Influence of diethylene glycol as a porogen in a glyoxal crosslinked polyvinyl alcohol hydrogel

Abstract

Pore structure plays a great role in determining the physical and mechanical properties of a gel. In this study, a polyvinyl alcohol (PVA) based hydrogel has been synthesized using glyoxal as crosslinker, and the influence of diethylene glycol (DEG) as a porogen in modifying its properties has been investigated. The resulting hydrogel possesses a unimodal distribution of nanopores, the content & sizes of which are controlled by using various dosages of porogen. The hydrogel has been characterized using FTIR spectroscopy which confirms the acetal linkage formation caused by the crosslinking reaction. The presence of DEG influences its rheological properties & decreases its glass transition temperature (T_g). Cyclical swelling–deswelling of the hydrogel samples leads to their improvement. DEG also assists in modifying the thermo-responsive nature of the crosslinked PVA matrix.

---

1. Introduction

Immense interest has been generated recently in various research areas concerning the usage of porogens in controlling the pore dimensions of different hydrogels. These find extensive applications in tissue engineering, bone & cartilage replacement and their generation, growth & repair.^1–3 Microfluidic sensors & lab-on-a-chip devices which are mostly based on porogens find extensive applications in biotoxin sensing, cell culture and biological engineering.⁴ The use of porogens has been extended to waste water treatment where cell immobilization techniques⁵ have been modified to a great extent by the presence of suitable porogens. Baraka et al. used solvent mixtures as porogens to increase porosity in a hydrogel matrix with the object of facilitating the transportation of ions inside the pores. This resulted in improvement in the ability to remove copper ions from aqueous solution.⁶

Microporous structures possess a large surface area and also are believed to create a platform for higher protein adsorption as well as ion exchange processes.⁷ Karageorgiou et al.⁸ claimed that low porosity stimulated osteogenesis by suppressing cell proliferation and forming cell aggregation. Clark et al. showed the influence of porogen loading on the microstructure of gel & the consequent mechanical properties.⁹ The control of pore dimension in a hydrogel can bring forth tremendous changes in properties and various techniques for introducing porogen in the hydrogel matrix have been developed. These include (i) freeze drying, (ii) solvent casting/particle leaching (introduction of porogen), (iii) gas foaming, and (iv) template method.¹⁰ The hazards due to organic solvent entrapment in most of the techniques used for controlling the pore dimension can be eliminated completely by the introduction of a suitable porogen.

Hou et al.¹⁰ employed a technique which involves the combination of the advantages of thermal processing & particulate leaching. They claimed to obtain greater porosity in this fashion. Iturralde et al. used toluene as a porogen along with propranolol as a template to control the porosity.¹¹ Michael S. Silverstein commented in his review paper that the generation of a secondary porous structure within the walls of polyHIPE (high internal phase emulsions) through the addition of porogens could be achieved easily. According to their proposals, the solvent acts as a co-surfactant which leads to a decrease in void size & interconnecting hole size as well. However the hole density also increases. Thus solvating porogen was found to be a better option than precipitating porogen in terms of creating an interconnected porous network of increased pore density.¹² Yu et al. studied the effect of porogenic solvents on the nature of the pores. They found that methanol & ethanol created a much larger pore diameter which was in the micrometer range. They also commented on the phenomenon of increase in pore size with poor solvent (used as porogen). Both pore size and pore density are found to have a proportional relationship with the concentration of porogen used, its solubility in the system & finally the evaporation rate of the porogen.¹³ Small pore structures appeared to be essential for ion removal in waste water treatment due to high specific surface area. Ethylene glycols and polyethylene glycols of various molecular weights have been investigated by many researchers but all of them showed the pore diameters of the gel to be in the micro range.¹⁴ Courtois et al. tried to establish a relationship between molecular weight of PEG (porogen) and pore diameter achieved. They found a way to adjust the porous network in the gel. It was concluded from their study that longer chains produced pores of larger diameter, while the surface area became lower.¹⁴

Thus from the above discussion it can possibly be predicted that (i) small pores can be created with a lower concentration of porogen having a low rate of evaporation and (ii) a reduction of pore size may be achieved by using lower homologues of polyethylene glycol. These findings provided the necessary impetus to use diethylene glycol as a porogen with water in small proportions. To be precise, our objective in the present work was to administer & control the pore structure and its distribution in a hydrogel. The consequence of such control on the physical, thermo-mechanical & morphological characteristics was also investigated.

Considering the necessity of well-controlled pore size, the present work attempted to use diethylene glycol (DEG) as a porogen to generate a porous structure in the crosslinked network of polyvinyl alcohol (PVA). PVA is the most readily available synthetic polymer. It is well known for its biocompatibility which can be exploited to develop biocompatible hydrogel.¹⁵ The review article of Rojas et al. clearly commented on the nontoxic behavior of glyoxal as it is produced during the normal cellular metabolism.¹⁶

2. Experimental

2.1 Materials

Poly(vinyl alcohol) (PVA) of molecular weight 115 000, 98–99% hydrolyzed was supplied by LOBA CHEMIE. Glyoxal of concentration 40%, supplied by LOBA CHEMIE, was used as crosslinker. Diethylene glycol (DEG) was used as porogen. Conc. H₂SO₄ was used as acid catalyst.

2.2 Method

An aqueous solution of 4% PVA was prepared and crosslinked with 0.022 moles of glyoxal using 100 μl conc. H₂SO₄ as acid catalyst at 98 ± 2 °C to ensure the completion of the reaction between polyol and glyoxal. This dose of glyoxal was chosen after a suitable trial and error method with PVA; maximum equilibrium percentage of swelling was observed with this dose while higher or lower doses led to a fall in water retention capacity. During the crosslinking reaction, controlled addition of DEG at varying percentages (50%, 75%, 100% & 125%) with respect to the crosslinker was made to engineer the cage structure in order to optimize the mechanical properties and the water retention capacity of the gel. The gel samples synthesized with 50%, 75%, 100% & 125% DEG (with respect to crosslinker) & the neat gel without any DEG are named as CPVD1, CPVD2, CPVD3, CPVD4 and CPVD0 respectively. Doses of DEG used here were chosen in a random manner to study the impact of such additions on the properties of the gel matrix. Addition of DEG was completed within 15 minutes after the addition of the crosslinker. The reaction was then allowed to continue for half an hour under isothermal conditions with constant stirring. The crosslinking time which was selected earlier by trial and error method was kept constant for the samples. This was taken as the time to reach optimum viscosity after glyoxal addition such that easy casting of the solution could still be made. The acid was then removed by dipping the product obtained in 1 : 1 ammonia solution followed by frequent washing with distilled water. Complete removal was ensured by testing the aqueous washing with a conventional pH meter repeatedly. The complete procedure of the porous hydrogel network development is described in Scheme 1. The samples were then dried at 25 °C under atmospheric pressure until no weight loss was observed.

Scheme 1

The completely dried samples are very clear and transparent. Structural, rheological and swelling–deswelling characterizations were performed using various techniques discussed below. Characterization of the cryo-fractured specimens by scanning electron microscopy was also carried out to determine the effect of DEG on the porous nature of the materials.

3. Characterization

3.1 Spectroscopic studies

The chemical composition of the gel was determined by Fourier transform infrared (FTIR) spectra (4000–600 cm⁻¹) taken on a FTIR spectrophotometer (Perkin Elmer, Japan) using an ATR attachment. The interaction between glyoxal and DEG was also monitored using the same FTIR equipment and the Perkin Elmer LAMBDA 25 UV/Vis spectrophotometer.

3.2 XRD studies

Identification of crystalline planes as well as changes in crystalline parameters like d-spacing, and the crystallite size were investigated by a wide angle X-ray scattering study (WAXS) using an XPERT-PRO Diffractometer system, Goniometer PW3050/60 (Theta/Theta) (XRD PANalytical, Netherlands). Studies were undertaken from 5–70° 2θ using a continuous scan with 0.0330° 2θ step size.

Air dried 0.1 ± 0.02 mm thick films of every sample were then used to perform WAXS. Crystallite size was calculated using the Scherrer equation¹⁷ τ = Kλ/β cos θ where θ = angle of incident radiation, K = shape factor = 0.9, λ = wavelength of incident ray in Å and β = FWHM of the peak (101 in case of polyvinyl alcohol).

3.3 Swelling and deswelling study

Swelling efficiency was determined using the equations Q = W_s/W_d, where W_s is the weight of water present in a swollen hydrogel at a certain temperature and W_d is the weight of the hydrogel after air drying. Percentage of swelling, P.S. = Q × 100. The equilibrium swelling percentage is derived from the plot of percentage of swelling using conventional techniques as described by El-Hamshary et al.¹⁸ A kinetic study of swelling was carried out in order to investigate the change in diffusion characteristics of the solvent (for hydrogel) through the polymeric gel network. Here nth order kinetics was assumed as indicated by Ruiz et al.¹⁹ Swelling–deswelling cyclic kinetic studies were performed on up to 4 cycles. Here gels were not allowed to attain an equilibrium state under swelling or deswelling (time required to attain equilibrium is 22 h on average). The exposure time was 1 h initially and varied up to 14 h to study the gel reusability in dynamic conditions for swelling and deswelling.

3.4 Studies on thermo-responsive characteristics

The effect of the temperature of the swelling medium on the swelling characteristics of the different hydrogels was studied by carrying out swelling at 22, 37 & 48 °C in water at pH 6.5. The samples were exposed for 3 hours at each temperature.

3.5 Rheological characteristics

Rheological studies were conducted using an Anton Paar Rheometer, Physica (MCR-102). Rheology of the air dried gel after attaining equilibrium was examined using DMTA with a CTD 450 attachment. The thermal sweep test was performed using samples of dimension 1 cm × 2 cm × 0.1 ± 0.02 cm (W × L × T) from −20 °C to 60 °C at a heating rate of 5 °C min⁻¹ under oscillation by ramp in shear stress from 0.05 to 0.01 MPa under a constant frequency of 1 Hz. Shear stress ramping from 0.1 to 0.02 MPa was used as a pretension force. The glass transition temperature was also determined from the peak of the tan δ plot using the same conditions. All the statistical analysis and fitting were done using Origin 8.5 software.

3.6 Morphology studies

Scanning electron micrographs of the gel surfaces were studied using ESEM apparatus, CARL ZEISS EVO18, Special Edition, Germany. CRYO-SEM was performed at −160 °C after fracturing the samples at −187 °C followed by platinum coating (8 nm) in an argon atmosphere at −155 ± 5 °C to analyze the pore structure. “Smart SEM V-05.04.05.00” software was used to measure pore diameter and plotted in Origin 8.5.

4. Results and discussion

4.1 Spectroscopic studies

Fig. 1, showing the FTIR spectra of different samples, exhibits a broad peak due to O–H stretching at 3233 cm⁻¹ (ref. 20) for CPVD (crosslinked polyvinyl alcohol along with diethylene glycol in different proportions). It is apparent that incorporation of DEG does not affect the hydrogen bonded skeleton in the crosslinked matrix. The peaks at 2917 cm⁻¹ & 2849 cm⁻¹ are due to methyl C–H stretching. There is also a peak at 1579 cm⁻¹ & this is attributed to C=C stretching. Here unsaturation in the gel may be attributed to the dehydration of the gel matrix, i.e. PVA in the presence of conc. H₂SO₄. The peaks at 1390 cm⁻¹ are assumed to appear for carboxylate salts developed mainly due to the partial oxidation of the polymeric matrix by conc. H₂SO₄. The very broad peak at 1127 cm⁻¹ suggests that C–O–C stretching developed due to acetal linkage formation during the crosslinking reaction.²¹ The peak at 880 cm⁻¹ corresponds to C–H bending of alkene. All the samples exhibit identical spectra, which indicates that there is no effect of DEG on crosslinked PVA as no new peaks are generated nor any peaks depleted on addition of DEG. This is observed for all CPVD gels 1 to 4. The acetal linkage formed is exclusively due to the interaction of glyoxal and the hydroxyl group of DEG. There has not been any modification or alteration of the acetal linkages with the increase in DEG percentage.

Fig. 1 Comparative FTIR spectra of dry CPVD gels.

In order to investigate the possibility of any interactions between DEG and glyoxal, an experiment was designed such that the above mentioned constituents could get an opportunity to interact in the absence of PVA under identical conditions in reference to those described earlier. The stoichiometric ratio of glyoxal & DEG was kept the same as that maintained in the presence of PVA.

Four samples were investigated through FTIR & UV spectral analysis. These are symbolized in the spectra as: T1 (only distilled water), T2 (solution of glyoxal), T3 (solution of DEG), and T4 (the aliquot taken after a reaction time of 30 minutes of the original mixture used during the gel preparation but without any PVA).

FTIR spectra of T2, T3 and T4 are shown in Fig. 2(a). All of the samples show a broad peak at 3000–3500 cm⁻¹ which appears due to O–H stretching and the presence of water molecules in the liquid under testing. At 1636 cm⁻¹ a sharp peak appears indicating the presence of water, which masks the peak supposed to appear due to C=O stretching of the aldehyde group (glyoxal). Another weak peak appears at 1071 cm⁻¹ mainly due to O–H stretching and at 1120 cm⁻¹ which is attributed to C–O stretching indicating the presence of DEG. For T2, a broad but weak peak (1030 to 1110 cm⁻¹) appears which can be ascribed to the C–O stretch for glyoxal. Similar peaks have been found to appear in the spectrum for DEG and for the aliquot of the reaction mixture after heating for 30 minutes. From the spectrum of T4 it was found that no new peak appears in the spectrum of the aliquot. Max et al.²² showed that in a typical FTIR spectrum of any acetal compound (assumed to be formed by the interaction of an alcohol with an aldehyde), a characteristic peak of acetal should appear at 993 cm⁻¹. Its complete absence indicates the inertness of the glyoxal & DEG towards each other under the reaction conditions used here.

Fig. 2 (a) FTIR spectra and (b) UV spectra of the individual reactants (glyoxal and DEG) and the product after reaction under the conditions used for gel preparation.

UV spectroscopy shows two peaks appearing for glyoxal (T2) as well as for aliquot (T4) at 214 cm⁻¹ and 273.5 cm⁻¹ in Fig. 2(b). The former peak appears due to π–π* and the later peak for n–π* transitions indicating the presence of glyoxal in the solution under testing.

This indicates that no reaction occurs between glyoxal and DEG, under identical conditions to those used for gel preparation. Thus it may be inferred that DEG here functions only as a porogen.

4.2 X-ray studies

A wide angle X-ray scattering (WAXS) study was undertaken to investigate the molecular packing & hence to get an idea of the d-spacing in the crystal arrangement of the gel forming polymer. The diffractograms & the calculated results are well described in Fig. 3 and 4 respectively. Fig. 3 shows a comparative spectral analysis of CPVD families which exhibit almost no change in peak patterns. This reveals the appearance of sharp crystalline peaks of (101̄), (101) at 2θ values of 19.02° and 20.21° respectively²³ for all the gels, however they differ in intensity of the peak (Fig. 3), crystallite size and d-spacing values (Fig. 4). The peaks indicate the orthorhombic structure of the polyvinyl alcohol which is not affected by the addition of DEG.²⁴ From Fig. 3, it is observed that with increasing DEG percentages, the peak heights also increase & reach their maxima for CPVD4. From this study it is observed that sharp, crystalline peaks of (101) at a 2θ value of 20.21° are present in all the four gel samples. From this study the crystallite sizes are calculated using the Scherrer equation. The crystallite size has maximum value for CPVD4 & for the other three, the crystallite sizes differ marginally as indicated in Fig. 4. The d-spacing values (in angstrom) show a marginal increase with the increase in DEG percentage. Thus the cage formation & the consequent increase in percent swelling due to the porogen DEG may be assumed to take place & it is also observed that the maximum swelling occurs for CPVD4.

Fig. 3 Comparative 3-dimensional WAXS spectra indicating the deviation of peak height of dry CPVD gels.

Fig. 4 Variation of crystallographic parameters with the percentage of DEG for the gels.

4.3 Swelling studies

The equilibrium swelling percentage data of the different gel samples reveals that CPVD4 reaches maximum swelling in distilled water at room temperature (22 °C) & the magnitude increases with the increment in DEG percentage incorporated in the matrix as indicated in Fig. 5. The maximum swelling observed with the gel CPVD4 may be ascribed to the combined action of crosslinking of the glyoxal and the porogen effect of DEG. This supports the cage formation in the gel enabling it to accommodate the maximum amount of water amongst all gels. Water retention capacity is enhanced with the increase in DEG percentage. Swelling rate also follows the same trend.

Fig. 5 Comparative swelling study of the CPVD gels with time in distilled water at 25 °C and variation of equilibrium swelling percentage of the same.

The kinetic study of swelling was performed by assuming nth order diffusion of water for this gel. It is an established fact that when the value of n is 0.5, then Fickian diffusion occurs & when the value is between 0.5 and 1.0 then non-Fickian diffusion takes place. In our experiment, for all four samples the n value lies between 0.5 and 1.0 which definitely indicates the non-Fickian type of diffusion. Fig. 6 shows the variation of the exponent n with the percentage of DEG, & it is observed that n increases a little up to 75% DEG, beyond which it remains constant.

Fig. 6 Analysis of swelling kinetics data with the percentage of DEG loading in the crosslinked PVA matrix.

A cyclic swelling–deswelling study was carried out by immersing the swelled gel with various percentages of DEG in a solution containing 0.5 mol% NaCl as described by Masuda et al.²⁵ and Patachia et al.²⁶ for deswelling. After 84 minutes of deswelling in this way, the gel was further immersed in water at pH 6.5 and at a temperature of 22 °C for re-swelling. This cycle of swelling & deswelling was carried out for all the hydrogel samples. This study was performed on up to 4 cycles and described in detail in Fig. 7. The gels were not allowed to attain an equilibrium state during the cycle either for swelling or for deswelling. The study was undertaken to characterize the gels' ability to swell and deswell in a non-steady state condition. This is why the swelling responses did not corroborate the findings as obtained in Fig. 5 which indicated an equilibrium condition. A differential rate of solvent uptake by the gel with time may be the reason for such observations. At the initial stage of swelling, it was found that CPVD4 shows slower uptake in water but uptake greatly enhances with time. From this study it is observed that some amount of water is retained by the gels after the second swelling–deswelling cycle is completed. However when these cycles are further repeated, all the gels except CPVD4 release almost all the water molecules that get entrapped during swelling due to the osmotic pressure difference at the solvent gel interface. Increasing the number of swelling cycles also leads to increase in swelling efficiency. This may be the reason for an increase in osmotic pressure difference when the swelled gel is immersed in NaCl solution, which in turn compels the gel to release most of the solvent entrapped in its cage structure. CPVD4 maintains the capacity to retain water which is not observed for any other gels. It is interesting to note that CPVD4 develops this unique capacity of retaining the maximum amount of water during deswelling in the presence of sodium chloride solution. The abrupt change in the capacity to retain water for the gel CPVD4 may possibly be ascribed to a combined effect of gel elastic recovery, extent of crosslinking, & the increased possibility of cage expansion which counteract the effusion of water.

Fig. 7 Cyclic swelling–deswelling dynamics study of the CPVD gels (up to 4 cycles).

4.4 Studies on thermo-responsive characteristics

Fig. 8 shows the influence of temperature on various gels under study. Here CPVD1 shows a slight increase in swelling with increase in temperature up to 48 °C and the reverse behavior is observed during cooling. This can be ascribed to the presence of the very low percentage of DEG (CPVD1) which is expected to have a negligible effect on the thermo-responsive cage expansion of the gel. Except for CPVD1, the gels swell at 22 °C and 37 °C (under investigation in the present study), but show deswelling at 48 °C (under investigation in the present study). To prove the efficiency & reproducibility, the gels were allowed to equilibrate at 48 and 37 °C in cyclic order and the results showed similar trends repeatedly.

Fig. 8 Variation of thermal response with increase in DEG percentage for CPVD gels.

4.5 Rheological characteristics

Fig. 9(a) to (e) describe the DMTA results of CPVD0 to CPVD4 respectively. CPVD0 shows a distinct peak corresponding to the glass transition temperature of the material. The DEG incorporated matrices exhibit no such clear peak in their tan δ plots. The presence of a multiphase structure (as exemplified in morphology as indicated in Fig. 10), obtained in the presence of DEG, may be responsible for such behavior of the material. The comparative study of the storage moduli shows that DEG resists a fall of storage modulus with an increase in temperature. When compared at a particular temperature, the storage modulus falls off with the increase in DEG percentage under shearing conditions but the loss modulus shows anomalous behavior when compared under the same conditions. From the peak of tan δ, T_g is estimated and the variation of the same with DEG percentage (%) is plotted in Fig. 9(f). According to the best fitted curve, the T_g (Y) decreases linearly with the DEG percentage, “x” (%), and this decrease can be approximated by the relationship given here, Y + 0.167x = 64.6. From the Pearson's r value of −0.94 we can statistically say that the two variables possess very strong negative correlation. The multiple maxima of the tan δ plots with the increase in temperature are responsible for the multiphase distribution in the matrix. Though a more uniform sharp peak is obtained in the case of hydrogel without any DEG content and with maximum DEG content, the rheology corresponds to a uniphasic structure for both of them. Pore size distribution, also responsible for such behavior, is found to be related to morphology.

Fig. 9 Variation of rheological parameters, storage modulus, loss modulus & tan δ with temperature for (a) CPVD0, (b) CPVD1, (c) CPVD2, (d) CPVD3 and (e) CPVD4; (f) comparative analysis of T_g with the variation of DEG percentage estimated using DMTA study of the same.

Fig. 10 Scanning electron micrographs of cryo-fractured specimens of (a) CPVD1, (b) CPVD2, (c) CPVD3 and (d) CPVD4.

4.6 Morphological study

Accurate pore structure and its distribution are obtained from CRYO-SEM images for the gels. The gel samples were cryo-fractured after attaining equilibrium swelling percentage and Fig. 10 exhibits images of the CPVD families. This shows irregular pore distribution all over the surfaces, but the pores seem to become more regular and ellipsoid in shape with the increase in DEG percentage. The surface shows a multiphase structure for all of them. This also supports the response in the DMTA study.

To obtain detailed information on pore size distribution an attempt has been made to count the pore average diameter for 500 micron² areas. The distribution is described by Fig. 11. The study indicates the shifting from polymodal to unimodal distribution in pore diameter with an increase in DEG percentage. The sizes are also skewed to lower values in the nanometer range with the increase in DEG. It is observed that the percentage of nanopores seems to face an increment on the basis of this count with the increase in DEG loading. This increment follows a linear relationship with DEG percentage, which is represented by the equation plotted in Fig. 12(a). Here “y” is the percentage of nanopores and “x” is the percentage of DEG (%) loading. Standard error for slope and intercept are 0.01 and 1.19 respectively. The Pearson's r value is 0.77295 which statistically indicates that the two variables concerned have a very strong positive relationship. The average pore size falls first for CPVD2, but shows marginal increase beyond it as indicated in Fig. 12(b).

Fig. 11 Analysis of the pore size distribution of (a) CPVD1, (b) CPVD2, (c) CPVD3 and (d) CPVD4 from the CRYO-SEM images.

Fig. 12 Variation of (a) percentage of nanopore and (b) average pore size with the percentage of DEG.

5. Conclusion

Our investigation shows that diethylene glycol (DEG) can be used as a porogen to obtain a nanoporous structure in a PVA gel matrix along with the incorporation of smart thermo-responsive activity in the gel. Thus a possible application of these types of gel may be in the tissue engineering field. An increase in nanopores with the increase in DEG percentage is clearly distinguishable in this study. A reduction of pore size also helps in increasing the percentage of pores in the matrix that in turn helps to increase the specific surface area which is beneficial in many areas as discussed in the introduction. The distribution of pores and their morphology are found to have an impact on the thermo-mechanical behavior of the material. Thermo-responsive swelling–deswelling may invoke interest in the application of such materials as mentioned in the introduction section. The future scope of work involves the exploration of the possibility of application of the gel in collaboration with a physiological/clinical laboratory.

Acknowledgements

The authors gratefully acknowledge the University of Calcutta for providing research facilities and the Fellowship. The authors extend their gratitude to Mrs Mahuya Biswas and Mr Abir Deb of DIC India Limited, Kolkata for assistance in FTIR characterization of the samples.

References

1. R. H. Schmedler, K. S. Masters and J. L. West, Biomaterials, 2002, 23, 4325–4332.
2. E. A. Scott, M. D. Nichols, R. Kuntz-Willits and D. L. Elbert, Acta Biomater., 2010, 6, 29–38.
3. S. A. Bencherif, T. M. Braschler and P. Renaud, J. Periodontal Implant Sci., 2013, 43, 251–261.
4. A. G. Lee, C. P. Arena, D. J. Beebe and S. P. Palecek, Biomacromolecules, 2010, 11, 3316–3324.
5. Y. Zhang and L. Ye, Polym.-Plast. Technol. Eng., 2011, 50, 776–782.
6. A. Baraka, P. J. Hall and M. J. Heslop, J. Hazard. Mater., 2007, 140, 86–94.
7. H. Yuan, K. Kurashina, J. D. de Bruijn, Y. Li, K. de Groot and X. Zhang, Biomaterials, 1999, 20, 1799–1806.
8. V. Karageorgiou and D. Kaplan, Biomaterials, 2005, 26, 5474–5491.
9. A. Clark, T. A. Milbrandt, J. Z. Hilt and D. A. Puleo, Acta Biomater., 2014, 10, 2125–2132.
10. Q. Hou, D. W. Grijpma and J. Feijen, Biomaterials, 2003, 24, 1937–1947.
11. I. Iturralde, M. Paulis and J. R. Leiza, Eur. Polym. J., 2014, 53, 282–291.
12. M. S. Silverstein, Polymer, 2014, 55, 304–320.
13. E. Salehia and S. S. Madaeni, Appl. Surf. Sci., 2014, 288, 537–541.
14. J. Courtois, E. Byström and K. Irgum, Polymer, 2006, 47, 2603–2611.
15. S. Yano, K. Kurita, K. Iwata, T. Furukawa and M. Kodomari, Polymer, 2003, 44, 3515–3522.
16. J. Rojas and E. Azevedo, Int. J. Pharm. Sci. Rev. Res., 2011, 8(1), 28–36.
17. N. Degirmenbasi, D. M. Kalyon and E. Birinci, Colloids Surf., B, 2006, 48, 42–49.
18. H. El-Hamshary, Eur. Polym. J., 2007, 43, 4830–4838.
19. J. Ruiz, A. Mantecón and V. Cádiz, Polymer, 2001, 42, 6347–6354.
20. H. S. Mansur, C. M. Sadahira, A. N. Souza and A. A. P. Mansur, Mater. Sci. Eng., C, 2008, 28, 539–548.
21. M. Ştefănescu, M. Stoia, O. Ştefănescu, C. Davidescu, G. Vlase and P. Sfîrloagǎ, Rev. Roum. Chim., 2010, 55(1), 17–23.
22. J. J. Max and C. Chapados, J. Phys. Chem. A, 2007, 111(14), 2679–2689.
23. R. Ricciardi, F. Auriemma, C. D. Rosa and F. Lauprêtre, Macromolecules, 2004, 37, 1921–1927.
24. J. Hana, T. Lei and Q. Wu, Carbohydr. Polym., 2014, 102, 306–316.
25. Y. Masuda, T. Tanaka and T. Nakanishi, Colloid Polym. Sci., 2001, 279(12), 1241–1244.
26. S. Patachia, A. J. Valente and C. Baciu, Eur. Polym. J., 2007, 43(2), 460–467.

---