an internal volume of 35 cm3, loaded with stainless steel perforated saddles, is used as pre-mixer. A cylindrical vessel with an internal volume of 500 cm3 is used as precipitation chamber and it is electrically heated using thin band heaters. The aqueous solution is delivered to the precipitator through a thin wall stainless steel nozzle with the diameter of 100 μm. A stainless steel frit, put at the bottom of the vessel, is used to collect the solid product, allowing the CO2/H2O/co-antisolvent solution to pass through. A second vessel located downstream the precipitation vessel, operating at a lower pressure (1.8–2 MPa) is used to recover the mixture of water and co-antisolvent. The organic solvent is recovered from this mixture, using a rotary evaporator. 2.3. ELAS procedure An ELAS experiment starts delivering supercritical CO2 at a constant flow rate to the pre-mixer and to the precipitation chamber, until the desired pressure is reached. Then, the co-antisolvent is pumped to the pre-mixer, where it comes in contact with CO2, forming an expanded liquid solution. When stable flow rates, temperature and pressure conditions in the precipitator are reached, water is sent through the injector to obtain steady state composition conditions of the fluid phase during solute precipitation. Afterwards, the flow of water is stopped and the aqueous solution is delivered through the nozzle at the same flow rate, producing the precipitation of the solute. At the end of injection, two washing steps are carried out: in the first one, the antisolvent mixture continues to flow in the chamber to eliminate water residues (for a time t1) and, in the second one, CO2 alone continues to flow to eliminate co-antisolvent residues (for a time t2). At the end of the second washing step, CO2 flow is stopped and the precipitator is depressurized down to atmospheric pressure. Eventually, it is possible to recover the dry powders from the porous filter. 2.4. Analytical methods Samples of the precipitated material were observed by a Field Emission Scanning Electron Microscope (FESEM, mod. LEO 1525, Carl Zeiss SMT AG, Oberkochen, Germany). Powders were dispersed on a carbon tab previously stuck on an aluminum stub (Agar Scientific, Stansted, United Kingdom); then, were coated with gold (layer thickness 250 Å) using a sputter coater (mod. 108 A, Agar Scientific, Stansted, United Kingdom). Particle size distribution (PSD) of the powders was measured from FESEM photomicrographs using the Sigma Scan Pro image analysis

142

V. Prosapio et al. / Powder Technology 292 (2016) 140–148

Fig. 1. Schematic representation of ELAS apparatus. S1: CO2 supply; S2: co-antisolvent supply; S3: aqueous solution supply; RB: refrigerating bath; P1, P2, P3: pumps; TC: thermocouple; M: manometer; PM: pre-mixer; PC: precipitation chamber; MV: micrometering valve; LS: liquid separator; BPV: back-pressure valve; R: rotameter and DM: dry test meter.

software (release 5.0, Aspire Software International Ashburn, VA). Approximately 1000 particles, taken at high enlargements and in various locations inside the precipitator, were analyzed in the elaboration of each particle size distribution. Histograms representing the particle size distributions were fitted using Microcal Origin Software (release 8.0, Microcal Software, Inc., Northampton, MA). X-ray diffractograms were recorded using an X-ray powder diffractometer (model D8 Discover; Bruker Corporation, Billerica, MA) with a Cu sealed tube source. The measuring conditions were: Ni-filtered CuKα radiation, λ = 1.54 Å, 2θ angle ranging from 10° to 60° with a scan rate of 4°/min [53]. Ethanol, acetone and isopropyl alcohol residues were measured using a headspace sampler (model 7694E, Hewlett Packard, Palo Alto, CA) coupled to a gas chromatograph equipped with a flame ionization detector (GC-FID, model 6890 GC-SYSTEM, Hewlett Packard, Agilent Technologies Mfg. Gmbh & Co. KG, Böblingen, Germany). The solvents were separated using two fused silica capillary columns connected in series by press-fit: the first column (model Carbowax EASYSEP, Stepbios, Bologna, Italy; 30 m length, 0.53 mm i.d., 1 μm film thickness) is connected to the detector, the second (model Cp Sil 5CB CHROMPACK, Stepbios, Bologna, Italy; 25 m length, 0.53 mm i.d., 5 μm film thickness) to the injector. GC conditions were: oven temperature at 160 °C for a total time equal to 8.80 min. The injector was maintained at 250 °C (split mode, ratio 5:1), and helium was used as the carrier gas (2 mL/min). Head space conditions were: equilibration time, 9 min at 170 °C; pressurization time, 0.3 min; loop fill time, 0.4 min. Head space samples were prepared in 20 mL vials filled with 50 mg of polymer dissolved in water. Analyses were performed in triplicates on each batch of processed polymer. 3. Experimental results The operating conditions used in the ELAS experiments performed in this work were selected on the basis of our previous experience in

this process [50–52]. Operating pressure was fixed at 15 MPa, temperature at 40 °C, aqueous solution flow rate at 1 mL/min and CO2 flow rate at 20 g/min. The first set of experiments on alginate and PVA was carried out varying the kind of co-antisolvent and its mole fraction. Then, some experiments were performed varying the concentration of the liquid solution. A list of the experiments performed, the obtained morphologies, mean diameter (m.d.) and standard deviation (s.d.) of the powders is reported in Table 1. 3.1. Effect of different co-antisolvents and their mole fraction 3.1.1. Ethanol Fixing EtOH flow rate at 10 mL/min (the corresponding mole fractions of the fluid components were: xCO2 =0.64, xEtOH = 0.27, xH2O =0.09; #1 in Table 1), sodium alginate precipitated in form of expanded microparticles with a mean diameter of about 13 μm, characterized by a discontinuous nanostructured surface (also called nanostructured balloons or nanostructured microparticles). A qualitative indication is given in the FESEM image reported in Fig. 2; quantitative data is reported in Table 1. Nanostructure is evident, observing the discontinuous surface of the microparticles reported in Fig. 2. Other experiments were performed, using EtOH flow rates of 15 mL/min (xCO2 = 0.57, xEtOH = 0.35,xH2 O = 0.08; #2 in Table 1) and 25 mL/min (x CO 2 = 0.46, x EtOH = 0.48, x H2 O = 0.06; #3 in Table 1), that confirmed the morphology previously observed: nanostructured microparticles were produced also in these cases. Comparing the particles obtained at different ethanol flow rates, it is possible to observe that, increasing EtOH flow rate, the mean diameter slightly decreased and the standard deviation (that is an indication of the sharpness of the particle size distribution) clearly decreased. The same experiments were carried out using PVA as solute. When an EtOH flow rate equal to 10 mL/min was used (#14 in Table 1), PVA precipitated in form of microparticles with a mean diameter of

V. Prosapio et al. / Powder Technology 292 (2016) 140–148

143

Table 1 ELAS experiments performed on sodium alginate and PVA (NP: nanoparticles; MP: microparticles; NMP: nanostructured microparticles; u.*: unsuccessful; m.d.: mean diameter; s.d.: standard deviation). #

Co-antisolvent

Sodium alginate 1 EtOH 2 3 4 AC 5 6 7 8 9 iPrOH 10 11 12 13 Polyvinyl alcohol 14 EtOH 15 16 17 AC 18 19 20 21 22 iPrOH 23 24 25

Qco-ant (mL/min)

conc (mg/mL)

xCO2

xH2O

xco-ant

Operating point

Morph.

m.d. (μm)

s.d. (μm)

10 15 25 10 15 25

10

0.64 0.57 0.46 0.68 0.61 0.51

0.09 0.08 0.06 0.09 0.06 0.07

0.27 0.35 0.48 0.23 0.33 0.42

A B C A B C

0.69

0.09

0.22

A

0.62 0.51

0.08 0.07

0.30 0.42

B C

NMP NMP NMP u.* u.* MP MP MP MP MP MP MP MP

12.98 11.00 10.37 – – 0.91 3.35 5.40 2.05 7.47 12.50 6.19 3.31

13.96 5.78 3.22 – – 0.59 1.61 3.04 0.49 1.75 2.81 1.81 1.92

0.64 0.57 0.46 0.68 0.61 0.51

0.09 0.08 0.06 0.09 0.06 0.07

0.27 0.35 0.48 0.23 0.33 0.42

A B C A B C

0.69 0.62 0.51

0.09 0.08 0.07

0.22 0.30 0.42

A B C

MP NP NP u.* MP MP MP MP MP MP MP MP

2.10 0.26 0.18 – 9.27 3.58 6.54 7.76 3.75 2.80 2.30 2.36

0.64 0.08 0.05 – 2.75 1.03 1.58 3.06 1.08 0.95 0.79 0.88

10

15 25

10 15 25 10 15 25

10 15 25

10 5 10 15 5 10 15 10

20

20 10 15 20 20 10 20

2.10 μm. Fig. 3a allows to qualitatively assess the observed morphology and size of microparticles. Increasing the co-antisolvent flow rate at 15 and 25 mL/min (#15–16 in Table 1), nanoparticles were obtained, as reported in Fig. 3b, that shows an example of this morphology. Therefore, SEM images show that, increasing the EtOH flow rate from 10 to 15 mL/min, a change of morphology occurred (microparticles to nanoparticles): the mean size of the particles decreased and the PSD became narrower, as shown in Fig. 4, where volumetric cumulative particle size distributions are reported. No relevant size variations were observed, changing co-antisolvent flow rate from 15 to 25 mL/min, as it is also possible to verify from data of mean diameter and standard deviation reported in Table 1 (experiments #15 and #16). 3.1.2. Acetone Acetone was the second solvent used as co-antisolvent in our experiments; when its flow rate was fixed at 10 and 15 mL/min (the corresponding mole fractions of the fluid components were: xCO2 = 0.68,

Fig. 2. FESEM image of nanostructured sodium alginate nanostructured microparticles obtained at 15 MPa, 40 °C, 10 mg/mL using an ethanol flow rate of 10 mL/min.

xAC = 0.238, xH2O = 0.09 and xCO2 = 0.61, xAC = 0.33, xCO2xH2O = 0.06 respectively; #4–5 in Table 1), sodium alginate was completely extracted by the mixture CO2 + AC; increasing the co-antisolvent flow rate at 25 mL/min (xCO2 = 0.51, xAC = 0.42, xH2O = 0.07; #7 in Table 1), microparticles with a rough surface and a broad PSD were obtained, as it is possible to observe in the FESEM image reported in Fig. 5a. In the case of PVA, using an AC flow rate of 10 mL/min (#17 in Table 1), as in the case of sodium alginate, the solute was completely extracted by the mixture CO2 + AC; whereas, in correspondence of AC flow rates of 15 and 25 mL/min (#18 and #21 in Table 1), well separated microparticles with a smooth surface were produced, as shown in the FESEM image reported in Fig. 5b. Comparing the PVA particle size distribution data of the particles obtained at different AC flow rates (Table 1), it is possible to observe that, increasing AC mole fraction, the mean diameter of the particles slightly reduced (experiments #18 and #21). 3.1.3. Isopropyl alcohol Using iPrOH as co-antisolvent and fixing its flow rate at 10 mL/min (mole fractions of the fluid components are: xCO2 = 0.69, xiPrOH = 0.22, xH2O = 0.09; #10 in Table 1), sodium alginate precipitated in form of microparticles with a rough surface, as it is possible to observe from the FESEM image in Fig. 6a. Increasing the iPrOH flow rate at 15 mL/min (xCO2 = 0.62, xiPrOH = 0.30, xH2O = 0.08; #12 in Table 1) and 25 mL/min (xCO2 = 0.51, xiPrOH = 0.42, xH2O = 0.07; #13 in Table 1), microparticles were still obtained, as shown in Fig. 6b, where a FESEM image of the particles processed using a iPrOH flow rate equal to 15 mL/min is reported. However, their surface is rough also at these conditions. The comparison of the volumetric cumulative particle size distributions, at different iPrOH flow rates, is reported in Fig. 7 and shows that, increasing the co-antisolvent flow rate, the mean diameter of sodium alginate particles largely reduced and the corresponding PSD became narrower. The same experiments were repeated using PVA. In correspondence of all iPrOH flow rates, PVA precipitated in form of well-separated spherical microparticles. An example of FESEM image is reported in Fig. 8.

144

V. Prosapio et al. / Powder Technology 292 (2016) 140–148

Fig. 3. FESEM images of PVA particles precipitated at 15 MPa, 40 °C, 20 mg/mL, using an ethanol flow rate of: (a) 10 mL/min and (b) 15 mL/min.

The comparison of the diameters of the obtained PVA particles, reported in Table 1 (experiments #22, #23 and #25) shows that the mean size only slightly reduced increasing iPrOH flow rate. 3.2. Effect of polymer concentration in the aqueous solution The effect of the concentration of solute in the aqueous solution on particles morphology, mean size and PSD was investigated in correspondence of some selected operating conditions. Using AC as co-antisolvent at a flow rate equal to 25 mL/min, the concentration of the polymer in water was varied from 5 to 15 mg/mL in the case of sodium alginate and from 10 to 20 mg/mL in the case of PVA. For both the polymers, the comparison of the volumetric particle size distributions obtained at different concentrations (data reported in Table 1), shows shat increasing polymer concentration, the mean diameter of the particles increased and the PSD enlarged. The effect of sodium alginate concentration in water was investigated also using iPrOH as co-antisolvent, using the same set of concentrations of the previous experiments. Comparing the volumetric particle size distributions, shown in Fig. 9, it is possible to observe that increasing alginate concentration in the liquid solution, the mean diameter largely increased and the particle size distribution enlarged. Using iPrOH as co-antisolvent with a flow rate equal to 25 mL/min, the effect of solute concentration in water was studied also for PVA. The experiments were performed fixing the concentration of the liquid

solution at 10 mg/mL and well separated microparticles were produced. Comparing the mean diameter of these particles with the one of the particles produced at 20 mg/mL, at the same operating conditions, it was found that reducing the concentration, the mean size of the particles and the PSD did not change, as it is possible to observe in Table 1 (experiments #24 and #25). 3.3. Analyses XRD analyses for untreated and ELAS processed sodium alginate and PVA were performed to study if the structure of the material was changed by processing. The X-ray diffractograms, in the case of both the polymers and using all the co-antisolvents, are shown in Fig. 10. As it is possible to observe from the spectra related to sodium alginate reported in Fig. 10a, the traces are very similar to the ones of the unprocessed material; in the case of PVA, the ELAS processed material is always amorphous, differently from the semicrystalline untreated PVA. It means that amorphous polymers are in all the cases obtained because of ELAS processing. According to Food and Drug Administration (FDA) guidelines, ethanol, acetone and isopropyl alcohol belong to class 3 residual solvent; therefore, their maximum acceptable concentration in the final product is 5000 ppm. A headspace sampler, coupled to a gas chromatograph, was used to verify the co-antisolvent residue content in the ELAS produced particles. The analysis revealed that the solvent residue was for both the polymers around 50 ppm in the case of AC, 300 ppm in the case of EtOH and 200 in the case of iPrOH; i.e., in all cases, lower than the FDA limits. 4. Discussion

Fig. 4. Volumetric cumulative PSDs of PVA particles obtained at 15 MPa, 40 °C, 20 mg/mL and different ethanol flow rates.

In order to give an explanation to the obtained results, it is advisable to summarize the general interpretation of SAS process from which ELAS derives. In SAS precipitation, several mechanisms are involved that regard high pressure vapor–liquid equilibria, mass transfer and fluid-dynamics and nucleation and growth [38,42,44,54]. From a thermodynamic point of view, if the solute is insoluble in the mixture solvent/antisolvent, its presence can be neglected [55] and the obtained morphology can be related to the position of the operating point with respect to the mixture critical point (MCP) of the binary system CO2/solvent. Considering only the operating points located in a single phase region, when SAS is operated at a pressure near above the MCP, microparticles are produced; when the operating point is located far above the MCP, nanoparticles can be observed; when the operating point is located below the MCP expanded microparticles are obtained. In this last case, two different surface

V. Prosapio et al. / Powder Technology 292 (2016) 140–148

145

Fig. 5. FESEM images of microparticles obtained at 15 MPa, 40 °C using an acetone flow rate of 25 mL/min: (a) sodium alginate (10 mg/mL); (b) PVA (20 mg/mL).

morphologies have been observed [56]: uniform surface, which is obtained when a slow nucleation and a fast diffusion characterize particles formation and nanostructured surface, formed by nanometric units connected on a spherical surface, that is obtained when fast nucleation and slow diffusion occur. Regarding fluid-dynamics aspects, the obtainment of a given morphology depends on the competition between two characteristic times: the time of jet break-up (τjb) and the time of complete surface tension vanishing (τts). When τjb b τts, jet break-up prevails, droplets are formed and, as a final result, microparticles are produced; when τjb N τts, gas mixing prevails and nanoparticles are obtained by gas-to-particles nucleation and growth. In previous works [50–52], it was observed that it is possible to extend the interpretation of SAS mechanisms to ELAS process; but, in this case, the high pressure phase equilibria (VLE) of the ternary systems CO2/H2O/co-antisolvent (Fig. 11) must be considered. In this work, the available VLE diagrams, adapted from the scientific literature [57–60], are referred to the same temperature used in ELAS experiments (40 °C), but at different pressures. For the systems CO2/H2O/ EtOH and CO2/H2O/iPrOH, data are available at 10 and 20 MPa and the experimental points of the miscibility hole (separating the regions in which one or two coexisting phases are present) at different pressures are very close to each other or even overlap. Therefore, in the range of 10–20 MPa, the miscibility hole remains substantially unchanged. Consequently, it is possible to assume that the VLE curve at 15 MPa superimposes the literature data too. For the system CO2/H2O/AC, experimental data are available only at 10 MPa (Fig. 11b). However,

literature reports that for the systems CO2/H2O/organic solvents (such as, for example, ethanol [59], isopropanol [57,58] and methanol [61]), in the range of 7–30 MPa, increasing the pressure, the biphasic region is not sensibly changed or, at least, it is slightly reduced. Therefore, it is possible to hypothesize that, also for the system CO2/H2O/AC, an increase of pressure to 15 MPa does not cause a substantial modification of the VLE. After these preliminary considerations, it is possible to look at the two regions in which the diagrams reported in Fig. 11a–c are divided: (a) an upper region, indicated with I, in which the system is homogeneous; one phase is formed and it is possible to suppose that the solute precipitates from a supercritical or an expanded liquid mixture, depending on the position of the operating point with respect to the mixture critical point; (b) a lower region, indicated with II, in which two phases in equilibrium coexist and solute can precipitate from the liquid-rich side of the corresponding tie-line. The amplitude of the biphasic region is different for each co-antisolvent and, in particular, it is wider for the system CO2/H2O/AC.

The operating points, investigated in this work, are indicated with the letters A, B, C in Fig. 11a–c; the corresponding mole fractions are shown in Table 1. Depending on the chosen co-antisolvent, different morphologies and mean diameters were obtained that can be related

Fig. 6. FESEM images of sodium alginate particles obtained at 15 MPa, 40 °C, 10 mg/mL using an isopropyl alcohol flow rate of: (a) 10 mL/min and (b) 15 mL/min.

146

V. Prosapio et al. / Powder Technology 292 (2016) 140–148

Fig. 7. Volumetric cumulative PSDs of sodium alginate particles obtained at 15 MPa, 40 °C, 10 mg/mL and different iPrOH flow rates.

with the position of the operating point with respect to the VLE curve. For the system CO2/H2O/EtOH, all the operating points lie in the upper region and the obtainment of microparticles or nanoparticles, both in the case of sodium alginate and PVA, can be explained considering the competition between τjb and τts, as in the SAS process: if the time of jet break-up τjb is higher than the time of the surface tension vanishing τst, gas mixing prevails and nanoparticles are obtained; otherwise, jet break-up prevails and microparticles are produced. When AC is used as co-antisolvent, the system CO2/H2O/AC has to be considered; in this case, the miscibility hole is wider with respect to the one of the system CO2/H2O/EtOH. Indeed, for both solutes, using an acetone flow rate of 10 mL/min the operating point is located inside the biphasic region, whereas at 25 mL/min the operating point is located outside. In correspondence of 15 mL/min, instead, the results were different for the two polymers: ✓ in the case of sodium alginate, the precipitation was unsuccessful; this experimental evidence can be explained hypothesizing that the presence of the solute induced a slight enlargement of the miscibility gap, causing the operating point fall in the biphasic region (with a result similar to the case corresponding to a flow rate equal to 10 mL/min);

Fig. 9. Volumetric cumulative PSD sodium alginate particles obtained at 15 MPa, 40 °C and different concentrations using iPrOH as co-antisolvent (25 mL/min).

✓ in the case of PVA, microparticles were obtained, indicating a successful precipitation; this result can be explained, considering that this polymer only slightly modified the VLE and, therefore, the operating point lies in the one-phase region.

Therefore, the morphology and the particle size obtained, once fixed the polymer and the co-antisolvent, is related to the polymer solubility in the mixture CO2/H2O/co-antisolvent and to the consequent modification of the VLE. In Fig. 11c, it is possible to observe that, using iPrOH as co-antisolvent, the miscibility hole is narrower with respect to the other two coantisolvents, therefore, all the operating points lie in the homogeneous region. In correspondence of all the flow rates, microparticles were produced and it was observed that, increasing the flow rate, the mean size of the particles reduced and PSD became narrower. This experimental evidence can be explained, as in the SAS process, considering the distance of the operating point from the VLE curve. Indeed, increasing the coantisolvent mole fraction, this distance increased; correspondingly, the surface tension reduction was faster and the generated droplets were smaller. Consequently, microparticles with smaller diameters were produced. It was, moreover, observed that, increasing sodium alginate concentration, the mean size increased and the PSD enlarged. This result can be explained considering that, increasing the polymer concentration, the cohesive forces increased and also the viscosity of the aqueous solution enlarged, with the consequent formation of larger droplets during the atomization process. 5. Conclusions

Fig. 8. FESEM image of PVA particles obtained at 15 MPa, 40 °C, 20 mg/mL using an iPrOH flow rate of 10 mL/min.

ELAS processing of polymers demonstrated to be very efficient in producing amorphous particles, extending the applicability of this technique to other products. A large variety of nanoparticles, microparticles and nanostructured microparticles were obtained, varying the kind of co-antisolvent, its mole fraction and the solute concentration in water; in all the experiments, the yield of the process was larger than 90%. The morphology and the dimension of the precipitates can be strictly correlated with the distance of the operating point with respect to the high pressure vapor liquid equilibria curve and to the competition between atomization and gas-mixing. The analyses on the obtained powders confirmed that the material did not undergo alterations because of ELAS processing and the negligible residual solvent makes these particles safe and useful for drug delivery applications.

V. Prosapio et al. / Powder Technology 292 (2016) 140–148

147

Fig. 10. XRD spectra: (a) sodium alginate powders; (b) PVA powders.

Fig. 11. High pressure phase equilibria for the systems CO2/H2O/co-antisolvent, at 40 °C and different pressures, adapted from literature [57–60]: a) ethanol; b) acetone; c) isopropyl alcohol. I indicates the homogeneous region, II the two-phase region.

148

V. Prosapio et al. / Powder Technology 292 (2016) 140–148

References [1] A. Dash, G. Cudworth, Therapeutic applications of implantable drug delivery systems, J. Pharmacol. Toxicol. Methods 40 (1998) 1–12. [2] S. Freiberg, X. Zhu, Polymer microspheres for controlled drug release, Int. J. Pharm. 282 (2004) 1–18. [3] R. Rani, N. Dilbaghi, D. Dhingra, S. Kumar, Optimization and evaluation of bioactive drug-loaded polymeric nanoparticles for drug delivery, Int. J. Biol. Macromol. 78 (2015) 173–179. [4] S. Kayal, R. Ramanujan, Doxorubicin loaded PVA coated iron oxide nanoparticles for targeted drug delivery, Mater. Sci. Eng. C 30 (2010) 484–490. [5] S.M. More, R.V. Kulkarni, B. Sa, N.V. Kayane, Glutaraldehyde-crosslinked poly (vinyl alcohol) hydrogel discs for the controlled release of antidiabetic drug, J. Appl. Polym. Sci. 116 (2010) 1732–1738. [6] F. Tuğcu-Demiröz, F. Acartürk, S. Takka, Ö. Konuş-Boyunağa, Evaluation of alginate based mesalazine tablets for intestinal drug delivery, Eur. J. Pharm. Biopharm. 67 (2007) 491–497. [7] P.F. Builders, O.O. Kunle, L.C. Okpaku, M.I. Builders, A.A. Attama, M.U. Adikwu, Preparation and evaluation of mucinated sodium alginate microparticles for oral delivery of insulin, Eur. J. Pharm. Biopharm. 70 (2008) 777–783. [8] S.K. Motwani, S. Chopra, S. Talegaonkar, K. Kohli, F.J. Ahmad, R.K. Khar, Chitosan– sodium alginate nanoparticles as submicroscopic reservoirs for ocular delivery: formulation, optimisation and in vitro characterisation, Eur. J. Pharm. Biopharm. 68 (2008) 513–525. [9] C.H. Goh, P.W.S. Heng, L.W. Chan, Alginates as a useful natural polymer for microencapsulation and therapeutic applications, Carbohydr. Polym. 88 (2012) 1–12. [10] B. Singh, V. Sharma, D. Chauhan, Gastroretentive floating sterculia–alginate beads for use in antiulcer drug delivery, Chem. Eng. Res. Des. 88 (2010) 997–1012. [11] T.W. Wong, Alginate graft copolymers and alginate–co-excipient physical mixture in oral drug delivery, J. Pharm. Pharmacol. 63 (2011) 1497–1512. [12] H. Shagholani, S.M. Ghoreishi, M. Mousazadeh, Improvement of interaction between PVA and chitosan via magnetite nanoparticles for drug delivery application, Int. J. Biol. Macromol. 78 (2015) 130–136. [13] T.-Y. Ting, I. Gonda, E.M. Gipps, Microparticles of polyvinyl alcohol for nasal delivery. I. Generation by spray-drying and spray-desolvation, Pharm. Res. 9 (1992) 1330–1335. [14] Y. Dong, Z. Zhou, H. Ding, S. Zhang, Preparation and properties of a pH sensitive carrier based on three kinds of polymer blend to control the release of 5-amino salicylic acid, Pharm. Dev. Technol. 19 (2014) 960–967. [15] C. Li, W. Wang, X. Wang, H. Jiang, J. Zhu, S. Lin, Fabrication of porous polymer microspheres by tuning amphiphilicity of the polymer and emulsion–solvent evaporation processing, Eur. Polym. J. 68 (2015) 409–418. [16] G. Nykamp, U. Carstensen, B. Müller, Jet milling—a new technique for microparticle preparation, Int. J. Pharm. 242 (2002) 79–86. [17] M.D. Blanco, R.L. Sastre, C. Teijón, R. Olmo, J.M. Teijón, Degradation behaviour of microspheres prepared by spray-drying poly(D,L-lactide) and poly(D,L-lactide-co-glycolide) polymers, Int. J. Pharm. 326 (2006) 139–147. [18] W. Wang, G. Liu, J. Wu, Y. Jiang, Co-precipitation of 10-hydroxycamptothecin and poly (L-lactic acid) by supercritical CO2 anti-solvent process using dichloromethane/ethanol co-solvent, J. Supercrit. Fluids 74 (2013) 137–144. [19] L.W. Chan, H.Y. Lee, P.W. Heng, Mechanisms of external and internal gelation and their impact on the functions of alginate as a coat and delivery system, Carbohydr. Polym. 63 (2006) 176–187. [20] A.J. Ribeiro, C. Silva, D. Ferreira, F. Veiga, Chitosan-reinforced alginate microspheres obtained through the emulsification/internal gelation technique, Eur. J. Pharm. Sci. 25 (2005) 31–40. [21] C.M. Silva, A.J. Ribeiro, I.V. Figueiredo, A.R. Gonçalves, F. Veiga, Alginate microspheres prepared by internal gelation: development and effect on insulin stability, Int. J. Pharm. 311 (2006) 1–10. [22] J.P. Paques, E. van der Linden, C.J. van Rijn, L.M. Sagis, Preparation methods of alginate nanoparticles, Adv. Colloid Interf. Sci. 209 (2014) 163–171. [23] F. Nan, J. Wu, F. Qi, Y. Liu, T. Ngai, G. Ma, Uniform chitosan-coated alginate particles as emulsifiers for preparation of stable Pickering emulsions with stimulus dependence, Colloids Surf. A Physicochem. Eng. Asp. 456 (2014) 246–252. [24] L. Baldino, S. Cardea, E. Reverchon, Supercritical assisted enzymatic membranes preparation, for active packaging applications, J. Membr. Sci. 453 (2014) 409–418. [25] R. Campardelli, L. Baldino, E. Reverchon, Supercritical fluids applications in nanomedicine, J. Supercrit. Fluids 101 (2015) 193–214. [26] R. Campardelli, G. Della Porta, L. Gomez, S. Irusta, E. Reverchon, J. Santamaria, Au– PLA nanocomposites for photothermally controlled drug delivery, J. Mater. Chem. B 2 (2014) 409–417. [27] G. Caputo, M. Scognamiglio, I. De Marco, Nimesulide adsorbed on silica aerogel using supercritical carbon dioxide, Chem. Eng. Res. Des. 90 (2012) 1082–1089. [28] G. Della Porta, N. Falco, E. Giordano, E. Reverchon, PLGA microspheres by Supercritical emulsion extraction: a study on insulin release in myoblast culture, J. Biomater. Sci. Polym. Ed. 24 (2013) 1831–1847. [29] S. Liparoti, R. Adami, E. Reverchon, Supercritical assisted atomization: effect of operative conditions on PVP microparticle size and morphology, J. Supercrit. Fluids 97 (2015) 31–35. [30] A. Montes, M. Gordillo, C. Pereyra, E.M. de La Ossa, Supercritical CO2 precipitation of poly(L-lactic acid) in a wide range of miscibility, J. Supercrit. Fluids 81 (2013) 236–244. [31] D. Perinelli, G. Bonacucina, M. Cespi, A. Naylor, M. Whitaker, G. Palmieri, G. Giorgioni, L. Casettari, Evaluation of P (L) LA-PEG-P (L) LA as processing aid for biodegradable particles from gas saturated solutions (PGSS) process, Int. J. Pharm. 468 (2014) 250–257.

[32] P. Pisanti, A.B. Yeatts, S. Cardea, J.P. Fisher, E. Reverchon, Tubular perfusion system culture of human mesenchymal stem cells on poly-L-lactic acid scaffolds produced using a supercritical carbon dioxide-assisted process, J. Biomed. Mater. Res. A 100 (2012) 2563–2572. [33] V. Prosapio, I. De Marco, M. Scognamiglio, E. Reverchon, Folic acid–PVP nanostructured composite microparticles by supercritical antisolvent precipitation, Chem. Eng. J. 277 (2015) 286–294. [34] V. Prosapio, E. Reverchon, I. De Marco, Coprecipitation of PVP/β-carotene by supercritical antisolvent processing, Ind. Eng. Chem. Res. (2015). [35] I. Smirnova, S. Suttiruengwong, W. Arlt, Feasibility study of hydrophilic and hydrophobic silica aerogels as drug delivery systems, J. Non-Cryst. Solids 350 (2004) 54–60. [36] R. Adami, L. Sesti Osséo, R. Huopalahti, E. Reverchon, Supercritical antisolvent micronization of PVA by semi-continuous and batch processing, J. Supercrit. Fluids 42 (2007) 288–298. [37] M.S. Costa, A.R.C. Duarte, M.M. Cardoso, C.M. Duarte, Supercritical antisolvent precipitation of PHBV microparticles, Int. J. Pharm. 328 (2007) 72–77. [38] I. De Marco, O. Knauer, F. Cice, A. Braeuer, E. Reverchon, Interactions of phase equilibria, jet fluid dynamics and mass transfer during supercritical antisolvent micronization: the influence of solvents, Chem. Eng. J. 203 (2012) 71–80. [39] I. De Marco, V. Prosapio, F. Cice, E. Reverchon, Use of solvent mixtures in supercritical antisolvent process to modify precipitates morphology: cellulose acetate microparticles, J. Supercrit. Fluids 83 (2013) 153–160. [40] I. De Marco, M. Rossmann, V. Prosapio, E. Reverchon, A. Braeuer, Control of particle size, at micrometric and nanometric range, using supercritical antisolvent precipitation from solvent mixtures: application to PVP, Chem. Eng. J. 273 (2015) 344–352. [41] H. Jin, S. Li, D. Hu, Y. Zhao, Preparation of PLA–PEG nanoparticles by the solution enhanced dispersion with enhanced mass transfer using ultrasound in supercritical CO2, Powder Technol. 227 (2012) 17–23. [42] F. Marra, I. De Marco, E. Reverchon, Numerical analysis of the characteristic times controlling supercritical antisolvent micronization, Chem. Eng. Sci. 71 (2012) 39–45. [43] D.L. Obrzut, P.W. Bell, C.B. Roberts, S.R. Duke, Effect of process conditions on the spray characteristics of a PLA+ methylene chloride solution in the supercritical antisolvent precipitation process, J. Supercrit. Fluids 42 (2007) 299–309. [44] E. Reverchon, I. De Marco, Mechanisms controlling supercritical antisolvent precipitate morphology, Chem. Eng. J. 169 (2011) 358–370. [45] A.N. Sabirzyanov, A.P. Il'in, A. Akhunov, F.M. Gumerov, Solubility of water in supercritical carbon dioxide, High Temp. 40 (2002) 203–206. [46] A. Bouchard, N. Jovanović, W. Jiskoot, E. Mendes, G.-J. Witkamp, D.J. Crommelin, G.W. Hofland, Lysozyme particle formation during supercritical fluid drying: particle morphology and molecular integrity, J. Supercrit. Fluids 40 (2007) 293–307. [47] A. Bouchard, N. Jovanović, Á. Martín, G.W. Hofland, D.J. Crommelin, W. Jiskoot, G.-J. Witkamp, Effect of the modifier on the particle formation and crystallisation behaviour during precipitation from aqueous solutions, J. Supercrit. Fluids 44 (2008) 409–421. [48] R.T. Bustami, H.-K. Chan, F. Dehghani, N.R. Foster, Generation of micro-particles of proteins for aerosol delivery using high pressure modified carbon dioxide, Pharm. Res. 17 (2000) 1360–1366. [49] R.T. Bustami, H.-K. Chan, T. Sweeney, F. Dehghani, N.R. Foster, Generation of fine powders of recombinant human deoxyribonuclease using the aerosol solvent extraction system, Pharm. Res. 20 (2003) 2028–2035. [50] I. De Marco, E. Reverchon, Supercritical carbon dioxide + ethanol mixtures for the antisolvent micronization of hydrosoluble materials, Chem. Eng. J. 187 (2012) 401–409. [51] V. Prosapio, E. Reverchon, I. De Marco, Antisolvent micronization of BSA using supercritical mixtures carbon dioxide + organic solvent, J. Supercrit. Fluids 94 (2014) 189–197. [52] V. Prosapio, E. Reverchon, I. De Marco, Production of lysozyme microparticles to be used in functional foods, using an expanded liquid antisolvent process, J. Supercrit. Fluids 107 (2016) 106–113. [53] N. Devi, D.K. Kakati, Smart porous microparticles based on gelatin/sodium alginate polyelectrolyte complex, J. Food Eng. 117 (2013) 193–204. [54] E. Reverchon, E. Torino, S. Dowy, A. Braeuer, A. Leipertz, Interactions of phase equilibria, jet fluid dynamics and mass transfer during supercritical antisolvent micronization, Chem. Eng. J. 156 (2010) 446–458. [55] E. Reverchon, G. Caputo, I. De Marco, Role of phase behavior and atomization in the supercritical antisolvent precipitation, Ind. Eng. Chem. Res. 42 (2003) 6406–6414. [56] E. Reverchon, I. De Marco, R. Adami, G. Caputo, Expanded micro-particles by supercritical antisolvent precipitation: interpretation of results, J. Supercrit. Fluids 44 (2008) 98–108. [57] A.P. Bünz, N. Zitko-Stemberger, R. Dohrn, Phase equilibria of ternary and quaternary systems containing carbon dioxide, water, isopropanol, and glucose, J. Supercrit. Fluids 7 (1994) 43–50. [58] B.-S. Chiu, G.T. Wilkinson, Ternary phase equilibria of the iso-propanol + water + carbon dioxide system at high pressure, Korean J. Chem. Eng. 16 (1999) 187–192. [59] N.E. Durling, O.J. Catchpole, S.J. Tallon, J.B. Grey, Measurement and modelling of the ternary phase equilibria for high pressure carbon dioxide–ethanol–water mixtures, Fluid Phase Equilib. 252 (2007) 103–113. [60] P. Traub, K. Stephan, High-pressure phase equilibria of the system CO2–water–acetone measured with a new apparatus, Chem. Eng. Sci. 45 (1990) 751–758. [61] J.H. Yoon, M.K. Chun, W.H. Hong, H. Lee, High-pressure phase equilibria for carbon dioxide–methanol–water system: experimental data and critical evaluation of mixing rules, Ind. Eng. Chem. Res. 32 (1993) 2881–2887.