INFLUENCE OF THE SULPHUR CONTENT IN THE CARBON ANODES IN ALUMINIUM ELECTROLYSIS - A LABORATORY STUDY Stanislaw Pietrzyk 1 and Jomar Thonstad 2 1 AGH-University of Science and Technology, 30 Mickiewicza Av., 30-059 Krakow (Poland) 2 NTNU, Norwegian University of Science and Technology, 7491 Trondheim (Norway) Keywords: aluminium electrolysis, carbon anodes, sulphur content, current efficiency, carbon consumption Abstract The chemistry of sulphur in carbon anodes is not fully understood, especially its influence on the electrolysis parameters. The results of this study are indicative of an important link between the sulphur content in the anode material and the carbon consumption as well as the current efficiency during aluminium electrolysis. By performing a laboratory scale investigation of different carbon anodes with sulphur contents ranging from 1.97 to 3.82 wt% S in addition to graphite anodes with sulphur content close to zero, it was found that increasing sulphur content contributes significantly to a decrease in the current efficiency and a rise in the carbon consumption. When going from 0 to 3.82 wt% S, the current efficiency decreased from 92 to 85% (1.8 % per 1 wt% S), and the carbon consumption rose from 108 to 128% (5.2 % per 1 wt% S). Introduction Aluminium is produced by electrolysis of alumina dissolved in cryolite-based melts. The electrolyte also contains some impurities, i.e. iron, silicon, phosphorus, sulphur, etc. [1]. The impurities are introduced into the electrolyte with the alumina or fluoride salts or they originate from the carbon anodes. In the Hall-Heroult process we know that sulphur originates as sulphur in the anode carbon plus some sulphur contained in the alumina and in the aluminium fluoride (1-5 wt.% S). Sulphur originates mainly from two sources. Petroleum coke used in the production of carbon anodes contains 0.7-3.5 wt.% sulphur (cokes with higher sulphur contents are usually blended with low- sulphur cokes). Cryolite and aluminium fluoride also contain sulphur, mainly as sulphate (up to 1 wt.%). The chemistry of sulphur in carbon anodes is not fully understood, especially its influence on the electrolysis parameters. Since the sulphur content in the crude oil used in the production of petroleum coke tends to increase with time, the effect of the sulphur content on the carbon consumption and the current efficiency was studied in the present work. The main parameters affecting the current efficiency have been known for a long time. The concepts of loss in current efficiency were developed more than 50 years ago. The primary electrochemical reaction producing aluminium is: Al 2 O 3 (dissolved) + 3/2 C(s, anode) = 2 Al(l) + 3/2 CO 2 (g) (1) and the main chemical back reaction causing loss of aluminium has traditionally been written as: 2 Al (dissolved) + 3 CO 2 (g) = Al 2 O 3 (dissolved) + 3 CO(g) (2) where Al(dissolved) is metal dissolved in the electrolyte [1]. In principle, the current efficiency can be determined from the weight of aluminium tapped from the cell, when knowing the quantity of electricity used. Reliable results for industrial cells can then be obtained for periods of several months, because the metal inventory in the cell is not known precisely. This method is also commonly used in laboratory cell experiments, where the weight increase of aluminium can be determined precisely after each short term experiment. With the assumption that CO 2 (g) is the only primary anode product and that the main back reaction causing a loss of aluminium is (equation 2) producing CO(g), the current efficiency (CE) may be calculated by the well known Pearson-Waddington equation: CE(%) = 100% - 0.5 [%CO(g)] = 50% + 0.5 [%CO 2 (g)] (3) This equation has traditionally been used to estimate current efficiency, with the use of various gas analysis techniques to determine the concentrations of CO 2 (g) and CO(g) in the anode gas. Thus, measurements of the CO 2 (g)/CO(g) ratios gives an instantaneous current efficiency determination. Prediction of current efficiency by this equation is usually believed to be accurate within a few per cent. Error limits are discussed in [1], and it is mainly due to difficulties in determining the exact extent of side reactions like: reaction between carbon, oxygen and CO 2 , electrolytic formation of CO, the back reduction of CO, oxidation of aluminium carbide, the effect of sulphurous gases etc. An alternative method of calculating the current efficiency, which is being used in the present work, is the oxygen balance method, based on a mass balance of the gaseous oxygen in the cell. This method (OxyB) calculates the current efficiency by comparing the total net oxygen production in the form of CO 2 and CO with the theoretical amount of oxygen that should be produced from the cell according to Faraday’s law. In a similar way a carbon mass balance can be made for the cell where the carbon consumption and the carbon dust formation can be determined independently of the current efficiency [2-5]. The theoretical carbon consumption (CC th ) per second can be calculated by Faraday's law: Light Metals 2012 Edited by: Carlos E. Suarez TMS (The Minerals, Metals & Materials Society), 2012 659