Determination of the Lithium Content of Molten Aluminum Using a Solid Electrolyte

  • PDF / 1,452,356 Bytes
  • 6 Pages / 614.28 x 794.28 pts Page_size
  • 103 Downloads / 195 Views

DOWNLOAD

REPORT


I.

INTRODUCTION

L I T H I U M salts are added to the electrolyte of many HallHeroult cells in order to increase the conductivity of the electrolyte, thereby lowering the IR loss in the electrolyte.l According to Dewing and Gilbert 2 an average cryolite electrolyte containing 2.5 pct LiF will be in equilibrium with about 25 ppm lithium and 80 to 120 ppm sodium. After siphoning off the aluminum from the bottom of the cell, the alkali levels decrease to 10 to 20 ppm lithium and 25 to 50 ppm sodium by oxidation and evaporation. The greater decrease in sodium content is due to the much greater vapor pressure of sodium compared to lithium. During casting it is advisable to have the lithium content below 5 ppm, and many techniques have been investigated in order to achieve this figure. 3 The usual method of analysis is to cast a sample and send it to the laboratory for emission spectrography, but this can take several minutes before the result is known. Ideally, an on-line system of analysis would be desirable. In recent years, solid electrolyte probes have been used extensively in industry for the instantaneous determination of oxygen in steel 4 and copper 5 and, in the laboratory, for the measurement of the sodium 6 and hydrogen 7 contents of molten aluminum. The basis of the method is a simple electrochemical cell which can be represented by m (ref) Im + conductor {m (metal) where M(ref) is the reference material of known activity, M + conductor is an ionic conductor of M + ions, and M is the activity of M in the metallic solution. The potential across the cell is given by -ZEF

= RT In aMImetall aMIref)

11]

where Z is the charge carried, E is the potential measured by a high impedance voltmeter, F is Faraday's constant, R is the gas constant. T is the temperature, a~4~me,.l~ is the activity of M in the metallic solution, and aM~,ef~is the activity of M in the reference. The activity can be related to the atom fraction through the relationship aM = YMXU

[2]

R C. YAO, formerly Research Student in the Department of Metallurgy and Materials Science, University of Cambridge, is now with Chung Shah Institute of Science and Technology, Taiwan, Republic of China. D.J. FRAY is University Lecturer, Department of Metallurgy and Materials Science, University of Cambrtdge, Cambndge, United Kingdom. Manuscript submitted May 30, 1984. METALLURGICALTRANSACTIONS B

where XM is the atom fraction and YM is the activity coefficient. The success of a particular measurement depends entirely upon the selection of a suitable electrolyte and reference. There are several lithium ion conductors s which have been investigated in connection with lithium batteries, but as there is likely to be quantities of sodium in the aluminum, it is important to choose an electrolyte which is stable in aluminum and does not readily interchange with sodium. A comparison of sodium and lithium conductors shows that there are some silicates which have very much better lithium ion conduction than sodium ion conduction. It is, therefore, likely that on placing