Reference Electrode for electrochemical investigations in cryolite-alumina melts
at 700-960 ºС
A. Suzdal’tsev, А. Khramov, Yu. Zaikov
Institute of High Temperature Electrochemistry,Ekaterinburg, Russia
The different electrode constructions were tested in cryolite-alumina at temperatures 700-960 ºС. A new aluminum reference electrode was proposed. It consists of porous alumina чехла and aluminum and it does not need other metals and inert atmosphere. The aluminum activity at this electrode is constant and equal 1. The electrode proposed was tested at cryolite-alumina melts at temperature range 700-960 ºС and its potential was found to be stable and reproducible during long-term experiments. At low temperature (700 ÷ 800 ºС) carbon reference electrode can be also used. The disadvantage of this electrode is non-stability of potential which increases with temperature over 750 ºС.
INTRODUCTION
At study of electrochemical processes an electrode potential is measured against reference electrode the value of which is supposed to be equal zero. The main requirements for reference electrode are stability, reproducibility and reversibility of its electrochemical potential. The last requirement means that e.m.f. of the element containing reversible reference electrode must correspond to the Gibbs energy change of current forming reaction. Carbon, aluminum and oxygen reference electrodes are the most used at study of cryolite-alumina melts. Piontelli (1) proposed the Al reference electrode construction shown in fig.1b. At our previous experiments we also used the Al reference electrode similar to that of Piontelli. Our electrode (see fig.1a) consisted of alundum case with liquid aluminum and cryolite on its bottom. Molybdenum or tungsten rod (2-4 mm diameter) was put into liquid aluminum. As distinct from the 1b construction in the 1a one the inert argon atmosphere was kept and there was no hole. In melts non saturated on alumina the alundum case was protected by carbon case in order to prevent its dissolution. The other authors (2, 3) put in doubt the quality of Piontelli’s electrode they supposed that the potential of such electrode could not be stable because of interaction tungsten with cryolite-alumina melt. They proposed (3) the electrode constructions where molybdenum or tungsten did not react with cryolite (fig. 1c, 1d). Thonstad (6) obtained the dependence of the carbon electrode potential on ratio [CO2]/[CO2+CO]in the reference electrode case. Experimental data obtained were in a good agreement with thermodynamic values calculated for summary reaction:
[1]
where х = – partial pressure of CO2, (1 – х) = – partial pressure of CO.
Our experience at work with open cells containing the carbon electrode and the aluminum reference electrode described above (fig.1a) gives the conclusion that the potential difference between these electrodes is close to thermodynamic value of e.m.f., , in galvanic element (7):
Al│ Cryolite + Al2O3│C, (CO2, CO)[2]
1) 2 Al + 3/2 CO2 = Al2O3 + 3/2 С (n = 6)[3]
2) 2 Al + 3 CO = Al2O3 + 3 С (n = 6)[4]
, V / , V1) / 1,342 / 1,192
2) / 1,342 / 1,075
Here e.m.f., , – the potential of carbon electrode against the Al reference electrode. At 960ºС in potassium cryolite with CR=2,85 (CR=[NaF+KF+LiF]/[AlF3])the experimental value of the standard potential for carbon electrode in open atmosphere was 1,12 ÷ 1,22 V, and at 700 ºС in potassium cryolite with CR=1,3 was in the rang of 1,29 ÷ 1,40 V. But the absence of e.m.f. reproducibility at 700 ºС and low reproducibility at 960 ºС put under suspicion of both the Al and carbon electrodes work. Thus there is a necessity in development of the new reference electrode for work in cryolite melts at low temperatures.
а b c d
Fig. 1.The schemes of different constructions of the Al reference electrode:а – this work; b – (1); c – (2); d – (3).
EXPERIMENTAL
Experiments were carried out at different temperatures:
K-cryolite KF-AlF3-Al2O3with CR = 1,3 at T = 700 and 750 ºС;
Na- cryolite NaF-AlF3-Al2O3with CR = 2,6 at T = 960 ºС.
Besides, Na-cryolite was investigated with some additions (see below).
The experimental cells are presented in fig.2.
The aluminum electrode (fig. 2a) was in form of alundum case-diaphragm with aluminum inside. Aluminum was in liquid state at lower part of the case and in solid state at upper part. The carbon electrode was rode made of “spectral pure” carbon (fig. 2b). The electrode was put into the melt on the bottom of alundum case- diaphragm with isolated inner part. СО2flow was kept during experiment. The alundum case was protected from dissolution in melts non saturated on alumina by thin (2 mm thickness) carbon case.
a b c
Fig. 2. The schemes of cells:
1 – aluminum / 11 – the rod made of “spectral pure” carbon2, 3, 6, 9, 15 – alumina
4 – graphite crucible / 12 – putty on the base of
Al2O3 and Na2O·SiO2,
5 – nichrome
7 – graphite case / 13 – melt
8 – vacuum rubber / 14 – graphite crumb
10 – carbon tube,
RESULTS
Carbon reference electrode
Carbon electrode can be used as reference electrode for measurements in cryolite-alumina melts. Its positive parameter in comparison with other electrodes is very fast forming the potential which ispractically equal to the value of anodic overvoltage. But the potential depends on the rate of the СО2 flow in internal electrode space that can be the disadvantage.
Let’s calculate the potentials of aluminum and carbon electrodes for reaction [1] . The thermodynamic value of the equilibrium potential of the carbon electrode against the aluminum electrode in dependence on the gas phase composition and alumina activity at constant temperature and aA1 = 1 can be written in the following way:
Here the standard change of for reaction [1] is expressed through the stantard changes of and for reactions [3] and [4] respectively.Taking into consedaration that
= х, = (1 – x), [5]
one can write
[6]
, n = 6,[7]
where – the equilibrium potential at corresponding T and x. Calculations were made for different temperatures and corresponding isotherms were expressed. Besides, the polytherm– was calculated. This polytherm shows relationship between the CO2 content in equilibrium mixture of gases according to [8] and the carbon electrode equilibrium potential
CO2 + C = 2 CO, ,[8]
where , , – the equlibrium constant and the equilibrium partial presseres of CO2 andCO for reaction [8]. Taking into consederation [5] it is possible to solve corresponding equaion regarding .
21/2,[9]
,[10]
The crossing points of polytherms with isotherms gives the values of equilibrium partial pressure of CO2 according to reaction [8] at corresponding temperatures.
At the CO2 insufflation through theclosed space of carbon electrode the gas phase composition can change from pure CO2 to its equilibrium composition according to reaction [8]. At that the electrode potential should be formed according to correspondent isotherm on the section between and =1. Calculated values of e.m.f. [2] are in the region ΔEshown in table1.
Тable 1. Calculated values of e.m.f in circuit [2].
T, ºС / E (V), in atmosphere of / ΔE,V
CO2 / equilibrium [8]
= 1
([Al2O3] ≈ 8 mas.% (10)) / 700 / 1,342 / 1,322 / 0,020
750 / 1,313 / 1,280 / 0,033
960 / 1,192 / 1,075 / 0,117 / [11]
= 0,05
([Al2O3] ≈ 2 mas.% (10)) / 700 / 1,384 / 1,364 / 0,020
750 / 1,357 / 1,324 / 0,033
960 / 1,245 / 1,128 / 0,117 / [12]
The values of ΔE manifest the stability increase with temperature decrease of the carbon electrode with non-controlled atmosphere. In theory, implementation of controlled atmosphere (CO, CO2) increases stability and reproducibility but in this case the gas flow rate effects greatly on the stability and the best way is to flow the equilibrium gas mixture at given temperature. This conclusion was confirmed experimentally. The experiments with sodium cryolite were carried out in a cell with carbon electrode 11 (fig.2a) with CO2 flow in closed space at 960 ºС
Al ║Na-cryolite + Al2O3(sat.)║ Na-cryolite + Al2O3(sat.)│C, (CO2)[13]
These experiments demonstrated the absence of the potential stability and reproducibility in spite of its values was in the region [11]. The values of E together with thermo-e.m.f are presented in fig.3. (The thermo-e.m.f measurements gave 1,2-1,7 mV). It should be noticed that the destruction of carbon parts of the cell was occurred relatively fast. Possibly the process of potential establishment was very fast and carbon interacts with carbon oxide since at T=960°C,= 0,0119 Bar.
Fig. 3. Potential of closed carbon electrode under CO2 flow (including thermo-e.m.f.) against the Al reference electrode, galvanic cell (13), cell (fig.2a).
Electrolyte composition: (mas.%):
1 –(50,12)NaF -(43,58)AlF3 -(6)CaF2 -(0,3)MgF2 + (6)Al2O3,
CR([NaF] / [AlF3]) = 2,3 ; T = 960 ºС.
2 – (51,21)NaF -(2)KF -(2,1)LiF -(39,4)AlF3 -(5)CaF2 -(0,3)MgF2 + (6)Al2O3,
CR([NaF+KF+LiF]/[AlF3]) = 2,85 ; T = 950 ºС.
Nichrome rod 5 (fig.2а) is served as current lead to cathode at aluminum electrolysis. The cell (fig.2b) differs from cell (fig.2а) by the presence of graphite diaphragm 7 above the aluminum diaphragm in order to preset the Al2O3 concentration lower than saturated one
Al ║Na-cryolite + Al2O3(sat.)║ C, (CO2, CO)│Ni(Cr).[14]
Though the electrode works in open cell it is possible that the oxygen pressure is negligible due to the carbon oxidation. The melt film between alumina case 2 of the Al-electrode and graphite diaphragm 7 plays role of electrolyte. (fig.2b) and the inner wall of the graphite diaphragm faced to the Al-electrode plays role of carbon electrode. It is due to the graphite diaphragm is placed on the bottom of graphite container 4 and it is in contact with container. The container itself is in contact with nichrome rod 5. At such construction the Al2O3 concentration in the melt 13 (fig.2b) is not important for e.m.f. of circuit [14] because it is determined by the Al2O3 concentration in the melt film which corresponds to saturation.
The potential between the Al-electrode and nichrome rod (fig.2b) in some experiments without current at 960 ºС and Al2O3 concentration 3-6 mas.% was 1,20–1,07 V including thermo-e.m.f. The comparison of these values with theoretical ones (11, 12) for 960 ºС in Al2O3 saturated and non-saturated melts let us conclude that nichrome rode in fact was a contact element to take readings from carbon electrode being in saturated melt. The carbon electrode potential change together with thermo-e.m.f. for circuit [14] (fig.2b) is shown in fig.4. The absence of the potential stability and reproducibility of the open carbon electrode (with nichrome current lead) at high temperature (960 ºС) can be observed.
Fig. 4.Potential of the open carbon electrode with N/Cr current lead (including thermo-e.m.f.) against the Al reference electrode, galvanic cell [14], cell (fig.2b),
T = 960 ºС
Electrolyte composition: (mas.%): (51,21)NaF-(2)KF-(2,1)LiF-(39,4)AlF3-(5)CaF2-(0,3)MgF2; CR ([NaF+KF+LiF]/[AlF3])= 2,85 ;
Content of Al2O3 in the melt 13 (fig.2 b), mas.%: 1 – 2, 2 – 4, 3 – 5
The stability of the carbon electrode potential was tested at low temperature in potassium cryolite in cell (fig.2a):
(CO2), C│K-cryolite+Al2O3(sat.)║K-cryolite + Al2O3(sat)│C, (CO2, CO)│Ni(Cr). [15]
During 10 hours the value of carbon electrode potential in open cell with N/Cr current lead against the carbon electrode under CO2 flow at T= 700 – 750 ºС did not exceed (+5)±5 mV (including thermo-e.m.f. between carbon and nichrome ≈5 mV, nichrome is positive). So the potentials of both electrodes can be considered as equal. It manifests the stability of carbon electrode at low temperature. There were no burning traces of closed electrode parts. Thus it may conclude that the potential corresponding to the atmosphere of pure CO2is formedboth as on the closed carbon electrode under CO2 flow as on the open one at low temperature.
The results of experiments carried out under CO2 flow (VCO2 = 1-2 ml/min, the inner space volume of carbon electrode is 5-6 cm3) and without CO2 flow at different temperatures are shown in fig.5. The intervals near the points are presented according to maximal and minimal values obtained at registration (25-1500 min); the point inside the interval is average). The calculated values of e.m.f. for equilibrium mixture CO2–CO [8] and pure CO2 are also presented in fig.5. It is seen that at high temperature under the CO2 flow the potential of carbon electrode is significantly shifted from equilibrium, it stability sharply decreases. The same was observed in experiments with opened carbon electrode. It was pointed out in paper (6) that at 1000 ºС the electrode potential C/CO2 did not depend on CO2consumption whenthe flow rate was higher than 30 ml/min. Probably that at such gas flow rate the composition of the gas phase has not reached the equilibrium composition according to reaction [8]. All experimental values of e.m.f. between carbon and Al-electrode are close to theoretical magnitudes. The most high dispersion of e.m.f. values was observed in sodium cryolite at 950-960 ºС (fig.5). In all other cases the e.m.f. fluctuation between carbon (under the СО2 flow and without) and the Al-electrode near the average value at certain temperature was not exceeded 5÷7 ± 5÷7 mV.
Fig. 5.Potential of the carbon electrode (under the СО2 flow and without) against the Al-electrode at 660 – 960 ºС.
Electrolyte: mas.%:
1 – (50,1)NaF -(43,6)AlF3 -(6)CaF2 -(0,3)MgF2 + (6)(initial) Al2O3,
CR = 2,3; T = 960 ºС.
2 – (51,2)NaF -(2)KF -(2,1)LiF -(39,4)AlF3 -(5)CaF2 -(0,3)MgF2 + (6)(initial) Al2O3, CR ([NaF+KF+LiF]/[AlF3]) = 2,85; T = 950 ºС.
3 – (46,7)KF -(48,3)AlF3 + (5)(initial) Al2O3
CR= 1,4; T = 660, 725, 766, 800, 837, 870, 910 ºС.
4 – (47,3)KF -(52,7)AlF3 + (5)(initial) Al2O3 ; CR = 1,3; T = 700, 750 ºС.
Al-reference electrode
The Al-reference electrode in some experiments on electrolysis in cryolite-alumina melts (fig.1a) was used earlier. The Mo or W rod was used as a contact element to take readings from electrode. The chemical analysis showed the presence of these metals up to 10 at.% in the melt of the reference electrode after long use. It is due to formation of intermetallides (11). Thus the quality of work of the earlier offered reference electrode designs raises the doubts (fig.1) (2, 3).It is more correctly to use such design of the Al-reference electrode in which as the contact element to take readings from electrode is using aluminum itself (fig.2).
As was expected, there was no stability and reproducibility of the Al-reference electrode in experiments with the tungsten rod in the alumina saturated cryolite-alumina melts at 960 ºС. The potential between two Al-electrodes with different contact rods (Al and W) in galvanic cell
Al ║Na-cryolite + Al2O3 (sat.)║ Na--cryolite + Al2O3 (sat )│Al │W, (Ar).[16]
can change greatly: e.m.f. of circuit [16] rises with time permanently, at that the electrode potential with tungsten contact rode is more positive. So, in one experiment e.m.f. increased on approximately 35 mV for 2 hours at continual contact of W and Al. The e.m.f. increasing rate in element [16] decreases in that a case, if the W-contact rod was immersed in aluminum only for short time (20 sec) to take the potential reading. Theoretically the potential must be zero for different contact rods but in this case it was 40 mV, and at other cases it was varied from 15 to 70 mV (the thermo-e.m.f. between Al and W is less 1 mV.
At usage of Al-electrode with Al-contact the impregnation of alundum case (diaphragm) of 2,5 mm thickness by sodium cryolite is taken place during 0,5-1,5 hours at 960 ºС. It potential almost doesn’t change at repeated use. The potential between two Al-electrodes after being using for long time (one electrode was used in 4 thermal cycle with total time more 40 hours, the second – in 2 thermal cycle with total time 10 hours) has changed on 20 mV for 14 hours: from –10 to +10. Thus it is obvious the advantage of this electrode at working with Na-cryolite at high temperature (960 ºС) in comparison with offered earlier.
There is a necessity to defend alumina case from dissolution at work with non-saturated alumina melts. Additional graphite case 7 was used in experiments with such melts (fig.2, cell c). It resulted in potential increase of two Al-electrodes up to –(25 ÷ 28) mV (the potential of electrode with graphite case is more negative).
Thus one can conclude that the Al-electrode can be realized in a form of only alumina case-diaphragm in such case when diffusion of the small amount of dissolved aluminum to working zone is allowed. Otherwise it is necessary to use two alumina cases. The usage of the graphite diaphragm is possible only above the second alumina case.
The experiment in potassium cryolite at low temperature was carried out. The Al-electrode potential against the carbon electrode (fig.2, cell a)
C, (CO2, CO)│K-cryolite + Al2O3 sat║Al[17]
in dependence on time of their location in the melt is shown in fig.6. The e.m.f. became constant after 2-3 hours. It can be explained by slow diffusion of melt through the alumina diaphragm pores. The value of e.m.f. was 1,303 0,003 V and it has not changed for long time. This value is close to theoretical one calculated for reaction [3]: 1,307 V for temperature 760 ºС. At the course of experiment the electrolyte was added into the cell. After that the electrodes were lifted above the melt and the cell was cooled. Then the same experiment with the same electrodes was repeated at second thermo cycle (interval 11-20 hours). As it is seen from fig.6 the Al-electrode potential was stable and reproducible.
.
Fig. 6.Potential of the Al-electrode (with the Ni(Cr)-contact) against the carbon electrode in potassium cryolite-alumina melt (including thermo-e.m.f.). Cell (fig.2a), circuit [17]. Electrolyte – alumina saturated. T = 760 ºС.
It is possible to conclude that the aluminum reference electrode after preliminary being in the melt for 2-3 hours can be used in potassium cryolite-alumina melts for long term measurements at low temperature.
CONCLUSION
The proposed aluminum reference electrode can be used for electrochemical measurements in cryolite–alumina melts at high temperatures (760-960 ºС). The porosity of alumina case is of a great importance for the forming stable values of the electrode potential.
Along with the aluminum electrode as reference electrode the carbon electrode with the CO2flow can be used for electrochemical measurements in cryolite –alumina melts at temperatures 700-800 ºС. At that it is necessary to know the values of aluminaactivity in the melt and the rate of CO2 flow to the electrode.
REFERENCES
1.Alabyshev А., Lantratov M., Morachevsky A. Reference electrodes for molten salts. Moscow, Metallurgia, 1965
2.Burgman J.W., Leistra J.A., Sides F.J. Aluminium. J. Electrochemical Society, 1986, vol.133 [3], p. 496-502.
3.D.R.Sadoway. Aluminum reference electrode, US Patent 4764257, Aug. 16, 1988, Assignee: Massachusetts Institute of Technology, Cambridge, Mass.
4.Rempel S. Anodic process at electrolytic production of aluminum. – Moscow, Metallurgia, 1961
5.Mashovets V., Revazyan A. Russian Journal of physical Chemistry, V.30, 1006, 1957
6.J.Thonstad, E.Hove.Can. J. Of Chemistry, 1964, 42, p. 1542.
7.Thermodynamic constants of pure substances. Handbook. Edited by Glushko ,V. 1-4, Moscow, Nauka, 1978-1982.
8. Kryukovsky et al. In: Proceedings of 12th International Conference “Aluminum of Siberia” 2006, Krasnoyarsk, pp. 46-49
9.A. V. Frolov, A. O. Gusev, Yu.P. Zaikov, A. P. Khramov, N. I. Shurov, O.Yu. Tkacheva, A. P. Apisarov, V. A. Kovrov. Light metals, 2007, p. 571-576.
10.Thonstad J., Fellner P., Haarberg G.M., Híveš J., Kvande H., Sterten Å. Aluminium Electrolysis. Fundamentals of the Hall-Héroult Process. 3rd edition. – Düsseldorf: by Aluminium-Verlag Marketing & Kommunikation GmbH, 2001. 354 р.
11.Phase equilibria in binary metallic systems. Handbook, edited by N.P. Lyakishev, V.1, Moscow, Mashinostroenie, 1996
3-1