Reactivity of suspended iron oxide particles in low temperature alkaline electrolysis

Abstract

Electrolysis of iron ore is a promising CO₂-free iron production route based on the electrolytic reduction of solid iron oxide particles suspended in 110 °C concentrated alkaline electrolyte. The reactivities of different iron compounds during their reduction into iron through this process have been compared using a model laboratory cell. Chronoamperometry experiments were performed on suspensions containing hematite (α-Fe₂O₃), magnetite (Fe₃O₄) or goethite (α-FeOOH) at a cell voltage of 1.66 V. Current density response, anode and cathode electrochemical potentials, faradaic efficiency and iron deposit morphology were compared. Hematite reduces to iron at 1100 A/m² with current yield near 85%. For goethite, the current density response was 33% lower and current efficiency dropped by 20% compared to hematite. Magnetite reactivity proved to be extremely low with eight-time lower current density and tenfold lower current efficiency than hematite. The weaker reactivity of goethite and magnetite particles could be ascribed to their more difficult adsorption on the cathode surface partly covered with metal iron, the far higher viscosity of goethite suspensions, although at the same solid concentrations as the other oxide particles and the occurrence of opposite electrode reactions with dissolved Fe(OH)₃ ⁻ and Fe(OH)₄ ⁻ ions.

Graphical Abstract

Appendix

The reactive adsorption reactions considered in this paper involve several solid compounds and liquid water in a 50 wt% NaOH–H₂O mixture. Gibbs free enthalpy of any of these reactions was obtained by summing the chemical potential of every compound involved in the balance reaction multiplied by their algebraic stoichiometric coefficient.

Each solid compound was considered a pure compound, therefore their chemical potential is equal to their standard chemical potential at considered temperature under 1 bar, that is to say to their molar Gibbs free energy at considered temperature under 1 bar:

$${\mu _i}(110\,^\circ {\text{C}},1\,{\text{bar}})=\mu _{i}^{o}\left( {110\,^\circ {\text{C}}} \right)=g_{i}^{o}\left( {110\,^\circ {\text{C}}} \right)=\mathop \int \limits_{{298.15}}^{{383.15}} CpdT - 383.15\mathop \int \limits_{{298.15}}^{{383.15}} \frac{{Cp}}{T}dT$$

Thermodynamic data used were IAPWS-95 [16] model for water thermodynamic properties and Barin’s data [17] for the heat capacities of solid compounds.

The chemical potential of water can be expressed as follows:

$${\mu _{{H_2}O}}\left( {110\,^\circ {\text{C}},1\,{\text{bar}}} \right)=\mu _{{{H_2}O}}^{o}\left( {110\,^\circ {\text{C}}} \right)+RTln({a_{{H_2}O}}(110\,^\circ {\text{C}},~25\,{\text{mol}}/{\text{kg}}))$$

IAPWS-95 model was used to represent pure water Gibbs free energy. In order to calculate pure liquid water Gibbs free energy at a temperature above standard boiling point, pressure was increased to water saturated vapour pressure at considered temperature. At 110 °C, the saturated vapour pressure calculated by IAPWS-95 model is 143 309 Pa, i.e., approx. 1.43 bar. For such a low pressure difference, thermodynamic properties of liquid phase will barely change.

$$\mu _{{{H_2}O}}^{o}\left( {110\,^\circ {\text{C}}} \right)=g_{{{H_2}O}}^{{o,liq}}\left( {110\,^\circ {\text{C}}} \right)~\sim ~g_{{{H_2}O}}^{{*,liq}}\left( {110\,^\circ {\text{C}},P_{{{H_2}O}}^{{sat}}} \right)$$

Akerlof and Kegeles semi-empirical model which has more recently described by Balej [18], was used to represent water activity. The data used for this model were for temperature lower than 70 °C and molality lower than 17 mol/kg, hence the equation of water activity was extrapolated to the conditions of this study. For the conditions of interest, the water activity was estimated to be near 0.15.