NOZZLE CONSTRUCTION FOR PARTICLE FORMATION USING SUPER CRITICAL ANTI SOLVENT PRECIPITATION

Size: px

Start display at page:

Download "NOZZLE CONSTRUCTION FOR PARTICLE FORMATION USING SUPER CRITICAL ANTI SOLVENT PRECIPITATION"

Esther Meryl Hoover
6 years ago
Views:

1 NOZZLE CONSTRUCTION FOR PARTICLE FORMATION USING SUPER CRITICAL ANTI SOLVENT PRECIPITATION Hubert C. Pellikaan 1*, Frank E. Wubbolts 1,2 1 Feyeon Development & Implementation B.V., Rijnkade 17 A, 1382 GS Weesp, The Netherlands 2 Laboratory for Proess Equipment, Delft university of Tehnology, Leeghwaterstraat 44, 2628 CA Delft, The Netherlands hubert@feyeon.om Fax: ABSTRACT In the development of drug delivery systems for pulmonal delivery and ontrolled release systems, small partiles (<5 mirometer) with a narrow partile size distribution are desired. Reent tehniques using a superritial fluid as an anti-solvent are onsidered to be useful alternatives to produe fine powders. Co-preipitation an be realized in one step. In the superritial fluid antisolvent tehniques, arbon dioxide is used as an antisolvent for the solute but as a solvent with respet to the organi solvent. Here one speifi version, SAS (Superritial Anti Solvent), is under investigation. For this proess several nozzle onfigurations are used in order to ontrol partile morphology, partile size and partile size distribution. A nozzle onfiguration is presented that enables ontrol over proess parameters to optimise partile morphology, partile size and partile size distribution. A nozzle onstrution was built for partile prodution using the SAS proess in a ontinuous proedure. Experiments were done using dextran and l-pla to establish the influene of onentrations and dynamis of the proess on the produt parameters. Proess parameters are identified that ontrol partile size between 500 nanometer and 20 mirometer. Partile size distribution of the produed partiles is narrow, with a standard deviation of less than 30% of the volumetri median diameter. Furthermore, o-preipitation of model ompounds with these polymers of protein and steroid is arried out without influene on partile morphology.

2 INTRODUCTION The mironisation based on the use of superritial antisolvents (SAS) has been suggested during the last deade as an alternative to urrently available mironisation tehniques for the prodution of drug delivery partiles [1]. The SAS tehnology has proved to be able to produe partiles from low moleular ative ompounds as well as from polymeri matrix materials. For the prodution of drug delivery systems it is neessary to ontrol morphology, partile size and partile size distribution to a high standard. The purpose of this study was twofold. First we wanted to improve the ontrol over partile size by modifying the SAS proess. We improved the partile formation proess by varying the solvent and antisolvent flow rates as well as introduing a delay time after mixing the solvent and the antisolvent. Thus the phase separation ould be performed at a relatively low super saturation. Seondly we wanted to find out if this proess is able to inorporate model ompounds into biodegradable miropartiles. The following polymers were used: Dextran and l-poly lati aid (l-pla). As model ompounds were used holesterol and Holus Lanatus extrat (protein mixture). We haraterised the miropartiles by sanning eletron mirosopy (SEM) and partile size measurements. We determined the degree of holesterol enapsulation as well as the presene of Holus Lanatus extrat. EXPERIMENTAL a. Materials Dextran (Mw ) and Cholesterol were purhased from Sigma Aldrih. L-PLA (Mw ) was kindly provided by Pura Biohem (Gorinhem, The Netherlands). The Holus Lanatus protein extrat was kindly provided by Fornix Biosienes (Lelystad, The Netherlands). Dihloromethane and DMSO were purhased at Sigma Aldrih. Carbon dioxide is provided by Hoek-Loos, The Netherlands. All hemials were used without further purifiation.

3 b. Proedure A shemati diagram of the SAS apparatus that was employed in the experiments is shown in Fig. 1. A detailed diagram of the nozzle onfiguration used is shown in Fig. 2. In all experiments a temperature of 40 C and a pressure of 120 bar are used. All SAS experiments start with the delivery of CO2 to the partile olletion vessel and the nozzle onfiguration until the desired pressure is reahed. A steady antisolvent flow is then set. Then simultaneously the solution and antisolvent flows through the nozzle onfiguration are started. Due to the plug flow harater in the nozzle onstrution steady state onditions are obtained. The experiment ends when the delivery of solution to the nozzle is interrupted. However superritial CO2 ontinues to flow through the nozzle and the partile olletion vessel to wash the apparatus and the produt from residual solvent. Washing is stopped after 10 Kg s of CO2 have passed the setup. Figure 1: Diagram of the used SAS set up Solution CO2 Pi Fi Ti Ti CO2 Pi

4 Figure 2: nozzle onstrution First CO2 inlet Solution inlet Nozzle 0,1mm saphire Mixing hamber Partile growth zone RESULTS In table 1 all performed experiments are summarised. Table 1: Experimental onditions and results No X1* Solution CO 2 1 Mixing Mixing CO 2 2 Growth Partile S.D. Yield flow rate flow rate energy time (s) flow rate time (s) size (µm) (mw) (g/min) (µm) % % % , % % *X: molefration of solvent / antisolvent in the mixing hamber Partile formation of dextrane from DMSO yields spherial, unagglomerated partiles (figure 3). Also there is a lear relationship between molefration of solvent and antisolvent at the first mixing step and the resulting partile size. In figure 4 this relationship is graphially depited.

Figure 3: Dextrane partiles Figure 4: molefration / partile size Partile size (mirometer) 5 4 3 2 1 0 0 0,5 1 Mole fration CO2 (X1) Enapsulation of holesterol in

Enapsulation of protein in dextrane yields spherial unagglomerated partiles ontaining the protein (Figure 6). The protein is analysed as unaltered by SDS Page.

protein in dextran DISCUSSION These experiments show that the presented proess has a number of possibilities in the prodution of partiles for drug delivery [2].

5 Figure 3: Dextrane partiles Figure 4: molefration / partile size Partile size (mirometer) ,5 1 Mole fration CO2 (X1) Enapsulation of holesterol in l-pla also yields spherial unagglomerated partiles (Figure 5). Spetrophotometri analysis shows that l-pla partiles are loaded with 1,2 % holesterol. Enapsulation of protein in dextrane yields spherial unagglomerated partiles ontaining the protein (Figure 6). The protein is analysed as unaltered by SDS Page. Figure 5: Cholesterol in l-pla Figure 6: Holus l. protein in dextran DISCUSSION These experiments show that the presented proess has a number of possibilities in the prodution of partiles for drug delivery [2]. The partile size an be adjusted, and enapsulation of model ompounds is suessful. With the formation of dextran partiles there is a linear relationship between the mole fration antisolvent/solvent and the resulting partile size. As a theory that explains this relationship the authors suggest the

6 following: the phase split that ours when solvent and antisolvent are first mixed yields both a superritial gas phase as a visous droplet still rih in solvent. The mole fration at first mixing determines the solvent ontent of these droplets and thus the size of this droplet. Due to the fat that the antisolvent /solvent ratio is relatively low when ompared to other authors, the super saturation is low and the phase split only slowly takes plae. Therefore a relatively long time is required before phase separation is omplete. After this stage the seond antisolvent flow is added to extrat solvent from the droplets in order to obtain dry and non agglomerating partiles. Enapsulation of protein in these experiments was suessful at low protein onentrations; higher onentrations lead to agglomeration of partiles. For several proteins with a high ativity this limitation is not a problem. Future work is addressing the enapsulation of proteins at higher onentrations. Enapsulation of holesterol in l-pla yielded partiles with a holesterol ontent of 1% where an initial loading of 10% was set. Probably the enapsulated holesterol is extrated from the partiles during the washing of the partiles at the end of the proess. In further experiments this an be avoided by washing the partiles with o2 saturated with holesterol. CONCLUSION The presented proess is shown to be able to produe partiles of polymers with an adjustable size and a narrow partile size distribution. The enapsulation of model ompounds was suessful. ACKNOWLEDGEMENTS This work was supported by the European Commission, Competitive and Sustainable Growth (GROWTH) Programme (Projet GRD ). Fornix Biosienes, Lelystad, The Netherlands for providing Holus Lanatus protein extrat and performing the SDS-Page protein analysis REFERENCES [1] JUNG J., PERRUT, M., Journal of Superritial Fluids, Vol. 20, 2001, p.179 [2] PELLIKAAN, H.C., WUBBOLTS F.E. Appliation No: : PROCESS FOR SMALL PARTICLE FORMATION (2002)

AT2G02060 AT2G40260 AT2G AT4G04580 AT2G06020 AT2G AT5G06800 AT3G13040 AT2G AT5G AT3G04030

AT2G02060 AT2G40260 AT2G AT4G04580 AT2G06020 AT2G AT5G06800 AT3G13040 AT2G AT5G AT3G04030 AT5G45580 75 AT3G10760 63 AT5G05090 92 AT2G40970 AT3G46640 LUX/PCL1 66 82 AT5G59570 BOA AT4G18020 PRR2 97 AT2G20570 GLK1 55 47 AT5G44190 GLK2 58 AT1G49560 HHO6 AT4G37180 HHO5 AT2G03500 HHO4 98 45 AT1G13300