Hemoglobin Recognition by Imprinting in Semi-Interpenetrating

cal Factory (Tianjin, China). N,N′-Methylenebisacrylamide. (MBA) was purchased from Tianjin Special Reagent Factory. (China). Potassium persulfate (...
0 downloads 0 Views 282KB Size
Download PDF
Biomacromolecules 2005, 6, 2601-2606

2601

Hemoglobin Recognition by Imprinting in Semi-Interpenetrating Polymer Network Hydrogel Based on Polyacrylamide and Chitosan Yong-qing Xia,† Tian-ying Guo,*,† Mou-dao Song,† Bang-hua Zhang,† and Bao-long Zhang‡ State Key Laboratory of Functional Polymer Materials for Adsorption and Separation, Institute of Polymer Chemistry, Nankai University, and Department of Chemistry, Nankai University, Tianjin, China, 300071 Received May 10, 2005

Semi-interpenetrating polymer network (semi-IPN) hydrogel was prepared to recognize hemoglobin, by molecularly imprinted method, in the mild aqueous media of chitosan and acrylamide in the presence of N,N′-methylenebisacrylamide as the cross-linking agent. The hydrogel obtained has been investigated by using thermal analysis, X-ray diffraction, differential scanning calorimetry (DSC), and environmental scanning electron microscope (ESEM). Langmuir analysis showed that an equal class of adsorption was formed in the hydrogel, and the adsorption equilibrium constant and the maximum adsorption capacity were evaluated to be 4.27 g/mL and 36.53 mg/g wet hydrogel, respectively. The imprinted semi-IPN hydrogel has a much higher adsorption capacity for hemoglobin than the nonimprinted hydrogel with the same chemical composition and also has a higher selectivity for the imprinted molecule. Introduction The synthesis of highly specific molecularly imprinted polymers (MIPs) has been the goal of many research groups recently.1-3 The main research of this field has included separation processes (chromatography, solid-phase extraction, membrane separations), artificial antibodies, and sensors recognition elements.4-6 Molecularly imprinting technique involves forming a pre-polymerization complex between the imprinted molecule and the functional monomer with specific chemical structures designed to interact with the imprinted molecule either by covalent7,8 or noncovalent9,10 interactions or by metal ion coordination.11-13 These complexes are fixed by polymerization with a certain degree of cross-linking. Removal of the imprinted molecules from the obtained polymer network affords complementary binding sites that can selectively rebind the same imprinted molecules. Since recognition requires three-dimensional orientation, most conventional molecularly imprinted polymers were highly cross-linked for greater affinity, capacity, and selectivity by limiting the movement of the memory site. To date, usually, only relatively low molecular weight compounds are used as imprinted molecules;14-16 biomacromolecule such as protein is seldom.17-20 Hydrogels are cross-linked, three-dimensional hydrophilic polymer networks, which swell but do not dissolve when brought into contact with water. Because of their significant water content, hydrogels possess a degree of flexibility very similar to natural tissue. From a biological viewpoint, it is essential to achieve molecularly imprinted hydrogels (MIHs) to recognize the imprinted protein from other proteins. * Corresponding author. Tel: +86-22-235-015-97; fax: +86-22-235015-97; e-mail: [email protected]. † Institute of Polymer Chemistry. ‡ Department of Chemistry.

Recently, chitosan used as an adsorbent has drawn attention because of its high contents of amino and hydroxy functional groups showing high potentials of the adsorption of dye, metal ions, and proteins.21-23 Recently, Aburto, J. et al. has used chitosan as an adsorbent for some organosulfur compounds by molecular imprinting method.24,25 Other useful features of chitosan include its abundance, nontoxicity, hydrophilicity, and biocompatibility.26 Polyacrylamide gel has been successfully used for protein imprinting.27,28 The main purpose of our work was to prepare hemoglobin-imprinted semi- interpenetrating polymer network hydrogel based on polyacrylamide/chitosan. IPN is a mixture of two or more cross-linked networks, which are dispersed or mixed together at a molecular segmental level. In this work, chitosan is used as a functional polymer chain to produce hemoglobin-imprinted semi-IPN, so that the affinity and thermal properties of the polyacrylamide gel could be improved. The thermal and morphological properties of the semi-IPN were investigated by thermogravimetry, differential thermogravimetry, differential scanning calorimetry, and environmental scanning electron microscope. Finally, the adsorptive behavior and specificity of the semiIPN were also discussed. Experimental Section Materials. Chitosan was purchased from Boao BioTechnology Company, Shanghai (China). Deacetylation degree is 90% and the viscosity average molecular weight of the chitosan was determined to be 503495 by viscometric method. Acrylamide (Am) was purchased from Miou Chemical Factory (Tianjin, China). N,N′-Methylenebisacrylamide (MBA) was purchased from Tianjin Special Reagent Factory (China). Potassium persulfate (KPS) was obtained from

10.1021/bm050324l CCC: $30.25 © 2005 American Chemical Society Published on Web 07/29/2005

2602

Biomacromolecules, Vol. 6, No. 5, 2005

Tianjin No.3 Chemical Reagent Factory. Bovine hemoglobin (Hb) and bovine serum albumin (BSA) were purchased from Sino-America Biotechnology Company (China), and the protein solutions were prepared using 0.01 M sodium dihydrogen phosphate buffer (pH 6.8). Am and KPS were recrystallized before used. Other chemicals were analytical grade and were used as received. Synthesis of Molecularly Imprinted Hydrogel Based on Polyacrylamide/Chitosan Semi-IPN (Named as MIH-sIPN). Chitosan was dissolved into 1% (v/v) acetic acid, and the solid content in solution was 4% (w/v); the solution was gently stirred for 4 h and then was stored for about 2 days at room temperature to allow swelling. For the preparation of the imprinted semi-IPN hydrogel, 26.0 g chitosan solution (pH 6.0), 1.9 g Am, 0.1 g MBA, 20.4 mg KPS, 600 mg Hb, and 18 mL 0.01 M sodium dihydrogen phosphate buffer (pH 6.8) were put into a 100-mL, four-necked flask equipped with a nitrogen inlet and a mechanical stirrer. The mixture (pH 6.8) was stirred continuously under a nitrogen atmosphere for 45 min, and then 5 mL sodium dihydrogen phosphate buffer containing 0.16% (w/v) NaHSO3 was added. The mixture was stirred for another 15 min and was stopped. The polymerization progress was under the nitrogen atmosphere for 4 h. The formed hydrogel was then pressed through a 26-mesh net to produce granules. The granules were washed with 10% (v/v) acetic acid containing 10% (w/ v) SDS solution to desorb the hemoglobin till the color was pale. Then, the granules were equilibrated with 0.01 M sodium dihydrogen phosphate buffer (pH 6.8) for 24 h. The nonimprinted hydrogel based on polyacrylamide/ chitosan semi-IPN (NIH-s-IPN) was prepared by the same procedure without addition of hemoglobin. To compare the adsorptive behavior of the MIH-s-IPN, we also prepared the imprinted polyacrylamide gel (named as MIH-polyacrylamide) and the nonimprinted polyacrylamide gel (named as NIH-polyacrylamide). The procedure and composition were the same as that above just without the addition of chitosan. The NIP-semi-IPN beads and NIP-polyacrylamide beads were prepared using inverse suspension polymerization also. The total monomer concentration and cross-linking concentration was as the MIP beads. The range of beads from 0.8 to 1.0 mm was selected to study mechanical strength. Characterization of the Hydrogels. The morphology of the hydrogels was studied using an environmental scanning electron microscope (ESEM Philips ×l30). Samples containing water without drying were mounted on metal stubs at a low vacuum degree (∼10-3 atm), and a relatively low temperature (near 0 °C) was observed. Thermogravimetry and differential thermogravimetry curves of samples were performed by a NETZSCH TG 209 (Germany) under nitrogen atmosphere from 25 to 600 °C at a heating rate of 10 °C/min. A NETZSCH DSC 204 (Germany) was used to measure the thermodynamic properties of the materials. Heating and cooling rates were 10 °C/min. DSC curves of each sample were obtained from the second heating run after a first run of heating to 150 °C. All samples were analyzed under continuous flow of dry nitrogen gas.

Xia et al.

X-ray diffraction patterns were measured by a D/max -2500 diffractrometer X-ray (Japan). A Cu Ka target was at 40 kV and 100 mA with the two-theta scanning mode between 3 and 60°. The water content of the MIH-s-IPN in distilled water was calculated by measuring the weight of the hydrogel as follows: water content ) (W - W0)/W The weight of the wet sample (W) was determined after removing the surface water by blotting with moistured filter paper. The weight of the dry sample (W0) was determined after drying the hydrogel in a vacuum oven for 3 days. Mechanical strength studies of the semi-IPN beads and polyacrylamide beads were completed using a mechanical stirrer, 100 beads, and 40 mL water, which were put in a 50-mL flask, and they were stirred at different stirring speeds for 30 min. The fragments were separated, and the beads were renumbered. The mechanical strength was determined by noting the percent loss in the number of the beads. Adsorption isotherm studies were conducted with a constant wet MIH-s-IPN and MIH-polyacrylamide weight (using filter paper to absorb the surface water) and with varying initial concentration of Hb (in 0.01 M sodium dihydrogen phosphate buffer, pH 6.8) in the range 0-0.7 mg/mL. The solutions in a conical flask were oscillated in an oscillator with a constant-temperature bath at 25 °C for 17 h; after filtration, the concentration of Hb in supernatant was analyzed by a criterion curve which was determined using a spectrophotometer at 280 nm. The amount of adsorption was calculated on the basis of the difference of Hb concentration in aqueous before and after adsorption, the volume of aqueous solution (25 mL), and the weight of the hydrogel (0.5 g) according to adsorption capacity (Q) ) (C0 - Ce)V/W where C0 is the initial Hb concentration (mg/mL), Ce is the equilibrium Hb concentration (mg/mL), V is the volume of Hb solution (mL), and W is the weight of the MIH (g). Results and Discussion Water Content. The MIH-s-IPN we prepared contains large amounts of water (about 94%). Compared to more traditional and well-established imprinting techniques with molecularly imprinted gels that were highly cross-linked for greater selectivity with a rigid approach rather than flexible recognition, the hydrogels we prepared have much water content and are similar to nature tissue while still retaining specific recognition. Thermal Analysis. Changes that take place during thermal degradation of the investigated gels were observed while performing thermogravimetric measurements. The TG and the corresponding DTG curves of chitosan, polyacrylamide, and polyacrylamide/chitosan semi-IPN recorded in nitrogen atmosphere from 25 to 600 °C are shown in Figure 1 and Figure 2, respectively.

Hemoglobin Recognition by Imprinting

Biomacromolecules, Vol. 6, No. 5, 2005 2603

Figure 3. DSC thermograms of chitosan, polyacrylamide, and polyacrylamide/chitosan semi-IPN. Figure 1. TG curves of chitosan, polyacrylamide, and polyacrylamide/ chitosan semi-IPN.

Figure 4. X-ray diffraction patterns of chitosan, polyacrylamide, and polyacrylamide/chitosan semi-IPN.

Figure 2. DTG curves of chitosan, polyacrylamide, and polyacrylamide/chitosan semi-IPN.

The first thermal event occurs in the temperature range 25-250 °C, where all samples present a mass loss ranging from 7% to 11%. This is attributed to the evaporation of water, whose content is a function of the morphology and crystallinity of the polymers. The TG and DTG curves of polyacrylamide/chitosan semi-IPN were different from those of polyacrylamide and chitosan and showed three stages. The first water-loss stage reached a maximum at 232 °C, with a weight loss of 10%. The temperature when weight loss reaches a maximum of the first stage was higher than that of chitosan (145 °C) and polyacrylamide (185 °C), respectively. These results demonstrated that it is difficult for the semi-IPN to lose water. The reason probably is that with the hydrophilic chitosan added, the interaction force between the water and the backbone of polyacrylamide/chitosan semiIPN was strengthened, and then the water in the polyacrylamide/chitosan semi-IPN hydrogel is more difficult to lose than that in the polyacrylamide hydrogel. The second stage was attributed to the thermal degradation of chitosan reaching a maximum at 298 °C, with weight loss of 16.6%. The third stage was the degradation stage of polyacrylamide reaching a maximum at 406 °C, with weight loss of 56%. These results demonstrated that the semi-IPN hydrogel has improved its ability to preserve water. It is well-known that the miscibility of the molecules in a blend of polymers can be judged by the properties of the

solid state such as the glass-transition temperature. DSC is one of the convenient methods to measure Tg. The DSC thermograms of chitosan, polyacrylamide, and polyacrylamide/chitosan semi-IPN were measured. Special care must be taken during DSC measurements since these samples are apt to adsorption moisture, which should strongly affect the DSC measurements. To eliminate the effect of moisture, two cycles of heating and cooling runs were adopted, considering that the temperature region will not bring about the thermal degradation in the first heating run. The results in the second run are shown in Figure 3. The DSC curve of chitosan showed a thermal decomposition temperature higher than 250 °C, and no glass transition (Tg) was exited before decomposition in the second heating scans of chitosan. The result just accords with the previous report.29 The glass-transition temperature of pure polyacrylamide was 97 °C, and the glasstransition temperature of polyacrylamide in the semi-IPN was 117 °C, which shifts to a higher temperature of about 20 °C than that of the pure polyacrylamide. The reason of this phenomenon may be that polyacrylamide and chitosan could form a weak hydrogen-bonding interaction between them, and then these interaction points can act as the physical crosslinking points to increase the chain entanglement; thus, the chain movement of polyacrylamide is limited and the glasstransition temperature of polyacrylamide is increased. X-ray Diffraction. The X-ray diffractograms of chitosan, polyacrylamide, and polyacrylamide/chitosan semi-IPN are shown in Figure 4. It can be seen that chitosan exhibits two

2604

Biomacromolecules, Vol. 6, No. 5, 2005

Xia et al.

Figure 5. ESEM microphotographs of (a) MIH-s-IPN (×2000 times) and (b) MIH-polyacrylamide (×2000 times). Table 1. Crumpling Ratio of Polyacrylamide Gel and Polyacrylamide/Chitosan Semi-IPN Gel stirring speed (r‚min-1)

200

beads polyacrylamide semi-IPN

400

600

800

crumpling ratio % 0 0

4 3

28 4

36 13

reflection falls at 2θ ) 11° and 2θ ) 20°, which were assigned to crystal forms I and II, respectively.30 The polyacrylamide/chitosan semi-IPN has one broad peak at around 2θ ) 21°, and the crystalline peak of chitosan at 2θ ) 11° disappeared as the semi-IPN was formed between polyacrylamide and chitosan, which could attribute to the weakening of hydrogen bonding between the amino groups and hydroxyl groups in the chitosan molecules. Figure 5 shows ESEM microphotographs of the MIH-sIPN (a) and MIH-polyacrylamide (b), respectively. The results showed that both hydrogels are porous but topologically different, and more uniform cavities may be observed in the MIH-s-IPN gel as compared with the MIH-polyacrylamide gel. The pore walls of the MIH-s-IPN gel become thicker than those of the MIH-polyacrylamide gel; maybe this was the reason that at the same condition the semi-IPN was stronger than the polyacrylamide gel (as seen in Table 1). Adsorption Isotherm. Figure 6 shows the experimental equilibrium isotherms for adsorption of Hb on the MIH-sIPN, MIH-polyacrylamide, NIH-s-IPN, and NIH-polyacrylamide. It can be seen that at the tested concentrations, the adsorption capacities of MIH-s-IPN and NIH-s-IPN are much higher than those of MIH-polyacrylamide and NIH-polyacrylamide. The reason may be that chitosan has plentiful -NH3+ and hydroxyl groups, and when Hb is added, -NH3+ and hydroxyl groups of chitosan will interact with Hb by electrostatic force or hydrogen bonding. So, in the MIH-sIPN hydrogels, the self-assembly process results in much more combined points than MIH-polyacrylamide gels after polymerization and removal of imprinted molecule Hb. Adsorption isotherms are important for describing how adsorbates will interact with adsorbent. Thus, the correlation of equilibrium data using either a theoretical or empirical equation is essential to practical adsorption process. Two isotherm equations were used in our study:

Figure 6. Adsorption isotherms of Hb on MIH-s-IPN, MIH-polyacrylamide, NIP-s-IPN, and NIP-polyacryamide. V ) 25 mL; C ) 0-0.7 mg/mL; the sample quantity: 0.5 g; T ) 25 °C; adsorption time ) 17 h.

1. Langmuir equation: Ce/Q ) Ce/Qmax + 1/bQmax Here, Qmax is the maximum adsorption (mg/g) and b is the Langmuir adsorption equilibrium constant (mL/mg). 2. Freundlich equation: Q ) QfCe1/n Here, Qf is roughly an indicator of the adsorption capacity and 1/n is the adsorption intensity. From linear plots of (Ce/Qe) versus Ce and ln Q versus ln Ce, the parameters in the Langmuir and Freundlich equations can be determined. The Langmuir constants and Freundlich constants with the correlation coefficients are given in Table 2. The Freundlich exponent n > 1 means that both the MIHs-IPN and MIH-polyacrylamides are good adsorbents for hemoglobin. It can also be seen that the linear fit with the Langmuir equation was comparably better (r > 0.98). Therefore, although the adsorption processes were complicated, there was a tendency for a chemical affinity to exist at the surface between the Hb and the MIH-s-IPN hydrogel.

Biomacromolecules, Vol. 6, No. 5, 2005 2605

Hemoglobin Recognition by Imprinting Table 2. Langmuir and Freundlich Isotherm Constants of MIH-s-IPN and MIH-Polyacrylamide Langmuir samples

Qmax b (mg/g) (mL/mg)

MIH-s-IPN 36.53 MIH-polyacrylamide 2.92 NIH-s-IPN 3.84 NIH-polyacrylamide 6.10

4.27 3.47 6.01 0.54

Freundlich

rL

Qf (mg/g)

n

rF

0.999 46.32 1.56 0.996 0.980 2.64 2.06 0.976 0.987 3.51 3.25 0.980 0.920 1.36 1.10 0.999

Table 3. Degree of Selectivity for Hb and BSA

KD (single component)

KD (mixture)

samples

Hb

BSA

Hb

BSA

MIH-s-IPN NIH-s-IPN MIH-polyacrylamide NIH-polyacrylamide

98.05 2.88 2.75 0.325

2.18 0.933 2.03 2.14

83.63 1.7 1.92 2.14

1.4 2.98

a

a

a

Too little to be detected.

Degree of Specificity on the MIH-s-IPN. The special selectivity test of MIH was carried out using BSA as a comparative substrate. The amounts of adsorption to MIH and NIH were determined with the equilibrium adsorption method. The distribution coefficients of the proteins between solution and MIHs are calculated and listed in Table 3. The distribution coefficient KD is defined as KD ) HbMIH/Hbsol where HbMIH is the concentration of protein on the MIH (mg/ g) and Hbsol is the concentration of protein in the solution (mg/mL). It can be seen from Table 3 that MIH-s-IPN exhibited high selectivity for the imprinting molecule Hb compared to the substrate BSA. However, NIH-s-IPN exhibited low values of KD for Hb and BSA. Since this analysis was carried out for the adsorption of a singlecompound solution, the binding of MIH for hemoglobin in the mixture may be interfered with by the presence of BSA. Therefore, the degree of specificity was further analyzed from the results of adsorption in mixture solutions. The concentrations of Hb and BSA were determined by means of equivalence absorbency double-wavelength ultraviolet spectrophotometry, and the determination wavelengths are se-

lected as 280 and 362 nm. From the results performed using single-compound solutions (either Hb or BSA) and twocompound solutions (Hb and BSA) that were listed in Table 3, it can be seen that in the presence of BSA, the degree of specificity of Hb by the MIH was reduced. Apparently, a competition between Hb and BSA for the specific binding sites on the MIH was developed. From Figure 7 which shows the ESEM microphotographs of MIH-s-IPN gel and NIH-s-IPN gel, it can be seen that the surfaces of the MIH-s-IPN hydrogel and the NIH-s-IPN hydrogel are not very different, and the evidence indicates that the imprinting methods create a microenvironment based on shape selection and position of functional groups that recognizes the Hb imprinted molecule. In the MIH-s-IPN gels, the self-assembly process results in the existence of stereocavity and combined points after polymerization and in the removal of imprinted molecule Hb. Though hemoglobin and bovine serum albumin are both globular proteins and their molecular weights are similar (64 500 and 68 000 Dalton, respectively), Hb is a tetrameric protein composed of pairs of two different polypeptides and has a biconcave shape, and the size of hemoglobin is about 65 Å; BSA consists of one polypeptide and has an ellipsoidal shape, and the size of BSA is about 154 Å, larger than Hb. Since the cavities formed of MIP are matched to the size of Hb, it is very difficult for the molecules with different dimensions to enter the cavities; therefore, the distribution coefficient of MIP to BSA is lower correspondingly. Conclusions The novel molecular imprinting technique was used in this work to imprint hemoglobin in aqueous solution using chitosan and acrylamide as the multiple functional monomers. Success has been demonstrated in recognizing between hemoglobin and bovine serum albumin at the same condition. The adsorption capacity in the hemoglobin-imprinted semiIPN hydrogel was as high as 36 mg hemoglobin per gram of wet hydrogel and was much higher than the nonimprinted semi-IPN hydrogel with the same chemical composition. However, there is still much work to be done to further understand the factors affecting the behavior of the MIPs.

Figure 7. ESEM microphotographs of (a) MIH-s-IPN (×1000 times) and (b) NIH-s-IPN (×1000 times) V ) 10 mL; C ) 1.0 mg/mL; the sample quantity: 0.5 g; T ) 25 °C; adsorption time ) 17 h.

2606

Biomacromolecules, Vol. 6, No. 5, 2005

Further work will include attempts at the preparation of reliable and predictable high-quality MIP and synthesis of intelligent, controlled release hydrogels imprinted for hemoglobin. Acknowledgment. The authors are grateful to National Nature Science Foundation of China for financial support (Proj.No.50473023). References and Notes (1) Wulff, G. Angew. Chem., Int. Engl. 1995, 34, 1812. (2) Vlatakis, G.; Andersson, L. I.; Mu¨eller, R.; Mosbach, K. Nature 1993, 361, 645. (3) Liu, J. Q.; Wulff, G. J. Am. Chem. Soc. 2004, 126, 7452. (4) Dhal, P. K.; Arnold, F. H. J. Chromatogr., A 1995, 708, 19. (5) Ansell, R. J.; RamstrO ¨ m, O.; Mosbach, K. Clin. Chem. 1996, 42, 1506. (6) Haupt, K.; Mosbach, K. Chem. ReV. 2000, 100, 2495. (7) Wulff, G.; Gross, T.; SchO ¨ nfeld, R. Angew. Chem., Int. Engl. 1997, 36, 1962. (8) Shea, K. J.; Sasaki, D. Y. J. Am. Chem. Soc. 1989, 111, 3442. (9) Whitcomb, M. J.; Rodriguez, M. E.; Villar, P.; Vulfson, E. N. J. Am. Chem. Soc. 1995, 117, 7105. (10) Mosbach, K.; Yu, Y.; Andersch, J.; Ye, L. J. Am. Chem. Soc. 2001, 123, 12420. (11) Dhal, P. K.; Arnold, F. H. J. Am. Chem. Soc. 1991, 113, 7417.

Xia et al. (12) (13) (14) (15) (16) (17) (18) (19) (20) (21) (22) (23) (24) (25) (26) (27) (28) (29) (30)

Dhal, P. K.; Arnold, F. H. Macromolecules 1992, 25, 7051. Krebs, J. F.; Borovik, A. S. J. Am. Chem. Soc. 1995, 117, 10593. Liu, J.; Wulff, G. J. Am. Chem. Soc. 2004, 126, 7452. Yoshida, M.; Hatate, Y.; Uezu, K.; Goto, M.; Furusaki, S. Colloids Surf., A 2000, 169, 259. Fisher, L.; Muller, R.; Ekberg, B. J. Am. Chem. Soc. 1991, 113, 9358. Kempe, M.; Mosbach, K. J. Chromatogr., A 1995, 691, 317. Shi, H. Q.; Tsai, W. B.; Garrison, M. D.; Ferrari, S.; Ratner, B. D. Nature 1999, 398 (6728), 593. Rachkov, A.; Minoura, N. Biochim. Biophys. Acta 2001, 1544, 255. Guo, T. Y.; Xia, Y. Q.; Hao, G. J.; Song, M. D.; Zhang, B. H. Biomaterials 2004, 25, 5905. Wong, Y. C.; Szeto, Y. S.; Cheung, W. H.; Mckay, G. Process Biochem. 2004, 39, 695. Inoue, K.; Yoshizuka, K.; Ohto, K. Anal. Chim. Acta 1999, 388, 209. Zeng, X.; Buckenstein, E. Biomaterials 2002, 23, 2561. Aburto, J.; Borgne, S. L. Macromolecules 2004, 37, 2938. Aburto, J.; Mendez-Orozco, A.; Borgne, S. L. Chem. Eng. Process. 2004, 43, 1587. Mi, F.; Tan, Y.; Liang, H.; Sung, H. Biomaterials 2002, 23, 181. Liao, J. L.; Wang, Y.; Hjerte´n, S. Chromatographia 1996, 42, 259. Hjerte´n, S.; Liao, J. L.; Nakazato, K.; Wang, Y.; Zamaratskaia, G.; Zhang, H. X. Chromatographia 1997, 44, 227. Jiang, H.; Su, W.; Caracci, S.; Bunning, T. J.; Cooper, T.; Adams, W. W. J. Appl. Polym. Sci. 1996, 61 (7), 1163. Samuels, R. J. J. Polym. Sci. Polym. Phys. 1981, 19, 1081.

BM050324L