Development of a Novel and Green Mannich Reaction-modified Chitosan Hydrogel for Dysprosium(III) Recovery

Abstract

In this dissertation, a two-step protocol for the fabrication of a novel, green, and sustainable chitosan-based hydrogel adsorbent was developed for the recovery of dysprosium (III) (Dy(III)), an important rare earth element (REE) used in the manufacture of magnets. The virgin chitosan’s physicochemical and physicomechanical properties were improved by crosslinking the primary amine of chitosan with glutaraldehyde and 4-hydroxycoumarin via a one-step three-component Mannich reaction to form a modified-chitosan hydrogel (Cs/MB). The utilization of 4-hydroxycoumarin as a component in the Mannich crosslinking reaction was justified by its efficient reactivity with variety of primary amines that produced a library of 4-hydroxycoumarin Mannich products in moderate to high yields (60-90%) and high purity (97.8 ± 3.1%) verified by Nuclear Magnetic Resonance (NMR) spectroscopy, without the need for a catalyst or chemical purification.

The functionalization of the Cs/MB with a phosphate group at the primary alcohol of the chitosan’s backbone was found to be imperative for the coordination of Dy(III) that was achieved by phosphorylation of chitosan with P₂O₅ in methanesulfonic acid by an established method. The determined sequence of this two-step protocol that optimized product’s physical and chemical quality was to first introduce the phosphate group to the virgin chitosan by the P₂O₅/methanesulfonic acid method, then crosslinking via the three-component Mannich reaction in aqueous ethanol at room temperature for 12 h. PCs/MB was characterized by Fourier Transform-Infrared Spectroscopy (FT-IR) verifying the presence of the key functional groups added to the chitosan’s backbone polymer namely the phosphate group with peaks at 1319, 972, 908, and 514 cm^-1 and the 4-hydroxycoumarin with a peak at ~754 cm^-1. Elemental mapping verified the presence of the Na in the PCs/MB, suggesting a successful conversion of the hydrogel to its Na-form.

Batch adsorption studies of Dy(III) with PCs/MB were conducted and obtained a maximum adsorption capacity of 34 mg/g at room temperature for 2 h, whereas the column adsorption attained a saturation adsorption capacity of 22 mg/g after 3 h with bed length of 1.8 cm, flow rate of 0.1 mL/min, and feed concentration and pH of 13 mg/L and pH = 5.2, respectively, at room temperature. The presence of the lanthanide-phosphate peak at ~630 cm^-1, absence of the glycosidic C-O-C stretching at ~1154 cm^-1, and weakening of the N-H peak at ~3400 cm^-1 in the FT-IR spectrum of the spent PCs/MB provided direct evidence of a tandem coordination of Dy(III) onto the phosphate group of one hexose unit and a subsequent complexation onto the free amino group of the next unit. The presence of Dy(III) and the absence of Na in the surface of the spent PCs/MB by elemental mapping confirmed an ion exchange phenomenon between Na of the PCs/MB and Dy(III). Kinetics and thermodynamics studies predicted that the transport of Dy(III) onto the PCs/MB is limited by surface reaction and the adsorption of Dy(III) onto the PCs/MB is an exothermic process with a Gibb’s free energy of adsorption of +1.88 kJ/mol and equilibrium constant of 0.47 at 25°C.

Competitive adsorption studies of Dy(III) and neodymium(III) (Nd(III)) of approximately equimolar concentration were carried out both in batch and continuous mode, which consistently achieved a separation factor of 2.2 ± 0.2 in preference to Dy(III). Complete recovery of Dy(III) and Nd(III) from PCs/MB was achieved by batch desorption in two subsequent 48-hr cycles using 2.0 g/L of 0.001 M HCl per cycle and a total REEs concentration factor of 1.8x was attained in the first cycle. The C-O-C glycosidic peak at ~1154 cm^-1 in the FT-IR spectrum reappeared after desorption, which provided direct evidence of Dy(III) liberation from PCs/MB by desorption.