|M.Sc Student||Aviezer Yaron|
|Subject||A New thermal-Reduction-based Approach for Producing|
Mg(s) from Seawater
|Department||Department of Civil and Environmental Engineering||Supervisors||Professor Ori Lahav|
|Dr. Liat Birnhack|
|Full Thesis text|
A new process is presented for producing metallic magnesium from seawater. The new process is based on extraction of Mg2 from seawater followed by precipitation of the produced magnesium hydroxide on the surface of fine FeSi powder, which is the reducing agent in the silicothermic production of metallic magnesium. Two different techniques were examined for the Mg2 extraction step from seawater. The first tested technique used an anion exchange resin to separate the sulfate anions from seawater RO brine. This step was followed by concentration of divalent cations using an NF membrane. Preliminary experiments showed that the concentrated Mg2 solution contained considerable amount of impurities which remained in the raw product cake after precipitation of the magnesium hydroxide. The impurities, which included mainly gypsum, sodium and potassium, were difficult to remove, leading to reduced reduction efficiency and product purity. Hence, a second Mg2 extraction technique was developed, which was based on selective ion exchange extraction of the Mg2 from seawater, without carrying Na, K and sulfate impurities to the proceeding production steps and further into the product Mg(s). The magnesium extracted from the ion exchange resin was reacted with lime and precipitated on the surface of fine ferrosilicon suspension to produce a Mg(OH)2-Ca(OH)2-FeSi solids cake, which, after further wash and decantation, yielded raw material at a quality similar to that used in the classical dolomite-based Pidgeon silicothermic reduction technique. The filtrate of the raw cake decantation step, comprising predominantly of CaCl2, was used as the ion-exchange regeneration solution. Ca2 ions are adsorbed on the resin and thereafter replenished through lime dosage, applied for the Mg(OH)2 precipitation. The close contact between the Mg(OH)2 and the FeSi in the dry cake enabled direct usage of the cake in ~1150 oC retort which was applied for single-step dehydration and thermal reduction of the Mg2, thereby minimizing the problem of heat and mass transfer through the briquettes and eliminating the need for separate calcination, briquetting and reduction. Feasibility dehydration and reduction tests yielded 99.0-99.5% pure Mg(s). Materials and energy cost assessment revealed that the proposed process is ~$0.75 more expensive per kg of magnesium than the dolomite-based process, assuming equal thermal reduction yields. The cost estimation stimulates further optimization and parametric study of the proposed process and suggests that production of Mg(s) in places where dolomite is not available, requiring seawater as the sole raw resource, is feasible.