Home Electrical Conductivity of Magnesium Oxide/Molten Carbonate Eutectic Coexisting System
Article Publicly Available

Electrical Conductivity of Magnesium Oxide/Molten Carbonate Eutectic Coexisting System

  • Elena V. Nikolaeva EMAIL logo , Andrey L. Bovet and Irina D. Zakiryanova
Published/Copyright: May 27, 2019

Abstract

The electrical conductivity of molten ternary alkali carbonate eutectic, coexisting with MgO particles, has been investigated. The conductivity was measured by the AC impedance method. The apparent activation energy ΔEa increased with the MgO content. This fact can be attributed to the effect of the solid phase. The specific conductivity of those systems could not be described using the Maxwell model over the solvation process of the carbonate ions on the particles of the magnesium oxide.

1 Introduction

The solid oxide/molten salt coexisting systems are widely used in electrochemical devices such as batteries, electrolysers, and capacitors [1]. Alkali carbonate melts and composites on their base have application as the electrolytes of molten carbonate fuel cells [2]. The interaction between solid and liquid phases has an effect on the physical properties of coexisting systems. A clarification of such anomalies is important from both fundamental and industrial points of view [1], [2], [3], [4], [5], [6].

There are many studies concerning the composition dependence and ionic transport phenomena of a coexisting system containing inorganic powder and molten carbonates. Particularly, the results of Mizuhata et al. [1], [3], [5], [6] revealed that the properties and structures of molten carbonate electrolyte at the interface of the solid oxide phase are different from those of the bulk.

In this article, we focused on the molten ternary alkali carbonate eutectic mixed with MgO powder. The electrical conductivity of MgO/molten Li2CO3 (43.5)–Na2CO3 (31.5)–K2CO3 (25) coexisting systems has been investigated. The aim is to clarify the effect of solid phase on the electroconductive properties of molten carbonate electrolyte. The thickened mixture at temperatures above the melting temperature of the carbonate eutectic is a suspension in which a dispersed phase (MgO powder) is distributed in a liquid dispersion medium (molten salt). If the dispersion medium is an electrolyte, the conductivity of the suspension is connected with the electrical conductivity of the pure liquid σL by the following relationship:

σ=σLF,

where F is the dimensionless coefficient of structural resistance, which depends only on the shape of the particles of the disperse phase and on its volume fraction in the suspension. Maxwell [7] derived an equation for the effective conductivity of a mixture consisting of spheres surrounded by a continuous medium with conductivity σL, which for the case of nonconducting inclusions can be written as

(1)σ/σL=13/(2/φ+1).

Here, φ is the volume fraction of the solid phase. The validity of the Maxwell formula was confirmed in a wide range of dispersed phase concentrations (up to 60 vol. %) [7] provided that the interaction with the dispersion medium had been absent. As we have seen earlier [8], the specific electrical conductivity of the α-Al2O3/(Li2CO3–Na2CO3–K2CO3)eut system can be adequately described by the Maxwell equation up to 30 vol. % of the α-Al2O3 content.

2 Experimental

2.1 Sample Preparation

Chemically pure-grade Li2CO3, Na2CO3, and K2CO3 (“Vekton,” St. Petersburg, Russian Federation) were preliminarily dehydrated under vacuum under stepwise increase of temperature and melted in argon atmosphere. The carbonate eutectic mixture Li2CO3 (43.5)–Na2CO3(31.5)–K2CO3(25) was prepared by fusion of the weighted portions of the prepared salts. The thoroughly mixed salt mixtures were dried under vacuum at 473 K for 1 h to remove traces of adsorbed water. After drying, the mixtures were heated up to 1073 K in an atmosphere of carbon dioxide. The melting point of the carbonate eutectic (670 K) determined by the DSC method (STA 449C Jupiter Thermal Analyzer; NETZSCH, Germany) agrees well with the literature data [9].

Analytically pure-grade MgO powder (“Chimreaktivsnab,” Ufa, Russian Federation) was used as the solid phase. The average crystallite size (coherent scattering region) of the MgO powder was estimated using the Scherrer formula [10] and half-widths of diffraction peaks obtained on a Shimadzu XRD-7000 diffractometer (Japan). Certified silicon powder was the standard that indicated the instrumental contribution to the peak width. Specific surface area of MgO powders was determined by gas absorption using a SORBI instrument (Meta, Russia). The morphology of MgO powder investigated by SEM (TESCAN MIRA 3 LMU; TESCAN, Czech Republic) is shown in Figure 1. The image reveals that the MgO powder consists of particles smaller than 100 nm. The average powder crystallite size was 95 nm. The specific surface area was 9.35 ± 0.07 m2/g. The lack of any sedimentation of finely dispersed powder in molten Li2CO3–Na2CO3–K2CO3 eutectic in a course of measurements was shown earlier [8].

Figure 1: The morphology of MgO powder.
Figure 1:

The morphology of MgO powder.

Samples were tested for possible impurities using infrared (IR) absorption spectroscopy on a Tensor 27 FT-IR spectrometer (Bruker, Germany). The IR spectrum of MgO (Fig. 2) showed a strong broadband in the range of 400–700 cm−1 that corresponded to Mg–O vibrations. Vibrational characteristic bands of adsorbed H2O or hydroxyls were not observed.

Figure 2: The IR spectrum of MgO.
Figure 2:

The IR spectrum of MgO.

The volume fraction of the solid phase, φ, was calculated as follows:

(2)φ=(ws/ds)/[(wl/dl)+(ws/ds)],

where wl, dl, ws, and ds are weight of liquid phase, density of liquid phase, weight of solid phase, and density of solid phase, respectively. The density of MgO and the density of the melt were calculated according to the equations given in [9], [11]. The average values of the MgO volume fraction for the temperature interval under study were used.

2.2 Electrical Conductivity Measuring Technique

AC impedance was measured under Ar atmosphere with the impedance meter Z-1500J (Elins, Russia) in a frequency range 20 Hz–1.5 MHz using platinum (Pt) electrodes. The melt resistance, which was defined via the impedance diagrams, was used to calculate the melt electrical conductivity. The cell layout and the measurement procedures were the same as those in our previous study [8]. The electrical conductivities of the electrolytes under study were calculated taking into consideration the temperature dependence of the cell constant, which values were (1.0–1.4) ⋅ 10−2 m−1 within the temperature range of 673–823 K.

Table 1:

Results of the electrical conductivity measurements and coefficients of (3) (confidence range 0.95).

MgO content, vol. %A ± ΔAB ± ΔB, KΔσ, S/cmσ, S/cm
673 K773 K
00.334 [8]0.682 [8]
54.32 ± 0.013692 ± 20.0010.3100.633
104.36 ± 0.023809 ± 130.0020.2730.567
204.03 ± 0.023713 ± 140.0020.2260.461
303.97 ± 0.014072 ± 70.0010.1240.272
403.67 ± 0.014880 ± 70.0010.0280.071
503.37 ± 0.065789 ± 480.0050.0050.016

The salt composition and compositions of frozen melts after high-temperature experiments were determined using XRD method. Certification of the samples for the presence of possible impurities was carried out by IR absorption spectroscopy.

3 Results and Discussion

3.1 Temperature and Composition Dependence of the Electrical Conductivity

The electrical conductivity values of the ternary carbonate eutectic melt obtained in our earlier work [8] agree well with the literature data [9] in the temperature range 673–800 K. The temperature dependences of the electrical conductivity (ln(s) vs. 1/T) of the MgO/molten (Li2CO3–Na2CO3–K2CO3)euc coexisting systems for various MgO content are shown in Figure 3. In all cases, the conductivity decreases with an increase of the solid content. Thus, the conductivity reduces at the order of magnitude when volume fraction of MgO is equal to 40 vol. %.

Figure 3: Temperature dependence of the electric conductivity for MgO/(Li2CO3–Na2CO3–K2CO3)eut systems with different MgO volume concentrations (ϕ).
Figure 3:

Temperature dependence of the electric conductivity for MgO/(Li2CO3–Na2CO3–K2CO3)eut systems with different MgO volume concentrations (ϕ).

3.2 Variation of Apparent Activation Energy ΔEa With MgO Content

The measured conductivity showed Arrhenius-type temperature dependence. The electrical conductivity as a function of temperature can be approximated by linear equations as follows:

(3)ln(σ)=A+B/T.
Figure 4: Variation of the activation energy of the electrical conductivity of the liquid phase with the MgO content.
Figure 4:

Variation of the activation energy of the electrical conductivity of the liquid phase with the MgO content.

Figure 5: Isotherm of specific electroconductivity of molten carbonate eutectic Li2CO3–Na2CO3–K2CO3, thickened with MgO, normalised by the specific conductivity of molten phase (T = 773 K).
Figure 5:

Isotherm of specific electroconductivity of molten carbonate eutectic Li2CO3–Na2CO3–K2CO3, thickened with MgO, normalised by the specific conductivity of molten phase (T = 773 K).

The coefficients of these equations for various solid contents are given in Table 1. Therefore, the apparent activation energy of the electrical conductivity ΔEa was calculated by the Arrhenius equation ΔEa = −BR. Here R is the universal gas constant. The variations of ΔEa with the MgO content are shown in Figure 4. The increase of ΔEa began at ca. 30 vol. % of solid phase content. These results indicate that the liquid phase is greatly perturbed by the solid phase in this region.

3.3 Implementation of Maxwell Model

Figure 5 demonstrates variations of the electrical conductivity with solid content for MgO/molten (Li2CO3–Na2CO3–K2CO3)euc coexisting system at 773 K. Beginning with the 20 vol. % of the MgO content, the experimental values of conductivity were less than those calculated according to (1). So the specific conductivity of those systems could not be described using the Maxwell model [7]. Consequently, the change in conductivity of the carbonate electrolyte when adding a finely dispersed MgO should be associated not only with the reduction of the charge carrier concentration but also with a certain kind of interaction between the components.

According to X-ray diffraction results for frozen fusions after the experiment, no chemical reactions in this system occurred, and the only phases of magnesium oxide and lithium, sodium, and potassium carbonates were found. The chemical inertness of MgO in (Li2CO3–Na2CO3–K2CO3)eut melt was demonstrated using in situ Raman spectroscopy [12].

Raman spectra [12] indicated that the symmetric stretching mode of carbonate is slightly perturbed by the solid phase. These peculiarities can be associated with the solvation process of the carbonate ions on the particles of the solid phase. Hence, the anions adsorbed on nanopowder-particle surfaces formed solvate shells that should considerably affect the physicochemical properties of the high-temperature disperse system. In particular, such kind of interaction between solid and liquid phases should have an appreciable effect on the electrical conductivity in the suspensions with relatively low oxide phase contents. The coagulation of the solid particles and their subsequent sedimentation were expected to be suppressed in the systems of similar type.

4 Conclusions

The electrical conductivity of molten Li2CO3–Na2CO3–K2CO3 eutectic, coexisting with nanodispersed MgO powder, has been investigated.

The temperature and the composition dependencies of the electrical conductivity have been analysed. The apparent activation energy ΔEa, increased beginning with the 30 vol. % of solid phase content. This fact indicates that the electrical conductivity in liquid phase is influenced by certain kind of interaction with the solid phase in this region. The impracticability of the Maxwell equation confirms this supposition. According to the available spectral data [12], these peculiarities can be associated with the solvation process of the carbonate ions on the particles of the solid phase.

Acknowledgement

We are grateful to V. N. Dokutovich for samples preparation. The granulometric and the XRD-analysis were performed using the facilities of the Shared Access Centre “Composition of Compounds” (Institute of High Temperature Electrochemistry, Ural Branch of RAS).

References

[1] M. Mizuhata, Y. Harada, G. Cha, A. Béléké, and S. Deki, J. Electrochem. Soc. 151, 179 (2004).10.1149/1.1688798Search in Google Scholar

[2] D. Shores, J. Selman, S. Israni, and E. Ong, in: Molten Carbonate Fuel Cell Technology (Eds. J. R. Selman, D. A. Shores, H. C. Maru, and I. Uchida), PV 90-16, The Electrochemical Society Proceedings Series, Pennington 1990, p. 290.Search in Google Scholar

[3] M. Mizuhata, A. Béléké, H. Watanabe, Y. Harada, and S. Deki, Electrochim. Acta 53, 71 (2007).10.1016/j.electacta.2007.06.020Search in Google Scholar

[4] M. Mizuhata, T. Ohashi, and A. Béléké, Int. J. Hydrogen Energy 37, 19407 (2012).10.1016/j.ijhydene.2011.09.109Search in Google Scholar

[5] H. Nafe, J. Solid State Sci. Technol. 3, 7 (2014).10.1149/2.0091407jssSearch in Google Scholar

[6] A. Béléke, M. Mizuhata, and S. Deki, Vib. Spectrosc. 40, 66 (2006).10.1016/j.vibspec.2005.07.002Search in Google Scholar

[7] D. Pletcher, Mathematical Methods for the Development of a Theory for Electrochemical Experiments. in: Instrumental Methods in Electrochemistry (Eds. D. Pletcher, R. Greef, and R. Peat et al.), Horwood Publishing Ltd., Coll House, Westergate, Chichester, England 2001, p. 388.10.1533/9781782420545Search in Google Scholar

[8] E. Nikolaeva, A. Bovet, and I. Zakiryanova, Z. Naturforsch. 73, 79 (2018).10.1515/zna-2017-0222Search in Google Scholar

[9] G. Janz and M. Lorenz, J. Chem. Eng. Data 6, 321 (1961).10.1021/je00103a001Search in Google Scholar

[10] R. Jenkins and R. Snyder, in: Introduction to X-ray Powder Diffractometry (Ed. Winefordner), John Wiley & Sons Inc., New York 1996, p. 89.10.1002/9781118520994Search in Google Scholar

[11] D. J. Duval, Mullite. in: Ceramic and Glass Materials. Structure, Properties and Processing (Eds. J.F. Shackelford and R.H. Doremus), Springer, New York 2008, p. 27.10.1007/978-0-387-73362-3_2Search in Google Scholar

[12] I. Zakir′yanova, J. Appl. Spectrosc. 85, 611 (2018).10.1007/s10812-018-0694-5Search in Google Scholar

Received: 2019-04-04
Accepted: 2019-04-26
Published Online: 2019-05-27
Published in Print: 2019-09-25

© 2019 Walter de Gruyter GmbH, Berlin/Boston

Downloaded on 27.9.2025 from https://www.degruyterbrill.com/document/doi/10.1515/zna-2019-0109/html?lang=en
Scroll to top button