Study of a Graphite Sensor for Alumina Concentration Measurements in Cryolite Melts
MetadataShow full item record
Alumina dissolution is one of the most crucial processes taking place in the Hall-Héroult process for the production of primary aluminium. One of the key issues for advanced aluminium electrolysis is therefore to understand and improve the dissolution process. In order to understand it, it is important to have a method for which the alumina concentration can be measured. The literature describes several methods that have been tested. Electromotive force, emf, measurements is one of the methods. In this thesis electromotive force measurements between a graphite sensor and two difference reference electrodes (alumina reference electrode and graphite reference electrode) have been studied in order to figure out if graphite material is suited for determination of alumina concentration in cryolite melts. The results showed a decreasing trend in emf values when alumina was added as expected from theoretical calculation. An important observation from the initial experiments was that it was observed sudden drops and jumps in emf immediately after addition. These drops and jumps are probably due to other changes in the system and a large part of the following work was dedicated to reveal the reasons for this behaviour. Immediate changes in emf in the moment alumina was added may be due to air which flowed together with the powder down to the melt and caused an immediate change of the potential on the graphite sensor due to the formation of CO and CO2. Introduction of CO2 gas to the system gave substantial impact on the emf. So the assumption of a constant partial pressure of CO/CO2 used in the theoretical calculations of potential might not be valid, and a solid buffer system of CO/CO2 may be needed to reduce the influence of variation in the partial pressures of CO and CO2. However, the sudden drops and jumps in the potential after alumina addition were still present even though CO was fed to the system. It seemed like gas evolution caused by reaction between moisture or hydoxyl containing species in the primary alumina in conjunction with the addition of primary alumina affected the emf immediately after addition, but the effect was reduced by the use of fully calcined alumina and a totally immersed probe. Drifting in emf was observed in all experiments and might have several explanations. The reaction between dissolved metal, possibly stemming from the aluminium reference electrode, and CO and CO2 is believed to be one of the main reasons for the drifting. To get the aluminium leakage from the aluminium reference electrode under control and to develop a solid buffer system of CO and CO2 seem important. The use of graphite sensors for determination of alumina concentration in cryolite melts is not appropriate as long as other variations in the system influence the measured potentials. To get control of the other variables in the system will be important in the further development of this method of measuring the alumina concentration.