Solubility in the ternary system MgCl2-FeCl2-H2O at 288 K by conductance method

Experimental studies on the solubility of the ternary system MgCl2-FeCl2-H2O at 288 K were determined using a synthetic method based on conductivity measurements. The corresponding solid-liquid phase diagram has been plotted. As can be concluded in the ternary system at 288 K, there are two eutectic points, three crystallization fields corresponding to MgCl2.6H2O, FeCl2.4H2O and solid solution Fe(1-x)MgxCl2.10H2O) with 0 < x ≤ 0.38 were found. However, no double salt has been formed at the studied temperature.


Introduction
The study of solubility equilibria between solid salts, salt hydrates and water provides important information for scientific and industrial applications [1][2][3] . Furthermore, phase diagrams permit us to study and control important processes such as phase separation, crystallization, solidification, purification... For these reasons, the study of the solidliquid equilibrium of the ternary system (MgCl2 -FeCl2 -H2O) is a promising route for the knowledge of the solubility behavior of magnesium and iron chlorides.
The present study represents a continuation of our previous work focusing on the establishment of the phase diagrams of the ternary systems M, Fe / Cl -H2O (M = K, Na) 4-6 . Our purpose is to examine deeply the ternary system MgCl2-FeCl2-H2O behavior at 288 K.
A large amount of experimental data related to the phase diagrams of the binary systems MgCl2-H2O and FeCl2-H2O 7 are available. In our latter study 8 , a semiempirical model in coherence with thermodynamic conditions of equilibrium was used and oriented to the calculation of phase diagrams of the binary systems H2O-MCln (M = Mg 2+ , Fe 2+ , Fe 3+ ). To estimate solubility, the databases on iron and magnesium mineral's solubility were taken from the compilations of Linke and Seidell 7 and other single determinations.
For each solid phase, the exploitation of the experimental and previously reported data gives a liquidus curve equation comprising a limited number of parameters. The liquidus curves for the hydrated salts FeCl2·nH2O (n=2, 4, 6) and MgCl2·nH2O (n=2, 4, 6, 8, 12) have been established. They allow estimating with precision the solubility of the stoichiometric solid phase in a large range of temperature and composition.
Concerning the experimental solubility of the ternary system MgCl2-FeCl2-H2O, the first study was conducted by Boeke in 1911 9 . He has determined the solubility of individual mixtures of these chlorides at 295. 8  In supplementary work, minor changes of the temperature and composition were made by Balarew and Spassov many years ago 18 . They studied the isothermal solubility of FeCl2-MgCl2-H2O at 298 K. In addition to the fields of simple MgCl2.6H2O and FeCl2.4H2O salts, Balarew and Spassov also found a very narrow field corresponding to the crystallization of the double salt MgCl2.FeCl2.8H2O. However, no solid solution was found. Christomir Christov has described a Pitzer ioninteraction for thermodynamic analysis model that calculates (solid + liquid) equilibria in the {m1MgCl2 + m2FeCl2}(aq) systems, where m denotes molality at 298 K 19 . All mixed solution parameters are evaluated using experimental solubility data in ternary systems taken from Balarew and Spassov 18 . This thermodynamic model was extended to other systems 20 .
Unfortunately, the results from these previous studies show contradictory observations. Significant differences are found in the diagrams of the ternary system at different temperatures. However, the crystallization fields are very different, especially the fields of the double salt and the existence of the solid solution which remain uncertain. No research has yet proved or disproved the previous studies results.
It can be concluded that further experimental studies are necessary in order to clarify this ambiguity of the solid phases present in this ternary system at 15 °C before establishing the quaternary diagram.

Experimental section
Reagents and experimental procedure The study of this system was carried out at 288 K under a pressure of nitrogen gas of 0.10 MPa, in order to prevent the oxidation of Fe 2+ to Fe 3+ . Table 1 summarizes the information of the key reagents used in the current study, their manufacturer (origins) and their corresponding purity All the sample solutions were prepared with bi-distilled water (conductivity ˂ 10 -4 Sm -1 ). The magnesium chloride used for this study as starting reagents was recrystallized from bidistilled water and was dried at 313 K under water-pump vacuum to get pure MgCl2.6H2O, and it was preserved in a desiccator in the presence of P2O5.
The aqueous solution of iron (II) chloride, also known as ferrous chloride, was prepared in the laboratory by addition of iron powder to a solution of concentrated hydrochloric acid. After filtering, iron chloride tetrahydrate was crystallized using vacuum evaporation at the temperature of about 321 K. The obtained crystals were washed with bi-distilled water, dissolved in it and recrystallized again by vacuum evaporation of aqueous solution at 311 K using the processes earlier published 4 .

Analytical Methods
The water content of the FeCl2·4H2O and MgCl2·6H2O was determined by chemical analysis. The techniques of analysis have been previously described 4 .
The recrystallized product was characterized using several analytical techniques, in order to check its purity. The chloride ions are analyzed by a potentiometric method using silver nitrate (precision ± 0.2%). The Mg 2+ concentration was determined with a precision of within ± 0.5% by complexometric titration at pH 9.5-10 (ammonia buffer), using Eriochrome black T as the indicator. The Fe 2+ solution was titrated with a precision of about 0.3% by standard potassium dichromate (K2Cr2O7) solution using sodium diphenylamine sulfonate as the indicator.

Synthetic Method
According to the phase equilibrium composition, an appropriate quantity of the salt was dissolved in bi-distilled water first to prepare a few starting solutions. The phase equilibrium was performed in a glass jacket. The jacket temperature was controlled by thermostatic circulating water bath within ± 0.1 K.
The experimental process, as described in the previous papers, consists of determining the solubility by a synthetic method based on conductivity measurements (uncertainty, ± 0.1 mScm -1 ). The proposed method was successfully applied for direct determination of salts solid-liquid equilibria 21,22 . This technique was used by Tenu et al. to delimit the boundary of the solid solution in the CaCl2-SrCl2-H2O system 23 . The electric conductivity of the solution in thermodynamic equilibrium is measured when small amounts of water are progressively added to an isothermal saturated mixture of given initial composition using a micro-burette. Each phase changes was indicated by discontinuities in the electric conductivity-composition curve of the saturated solution. The last break corresponds to the dissolution of the last salt crystal. An invariant equilibrium is characterized by a plateau in the curve.

Results and Discussions
Under a pressure of nitrogen gas of 0.10 MPa, the solid-liquid equilibrium solubility data for the ternary system MgCl2-FeCl2-H2O at 288 K are listed in Table 2. The corresponding phase diagram of the system is built and illustrated graphically in Figure 2 with the Jänecke coordinate. Figure 3 is a righttriangular diagram for the same system with the weight percent. The composition values of the equilibrium solution phase and solid phase were expressed in terms of Jänecke coordinate. The composition of a mixture is related to the total amount of ions P taking into account their charge. Multiple measurements were performed to determine the uncertainty of the measurement. The evolution of the electric conductivity, when the saturated solution is diluted, can be performed through an indirect method. In particular, the use of the Jänecke coordinate Z, to represent the water volume, displays data in its correct position. This mode of representation has been found to be the most convenient. By way of example, Figure 4 plots the Z (water content) value, added to an initial mixture M, against electric conductivity. The curve gives valuable information about how and where the global composition takes changes.
The results of the conductivity measurements obtained for three ternary mixtures M8(U=50), M12(U=80) and M16(U=94.97) were presented on Table 3 and reported in Figures 4a, 4b and 4c respectively. The water content vs. conductivity graph of a mixture in a ternary system presents a break at every appearance/disappearance of any phase and a plateau when an invariant equilibrium is reached.
The phase changes was indicated by discontinuities in the graph of water content versus conductivity. The phase boundaries was therefore estimated from the water content of each phase change, and the dotted lines symbolize these boundaries.
These Figures show a typical shape of the curves, which bound:  One phase field "liquid",  Two phases liquid+solid "solid solution, FeCl2.4H2O or MgCl2.6H2O"  Two invariant phase fields liquid + FeCl2.4H2O + "solid solution or MgCl2.6H2O" The analysis of the different ternary mixtures, suitably chosen, allowed us to confirm the nature of the solid phases, which are manifested, to delimit the mono-, di-and three-phase stability domains and to determine the composition of the invariant solutions. This procedure permitted the determination of the entire phase diagram of MgCl2-FeCl2-H2O at 288 K. The solubility results are in general in good agreement, especially in the FeCl2 rich side, with those previously reported by literary 10,16 . The synthetic method allowed us to draw, in the FeCcl2 rich side, the tie lines of the solid solutions and to determine its chemical formula, which is i.e. (Fe(1-x)MgxCl2.10H2O) with 0 <x ≤ 0.38. However, our investigations indicate that no solid solution areas were defined in the system at a high concentration of magnesium chloride. Also, the existence of the double salt was not found in this isothermal phase diagram at a temperature of 288 K. These results contradict those published in previously phase diagrams in which the double salt FeCl2.MgCl2.8H2O have been observed at 298 K 18 .

Conclusions
In this study, the solubility isotherm of MgCl2-FeCl2-H2O system at 288 K was established using the synthetic method based on conductivity measurements. It was found that only three solubility branches are corresponding to MgCl2. 6H2O Three diphasic fields, where a solid phase coexists in equilibrium with liquid: "liquid + solid solution", "liquid + FeCl2.4H2O" and "liquid + MgCl2.6H2O". The smallest crystallization field is MgCl2·6H2O. According to our experimental data, no double salt was formed at 288 K.
It seems that the synthetic method based on conductivity measurements is an important tool to investigate the solid solution in ternary aqueous phase diagrams.