Table of Contents
3. Properties of the Electrolyte
3.5 Liquidus Temperature
The liquidus temperature (TL) is the temperature at which solid material is precipitating when a liquid is cooled.
The bath temperature (TB) is the temperature of the liquid electrolyte in a electrolysis cell that is measured
during normal pot operation. The difference between bath temperature and liquidus temperature is called
The electrolyte is a multicomponent system of cryolite (Na3AlF6) with additions of aluminum fluoride (AlF3),
calcium fluoride (CaF2), lithium fluoride (LiF), magnesium fluoride (MgF2) and potassium fluoride (KF). In the literature
several relations of the cryolite liquidus temperature for this multicomponent system are found.
Remark:You should be aware that NOT all equations take care of all the possible components of the electrolytic bath.
Only the relations of Kolås and Solheim, for instance, use the potassium fluoride content. The lithium fluoride concentration is lacking in the relations
of Peterson and Lee and magnesium fluoride in the equations of Røstum, Peterson and Lee. If a concentration term is lacking the calculated
liquidus temperature is obviously not influenced by the corresponding concentration.
AlWeb does NOT check the limits of the input values as demanded by some authors. Therefore the calcution results may be misleading.
Alweb and AlPrg apply the following relations:
The expression of Dewing uses the concentration of lithium cryolite (Li3AlF6). Lithium cryolite
is formed according to the following reaction equation from lithium fluoride and aluminum fluoride:
3LiF + AlF3 = Li3AlF6
To use Dewing's equation the concentration of lithium cryolite must be calculated if the lithium fluoride concentration
is given and the input aluminum fluoride concentration must correspondingly be corrected.
If one calculates the liquidus temperature of the electrolyte using its chemical analysis and one of the shown relations, one finds that some cells operate
successfully for extended periods (weeks) at temperatures below the calculated liquidus.
Haupin [Lit.] has analyzed this so called Liquidus Enigma.
He offers the following explanations:
(1) impurities not determined in the industrial bath analysis lower the true liquidus,
(2) an analytical bias of bath composition may exist,
(3) a bias in how bath temperature is measured may exist,
(4) bath may remain supersaturated with alumina for extended period of time,
(5) industrial cells may be operating with suspended precipitate.
Haupin considers explanation 1, 3 and 5 as most likely.
Tarcy et al. [Lit.]
have analyzed the analytical methods to determine the bath composition especially concerning the aluminum
oxide concentration. They have shown that the extraction method which is normally used to determine the alumina content of bath
samples delivers systematically too low values. The observation of cells operating below the calculated liquidus temperature is,
to a large extend, accounted for a systematic alumina measurement error. Some cases of liquidus enigma may also be due to impurities
or measurement errors especially of the bath temperature.
Recently Moxnes et al. [Lit.]
looked again at the problem of the liquidus enigma. The superheat calculated from measured bath temperatures and
bath sample analyses using Solheim's equation often show negative values, although
measurements made with the superheat-sensor from
Heraeus Electro-Nite (Lit.
always show positive values. Simultaneously, bath samples were taken for aliminum carbide (Al4C3) analysis,
XRF (x-ray fluorescence), XRD (x-ray diffraction) and ICP (inductively coupled plasma) spectrometer analysis
(see Teledyne Instruments, for instance), as well as alumina (Al2O3) analysis by
LECO and thermal analysis. It turned out that Al4C3 content in the bath was very low. By comparing all the measured
temperatures, the Heraeus probe seemed to give a somewhat high superheat. The key factor, however, appeared to be the precision of the bath analysis. Furthermore, it is possible that some
of the trace elements in the bath will reduce the liquidus temperature more in an acidic bath than in pure cryolite.
next page |