Irradiation Stability of Saturated and Unsaturated ... - ACS Publications

Biomacromolecules 2001, 2, 1015-1022

1015

γ-Irradiation Stability of Saturated and Unsaturated Aliphatic Polyanhydrides-Ricinoleic Acid Based Polymers Doron Teomim,† Karsten Ma¨der,‡ Alfonso Bentolila,† Amir Magora,†,§ and Abraham J. Domb*,† Department of Medicinal Chemistry and Natural Products, School of Pharmacy, Faculty of Medicine, The Hebrew University of Jerusalem, Jerusalem 91120, Israel; and Department of Pharmacy, Humboldt University, Berlin, Germany Received April 19, 2001; Revised Manuscript Received May 22, 2001

The effect of terminal sterilization by γ-irradiation on several ricinoleic acid based polyanhydrides was investigated. The following polymers were used: poly(ricinoleic acid maleate) [P(RAM)], poly(ricinoleic acid succinate) [P(RAS)], poly(hydroxy stearic acid succinate) [P(HSAS)], poly(hydroxy stearic acid maleate) [P(HSAM)], and their copolymers with sebacic acid. The polymers were irradiated with an absorbed dose of 2.5 or 10 Mrad by means of a 60Co source under dry ice or at room temperature. No differences were found between samples irradiated under dry ice and at room temperature. Polymers prepared from monomers containing maleate residues, which contain double bonds adjusted to the anhydride linkage along the polymer chain, decreased in molecular weight, became insoluble, and showed fast hydrolytic degradation. For example, p(RAM), p(HSAM), and their copolymers with sebacic acid decreased in Mw from about 10 000 to about 2000, and from about 30 000 to about 5000, respectively, while polymers based on RAS and HSAS remained stable. This phenomenon was explained by an anhydride interchange-self-depolymerization process of the unsaturated anhydride bonds induced by γ-irradiation. This explanation was supported by the depolymerization of another class of polymers having an anhydride bond between two double bonds, fumaric acid anhydride polymers. The anhydride bond that lies between two double bonds was found to be more sensitive to γ-irradiation. This anhydride bond may be cleaved to form two radicals that further react with aliphatic anhydride bonds along the polymer chain to form inter- and/or intracyclization products. Introduction Poly(anhydrides) are hydrolytically degradable polymers which have been used as vehicles for controlled delivery of drugs.1,2 Drug release rate and hydrolytic degradation of polyanhydrides can be tailored by using various hydrophobic and hydrophilic diacid monomers.3 A new class of biodegradable polyanhydrides based on ricinoleic acid, has been synthesized.4 These polymers were prepared from ricinoleic acid maleate or ricinoleic acid succinate diacid half esters, prepared by reacting ricinoleic acid with succinic or maleic anhydride. Pure diacid monomers (>99%) were synthesized and copolymerized with sebacic acid by melt condensation to yield film-forming polymers with molecular weights exceeding 40 000.4 These polymers are soluble in common organic solvents and melt at temperatures below 100 °C, which allows their melt fabrication into microspheres and implants. These polymers hydrolyze into their ricinoleic acidcontaining acid ester monomers, which further hydrolyze to * Corresponding author, Tel: 972-2-6757573, Fax: 972-2-6757629, E-mail: [email protected]. Affiliated with the David. R. Bloom Center for Pharmacy at the Hebrew University, and the Alex Grass center for Drug Design and Synthesis, The Hebrew University. † The Hebrew University of Jerusalem. ‡ Humboldt University. § Current address: Division of Identification and Forensic Science, Israel Police Headquarters, Jerusalem 91906, Israel.

ricinoleic acid and succinic or maleic acid. The inflammatory response after subcutaneous implantation of polymer samples following 21 days post-implantation was only minimal to mild, comparable to that noted with the clinically used Vicryl-absorbable surgical suture.4 The main application of these polymers is as implantable drug carriers. It is therefore essential to properly sterilize the device prior to administration. Sterilization should inactivate any active biological substances throughout the polymer mass without affecting the device properties; thus, common sterilization methods such as autoclaving or ethylene oxide, are not relevant.5 During the course of introducing this class of polymers for clinical use, polymers based on ricinoleic acid4 were sterilized by γ-irradiation. It was found that although most polymers did not change their molecular weight and the hydrolytic degradation and drug release, one polymer containing ricinoleic acid maleate moieties showed a significant decrease in molecular weight along with rapid degradation and drug release from the polymer. These results indicated that some polyanhydride structures are affected by γ-irradiation. It is therefore the purpose of this study to investigate the γ-irradiation effect on saturated and unsaturated polyan-

10.1021/bm010078n CCC: $20.00 © 2001 American Chemical Society Published on Web 07/06/2001

1016

Biomacromolecules, Vol. 2, No. 3, 2001

hydrides using ricinoleic acid based polyanhydrides as representative polymers. The most effective method for the terminal sterilization of moisture- and heat-sensitive polymer articles is exposure to ionizing radiation.6 The advantages of high-energy ionizing radiation sterilization are high efficiency, negligible thermal effects, and sterilization of ready-to-use packaged products. The irradiation resistance of various microorganisms varies with the type of organism;7 it is currently thought that a radiation dose of 2.5 Mrad is generally sufficient for the sterilization of medical implants.7 On this premise, it is important to investigate the influence of irradiation on the physicochemical properties of the chosen polymer because radiation-induced changes might affect the mechanical properties, hydrolytic degradation behavior, drug-release properties, and even the biocompatibility of implantable polymers.8 A number of studies have addressed the effects of γ-sterilization on biodegradable polymers. It was reported that biodegradable polyesters undergo chain scission and cross-linking after exposure to γ rays.9 Usually, γ-irradiation of biodegradable poly(lactic-co-glycolic acid) (PLGA) or poly(lactic acid) (PLA) reduces the molecular weight, and eventually accelerates the polymer hydrolysis rate. Therefore, γ-irradiation affects both device hydrolysis and release profile of incorporated drugs.10 Previous publications indicated that polyanhydrides are stable to γ-irradiation for an extended time period.2,11 This work describes the effect of g-irradiation on saturated and unsaturated polyanhydrides based on ricinoleic acid with an emphasis on the general consequences of this study on the irradiation stability of polyanhydrides in general. Experimental Section Materials. Ricinoleic acid, 85% pure (Fluka, Buch, Switzerland), maleic, succinic, and acetic anhydride (Merck, Darmstadt, Germany), Meth-Prep II, methanol esterification reagent (Alltech, Deerfield, IL), and 3% palladium on carbon (Aldrich, Milwaukee, WI) were used in this study. All solvents were analytical grade from Aldrich or Frutarom (Haifa, Israel). Instrumentation. Infrared (IR) spectroscopy (PerkinElmer System 2000 FT-IR) was performed on monomer, prepolymer, and polymer samples cast on NaCl plates from dichloromethane solution. Thermal analysis was determined on a Mettler TA 4000-DSC differential scanning calorimeter (DSC), calibrated with zinc and indium standards, at a heating rate of 10 °C/min. Molecular weights of the polymers were estimated on a size exclusion chromatography (SEC) system consisting of a Spectra Physics (Darmstadt, Germany) P1000 pump with UV detection (Applied Bioscience 759A Absorbance UV detector) at 254 nm, a Rheodyne (Coatati, CA) injection valve with a 20 µL loop, and a Spectra Physics Data Jet integrator connected to a computer. Samples were eluted with CHCl3 through a linear Styrogel column (Waters, 10 Å pore size) or a Phenogel column for the determination of molecular weights in the range 100-1000 (Phenomenex, CA) at a flow rate of 1 mL/min. The molecular weights were

Teomim et al.

estimated relative to polystyrene standards (Polyscience, Warrington, PA) with a molecular weight range of 2101 500 000 using a WINner/286 computer program. 1H NMR spectra (CDCl3) were obtained on a Varian 300 MHz spectrometer using CDCl3 as solvent and tetramethylsilane (TMS) as reference. Catalytic hydrogenation was carried out with 3% palladium on carbon powder using a Parr 3911EK hydrogenation apparatus (Parr Instruments Moline IL). Fatty acid degradation products were esterificied using the commercial esterification reagent, Meth-Prep II, and determined by gas chromatography using a Perkin-Elmer 1020 plus GC, with a Silar 10C chromatographic column (Alltech) and FID detector using a temperature gradient of 5 °C/min from 140 to 240 °C. Ricinoleic Acid Based Monomer Synthesis. Ricinoleic acid maleate (RAM) or succinate (RAS) was synthesized as previously described.4 Briefly, a solution of ricinoleic acid (0.12 mol) and maleic or succinic anhydride (0.24 mol) in toluene (70 mL) was stirred at 90 °C overnight. The solution was washed four times with distilled water, dried over MgSO4, and evaporated to dryness to yield an orange-colored oil (84%). Purification by silica gel column (mesh 70-230), using a petroleum ether/ethyl acetate/acetic acid (80/30/1 v/v/ v) mixture as eluent, yielded a clear oil or a white solid of ricinoleic acid maleate (RAM) and ricinoleic acid succinate (RAS), respectively. Hydrogenation of RAS gave 12hydroxystearic acid succinate (HSAS) using catalytic hydrogenation. Hydroxystearic acid maleate (HSAM) was synthesized by the reaction of 12-hydroxystearic acid and maleic anhydride. Prepolymer Synthesis. Anhydride prepolymers based on sebacic acid and ricinoleic acid monomers were prepared by refluxing the individual monomers in excess acetic anhydride (1:5 w/v) for 30 min and evaporating to dryness. The hot, clear, viscous prepolymer was dissolved in dichloromethane (1:1 v/v) and poured into a mixture of ether/ petroleum ether (1:1). The oily (RAM and HSAM) or solid (SA, RAS, and HSAS) prepolymers were collected and dried under vacuum at room temperature (>96% yield). Fumaric acid prepolymer was prepared by the reaction of fumaric acid with acetic anhydride. In brief, fumaric acid (20 g, 0.17 mol) was added to acetic anhydride (250 mL) and heated at reflux for 5 h. The product was crystallized at 20 °C for 3 days. The white crystalline precipitate was filtered and washed with dry ether to give 12 g (45% yield). Polymer Synthesis. Polyanhydrides were prepared by melt condensation of the prepolymers under vacuum.12 In a typical experiment, RAM and SA anhydride prepolymers (50:50 weight ratio) were melt polymerized at 150 °C for 4 h under a vacuum of 0.3 mmHg with constant stirring. The resulting polymers were dissolved in CH2Cl2 (1:5 w/v) and precipitated into excess of diisopropyl ether. The precipitate was collected by filtration and dried. Homopolymers were synthesized similarly to give viscous oils or semisolid products. P(CPP: SA) 20:80,13 P(FA:SA),14 and P(L-lactide)15 were synthesized as previously described. The molecular weights of the polymers as estimated by SEC were as follows: P(RAM:SA) 50:50, Mw 31 200, Mn 12 800; P(RAS:SA) 50:50, Mw 48 700, Mn 21 700; P(HSAS:

Polyanhydrides-Ricinoleic Acid Based Polymers

SA) 50:50, Mw 41 000, Mn 19 700; P(HSAM:SA) 50:50, Mw 30 000, Mn 18 000; P(RAM), Mw 10 000, Mn 7600; P(RAS), Mw 19 500, Mn 9600; P(HSAS), Mw 19 000, Mn 12 000; P(HSAM), Mw 10 000, Mn 7000; P(CPP:SA) 20:80, Mw 44 400, Mn 14 800; P(L-lactide), Mw 51 300, Mn 26 800; P(FA:SA) 10:90, Mw 8000, Mn 4500; P(FA:SA) 30:70, Mw 5000, Mn 3000; and P(FA:SA) 50:50, Mw 5700, Mn 3000. γ-Irradiation. Rectangular polyanhydride samples, 3 × 5 × 11 mm (300 mg), were prepared by melt casting at 5 °C above the melting temperature of the polymer for 5 min into a rubber mold. L-PLA (mp ) 180 °C) was irradiated as a powder. All polymers were irradiated with an absorbed dose of 2.5 Mrad by means of a 60Co source (450 000 Ci; 9.5 h). The irradiation was conducted at Sor-Van Radiation Ltd. (Kiryat Soreq, Yavne, Israel) Three sets of experiments were conducted. 1. Ricinoleic acid based copolymers with sebacic acid, P(CPP:SA), and L-PLA were divided into one of three groups: irradiated under dry ice, irradiated at room temperature, and not irradiated as control. 2. Ricinoleic acid based homopolymers, diacid monomers, and ricinoleic acid were irradiated at room temperature with an absorbed dose of 2.5 or 10 Mrad. 3. P(FA:SA) with three different FA:SA ratios (10:90, 30: 70, and 50:50) were irradiated at room temperature with an absorbed dose of 2.5 Mrad. All polymer samples were analyzed for molecular weight (estimated by SEC), viscosity, IR spectra, solubility in dichloromethane, and physical appearance before and after irradiation. The molecular weights of ricinoleic acid based monomers were calculated from the titration of the diacid in THF with NaOH using phenolphthalein as indicator, and by SEC using a column with molecular weight separation range of 100-1000. EPR Measurements. EPR experiments were performed at room temperature, using a Varian E109 spectrometer with the following settings: microwave frequency, 9.25 GHz; modulation amplitude, 0.1 mT; center field, 330 mT; scan range, 20 mT. In Vitro Hydrolytic Degradation of Polymers. The in vitro hydrolysis of the irradiated and nonirradiated polymers was carried out by placing rectangular samples of polymers (3 × 5 × 5 mm) in 10 mL of 0.1 M phosphate buffer at pH 7.4 and 37 °C with constant shaking (100 rpm). At each sampling time point, a polymer sample was taken out of the buffer and dried at room temperature for 2 h, and the hydrolysis of the polymer was monitored by (a) weight loss of the sample, (b) disappearance of the anhydride bonds by IR spectroscopy, and (c) changes in molecular weight of the polymer by SEC. The degradation products (fatty acids) were extracted from the aqueous phase by adding 1 mL of chloroform and vortexing for 1 min. The organic phase was collected and dried by a stream of nitrogen, followed by 2 h of lyophilization to remove all traces of water. The dry residue was dissolved in a mixture of 50 µL of toluene, 10 µL of methanol, and 20 µL of Meth-Prep II methanol esterification reagent. The solution containing the derivatized acids was then analyzed by gas chromatography16

Biomacromolecules, Vol. 2, No. 3, 2001 1017

Figure 1. Structure of irradiated polymers: (a) ricinoleic acid based polymers; (b) reference polymers.

Polymer Stability. Stability testing was conducted by placing rectangular samples of polymers (3 × 5 × 5 mm) at -10 °C under an argon atmosphere and monitoring the changes in molecular weight for a period of 3 months. Results and Discussion Polymer Characterization. The stability of ricinoleic acid based polyanhydrides to γ-irradiation was evaluated after irradiating rectangular samples with an absorbed dose of 2.5 Mrad. The clinically used polymers P(CPP:SA) 20:80 and P(L-lactide) were used as controls. The structures of the polymers are shown in Figure 1. The color and appearance

1018


Teomim et al.

Table 1. Molecular Weight and DSC Data Changes of Irradiated Polymersa 2.5 Mrad (25 °C)

before irradiation

2.5 Mrad (-20 °C)

polymer

Mw

Mn

Tm, °C

∆H, J/g

Mw

Mn

Tm, °C

∆H, J/g

Mw

Mn

Tm, °C

∆H, J/g

P(RAS:SA)50:50 P(HSAS:SA)50;50 P(RAM:SA)50:50 P(HSAM:SA)50:50 P(CPP:SA)20:80 l-PLA

48 700 41 000 31 200 30 000 44 000 51 300

21 700 20 000 13 000 19 000 15 000 27 000

61.1 70.4 59.3 67.3 68.0 179

65.8 78.4 37.9 50.6 64.0 43.0

34 000 33 500 2 500 8 000 40 000 57 000

13 000 16 000 1 800 7 300 15 000 31 000

60.8 71.0 58.9 68.1 68.7 181

66.0 79.0 38.4 51.2 64.6 44.2

50 000 37 000 3 300

18 000 17 000 2 500

61.5 70.6 60.2

65.5 78.5 37.4

40 000 54 000

15 000 33 000

67.0 181.2

65.1 45.0

a Irradiation was carried with an absorbed dose of 2.5 Mrad by means of a 60Co source (450 000 Ci; 9.5 h), M was determined by SEC, T and ∆H w m were recorded by DSC at 10 °C/min. All polyanhydrides had typical IR absorptions at 1740 and 1810 cm-1 (symmetrical and asymmetrical anhydride CdO stretching bands) before and after irradiation.

of all samples were the same before and after irradiation. Slight differences in molecular weight, solubility, and IR and NMR spectra were noted between samples irradiated in dry ice and at room temperature. The solubility of the polymers was determined in dichloromethane (100 mg polymer in 1 mL of solvent). All samples except P(RAM:SA) and P(HSAM:SA) dissolved immediately at room temperature to form a clear solution. P(RAM: SA) and P(HSAM:SA) formed insoluble gels, which gradually solubilized after continuous stirring for 24 h at room temperature. No difference was observed in thermal behavior (Tm, ∆H) of the irradiated and nonirradiated polymers. To determine the changes in molecular weight of the polymers before and after irradiation, polymer samples were dissolved in chloroform and analyzed by SEC. Both P(RAS: SA) and P(HSAS:SA) were slightly affected by γ-irradiation. In both cases, the molecular weight obtained after irradiation at room temperature was lower than that obtained in dry ice. Similar results were obtained for the P(CPP:SA) reference polymer. P(RAM:SA) and P(HSAM:SA) were analyzed after 24 h of stirring in chloroform and showed a significant lowering of Mw (from 31 200 to 2500, and from 30 000 to 8000, respectively) (Table 1). When the same nonirradiated polymers were stirred in chloroform for the same time period, the molecular weight dropped from about 30 000 to 15 000 and 20 000 for P(RAM:SA) and P(HSAM:SA), respectively. It should be noted that the nonirradiated polymers dissolved immediately in chloroform and the decrease in molecular weight due to depolymerization started only after 8 h of stirring. This phenomenon is known for other polyanhydrides.17 The gel formation and the drastic decrease in molecular weight are attributed to γ-irradiation effects. Since low molecular weight polymers were obtained after the irradiation of P(RAM:SA) and P(HSAM:SA), with no signs of formation of an acid as determined by IR, it can be concluded that the process that occurred is not hydrolysis of the polyanhydride to its component acids. The only chemical difference between unstable P(RAM:SA) and P(HSAM:SA) to γ-irradiation and the stable P(RAS:SA) and P(HSAS:SA) to γ-irradiation is the presence of double bonds conjugated to the anhydride bond of the maleate residue in ricinoleic or hydroxystearic acid maleate. We hypothesized that the conjugated double bonds (to the anhydride carbonyl) weaken the anhydride bond. Upon irradiation, the anhydride bond breaks, froming two carboxyl radicals that further react with other anhydride bonds by anhydride interchange to form

Table 2. Molecular Weight Changes of Irradiated Ricinoleic Acid Based Polymersb before irradiation P(RAS) P(HSAS) P(RAM) P(HSAM)

RAS HSAS RAM HSAM

after irradiation (2.5 Mrad)

Mw

Mn

Mw

Mn

19 500 19 000 10 000 10 000

9600 12 000 7600 7000

18 900 18 300 2000 3000

9000 11 000 1800 2700

calculated

titration

SEC

titration

398 400 396 398

397.8 400.1 396.5 398.4

400a 400a 400a 400a

398.2 400.5 396.8 398.6

a Data fit the estimated 400 molecular weight polystyrene standard. Irradiation was carried with an absorbed dose of 2.5 Mrad by means of a 60Co source (450 000 Ci; 9.5 h), Mw of homopolymers was determined by SEC. Mw of monomers was determined by SEC and by titration. b

more stable anhydride bonds. According to this hypothesis the possible anhydride bonds along the polymer chain should be arranged in the following stability order:

To test this hypothesis, a second irradiation experiment was conducted to determine the effect of conjugated double bonds on the stability of anhydrides to irradiation. In this experiment, monomers and homopolymers of RAM, HSAM, RAS and HSAS (Table 2), and ricinoleic acid were γ-irradiated at room temperature with a dose of 2.5 or 10 Mrad. These samples were characterized before and after irradiation, using the same methods used for the copolymers with sebacic acid. Again, it was found that the only affected polymers were P(RAM) and P(HSAM), but with no effect observed on their corresponding monomers, RAM and HSAM. At both doses, the polymers changed from a viscous oil or semisolid before irradiation to a solid material, which formed a gel when immersed in dichloromethane. The gel was gradually solubilized after 24 h to low molecular weight oligomers (Mw < 2000), with IR spectra similar to the original polymer. 1H NMR spectra of P(RAM) before and after irradiation are shown in Figure 2. It can be seen that the only difference in the polymer spectra is the disappearance of the double doublet at 6.85 ppm, with a corresponding increase in the peak ratio at 6.8 ppm. These peaks are attributed to the olefin protons of the maleic anhydride moiety in P(RAM), when


Figure 2.

1H


NMR spectra of P(RAM) before and after irradiation.

Figure 3. Suggested mechanism of a self-depolymerization process via inter- and/or intramolecular anhydride interchange.

the polymer units are linked in a maleate-maleate anhydride bond (6.85 ppm, head-to-head) or maleate-ricinoleate anhydride bond (6.8 ppm, head-to-tail). In the head-to-head linkage, the anhydride bond lies between two double bonds, which apparently makes it more sensitive to γ-irradiation,

resulting in self-depolymerization via inter- and/or intramolecular anhydride interchange to form low molecular weight polymeric rings that interlink to form a gel when immersed in chloroform (Figure 3). According to this proposed mechanism, the only change that occurred is

1020


Teomim et al.

Table 3. Dependency of the Molecular Weight of Irradiated P(FA:SA) on the Initial FA:SA Ratio (Mw was determined by SEC) FA:SA ratio (w/w) 10:90 30:70 50:50

before irradiation

after irradiation (2.5 Mrad)

Mw

Mn

Mw

Mn

8000 5000 5700

4500 3000 3000

8600 2700 1600

4900 1700 1200

Figure 5. Hydrolysis of ricinoleic acid based polymers before and after irradiation, monitored by weight loss. Hydrolysis was conducted in phosphate buffer pH 7.4 at 37 °C.

Figure 4. Repolymerization of irradiated P(RAM) and P(HSAM); repolymerization was conducted at 150 °C for 30 min.

rearrangement of the anhydride bonds from the less stable head-to-head bond to head-to-tail anhydride linkages, which are not expected to significantly affect the IR and H NMR spectra (besides the changes reported above on the unsaturated anhydride) or the DSC. A similar mechanism was suggested in the thermal and solution decomposition of several polymers including poly(anhydrides)17-20 To confirm the hypothesis that the decrease in Mw is due to a reversible anhydride interchange, repolymerization of the depolymerization products of P(RAM) and P(HSAM) on obtained samples was conducted at 150 °C for 30 min with constant stirring.17,18 As can be seen from Figure 4, polymers with the initial Mw were obtained. Repolymerization of the respective SA copolymers, P(RAM:SA) and P(HSAM:SA), resulted in a significant increase in molecular weight. These data show that the depolymerization process is reversible, and that the cyclic anhydride can be opened to form linear high molecular weight polymers again. On the basis of these data, a third experiment was conducted to study the effect of γ-irradiation on a different class of unsaturated polyanhydride, P(FA:SA).14 Copolymers of fumaric and sebacic anhydride in three weight ratios (10: 90, 30:70, and 50:50) were irradiated, and the properties of the polymers were determined. The increase in the fumaric acid content from 10 to 50% increased the probability of FA-FA anhydride bonds, i.e., the number of anhydride bonds with two conjugated unsaturated bonds.14 With the 10:90 ratio, the polymer dissolved immediately in dichloromethane, and no change in molecular weight was observed. This was not the case for the 30% and 50% FA content. Both formed an insoluble gel, which eventually dissolved after continuous stirring for 24 h at room temperature. The Mw dropped from about 6000 to 3000 for the 30:70 ratio and to 1500 for the 50:50 ratio (Table 3).

Figure 6. Gas chromatography data of P(RAM:SA) degradation products after 8, 24, and 96 h in buffer solution.16

In Vitro Degradation of Irradiated Polymers. The effect of irradiation on polymer degradation was studied by monitoring polymer mass loss and hydrolysis products when placed in pH 7.4 buffer solution. Irradiated P(RAS:SA) lost about 50% of its initial mass within 12 h while the nonirradiated sample and the HSAS- and RAS-based polymers lost only 25% of their mass during this period (Figure 5). Gas chromatography analysis of the degradation solution of nonirradiated P(RAM:SA) after 8 and 24 h indicated the existence of only sebacic acid. However, for the irradiated polymer, ricinoleic acid, maleic acid, ricinoleic acid maleate (RAM), and sebacic acid were detected after 24 h (Figure 6). This difference can be explained by the fact that the



Figure 7. Suggested mechanism for the degradation process of irradiated and nonirradiated P(RAM:SA).

irradiated RAM-based polymer formed low Mw cyclic fragments that are more susceptible to hydrolysis and rapidly degrade to RAM monomers, which further hydrolyze to the respective ricinoleic and maleic acids (Figure 7). These hydrolysis products were formed in the solution of the nonirradiated sample only at the 96 h time point (Figure 6). Free Radical Formation. Free radicals were detected by means of 9.25 GHz EPR for all γ-irradiated samples. No differences were found between the samples irradiated at room temperature or in dry ice. The EPR spectra were found to be a superposition of several radical species. Different levels of microwave power were applied in order to distinguish radicals having distinct relaxation properties. For peroxyl radicals, the signal intensity increases in proportion to the square root of the microwave power. For carboncentered radicals, the increase in microwave power will result in a decrease of signal intensity. This dependency of the EPR spectra on the microwave power suggests the presence of different species (Figure 8). Another important point is that the composition of the polymer strongly influences both the formation of radicals by ionizing radiation and their reactivity.11 Polymers with high melting points and crystallinity give the highest yield of observable radicals at room temperature. When comparing the radical formation in irradiated ricinoleic acid based copolymers with sebacic acid (crystallinity of 15-20%) and poly(sebacic anhydride) (crystallinity of 60%), a larger amount of radicals were observed for poly(sebacic anhydride).11 This can be explained by a decrease in the stabilization of the radicals by the polymer matrix.

Figure 8. Influence of microwave power on the shape of the EPR spectra of γ-irradiated P(RAM:SA).

Polymer Stability. Irradiated polymers are susceptible to continuous changes in properties long after irradiation.21 These post-irradiation effects may be attributed to the stable radicals formed during irradiation.21 γ-Rays induce the formation of free electrons and cationic radicals that may recombine or react further to secondary or tertiary radical species including carbon radicals.22 Due to the fact that it is hard to completely remove oxygen from the irradiated device, oxygen may have reacted with the carbon radicals to form peroxyradicals that can obtain a hydrogen radical and form hydroperoxides which react slowly with sites along the polymer chains, leading to scission or cross-linking.21,22 γ-Irradiated polymer samples were placed at -10 °C under argon atmosphere, and the changes in weight and molecular weight were monitored for a period of 3 months. No changes were observed during this period. When these polymers were stored at room temperature under argon atmosphere, the same process causing molecular weight loss was found as for nonirradiated polymers.

1022


Conclusions γ-Irradiation of aliphatic polyanhydrides at a dose of 2.5 Mrad has no effect on the polymer properties. However, polymers containing anhydride conjugated to unsaturated bonds decrease in molecular weight via apparently internal anhydride interchange which involves the conversion of the fully unsaturated anhydride bond into the more stable halfsaturated anhydride bond. As a result of the new arrangement of anhydride bonds within the polymer induced by the γ-irradiation, low molecular weight polymer rings were formed. These low Mw polymers hydrolyze much faster than the nonirradiated polymers. Other than the effect on the anhydride bond, no new C-C bonds or chain scissions commonly found in irradiated polymers were observed. The practical outcomes of this study are that polyanhydrides are stable to terminal sterilization unless unsaturation conjugated to the anhydride bond exists along the polymer chains and that unsaturated anhydride molecules may serve as radical scavengers due to their tendency to form less saturated derivatives. Acknowledgment. This work was supported in part by grants from NIH NCDDG U01 CA52857, the Israel Academy of Science, and the US-Israel Binational Science Foundation (BSF). References and Notes (1) Domb, A. J.; Amselem, S.; Langer, R.; Maniar, M. In Designed to Degrade Biomedical Polymers; Shalaby, S., Ed.; Carl Hauser Verlag: Munich, Germany, 1994; p 69. (2) Domb, A. J.; Elmalak, O.; Shastry, V. R.; Ta-Shma, Z.; Masters, D. M.; Ringel, I.; Teomim, D.; Langer, R. In Handbook of biodegradable polymers; Domb, A. J., Kost, J., Weiseman, D. M., Eds.; Harwood Academic Publishers: Amsterdam, 1997; p 135.

Teomim et al. (3) Domb, A. J.; Maniar, M. J. Polym. Sci., Polym. Chem. 1993, 31, 1275. (4) Teomim, D.; Nyska, A.; Domb, A. J. J. Biomed. Mater. Res. 1999, 45, 258-287. (5) Athanasiou, K. A.; Niederauer, G. G.; Agrawal, C. M. Biomaterials 1996, 17, 93-102. (6) Bruck, S. D.; Muller, E. P. J. Biomed. Mater. Res. 1998, 22, 133144. (7) Rubbo, S. D.; Gardner, J. F. A reView of sterilization and disinfection; Lloyd-Duke, Ltd.: London, 1965. (8) Charlesby, A. In Radiation Effect on Polymers; Clough, R. L., Shalaby, S. W., Eds.; American Chemical Society: Washington, DC, 1991; p 19. (9) Gupta, M. C.; Deshmukh, V. G. Polymer 1983, 24, 827-830. (10) Bittner, B.; Mader, K.; Kroll, C.; Borchert, H. H.; Kissel, T. J. Controlled Release 1999, 59, 23-32. (11) Mader, K.; Domb, A.; Swartz, H. M. Appl. Radiat. Isot. 1996, 47, 1669-1674. (12) Domb, A. J.; Langer, R. J. Polym. Sci., Polym. Chem. 1987, 25, 3373. (13) Domb, A. J.; Gallardo, C. E.; Langer, R. Macromolecules 1989, 22, 3200. (14) Domb, A. J.; Mathiowitz, E.; Ron, E.; Giannos, S.; Langer, R. J. Polym. Sci., Polym. Chem. 1992, 29, 571-579. (15) Hyon, SH.; Jamshidi, K.; Ikada, Y. Biomaterials 1997, 18, 15031508. (16) Tirosh, O.; Cohen, R.; Katzhendler, J.; Alon, A.; Barenholz, Y. Chem. Phys. Lipids 1997, 87, 17-22. (17) Domb, A. J.; Langer, R. Macromolecules 1989, 22, 2117-2122. (18) Carothers, W. H.; Dorough, G. L.; Van Natta, F. J. J. J. Am. Chem. Soc. 1933, 55, 5043. (19) Staubli, A.; Mathiowitz, E.; Langer, R. Macromolecules 1991, 24, 2291. (20) Staubli, A.; Mathiowitz, E.; Lucarelli, M.; Langer, R. Macromolecules 1991, 24, 2283-2291. (21) O’Donnell, J. H. In The Effects of Radiation on High Technology Polymers; Reichmanis, E., O’Donnell, J. H., Eds.; American Chemical Society: Washington, DC, 1989; p 1. (22) Sintzel, M. B.; Scwach-Abdellaoui, K.; Mader, K.; Stosser, R.; Heller, J.; Tabatabay, C.; Gurny, R. Int. J. Pharm. 1998, 175, 166-176.

BM010078N

Irradiation Stability of Saturated and Unsaturated ... - ACS Publications

Recommend Documents