The use of chitosan beads for the adsorption and regeneration of heavy metals

Abstract

This work studied the removal of heavy metals from wastewater through the use of South African chitosan beads produced from locally available raw materials. For this purpose, chitosan beads were prepared from chitosan flakes that were synthesized from the chitin derived from the exoskeleton of the Jasus lalandii. The molecular weight and degree of deacetylation of the chitosan flakes were 9.4-10 4 g/mol and 83% respectively. When the flakes were converted into non-cross-linked beads, the molecular weight decreased slightly to 7.8-10 4 g/mol. Different beads were prepared ranging in size from 0.9 to 3.8 mm and the amount of glutaraldehyde used to crosslink the beads was varied between 0 and 4 vol%, in order to obtain beads with a different degree of cross-linking. The beads were used as an adsorbent for heavy metals and were characterized for equilibrium and kinetic adsorption studies. The mine concentration, which is in direct relation to the adsorption capacity of non-cross-linked beads was determined as 4.9 mmol/g. The amine concentration decreased with an increasing glutaraldehyde concentration and a decreasing bead size. Cross-linking was however necessary to make the chitosan stable in acidic media, and a degree of cross-linking larger than 18% made the chitosan beads insoluble at a pH of 2. Two models, the Langmuir isotherm model and a pH-model were used to fit equilibrium adsorption data. Although the Langmuir model gave good fits, the obtained parameters were pH dependent. On the other hand, the pH-model, which was derived from: i) the adsorption equilibrium reaction between the chitosan and the metal; ii) the acid base properties of chitosan; and iii), a mass balance of the different forms of nitrogen in the chitosan, could satisfactory describe the adsorption using pH independent variables. When deriving the pH-model the effect of pH on the degree of protonation of the adsorbent was considered. The model was fitted with the maximum adsorption capacity, and the fitted values were in close agreement with the amine concentration. The desorption of the metal from the chitosan could also be predicted well with this model, indicating a reversible complexation of the metal on the chitosan, making the recovery and possible re-use of the metal possible. The kinetics of the adsorption process were described with a shrinking core model, where an instantaneous adsorption reaction was assumed. From this model, effective diffusion coefficients were determined from batch experiments. The adsorption was also studied in a column and the experiments were modelled with a CSTR's in series model, using the experimentally determined adsorption equilibrium data. The breakthrough curve could be described reasonably well with this model, and the fitted effective diffusion coefficient was close to the one determined in the batch experiments. The adsorption capacity of the locally sourced and produced chitosan beads was high in comparison to the values indicated in the literature for other adsorbents. It was also found to be higher than that of either the commercially produced chitosan or the ion-exchange resin. The regeneration of the metal from the chitosan was effective. Multiple adsorption/desorption experiments were also carried out, and it was found that the adsorption increased for the second and third cycle, but decreased for the fourth and fifth ones. After the fifth cycle, the chitosan was physically damaged and could not been used anymore. This degeneration of the beads across multiple adsorption/desorption cycles was found to be the major concern blocking the uptake of the studied chitosan beads in industrial applications.