Abstract
The electrical conductivities of molten KCl-ZrCl4 and CsCl-ZrCl4 mixtures, including their heterogeneous (melt+crystals) ranges, were measured for the first time. The concentration ranges were 65–72 and 66–75 mol.% of ZrCl4, and the temperature ranges were 482–711 and 548–735 K, respectively. The measurements were carried out in cells of an original design.
1 Introduction
Electrical conductivity is a very important property that is necessary for the correct organisation of electrolytic processes. For the development of new technologies and improvement of the existing electrolytic zirconium extraction and refining processes, the data on the physico-chemical properties of ZrCl4 solutions with a low ZrCl4 vapour pressure over solutions [1], [2], [3], [4] are needed.
The first experimental data on the electrical conductivity of low-melting zirconium-containing melts were obtained by Howell and Kellogg [5]. They measured the electrical conductivity of NaCl-ZrCl4 and NaCl/KCl (1:1)-ZrCl4 molten systems.
Recently we have measured the electrical conductivity of low-melting NaCl-KCl-ZrCl4 mixtures in the range 54.3–75.2 mol.% ZrCl4, with the NaCl/KCl molar ratio of 8:29 (Teut.=491 K) [6].
In general, molten mixtures of zirconium tetrachloride with alkali metal chlorides are characterised by a high saturated vapour pressure even at temperatures close to the liquidus temperature [1], [2], [4], [7], [8]. However, there are two concentration “windows” applicable for technical purposes because the vapour pressure of these melts is below the atmospheric pressure [1], [2], [4], [6], [7], [8], [9], [10]. They are located around the eutectics of the corresponding salt systems. The high-temperature window is 0–30 mol.% of MCl4, and the low-temperature one is 60–75 mol.% of MCl4, see Figure 1.
![Figure 1: Phase diagrams of (a) KCl-ZrCl4 and (b) CsCl-ZrCl4 systems [9].](/document/doi/10.1515/zna-2017-0396/asset/graphic/j_zna-2017-0396_fig_001.jpg)
Phase diagrams of (a) KCl-ZrCl4 and (b) CsCl-ZrCl4 systems [9].
The aim of this work is to measure the electrical conductivity of molten KCl-ZrCl4 and CsCl-ZrCl4 mixtures in their low-temperature concentration regions, which are bounded by eutectic valleys.
2 Experimental
We used reagent grade KCl and CsCl, which were carefully dried and additionally purified by the three-fold zone recrystallisation with the boat speed of 2–3 cm/h. Commercial analytical grade zirconium tetrachloride was subjected to vacuum sublimation 2–3 times and then sublimation in a pure helium atmosphere. For the measurements, we used pre-prepared ZrCl4-KCl or ZrCl4-CsCl melts of the desired compositions. The salts, taken in the required proportions, were melted together in evacuated and sealed quartz ampules and were aged at 713–743 K for 5–8 h and then were cooled to room temperature [6]. The fusion cakes were additionally subjected to double or triple melting for better mixing of the components. All weighing as well as loading into ampoules and measuring cells were carried out in a nitrogen atmosphere (water content less than 3 ppm) in a dry box containing P2O5 as an additional desiccant.
To measure the electrical conductivities of KCl-ZrCl4 and CsCl-ZrCl4 melts, we used special hermetically sealed cells made of quartz with a vertical quartz measuring capillary and tungsten wire electrodes [6]. Both the melt and the small gas space over the melt were located within the isothermal zone in the furnace. And, because of the fact that the gas space was small and had the same temperature as the melt, changes in the composition of molten salts, which could result from the selective evaporation of ZrCl4, were almost completely excluded. Nevertheless, the salt composition was analysed before and after the experiments for the Zr, K, and Cs contents by atomic emission spectrometry using an Optima 4300DV ICP-OES (Perkin Elmer, USA) spectrometer. The maximum difference did not exceed 2 rel.%.
The charged cells were heated in an electrical resistance furnace supplied with a massive metallic block and a high-precision temperature controller VRT-3. The temperature was measured using a Pt/Pt-Rh thermocouple, whose junction was placed near the measuring capillary of the cell. The measurements accuracy was ±1 K. The electrical conductivity of the melt was measured by an AC bridge at the input frequency of 10 kHz. In previous work [11], it had been established that the melt resistance remained constant at frequencies above 8 kHz.
The cells designed by us were calibrated against molten KCl [12] and the standard 1-molal KCl aqueous solutions [13], and identical results were obtained. Therefore, subsequent calibrations were performed against reagent grade 1-molal KCl solutions in distilled water. The measured cell constants were high enough (63–94 cm−1) to allow us to perform electrical conductivity measurements with a sufficiently high accuracy. The total conductivity measurement error did not exceed 3%. The salts preparations, design of the cells, and measurement procedure are described in more detail elsewhere [6].
3 Results and Discussion
The electrical conductivity polytherms of KCl-ZrCl4 and CsCl-ZrCl4 mixtures are shown in Figures 2 and 3.

Specific conductivity (κ) polytherms of the molten and heterogenous (melt+solid) KCl-ZrCl4 mixtures. Vertical dashes denote the liquidus temperatures.

Specific conductivity (κ) polytherms of the molten and heterogenous (melt+solid) CsCl-ZrCl4 mixtures. Vertical dashes denote the liquidus temperatures.
The high-temperature parts of these curves correspond to the electrical conductivity of homogeneous melts. The electrical conductivity of these mixtures increases smoothly with temperature (Figs. 2 and 3). These curves look almost linear because of the rather narrow temperature ranges (ΔT=50–115 K). Here, the increase in electrical conductivity with temperature is mainly due to the increase in the mobility of the ions. The numerical data are approximated to an adequate accuracy (determination coefficient R2>0.999) by equations of the following type:
where κ is the specific conductivity (S/cm); T is the absolute temperature (K); a, b, c are empirical coefficients, and Δ is the root-mean-square deviation of the experimental points at the confidence level of 0.95. The coefficients of (1) are given in Table 1.
Electrical conductivity of molten KCl-ZrCl4 and CsCl-ZrCl4 mixtures at temperatures above the liquidus.
[ZrCl4], mol % | –a | b×103 | –c×107 | Δ | κ (673.15 K), S/cm | Temperature range, K |
---|---|---|---|---|---|---|
KCl-ZrCl4 | ||||||
65 | 0.69617 | 1.9464 | 6.8810 | 0.0002 | 0.302 | 661–711 |
69 | 0.64892 | 1.9917 | 8.7329 | 0.0001 | 0.296 | 601–711 |
72 | 0.58087 | 1.8245 | 8.0248 | 0.0001 | 0.284 | 646–707 |
CsCl-ZrCl4 | ||||||
66 | 0.32542 | 0.84773 | 1.7689 | 0.0001 | 0.165 | 628–735 |
70 | 0.34393 | 0.92179 | 2.7521 | 0.0001 | 0.152 | 607–722 |
75 | 0.38516 | 1.00091 | 3.3516 | 0.0001 | 0.137 | 621–723 |
Coefficients of the equation κ=a+bT+cT2±Δ.
The lower temperature parts of the polytherms shown in Figures 2 and 3 correspond to heterogeneous (melt+crystals) systems. Here, the composition of the electrolyte changes as the temperature decreases. The change in the slope of the polytherms denotes the beginning of the solid phase (K2ZrCl6, Cs2ZrCl6, or ZrCl4) [4], [8], [9] precipitation. The points at the lowest temperatures, where the electrical conductivity drops sharply, correspond to the intersection of the solidus line.
Using the data on the electrical conductivity polytherm breakpoints, we determined the approximate coordinates of the eutectic points. The data were obtained at the cooling rate of 0.5–1.0 K/min. The results are listed in Table 2, together with the available literature data. Literature data are only partly consistent both with our data and with each other. For example, according to [7], in the KCl-ZrCl4 system, Teut.=493 K. This is practically the same as that found by us, i.e. Teut.=492 K, but far from Teut.=508 K according to [4].
Coordinates of the eutectic point in the KCl-ZrCl4 and CsCl-ZrCl4 systems.
System | [ZrCl4]eut., mol.% | Teut., K | Method | Ref. |
---|---|---|---|---|
KCl-ZrCl4 | 65–69 | 492 | Electrical conductivity | Present study |
65.5 | 508 | Visual polythermal method | [4] | |
57.8 | 493 | Thermal analysis by cooling curves | [9] | |
CsCl-ZrCl4 | 66–70 | 550 | Electrical conductivity | Present study |
67.2 | 559 | Thermal analysis by cooling curves | [9] |
All data on the electrical conductivity of MCl-ZrCl4 (M=alkali metal) systems are compared in Figure 4. At comparable temperatures and ZrCl4 concentrations, the electrical conductivity decreases naturally in the series:

Specific conductivity (κ) polytherms of the molten and heterogenous (melt+solid) MCl-ZrCl4 mixtures (M=Na, K, Cs).
In this row, the average radius of salt-solvent cations increases, and their ionic potentials (charge-to-radius ratio) decrease. This means that the mobility of the alkali metal cations decreases, and the strength of the complexes formed by Zr4+ ions increases. This, in turn, leads to the decrease in the number of mobile Cl− anions. All these factors cause the decrease in the electrical conductivity in the row (2). According to Raman spectroscopy [14], [15], [16] studies on molten mixtures with a high concentration of zirconium tetrachloride (about 60–70 mol.%), the main complex groupings are predominantly face-sharing bi-octahedral ions Zr2Cl9–.
The electrical conductivity isotherms of homogeneous low-temperature KCl-ZrCl4 and CsCl-ZrCl4 molten mixtures are shown in Figure 5. In the investigated composition ranges, the electrical conductivity decreases almost linearly with ZrCl4 concentration. The same was noted in the NaCl-KCl (8:29 mol.) – ZrCl4 system [6]. These isotherms were approximated by (3):

Specific conductivity (κ) isotherms of the molten KCl-ZrCl4 and CsCl-ZrCl4 mixtures vs. the ZrCl4 concentration.
where κ is the specific conductivity (S/cm), [ZrCl4] is the concentration of ZrCl4 (mol.%), a and b are the empirical coefficients, and Δ is the root-mean-square deviation of the experimental points at the confidence level of 0.95. The coefficients of (3) are given in Table 3.
Concentration dependence of the electrical conductivities of molten KCl-ZrCl4 and CsCl-ZrCl4 mixtures.
T, K | a | –b×103 | Δ | [ZrCl4], mol.% |
---|---|---|---|---|
KCl-ZrCl4 | ||||
673 | 0.47182 | 2.5898 | 0.007 | 65–72 |
713 | 0.61689 | 4.2198 | 0.004 | |
CsCl-ZrCl4 | ||||
633 | 0.33304 | 2.9153 | 0.001 | 66–75 |
673 | 0.37231 | 3.1433 | 0.001 | |
713 | 0.41464 | 3.4269 | 0.003 |
Coefficients of the equation κ=a+b·[ZrCl4]±Δ.
In molten MCl-ZrCl4 mixtures at 33.3<[ZrCl4]<100 mol.% of ZrCl4, the equilibrium involves ionic [ZrCl62–, Zr2Cl102– (or ZrCl5–), Zr2Cl9–], molecular (ZrCl4 monomers), and (ZrCl4)n polymer-like species with n=2 or 6 [14], [15], [16], [17]:
As the temperature increases, the electrical conductivity increases as a result of the increase in the mobility of the ions. The increase in the concentration of zirconium tetrachloride shifts equilibrium (4) to the right, which leads to a decrease in the electrical conductivity due to the decrease in concentration of charged particles, both alkali metal cations and chloride anionic complexes of zirconium(IV) (Figs. 2–5).
Individual zirconium tetrachloride is a molecular melt consisting mainly of the ZrCl4 monomer and Zr2Cl8 dimer molecules [14], [17]. It exists in the form of a liquid within the narrow temperature range from Tm=710 K to Tcr=778 K and only at high vapour pressures from 22 to 58 atm. [1], [2], and has a very low electrical conductivity (~1×10−4 S/cm [18]).
The mixtures of MCl and ZrCl4 studied by us and in [5] are ionic liquids, and their electrical conductivities are approximately three orders of magnitude higher than that of ZrCl4. In this respect, MCl-ZrCl4 mixtures are suitable as electrolytes for the electrorefining or electrodeposition of metallic zirconium.
4 Conclusion
For the first time, the specific electric conductivities of molten KCl-ZrCl4 and CsCl-ZrCl4 mixtures with a high concentration (65–75 mol.%) of volatile ZrCl4 were measured in the region of the low-temperature eutectics of the corresponding salt systems.
Above the liquidus line, the electrical conductivity almost linearly increases with the rise in temperature and the lowering of the concentration of alkali metal chlorides. The electrical conductivity also increases in the row of salt solvents as CsCl>KCl>(NaCl-KCl)>NaCl. The revealed regularities were explained from the point of view of the complex formation in these melts.
References
[1] I. A. Sheka and K. F. Karlysheva, Chemistry of Hafnium, Naukova Dumka, Kiev 1972.Suche in Google Scholar
[2] N. V. Baryshnikov, V. E. Geger, N. D. Denisova, A. A. Kazain, V. A. Kozhemyakin, L. G. Nekhamkin, V. V. Rodyakin, Yu. A. Tsylov. Metallurgy of Zirconium and Hafnium, in: L. G. Nekhamkin (Ed.), Metallurgiya, Moscow 1979 (in Russian).Suche in Google Scholar
[3] D. V. Drobot, E. I. Lysakova, and A. M. Reznik, Selected Chapters of Chemistry and Technology of Rare and Trace Elements. Chemistry and Technology of Zirconium and Hafnium, MITHT, Moscow, 2013.Suche in Google Scholar
[4] L. J. Howell, R. C. Sommer, and H. H. Kellogg, J. Metals 209, 193 (1957).Suche in Google Scholar
[5] L. J. Howell and H. H. Kellogg, Trans. Metallurg. Soc. AIME 215, 143 (1959).Suche in Google Scholar
[6] A. B. Salyulev, V. A. Khokhlov, and A. A. Red’kin, Russ. Metallurgy (Metally) 2014, 659 (2014).10.1134/S0036029514080114Suche in Google Scholar
[7] J. E. Dutrizac and S. N. Flengas, in: Proceedings on Advances in Extractive Metallurgy Symposium, Institution of Mining and Metallurgy, London 1968, p. 572.Suche in Google Scholar
[8] I. S. Morozov, Application of Chlorine in Metallurgy of Rare and Non-Ferrous Metals, Nauka, Moscow 1966.Suche in Google Scholar
[9] I. S. Morozov and S. In-Chzhu, Zhurn. Neorg. Khim. (Russ. J. Inorg. Chem.) 4, 678 (1959) (in Russian).Suche in Google Scholar
[10] J. D. Kim and D. R. Spink, J. Chem. Eng. Data 19, 36 (1974).10.1021/je60060a012Suche in Google Scholar
[11] A. Salyulev, A. Potapov, V. Shishkin, and V. Khokhlov, Electrochim. Acta 182, 821 (2015).10.1016/j.electacta.2015.09.152Suche in Google Scholar
[12] G. J. Janz, J. Phys. Chem. Ref. Data 17, 1 (1988).10.1063/1.555819Suche in Google Scholar
[13] CRC Handbook of Chemistry and Physics, in: D. R. Lide (Ed.), A Ready-Reference Book of Chemical and Physical Data, CRS Press, London, New York, Washington 2004.Suche in Google Scholar
[14] G. M. Photiadis and G. N. Papatheodorou, J. Chem. Soc., Dalton Trans. 10, 981 (1998).10.1039/a707126dSuche in Google Scholar
[15] A. B. Salyulev, E. G. Vovkotrub, and V. N. Strekalowsky, Rasplavy (Melts) 3, 45 (2008) (in Russian).Suche in Google Scholar
[16] A. B. Salyulev, I. D. Zakiryanova, and E. G. Vovkotrub, Rasplavy (Melts) 5, 53 (2012) (in Russian).Suche in Google Scholar
[17] A. B. Salyulev and I. D. Kornyakova, Rasplavy (Melts) 2, 60 (1994) (in Russian).Suche in Google Scholar
[18] A. B. Salyulev and A. A. Red’kin, Rasplavy (Melts) 3, 20 (1996) (in Russian).Suche in Google Scholar
©2018 Walter de Gruyter GmbH, Berlin/Boston
Artikel in diesem Heft
- Frontmatter
- Ferromagnetic Convection in a Densely Packed Porous Medium with Magnetic-Field-Dependent Viscosity – Revisited
- Correlations, Nonlocality and Usefulness of an Efficient Class of Two-Qubit Mixed Entangled States
- Painlevé Test, Bäcklund Transformation and Consistent Riccati Expansion Solvability for two Generalised Cylindrical Korteweg-de Vries Equations with Variable Coefficients
- Dust Ion-Acoustic Shock Waves in a Multicomponent Magnetorotating Plasma
- Exact Solutions of the Nonlocal Nonlinear Schrödinger Equation with a Perturbation Term
- Numerical Investigation of a Two-Phase Nanofluid Model for Boundary Layer Flow Past a Variable Thickness Sheet
- A New Chaotic Flow with Hidden Attractor: The First Hyperjerk System with No Equilibrium
- Viewing the Cosmological Consequences of Modified Holographic Dark Energy in Various Interaction Scenarios
- Electrical Conductivities of Low-Temperature KCl-ZrCl4 and CsCl-ZrCl4 Molten Mixtures
- Flow and Heat Transfer Analysis of an Eyring–Powell Fluid in a Pipe
Artikel in diesem Heft
- Frontmatter
- Ferromagnetic Convection in a Densely Packed Porous Medium with Magnetic-Field-Dependent Viscosity – Revisited
- Correlations, Nonlocality and Usefulness of an Efficient Class of Two-Qubit Mixed Entangled States
- Painlevé Test, Bäcklund Transformation and Consistent Riccati Expansion Solvability for two Generalised Cylindrical Korteweg-de Vries Equations with Variable Coefficients
- Dust Ion-Acoustic Shock Waves in a Multicomponent Magnetorotating Plasma
- Exact Solutions of the Nonlocal Nonlinear Schrödinger Equation with a Perturbation Term
- Numerical Investigation of a Two-Phase Nanofluid Model for Boundary Layer Flow Past a Variable Thickness Sheet
- A New Chaotic Flow with Hidden Attractor: The First Hyperjerk System with No Equilibrium
- Viewing the Cosmological Consequences of Modified Holographic Dark Energy in Various Interaction Scenarios
- Electrical Conductivities of Low-Temperature KCl-ZrCl4 and CsCl-ZrCl4 Molten Mixtures
- Flow and Heat Transfer Analysis of an Eyring–Powell Fluid in a Pipe