Main reason for the current losses in electrolytic magnesium production is the reaction between electrode products. Present study was devoted to effective separation of chlorine gas from the electrolysis environment by a new cell design and thus reducing the extent of back reaction between magnesium and chlorine to decrease energy consumption values. The new cell design was tested by changing temperature, cathode surface, current density, anode cathode distance and electrolyte composition. Both the voltages and the current efficiencies were considered to be influenced by the amount and hydrodynamics of chlorine bubbles in inter-electrode region. Cell voltages were also found to be affected from the nucleation of magnesium droplets and changes in electrolyte composition that took place during the electrolysis. A hydrodynamic model was used to calculate net cell voltage by including the resistance of chlorine bubbles on anode surface to theoretical decomposition voltage during electrolysis. Good correlations were obtained between experimental and calculated voltages. The same model was used to calculate current efficiencies by considering chlorine diffusion from bubble surfaces. A general agreement was obtained between calculated and experimental current efficiencies. Desired magnesium deposition morphology and detachment characteristics from cathode were obtained when MgCl2-NaCl-KCl-CaCl2 electrolytes were employed. Current efficiencies higher than 90% could be achieved using the above electrolyte. The cell consumes around 8 kWh&amp / #903 / kg-1 Mg at 0.43 A&amp / #903 / cm-2 as a result of high chlorine removal efficiency and capability of working at low inter-electrode distances. Furthermore, the cell was capable of producing magnesium with less than the lowest energy consumption industrially obtained, at about double the commonly practiced industrial current density levels.

No description available.

This master's thesis covers an investigation of the reduction behavior of different iron oxides when electrolytically reduced with molten salt electrolysis (MSE). The tested iron oxides were wüstite (FeO), hematite (Fe2O3), magnetite (Fe3O4) and magnetite ore concentrate from the Tapuli deposit in Pajala, Norrbotten, Sweden. The properties of the Tapuli magnetite ore and magnetite ore concentrate were analysed from a mineralogical perspective to evaluate how the raw ore material influences the concentrate and its reduction by the MSE technology. The electrolytic experiments were performed in an Inconel 625 cell body within a pit-furnace. The different iron oxides were tested separately. The reduction samples were constructed of one or three iron oxide briquettes of 20 g each within a molybdenum mesh attached on a molybdenum tray with molybdenum wires. The molten electrolyte was kept at 500°C with an applied voltage of 1.7 or 2.1 V. The used electrolyte was sodium hydroxide (NaOH). The mineralogical examination shows that the Tapuli ore varies in composition between different locations of the deposit with respect to magnetite grain size and skarn composition and grain size. Point analyses with Laser Ablation Single Collector Inductively Coupled Plasma Mass Spectrometry (LA-SC-ICP-MS) on magnetite grains in thin sections from five drill cores fromdifferent parts of the deposit show that the element composition in the magnetite grains vary between the samples. Core-to-rim analyses for Fe, Mg, Mn and Al reveal relatively homogenous grades throughout the grains, with a few exceptions. Phase analysis with XRD shows that reduction has occurred in all tested iron oxides. Without prevention, the reduction products from trials on Fe2O3, Fe3O4 and magnetite ore concentrate show distinct XRD peaks of the by-product NaFeO2. According to XRD, the addition of Na2O seems to have reduced the NaFeO2 formation. Interestingly, no NaFeO2 was formed in the FeO trials. This might be explained by the absence of Fe3+ in FeO. The variation of the current-time curves of the trials is interpreted to depend on the voltage applied, the number of briquettes, briquette decomposition and Na2O addition. Electrolysis in molten NaOH can be used to reduce iron oxides. Despite this, NaOH might not be a suitable electrolyte for this process due to its interaction with Fe2O3 and Fe3O4 resulting information of NaFeO2. Na2O can be used as an additive to prevent formation of NaFeO2 but sharply decreases the current response, thus having an apparent negative effect on the process efficiency. Another preventive measure that can be tested is to calibrate the process voltage to decompose the NaFeO2 but not NaOH. Due to the shown interaction tendency of NaOH, other electrolytes should however be considered for this process. Regarding the Tapuli ore concentrate, more tests are needed to draw conclusions about how the trace elements effects its electrolytic behavior. / Denna masteruppsats avhandlar en undersökning av reduktionsbeteendet hos olika järnoxider vid elektrolytisk reduktion i saltsmälta (molten salt electrolysis (MSE)). Järnoxiderna som har testats är wüstit (FeO), hematit (Fe2O3), magnetit (Fe3O4) och magnetitmalmkoncentrat från malmfyndigheten Tapuli i Pajala, Norrbotten, Sverige. Malmkoncentratets egenskaper har analyserats ur mineralogisk synvinkel för att utvärdera hur den råa malmens mineralogi påverkar koncentratet och dess reduktionsbeteende vid elektrolys i saltsmälta. Elektrolysexperimenten utfördes i cellkropp av Inconel 625 placerad i en gropugn. De olika järnoxiderna testades separat. Reduktionsproverna utgjordes av en eller tre järnoxidbriketter på 20 g inuti ett molybdennät, fastvirade på en molybdenbricka med molybdentråd. Den smälta elektrolyten hölls vid en temperatur av 500°C med en applicerad spänning av 1.7 eller 2.1 V. Elektrolyten som användes var natriumhydroxid (NaOH). Den mineralogiska undersökningen visar att tapulimalmens sammansättning varierar mellan olika delar av fyndigheten med avseende på magnetitens kornstorlek och skarnets sammansättning och kornstorlek. Punktanalyser med Laser Ablation Single Collector Inductively Coupled Plasma Mass Spectrometry (LA-SC-ICP-MS) på magnetitkorn i tunnslip från fem olika borrkärnor visar att elementkoncentrationerna i magnetitkornen varierar mellan proverna. Core-to-rim-analyser på magnetitkornen visar att halterna av Fe, Mg, Mn och Al är tämligen homogena genom hela magnetitkornet med undantag av några få avvikande punkter. Fasanalys med XRD indikerar att reduktion har skett i alla försök. Utan prevention visar reduktionsprodukterna från försöken på Fe2O3, Fe3O4 och magnetitmalmkoncentrat klara indikationer av biprodukten NaFeO2. Enligt XRD verkar tillsats av Na2O ha minskat bildningen av Na2O för Fe2O3, Fe3O4 och Tapuli magnetitmalmkoncentrat. Intressant är att ingen NaFeO2 bildades i försöken med FeO. Förklaringen till detta skulle kunna vara avsaknaden av Fe3+ i FeO. De varierande ström-tidkurvorna från försöken tolkas bero på den applicerade spänningen, antalet briketter, brikettsönderdelning och tillsats av Na2O. Elektrolys i smält NaOH kan användas för att reducera järnoxider. Trots detta kanske NaOH inte är lämplig som elektrolyt i denna process, detta på grund av dess påvisade interaktion med Fe2O3 och Fe3O4 som resulterar i bildning av NaFeO2. Na2O kan tillsättas för att förhindra bildning av NaFeO2 men har en kraftigt negativ effekt på strömstyrkan i processen vilket minskar processens effektivitet. En annan preventiv åtgärd som kan testas är att kalibrera processens spänning så att NaFeO2 sönderdelas men inte NaOH. På grund av den konstaterade interaktionstendensen hos NaOH bör andra elektrolyter tas i beaktande för denna process. Angående magnetitmalmskoncentratet från Tapuli behövs fler tester för att dra slutsatser kring hur spårelementen påverkar dess uppförande vid smältelektrolys.