| Abstract [eng] |
This dissertation focuses on the synthesis of different composite materials as adsorbents for caesium, cobalt, and europium, their characterisation, adsorption studies, and the use of an adaptive neuro-fuzzy inference system to predict the adsorption capacity of composites. It was found that the muscovite mica clay-graphene oxide-maghemite-magnetite composite has a sufficiently high adsorption capacity for Cs(I) and Co(II). The composites Prussian blue-graphene oxide 2.1, Prussian blue-graphene oxide 2.2, and magnetite-Prussian blue-graphene oxide 3.2 adsorb Cs(I) better than magnetite-Prussian blue-graphene oxide 2.3. Among the chitosan-mineral composites, the chitosan-muscovite mica clay, chitosan-muscovite mica clay cross-linked by epichlorohydrin, and chitosan-montmorillonite modified by glycerol composites showed the highest adsorption capacity for Cs(I) and Co(II), and chitosan-zeolite for Eu(III). Based on experimental and literary data, the suggested mechanism for the adsorption of Cs(I) and Co(II) on muscovite mica clay-graphene oxide-maghemite-magnetite and of Cs(I), Co(II), and Eu(III) on chitosan-mineral composites is ion exchange, complexation, and electrostatic attraction. The mechanism of Cs(I) adsorption on Prussian blue-graphene oxide and magnetite-Prussian blue-graphene oxide composites is complexation, ion exchange, and ion trapping. The adaptive neuro-fuzzy inference system shoved good performance and generalisation ability to predict the adsorption capacity of composites. |
