Langmuir, Vol.22, No.20, 8516-8522, 2006
Gelatin hydrogels cross-linked with bisvinyl sulfonemethyl. 2. The physical and chemical networks
This paper deals with the physical and the chemical gelation of gelatin in the presence of a reactant, bisvinyl sulfonemethyl (BVSM). The strategy of this investigation is to separate the contributions of the two types of cross-links in order to deduce the resultant elasticity of the network. In addition, the question raised by several authors concerning an increase of the thermal stability of the triple helices in the presence of cross-links was examined by using several techniques. In this study, the concentration of gelatin and BVSM were kept constant, while the influence of the thermal protocols was put in evidence. The gel formation was followed by rheological, thermodynamic (microcalorimetry), and optical spectroscopy (optical rotation) measurements. The results demonstrate the large differences which arise on the storage moduli by changing the thermal protocols. Cross-linking of the networks in the presence of the triple helices induce a heterogeneous repartition of the bonds, which can form along the triple helices and at the end of the sequences. Consequently, the rubber like network obtained by denaturation of the triple helices is still reminiscent of the initial twist of the chains, and a large modulus is observed, as if rigid segments were still present (storage modulus 10 times larger than for random cross-linking). The hydrogels have an elastic modulus which is larger that the addition of the physical and chemical contributions. The interpretation of the network elasticity is based on the predominant role of the rigid rods of triple helices, where the BVSM cross-links can either modify the ratio between the apparent length and distance between rods, l/d, and/or increase the rigidity of the interchain connections, which are loose coils for the physical gels. The hydrogels investigated have a network which is still close to the percolation threshold of the physical gel, and therefore, the statistical models known for well developed networks cannot be directly validated in these experimental conditions.