Aluminium dissolution in NaF-AlF3-Al2O3 systems

The rate of dissolution of electrolytically deposited aluminium was determined by the method of current reversal chronopotentiometry at a tungsten electrode in NaF?AlF₃?Al₂O₃ melts of varying NaF/AlF₃ molar ratios or cryolite ratios (CR). The temperature was maintained at 1031±3°C and the alumina content at 4 wt%. More accurate data were obtained by introducing delay times of various lengths (at zero current) between the cathodic and anodic current pulses, compared to direct current reversal chronopotentiometry with varying forward (deposition) times. The rate of aluminium dissolution increased with increasing NaF/AlF₃ molar ratio, the curve showing an inflexion in the vicinity of CR=3. This inflexion indicates two dissolution mechanisms, one being predominant depending on the CR. The main reaction in acidic melts (CR<3) may be represented by $$2Al(l) + AlF_6^{3 - } \rightleftarrows 3Al(I)F_x^{1 - x} + (6 - 3x)F^ - $$ while in basic melts (CR>3) $$Al(l) + 3Na^ + \rightleftarrows 3Na(soln) + Al(III)$$ is the likely dominant mechanism. For 0.8?7 mol cm^?2s^?1.     

Influence of CaF2 and AlF3 on the kinetics and mechanism of the Al electrode reaction in cryolite melts with various alumina contents

Electrochemical techniques were used to study the kinetics and mechanism of the aluminium electrode reaction in two cryolite-based melts containing cryolite with either 11 wt % AlF₃ or 5 wt % CaF₂ additions and variable alumina contents at 1000 °C. A three step electrode process was observed in both melts, comprising a preceding chemical reaction followed by two charge transfer steps. The exchange current density of the cathodic reaction was found to be dependent on the concentration of aluminium fluoride. By a combination of electrochemical impedance spectroscopy (EIS) and galvanostatic relaxation methods (GRM), the exchange current density of the first (slower) charge transfer step, the Warburg diffusion impedance, the double layer capacitance of the aluminium electrode and the rate of the preceding chemical step, were evaluated in the range of 2–8 wt % alumina. The role of the two additives, AlF₃ and CaF₂, was evaluated.     

Effect of fluoride mineralizers on alumina sintering

Conclusions Additions to ground commercial alumina of fluorides (AlF₃ or CaF₂) in amounts of 1% exert a retarding and refining action during the sintering of alumina fired in the temperature range 1200–1700°C. The AlF₃ is the most effective additive. Retardation of sintering is explained by the healing of the surface defects in the grains of Al₂O₃ under the influence of the mineralizers.Translated from Ogneupory, No. 1, pp. 37–40, January, 1970.     

Alkali and earth alkaline metal contents in the aluminium cathode in aluminium electrolysis

The contents of sodium, lithium, calcium and magnesium in aluminium in contact with NaF–AlF₃-based melts in laboratory and in industrial aluminium cells were investigated in the temperature range 950–1030 C. The experimental data were compared with a thermodynamic model. It was found that the addition of alumina or CaF₂ to the NaF–AlF₃ melts has only a minor effect on the equilibrium content of sodium in aluminium. Cathodic polarization enhances the content of sodium in aluminium. However, polarization has a smaller effect on the concentrations of lithium, calcium and magnesium in aluminium in industrial cells.     

Decomposition voltage for the electrolysis of alumina at low temperatures

The apparent decomposition voltage for the electrolysis of alumina in an equimolar Na₃AlF₆-Li₃AlF₆ electrolyte was measured over a temperature range of 800 to 1000° C by the extrapolation of voltagecurrent plots to zero current. Temperature coefficients of –1.9 and –2.4 mV° C^–1 were determined for conditions of variable alumina activity (constant concentration) and unit activity (saturated), respectively. The overvoltage contribution to the temperature dependency was estimated to be about –1.6mV° C^–1 (versus a –0.6 mV° C^–1 dependency for the reversible decomposition voltage). Reduced alumina solubility at low temperatures also appeared to increase the overvoltage, but was of secondary importance.Based on work partially supported by the US Department of Energy (Contract DE-AC03-76CS40215: Energy Savings Through the Use of an Improved Reduction Cell Cathode).     

Aluminium electrodeposition from AlCl3-NaCl melts on glassy carbon,platinum and gold electrodes

The electrochemical deposition and dissolution of aluminium on glassy carbon, platinum and gold electrodes in chloraluminate melts have been investigated using linear sweep voltammetry and potentiostatic pulse techniques. It was shown that deposition of aluminium on the glassy carbon electrode at low overpotentials takes place by 3-D progressive nucleation and growth, with the incorporation of atoms in the crystal lattice as the rate-determining step. At overpotentials higher than –100 mV vs Al, in the melts containing more than 52 mol % of AlCl₃, diffusion of Al₂Cl ₇ ^– , takes over the control of deposition of aluminium. Alloying of platinum and gold electrodes with aluminium from the melt occurs in the underpotential region.     

L'electrolyse de l'alumine dans la cryolithe fondue. Etude du mecanisme de decharge cathodique a l'aide de l'electrode a disque tournant

The use of a rotating disk platinum electrode, on which aluminium has been previously deposited, has allowed us to determine the mechanism of the cathodic reaction during the electrolysis of cryolite—alumina melts. This mechanism requires three stages within the Nernst's diffusion layer. The first of these is the dissociation of the ion AlF^3?₆, at δ′ from the electrode, forming the ions Al³⁺ and F^?. The second one is the discharge of ions Al³⁺ diffusing towards the cathode and the third one, at δ″ from the electrode, is the reaction of the ions F^? with the ions AlF^?₄ to give again the ions AlF^3?₆. The ions Na⁺, which transport most of the current serve to maintain electro-neutrality.     

Thermal stability of palladium on ceria-doped alumina

Several palladium on alumina and ceria/alumina catalysts were prepared and oxidized in air between 400 and 1000°C. The metal dispersion was determined by hydrogen titration of adsorbed oxygen. Dispersions above 50% were maintained on 0.2% Pd/Al₂O₃ up to 900°C. Adding 5.0% ceria, or increasing the metal loading to 2.5%, greatly reduces the thermal stability of the palladium, such that the dispersion falls rapidly at 600°C. The rates of methane oxidation (moles of CO₂/g Pd h) at 250°C and 5% excess oxygen are nearly equal on 0.22–2.50% Pd/3.5–5.2% CeO₂/Al₂O₃, dispersion 14–42%, and 0.20–0.46% Pd/Al₂O₃, dispersion 59–86%, but are 10 to 20 times lower than the rate on 2.3% Pd/Al₂O₃, dispersion 11%. The lower rate of methane oxidation on ceria-promoted and highly dispersed palladium on alumina might be due to the conversion of the palladium into less active palladium oxide during reaction.     

Exploration of molten hydroxide electrochemistry for thermal battery applications

The electrochemistry of molten LiOH–NaOH, LiOH–KOH, and NaOH–KOH was investigated using platinum, palladium, nickel, silver, aluminum and other electrodes. The fast kinetics of the Ag⁺/Ag electrode reaction suggests its use as a reference electrode in molten hydroxides. The key equilibrium reaction in each of these melts is 2 OH^– = H₂O + O^2– where H₂O is the Lux-Flood acid (oxide ion acceptor) and O^2– is the Lux–Flood base. This reaction dictates the minimum H₂O content attainable in the melt. Extensive heating at 500 °C simply converts more of the alkali metal hydroxide into the corresponding oxide, that is, Li₂O, Na₂O or K₂O. Thermodynamic calculations suggest that Li₂O acts as a Lux–Flood acid in molten NaOH–KOH via the dissolution reaction Li₂O_(s) + 2 OH^– = 2 LiO^– + H₂O whereas Na₂O acts as a Lux–Flood base, Na₂O_(s) = 2 Na⁺ + O^2–. The dominant limiting anodic reaction on platinum in all three melts is the oxidation of OH^– to yield oxygen, that is 2 OH^– 1/2 O₂ + H₂O + 2 e^–. The limiting cathodic reaction in these melts is the reduction of water in acidic melts ([H₂O] [O^2–]) and the reduction of Na⁺ or K⁺ in basic melts. The direct reduction of OH^– to hydrogen and O^2– is thermodynamically impossible in molten hydroxides. The electrostability window for thermal battery applications in molten hydroxides at 250–300 °C is 1.5 V in acidic melts and 2.5 V in basic melts. The use of aluminum substrates could possibly extend this window to 3 V or higher. Preliminary tests of the Li–Fe (LAN) anode in molten LiOH–KOH and NaOH–KOH show that this anode is not stable in these melts at acidic conditions. The presence of superoxide ions in these acidic melts likely contributes to this instability of lithium anodes. Thermal battery development using molten hydroxides will likely require less active anode materials such as Li–Al alloys or the use of more basic melts. It is well established that sodium metal is both soluble and stable in basic NaOH–KOH melts and has been used as a reference electrode for this system.     

Structural entities in NaF---AlF3 melts containing alumina

An ionic structure model is developed for NaFAlF₃ melts containing alumina. The model can be described by the following dissociation equilibria,

Na⁺ is proposed to be the only cation present in the system. The model is derived from activity data for NaF and AlF₃ in melts saturated with alumina at 1285 K and the assumption of an ideal ionic mixture, with randomly distributed anions on the anion positions in the melt. The available data for the system seem to fit the model reasonably well. The entity Al₂F^?₇ will only be present in AlF₃ rich melts. Oxygen atoms in the complexes are most probably involved in bridging bonds of the type AlOAl and

. It appears that aluminium atoms are involved in 4 bonds in AlF₃-rich melts. The number of bonds changes partly towards 5 and/or 6 with increasing NaF content of the melt.     

Interfacial tension between aluminum and cryolite melts during electrolysis of the systems Na3AlF6–AlF3 (NaF)–Al2O3

The interfacial tension between aluminum and cryolite melts containing different salt additions has been measured by the capillary depression method. The technique is based on the measurement of the capillary depression occurring when the capillary, which is moved vertically down through the molten salt layer, passes through the salt/metal interface. The depression is measured by simultaneous video recording of the immersion height of the alumina capillary. The interfacial tension was found to be strongly dependent on the n(NaF)/n(AlF₃) ratio (cryolite ratio, CR). At the cryolite ratio 2.28 (80 wt.% Na₃AlF₆ + 10 wt.% AlF₃ + 10 wt.% Al₂O₃ // Al, t = 1000 °C) the interfacial tension was 546 mN m⁻¹, while it was 450 mN m⁻¹ at the cryolite ratio 4.43 (80 wt.% Na₃AlF₆ + 10 wt.% NaF + 10 wt.% Al₂O₃ // Al, t = 1000 °C). Experiments under current flow conditions were also performed. During the electrolysis the interfacial tension at n(NaF)/n(AlF₃) ratio 2.28 decreased from 546 mN m⁻¹ at zero current to 518 mN m⁻¹ at 0.112 A cm⁻². The same trend was observed in the system with a cryolite ratio 4.43. The interfacial tension decreased from 450 mN m⁻¹ at zero current to 400 mN m⁻¹ at 0.112 A cm⁻². The consequent increase in interfacial tension of these systems caused by interruption of electrolysis was observed. Electrolysis of the system 25 wt.% NaF + 75 wt.% NaCl (eutectic mixture)/Al indicated no influence of applied current on the interfacial tension at 850 °C.     

The Viscosity of Molten Material in Pores and Capillary Channels of the Matrix Phase of High-Alumina Castables Employed in Thermal Power Units of Ferrous Metallurgy. 1. Viscosity Analysis of Melts in the CaO – Al2O3 –FeO – Fe2O3 – SiO2 System

A new approach to the viscosity analysis of high-alumina, low-calcium melts of the CaO – Al₂O₃ – FeO – Fe₂O₃ – SiO₂ system is used. A mathematical formalism for handling experimental viscosity data on the CaO – FeO – Fe₂O₃ – SiO₂ system is proposed that allows viscosity of the system to be evaluated within _av = 0.004 Pa · sec in the temperature range of 1500 – 1700°C at concentrations (mol.%) of CaO = 15.3 – 50.3, Fe_xO_y = 2.3 – 38.5, and SiO₂ = 17.7 – 57.7. An equation describing viscosity as a function of temperature is derived. Using this equation, the viscosity of model melts compositionally analogous to the products of interaction between molten slag and the glassy matrix phase of castables employed in thermal power units of ferrous metallurgy is considered.     

Anodic Dissolution of Metals in Oxide-Free Cryolite Melts

The anodic behavior of metals in molten cryolite-alumina melts has been investigated mostly for use as inert anodes for the Hall–Héroult process. In the present work, gold, platinum, palladium, copper, tungsten, nickel, cobalt and iron metal electrodes were anodically polarized in an oxide-free cryolite melt (11%wt. excess AlF₃ ; 5%wt. CaF₂) at 1273 K. The aim of the experiments was to characterize the oxidation reactions of the metals occurring without the effect of oxygen-containing dissolved species. The anodic dissolution of each metal was demonstrated, and electrochemical reactions were assigned using reversible potential calculation. The relative stability of metals as well as the possibility of generating pure fluorine is discussed.     

High strength alumina joints via transient liquid phase bonding

Low melting boron oxide, instead of metallic materials in other methods of transient liquid phase bonding, was taken as braze in joining alumina in this paper. Pure boron oxide melts at low temperature and reacts with alumina matrix to form a stable high melting compound. This transient liquid phase bonding has the advantage of producing a ceramic joint for high temperature applications at low processing temperature. In this study, alumina pieces coated with boron oxide layers in various thicknesses were bonded at 800 °C for various times in air under minor loading. The average flexural strength of joints were measured by means of four point bending, while the microstructure of the cross-section and fractured surface was observed by means of scanning electron microscopy. Phases at joints were identified by low angle X-ray diffraction. The maximum flexural strength reaches a value of 155 MPa after joining at 800 °C for 15 h with a 21 μm interlayer. Three compounds, 3Al₂O₃–B₃O₃, 2Al₂O₃–B₃O₃ and 9Al₂O₃–2B₃O₃ have been found at the joint. It is also found that 2Al₂O₃–B₃O₃ whiskers dominate at the joint with the maximum strength.     

Properties of RO-Y2O3-Al2O3-SiO2 glasses

The coefficient of linear expansion, glass-transition temperature, temperature at the orset of deformation (strain point), density. Young modulus, microhardness, crystallizability, and contact angle are studied as a function of the composition in RO–Y₂O₃–Al₂O₃–SiO₂ (R=Ca and/or Mg) glass systems. The composition ranges for glasses (with strain point >900°C and coefficient of linear expansion of (32–45)×10^–7°C^–1) that can be used for soldering silicon-nitride ceramics were established.Translated from Steklo i Keramika, No. 12, pp. 5–7, December, 1996     

Effect of the conditions of obtaining aluminum hydroxide and oxide on the properties of powders and their sintering

Conclusions A study of the effect of the conditions of obtaining aluminum hydroxide and alumina on the properties of the powders showed that fine milling and washing with water increase their dispersion and decreases the content of the alkali compounds. After fine milling and washing with water, the phase composition of the commercial-grade alumina undergoes a change: its a-Al₂O₃ content decreases due to the transformation into the hydrated forms of Al₂O₃.Fine milling and washing of the alumina obtained from anode slimes lead to a significant reduction of the content of the alkali compounds in it and significantly improve the degree of sintering at 1650°C at a specific surface less than that of the milled commercial-grade alumina, i.e., the sintering temperature is 100°C lower than that required for the milled alumina of KirAF.Translated from Ogneupory, No. 1, pp. 15–18, January, 1989.     

Synthesis of Glass-Ceramic Materials in the BaO–PbO–B2O3–Al2O3–TiO2 System

The influence of PbO, B₂O₃, and Al₂O₃ additives on the glass formation and crystallization of glasses with a high total content of BaO and TiO₂ (65–75 wt % or 76–86 mol %) is investigated. It is shown that glasses of the compositions (wt %) 31–35 BaO, 12–17 PbO, 34–42 TiO₂, 10–13 Al₂O₃, and 2–3 B₂O₃ are promising materials for use in preparing glass-ceramic ferroelectrics based on the melting–molding–crystallization technology. These compounds are characterized by a relatively low melting temperature (1450°C), the absence of spontaneous crystallization during molding, and the possibility of controlling the phase composition of the material through the appropriate choice of the crystallization temperature.     

Double-layer capacitance and polarization potential of baked carbon anodes in cryolite-alumina melts

Baked carbon anodes with varying apparent densities and baking temperatures were tested in Na₃AlF₆–Al₂O_3(sat) melts at 1010° C. The double-layer capacitance (C _dl) was used as an indicator of the wetted surface area. For unpolarized anodes,C _dl increased with increasing time of immersion and reached a constant level after 1.5–2h. The values decreased with increasing polarization potential in the range 1–1.5 V positive to aluminium. TheC _dl of polished samples increased markedly during electrolysis, particularly at low current densities. No clear correlation was found betweenC _dl and apparent density. Semi-logarithmic plots of potential versus current could be divided into three segments. The lower two were linear, the ranges and slopes being 0.01–0.1 A cm^–2, 0.20–0.44 V per decade and 0.1–0.5 A cm^–2, 0.18–0.24 V per decade, respectively. At higher current densities the curves bent upwards. The current density corresponding to an overpotential of 0.5 V increased slightly with increasing apparent density, whereas the ohmic voltage drop. at constant current density decreased. The current densities were corrected for differences in wetted surface area on the basis of theC _dl data. The change in baking temperature from 970 to 1100°C had no appreciable effect on the overpotential, whereas samples baked at 1250°C showed a somewhat lower overpotential.