Gamma irradiation effects on poly(dl -lactictide-co-glycolide) microspheres

Journal of Controlled Release 56 (1998) 219–229

Gamma irradiation effects on poly( DL-lactictide-co-glycolide) microspheres a, a b b Luisa Montanari *, Monica Costantini , Elena Ciranni Signoretti , Luisa Valvo , b b b b Mara Santucci , Monica Bartolomei , Paola Fattibene , Sandro Onori , Antonio Faucitano c , Bice Conti d , Ida Genta d a

b

Istituto di Chimica Farmaceutica, Universita` degli Studi di Milano, V. le Abruzzi 53, 20131 Milano, Italia Laboratorio di Chimica del Farmaco, Laboratorio di Fisica, Istituto Superiore di Sanita` , V. le Regina Elena 299, 00161 Roma, Italy c Dipartimento di Chimica Generale, Universita` di Pavia, V. le Taramelli 11, 27100 Pavia, Italy d Dipartimento di Chimica Farmaceutica, Universita` di Pavia, V. le Taramelli 12, 27100 Pavia, Italy Accepted 19 May 1998

Abstract Gamma radiation treatment plays an increasingly important role in the sterilization / sanitization of pharmaceutical products. However, irradiation may affect the stability of the product and thus its safety of use. We investigated the influence of ionizing radiation on modified release microparticulate drug delivery systems made of two types of polylactide-coglycolide copolymers (PLG): RG 503 and RG 503H; these polymers have identical molecular weights but different chemical structures. The effect of g radiation on polymer stability of the raw polymers (P) and related microspheres (Ms) was evaluated. Samples were irradiated at different irradiation doses (5, 15 and 25 kGy) using 60 Co as radiation source. The microspheres were prepared using the spray drying technique. Degradation of PLG and related microspheres was evaluated during six months in terms of average molecular weight (Mw ) loss by gel permeation chromatography (GPC) and variation in glass transition temperature (T g ) using differential calorimetry (DSC). The presence of free radicals in the product was tested by electron paramagnetic resonance (EPR). Both P and Ms showed a trend in decreasing their Mw at time 0 as a function of irradiation dose. For RG503 the decay in Mw is always negligible for doses below 15 kGy while it is about 10% for 25 kGy. After 150 days Mw decay was 25% in the microspheres and 20% in the raw polymer. It was not possible to evaluate the radiation effect, at different storage times, for RG503H because this polymer resulted to be unstable even in the regular storage conditions without being irradiated. The concentration of radiation-induced free radicals was higher in RG 503H (both P and Ms) and they were more stable than the free radicals species observed in the case of polymer RG 503. Alterations and / or production of new radicals were observed on exposure of RG 503H microspheres to the light. Radiolytic degradation of RG 503 under vacuum is characterized by a prevalence of the chain scission events leading to a decrease of Mw . Some crosslinking can occur mainly in the post irradiation stage through the decay and coupling of the hydrogen abstraction radicals. A hydroperoxydative cycle, whose mechanism is suggested, is generated in the presence of oxygen.  1998 Elsevier Science B.V. All rights reserved. Keywords: Y irradiation; Biodegradable polymers; Polylactide-co-glycolide; Microspheres

*Corresponding author. Tel.: 139 02 29403194; E-mail: [email protected]. 0168-3659 / 98 / $ – see front matter  1998 Elsevier Science B.V. All rights reserved. PII: S0168-3659( 98 )00082-0

220

L. Montanari et al. / Journal of Controlled Release 56 (1998) 219 – 229

1. Introduction Gamma radiation processing plays an increasingly important role in the sterilization / sanitisation of pharmaceutical products. Sterile drug delivery systems can be terminally sterilized in their final container or manufactured by aseptic process. The first method is easier from an operative point of view and economically more convenient, but the absence of harmful side effects must be experimentally verified. Application of the ionizing radiation as sterilization method is permitted when it has been experimentally proved and documented that no harmful side effects occur, due to possible qualitative and quantitative alterations of product. Moreover, essentially for economic profits an unauthorized and uncontrolled use of radiation process could be imagined. Microparticulate drug delivery systems for parenteral administration are candidates for gamma sterilization. They are designed to control drug release and consist of biocompatible and biodegradable polymeric micromatrices containing drugs for chronic disease therapies (e.g. hormones, antitumoral drugs, antibiotics) [1]. The interaction of polymers with ionizing radiation varies greatly depending on the nature and the composition of the polymers [2] and it can affect their technological performances. The literature especially addressed to the effects of g radiation on microparticulate drug delivery systems is scanty and quite controversial [3–7]. The purpose of this research is to evaluate the influence of gamma radiation on stability and safety of microparticulate drug delivery systems constituted by polylactide-co-glycolide 50:50 (PLG) microspheres. Two 50:50 copolymers have been employed: RG 503 and RG 503H; they had the same molecular weight but different structure in such a way that RG 503H carries predominantly free carboxylic acid groups on one of the chain ends and therefore it is more hydrophilic. The microspheres have been prepared by the spray drying technique. Polymers and microspheres have been irradiated with 5, 15, 25 kGy doses. The effects of g irradiation on polymer weight average molecular weight (Mw ), both on raw poly-

mer (P) and microspheres (Ms), have been evaluated by gel permeation chromatography (GPC) and by differential scanning calorimetry (DSC). Morphological characterization of microspheres has been performed by light blockage apparatus and scanning electron microscopy (SEM). High energy irradiation is often connected with formation of radical species promoting related phenomena, and among them polymer degradation. The radicals can be interesting either from a toxicological and analytical point of view as irradiation indicators. The formation of stable free radicals both in the polymers and the microspheres has been evaluated by electronic paramagnetic resonance technique (EPR) with the aim to explain the radiolysis mechanism for polymer degradation. Moreover, the radicals can be interesting both from a toxicological and analytical point of view, as irradiation indicators.

2. Materials and methods

2.1. Materials Polylactide-co-glycolide (PLG) 50:50, RG 503, inherent viscosity 0.39 dl / g, 34 000 Mw , polydispersity index 3, range 2.63 to 3.21; Polylactide-coglycolide (PLG) 50:50, RG 503H, inherent viscosity 0.39 dl / g, 34 000 Mw , polydispersity index 3.5, range 3.21 to 3.82, Boehringer Ingelheim KG, Ingelheim am Rheim, G. All solvents employed, unless specified, were of reagent grade.

2.2. Methods 2.2.1. Preparation of microspheres The microspheres have been prepared by the spray-drying method. A solution of polymer (4% w / w) in methylene chloride has been sprayed through the standard nozzle (0.7 mm inner diameter) of a spray-dryer Lab-Plant model SD04 (Lab-Plant LTD, West Yorkshire, UK). The experimental conditions were set as follows: inlet temperature 47– 488C, outlet temperature 31–328C, flow rate 20 ml /

L. Montanari et al. / Journal of Controlled Release 56 (1998) 219 – 229

min. The microspheres were harvested and kept under vacuum for 24 h before use.

2.2.2. g irradiation Part of the irradiations were performed under air at room temperature in a 60 Co source (Gamacell Nordion) with total dose of 5, 15, 25 kGy at dose rates of 0.64 kGy / h. By thermometric control it was checked that the sample temperature during the irradiation did not appreciably increase above room temperature. This was expected because of the low dose rate of the irradiation (0.64 kGy / h). The irradiation pertaining that part of the ESR measurements devoted to the identification and characterization of the primary species were performed at 77 K in a 60 Co source with total doses of 25 kGy and a dose rate of 1.8 kGy / h. To this purpose the samples were sealed under high vacuum (5–10 torr) in quartz tubes and kept in a device filled with liquid N 2 during the irradiation. 2.2.3. Microsphere characterization 2.2.3.1. Morphology. The microspheres, before and after irradiation, were characterized for shape and surface characteristics by scanning electron microscopy (JSM-T 200-Jeol Italia S.p.a. – Pieve Emanuele, I). Microsphere size was determined before and after irradiation at 25 kGy. Granulometric analysis was performed by a light blockage apparatus HIAC / ROYCO model 3000 equipped with a HC60 sensor. Samples of microspheres were suspended in filtered double distilled water and analysed under gentle stirring. The results are the average of five determinations. 2.2.3.2. Molecular weight analysis. Weight average molecular weights (Mw ) of polymers and microspheres, before and after irradiation were determined by a Waters Model 5900 GPC instrument (Milford, MA, USA), equipped with three ‘Ultrastyragel’ ˚ porosity, at 408C, using columns of 10 4 , 10 3 , 500 A THF as eluent. An IR detector Waters 410 (Milford, MA, USA) was used for analyses. Mw were determined by the universal calibration method, based on polystyrene standards, Mw 90 000–2600 Daltons,

221

Polisciences (Polisciences, Warrington, PA, USA), using the Mark–Houwink parameters k55.49310 24 and a50.64 [8]. GPC was performed on polymers and microspheres immediately after irradiation (time 0) and 15, 30, 60 and 150 days after irradiation, for all radiation doses. Before GPC analysis all samples were stored in sealed vials as dry powder at 48C

2.2.3.3. Determination of glass transition temperature. Glass transition temperature (T g ) of the polymers and related microspheres has been determined before and immediately after irradiation by a differential calorimetry DSC 7 (Perkin Elmer, Langen, Germany) connected to a cooling system and maintained in inert atmosphere by a constant N 2 flow. The equipment has been calibrated with Indio (T m 5 156.68C). The samples (about 4 mg) have been placed in sealed aluminium pans and pre-treated in a heating– cooling cycle between 30–808C at 58C / min heating rate, and 80–308C at 108C / min cooling rate. T g determination has been performed subsequently by heating the samples from 308C to 658C at heating rate of 58C / min [9].

2.2.4. EPR measurements The EPR measurements were carried out with microspheres and powder polymer samples by using Varian X-band E 4 , E 8 and a Brucker EMS 104 spectrometers. Two sets of experiments have been made starting with samples g irradiated at room temperature under air (g dose 25 kGy) and with samples g irradiated at the liquid N 2 temperature. In the latter case the EPR spectra were first recorded at 77 K, in order to identify the primary species, and subsequently after stepwise increases of the temperature above 77 K up to room temperature. This latter procedure was aimed to obtain suitable conditions for the reactions of the primary species and for the identification of the secondary radicals. The EPR spectra were analyzed by computer simulation using a Hamiltonian including the Zeeman electronic and nuclear terms and the hyperfine terms with isotropic and anisotropic components for the g and hyperfine tensors: H 5 b S ? g ? H 1 SiS ? A ? Ii 2 SiIi ? H

L. Montanari et al. / Journal of Controlled Release 56 (1998) 219 – 229

222

Table 1 Weight average molecular weights (Mw ) of PLG RG 503 polymer (P) and of PLG RG 503 microspheres (Ms) before and after irradiation Storage time (days)

Radiation doses (kGy) 0

0 15 30 60 150

5

15

25

P

Ms

P

Ms

P

Ms

P

Ms

19 800 – – 18 700 18 400

19 100 19 300 19 200 19 400 18 500

18 800 18 300 18 600 18 700 17 300

19 000 18 500 20 400 18 900 17 500

18 700 18 000 18 000 16 800 15 900

17 800 17 400 18 400 17 600 15 900

17 900 16 900 16 700 17 000 15 600

17 500 16 700 16 200 16 100 14 600

The microwave power level was set below 1 mW to avoid power saturation.

3. Results and discussion

3.1. Polymer characterization before irradiation Weight average molecular weights of the polymers determined by GPC analysis, before irradiation, are reported in Table 1 and Table 2. The values obtained for not treated polymers at time 0 were significantly lower than those given by the manufacturer because of the different solvent (THF instead of CHCl 3 ) and calibration method (universal calibration method) used in GPC analysis [10]. RG 503H polymer revealed to be more unstable than RG 503 polymer at room conditions. Mw decreases of about 7% in RG 503 after 5 months, while it decreases of about 45% in RG 503H polymer.

3.2. Microsphere characterization before irradiation Taking into account GPC measurements accuracy, Mw of PLG microspheres before irradiation resulted to be equivalent to those of raw polymers. The microspheres obtained were of good morphological characteristics, spherical shape and smooth surface. Fig. 1a shows as an example the scanning photomicrograph of RG 503H microspheres before irradiation. Microsphere size, as determined by granulometric analysis, resulted in the same range between 10–25 mm for both polymers (Fig. 2a). Thus no differences have been highlighted in the particle size distribution of the two batches of microspheres made of RG 503 and RG 503H respectively. The microspheres resulted always in a highly electrostatic powder: this characteristic is due to spray-drying process. This process, like other processes involving fluidization of particles by means of high volumes of dry air, determines impacts among

Table 2 Weight average molecular weights (Mw ) of PLG RG 503H polymer (P) and of PLG RG 503H microspheres (Ms) before and after irradiation Storage time (days) 0 15 30 60 150

Radiation doses (kGy) 0

5

15

25

P

Ms

P

Ms

P

Ms

P

Ms

19000 – – 18200 10200

20700 – – 19500 8900

19900 18300 18900 18600 12500

19500 18700 19300 18700 13000

18300 18000 16800 16000 11100

18200 17000 17900 17100 10900

16500 16600 16500 16000 8500

16 300 16 900 15 700 14 900 9300

L. Montanari et al. / Journal of Controlled Release 56 (1998) 219 – 229

Fig. 1. Photomicrographs of RG 503H (a) before irradiation, (b) after irradiation.

the solid microsparticles passing through the drying chamber that generate considerable static electricity. especially when the solid particles have poor electrical conductivity [11].

3.3. Y irradiation effect on polymers and microspheres Fig. 1b shows, as an example, the photomicrograph of the RG 503H microspheres after irradiation at 25 kGy. No changes in microsphere shape and size have been detected, for all irradiation doses, that can be ascribed to g irradiation. Furthermore, the microspheres in bulk resulted to be more closepacked because of some aggregation occurred after irradiation that can be highlighted above all in RG

223

503H microspheres. No change in the morphology of RG503 microspheres has been highlighted by SEM (data not reported). Granulometric analysis of the microspheres made of both polymers (Fig. 2b, c, d) showed an increase in microparticle size that can be related to irradiation dose. This is evident especially in the microspheres made of RG 503H: before irradiation 70% of particles had diameter of 15 mm while after irradiation at 25 kGy 80% of particles had diameter of 25 mm. This phenomenon can be ascribed to aggregates observed by SEM. Table 1 and Table 2 show the Mw of PLG raw polymers, and related microspheres. Radiation effect on Mw can be considered equivalent for the polymer and the related microspheres in spite of greater specific surface area of the microspheres. Both polymers and microspheres showed a trend in decreasing their Mw , at time 0, as a function of irradiation dose. For RG 503 the decay in Mw is always negligible for doses below 15 kGy while it is about 10% for 25 kGy. RG 503H polymer, that has prevalently free COOH end groups, seems to be more sensitive to g radiation with respect to the homologous RG 503 carrying esterified COOH end groups. For the latter, Mw decay is about 20% in the microspheres and 13% in the raw polymer when they have been irradiated at 25 KGy. It was not possible to evaluate the radiation effect on storage stability for RG 503H because this polymer resulted to be unstable even in the regular storage conditions without being irradiated. Thus Mw decay in the not irradiated sample is equivalent to that of irradiated samples. Further, after 150 days, Mw decay in RG 503 led to a total decrease of 25% in the microspheres and 20% in the raw polymer.

3.4. Evaluation of polymer thermal behaviour Table 3 lists the glass transition temperatures (T g ) of polymers and microspheres before and after irradiation. T g does not change either for the polymers or the microspheres up to the highest radiation dose administered (25 kGy). T g determinations have been repeated 150 days after irradiation and no

224

L. Montanari et al. / Journal of Controlled Release 56 (1998) 219 – 229

Fig. 2. Particle size distribution of RG 503 and RG 503H microspheres (a) before irradiation, (b) after irradiation at 5 kGy, (c) after irradiation at 10 kGy, (d) after irradiation at 25 kGy.

significant changes of the values obtained at zero time have been highlighted (data not reported).

Table 3 Gamma radiation effect on glass transition temperatures

RG 503 polymer RG 503H polymer RG 503 microspheres RG 503H microspheres

T g (8C) 0 kGy

T g (8C) 25 kGy

37.2 37.3 37.5 37.4

36.7 36.7 37.4 37.1

3.5. EPR analysis 3.5.1. Y irradiation at room temperature Following g irradiation at room temperature under air with 25 kGy the RG 503H copolymer yields a

L. Montanari et al. / Journal of Controlled Release 56 (1998) 219 – 229

225

weak signal having the appearance of a doublet of about 22 G. This doublet, by comparison with the EPR spectra of Fig. 3 and Fig. 4, is suggested to correspond to the central part of the 1:3:3:1 quartet due to the tertiary macroalkyl radical –C ? (CH 3 )–. This signal decays rather rapidly at room temperature under air with t ]12 of about 5 h (Fig. 5). The microspheres made of RG 503H did not give any signal before g irradiation. The signal after irradiation (Fig. 6) was different from that of the corresponding raw polymer: the microspheres give a more stable signal (60% decay in 30 days) made of several components. EPR signal related to the microspheres changed after exposing them to the light: successive exposures to the light highlighted further changes in the signal that are stable for shape and intensity (Fig. 7).

Fig. 4. EPR spectrum of RG 503H polymer after irradiation at 77 K under vacuum, gdose 27.6 kGy. A – recorded at 77 K after irradiation; B – recorded at 77 K after annealing; C – recorded at 298 K after admission of air.

These changes in EPR signal, after several successive exposures to the light, are typical of the RG 503H microspheres and they have not been detected in the EPR spectrum of the raw polymer. It is hypothesized that the signal recorded on the RG 503H microspheres could be caused by their preparation process.

Fig. 3. EPR spectra of RG 503 polymer after irradiation at 77 K under vacuum. Ydose 27.6 kGy; A – recorded at 77 K after irradiation; B – recorded at 77 K after annealing; C – recorded at 298 K after admission of air.

3.5.2. Y irradiation at 77 K and oxygen effect The EPR spectra of the RG503 and RG503H (Figs. 3 and 4) polymers are identical thus suggesting that the structural differences of the chain ends play a minor role in the overall radiolysis mechanism. The major species contributing to the 77 K spectra

226

L. Montanari et al. / Journal of Controlled Release 56 (1998) 219 – 229

previously annealed to room temperature, the build up of the EPR signal showing the characteristic anisotropy of a peroxyl radical was observed (Fig. 4c). No such a signal could be detected instead with the RG 503H sample. The formation of the peroxyl radical is controlled by the oxygen diffusion according to the scheme: O 2 (out ) → O 2 (in) -C ? (CH 3 )- 1 O 2 → -C(CH 3 )(OO ? )-

Fig. 5. Room temperature post irradiation fading of RG 503H copolymer EPR signal after irradiation at room temperature under air. The two curves refer to the intensity of the two more intense central peaks.

are the chain scission radicals –CH ? (CH 3 ). Also present are minor concentrations of the other chain scission radical –OCH ?2 and the radical. –C ? (CH 3 )– formally arising from the rupture of the tertiary C–H bond in the chain. On warming above 77 K the species –CH ? (CH 3 ) and –OCH ?2 decay partially generating the radical –C ? (CH 3 )–. The latter species decay slowly at room temperature affording evidence of the presence of another weak EPR pattern not yet identified. By opening to air the sample of RG 503

It is thus conceivable that the lack of observation of the peroxyl radical with the RG503H polymer sample be due to a slower oxygen diffusion rate caused by a different sample morphology and powder particle size.

3.5.3. Radiolysis mechanism The reaction scheme reported here below is based on the assumption of a prevalence of the ionic component in the radiolysis mechanism due to the polarity of the polymers. Following the primary ionization radical cations and radical anions are formed by electron loss and electron capture respectively. The radical anions decompose already at 77 K during the irradiation generating chain scission radicals. The most probable fate for the cation-radicals is the proton loss generat-

Fig. 6. EPR spectrum of RG 503H microspheres after irradiation at 25 kGy at room temperature under air.

L. Montanari et al. / Journal of Controlled Release 56 (1998) 219 – 229

ing neutral free radicals. With the RG503 and RG503H structures the favored site for the deprotonation is expected to be the tertiary C–H bond yielding the radical –C ? (CH 3 )–. The evolution of the EPR spectra on warming to room temperature is interpreted in terms of H abstraction reactions by the chain scission radicals at the weakest tertiary C–H bonds in the chain leading to an increase of the –C ? (CH 3 )– radical concentration. The picture stemming from these results is that the radiolytic degradation of the RG503 and RG503H polymers under vacuum is characterized by a prevalence of the chain scission events leading to a decrease of the average molecular weight. Some cross linking can occur

227

mainly in the post irradiation stage through the decay and coupling of the hydrogen abstraction radicals –C ? (CH 3 )–.

3.5.4. Post irradiation reaction with oxygen A simplified reaction mechanism initiated under O 2 atmosphere by the –C ? (CH 3 )– radicals is shown here below. After the admission of oxygen an hydroperoxydative cycle can take place leading to the formation of hydroperoxides –C(OOH)(CH 3 )–. The thermal and / or photolytic decomposition of these compounds will cause branching through the formation of alkoxy

228

L. Montanari et al. / Journal of Controlled Release 56 (1998) 219 – 229

(and OH) radicals and their favoured reaction modes which are the hydrogen abstraction and the b-scission. In the presence of water, the formation of CO 2 is expected to promote the hydrolytic degradation of the polymer through the decrease of the pH. The role of the chain scission radicals in the scheme is limited to a contribution to the main hydroperoxidative cycle through hydrogen abstraction reactions. For these species however also other reaction paths can be foreseen affording peracids and carboxylic acids which can play catalytic roles especially in the in source hydrolytic degradation of the polymers. The nature of the radicals identified are suggesting crosslinking events occurring through the actual couplings and also chain scissions

The competition among these two mechanisms will attenuate the effect on the Mw . The observation of a consistent decrease of Mw at the onset dose of 15 kGy suggests that chain scission is slightly prevalent.

4. Conclusions Both polymers and related microspheres show degradation effects depending on irradiation dose and storage time. Different sample morphology and powder particle size did not affect the Mw decay of the irradiated samples. RG503 polymer is more stable and irradiated up to 15 kGy it is stable for 5 months. After irradiation at 25 kGy Mw decay led to a total decrease of 25% in the microspheres and 20% in the raw polymer. RG503H polymer resulted unstable even in the regular storage conditions without being irradiated,

L. Montanari et al. / Journal of Controlled Release 56 (1998) 219 – 229

229

Fig. 7. EPR spectrum of RG 503H microspheres after irradiation at 25 kGy at room temperature under air followed by exposure to daylight.

thus it was not possible to quantitate the radiation effects at different radiation doses and storage times. The EPR analysis of the radical intermediates performed as a function of the temperature starting from 77 K has afforded basic information on the main features of the radiolysis mechanism. This mechanism, judging from the nature of the primary and secondary radical species, is not appreciably influenced by the differences of chemical structure between RG 503 and RG 503H copolymers. However the concentration of radiation-induced free radicals was higher in RG 503H (Ms and P) and they were more stable than the free radicals species observed in the case of polymer RG 503. Alteration and / or production of new radicals were observed on exposure of RG 503H microspheres to the light.

Acknowledgements This research is supported by a grant for the project ‘Sterilization processes and their influence on ` drug properties’, Istituto Superiore di Sanita.

References [1] H. Okada, H. Toguchi, Biodegradable microspheres in drug delivery, Crit. Rev. Ther. Drug Carrier Systems 12(1) (1995) 1–99.

[2] W.E. Skiens, Sterilizing radiations effects on selected polymers, Radiat. Phys. Chem. 15 (1980) 47–57. [3] H. Mitomo, Y. Watanabe, F. Yoshii, K. Makuuchi, Radiation effect on polyesters, Radiat. Phys. Chem. 46 (1995) 233– 242. [4] A.G. Hausberger, R.A. Kenley, P.P. DeLuca, Gamma irradiation effects on molecular weight and in vitro degradation of poly( DL-lactide-co-glycolide) microparticles, Pharm. Res. 12 (1995) 851–856. [5] J.H. Collett, L.Y. Lim, C. Booth, Drug release from irradiated poly( L-lactic-co-glycolic acid) microspheres, Proceed. Intern. Symp. Control. Rel. Bioact. Mater. Contr. Rel. Soc. 17 (1990) 285–286. [6] S.R. Hartas, J.H. Collett, C. Booth, The influence of gamma irradiation on the release of melatonin from poly(lactide-coglycolide) microspheres, Proceed. Intern. Symp. Control. Rel. Bioact. Mater. Contr. Rel. Soc. 19 (1992) 321–322. [7] M.B. Sintzel, A. Merkli, C. Tabatabay, R. Gurny, Influence of irradiation sterilization on polymers used as drug carriers, Drug Dev. Ind. Pharm. 23 (1997) 857–878. [8] J.A.P.P. Van Dijk, J.A.M. Smit, Characterization of poly( DLlactic acid) by gel permeation chromatography, J. Polym. Sci. 21 (1983) 197–208. [9] A.G. Hausberg, P.P. DeLuca, Characterization of biodegradable poly( DL-lactide-co-glycolide) polymers and microspheres, J. Pharm. Biomed. Anal. 13 (1995) 747–760. [10] D. Benedix. Some problems in the molecular weight determination of resorbable polymers by GPC and viscosimetry. Symposium on Characterization of macromolecules used as Pharmaceutical excipient. Gothenburg, Sweden, March 7–9 (1990). [11] C.Y. Wen, A.F. Galli. Dilute phase systems, in: J.F. Davidson, D. Harrison (Eds.), Fluidization, Chpt. 16, 1971, pp. 677–710.