Integrating diffusion dialysis for sustainable acid recovery from ion exchange regeneration stages: Characterization of metal and non-metal ions migration

Abstract

Seawater mining presents a potential option for recovering the European Union’s Critical Raw Materials (CRMs), but direct extraction from seawater is challenging due to their low concentrations, as most of them are Trace Elements (TEs) (at levels of mg/L or µg/L). Saltworks bitterns (ultraconcentrated brines resulting from the sea salt production process) offer an alternative solution, naturally concentrated up to 40 times more than seawater. These bitterns can be further processed with chelating Ion Exchange (IX) sorbents to selectively extract TEs. However, this process requires an acidic elution stage with strong acids, followed by neutralization, to recover TEs through precipitation, demanding extensive chemicals consumption. Diffusion Dialysis (DD) could be used to recover the excess acid without external reagents, using an acid-resistant Anion Exchange Membrane (AEM). This study evaluates DD through batch and once-through tests for acid recovery from simulated IX eluate generated in the elution stage of TEs (B, Ga, Ge, Co, Sr) recovery from saltworks bitterns. Batch tests achieved high recoveries for HCl (45–50 %) and H2SO4 (30–37 %), being the theoretical maximum attainable recovery equal to 50 %. B and Ge only partially permeated through the membrane (82 % rejection) by a diffusion mechanism in their neutral form (H3BO3(aq), H4GeO4(aq)). Ga, Co and Sr, in cationic form, were highly rejected (>96 %). Permeability followed the order Ga < Sr < Co < B, due to the relevant charge and size. HCl permeability correlated linearly with concentration, while H2SO4 was inversely proportional. Once-through tests showed higher acid (74 % HCl, 62 % H2SO4) and oxoacid (66 % H3BO3(aq), 52 % H4GeO4(aq)) recovery at a low specific flow rate, or apparent flux, (0.38 L/(m2membrane·h)) due to increased residence time. Water to acid flow rate ratios did not affect species transport when an excess of water was guaranteed. Conversely, an influence was observed when the ratio was below 1, with a minimum at 0.18, where a very low passage of species was observed due to the reduced dilution volume of the dialysate solution (water). A 1D transport model, incorporating the solutes permeabilities determined experimentally, effectively described the system performance, especially for HCl and B, albeit slightly overestimating the other TEs’ transport