On the Distillation Separation of Aluminum–Tellurium System Melts under Equilibrium Condition

Abstract

The problem to purify secondary aluminum raw materials from tellurium can be solved by the distillation method based on phase diagrams with liquid and vapor coexistence fields. Similar diagrams can be generated based on the vapor pressure values of the components. In this regard, the vapor pressure values of tellurium and aluminum telluride were determined by the boiling point method. The aluminum vapor pressure values are found by integration of the Gibbs-Duhem equation. The boundaries of the system vapor-liquid equilibrium fields for the Al-Te system at 101.32 kPa and 6.67 kPa were calculated based on the vapor pressure values of the components. The following conclusion can be made from the consideration of the position of the liquid and vapor coexistence field boundaries under atmospheric pressure and in a vacuum. Aluminum can be quite completely purified from Al₂Te₃ and Te by distillation in a vacuum in one operation at temperatures above 1273 K. Tellurium will be in a complete vapor state under these conditions—above the boiling line in the Al₂Te₃-Te system.

1. Introduction

The problem of aluminum purification from tellurium arises when secondary raw materials are processed. A distillation process in a vacuum can be one of the ways to remove tellurium from melts with aluminum. Tellurium is a semiconductor purified by a combination of various operations, including vacuum distillation, to high purity. It is possible to judge on the distribution of aluminum in the Al-Te system between liquid and vapor based on a complete equilibrium diagram that contains the boundaries of the condensed phase fields and the melt and vapor coexistence fields. The boundaries of the liquid and vapor coexistence fields at atmospheric pressure and in vacuum enable us to judge the possibility and completeness of the separation of elements, the number of technological cycles, the composition of the condensate and distillation residue, and reduced energy costs [1,2,3,4].

There are many works devoted to the study of tellurides of various metals [5,6,7,8,9] but there are only very few works devoted to the study of aluminum chalcogenides [10].

The authors of [11] found the formation heat of aluminum sesquichalcogenides equaled to 326.3 ± 21 kJ/mol in the Berthelot-Roth microbomb. Similar data were obtained [12] for the aluminum telluride formation enthalpy (−318.8 ± 4 kJ/mol) based on calorimetric measurements. The formation heat of gaseous Al₂Te₃ was −364 kJ/mol in [13].

The authors of [14] found the composition of the vapor phase over aluminum telluride. Mass spectrometric studies showed the presence of low intensity ions. The authors also obtained thermodynamic functions at 1292 K.

The authors of [15] found the congruent nature of Al₂Te₃ melting by the calorimetric method. Then, the authors of [16] established the presence of a stratification region of liquid solutions by differential thermal analysis. The presence of two liquid phases was confirmed by mass spectrometry.

Mass-spectroscopic, static and torsion-effusion methods [17] showed the gas phase composition over aluminum telluride within 538–760 K. The studies determined the presence of Te₂.

The partial and integral thermodynamic functions of metals were determined by measurement of the electromotive forces in concentration circuits at 700–820 K in [18].

The authors of [19] found a superstructure in the α-Al₂Te₃ compound. This compound was discovered in the phase equilibria study in the Al-Te system.

Data obtained in [20] indicates small thermodynamic activity values within the concentration range of 0–60 atm. % of tellurium at 1190 K.

The study [21] was devoted to thermodynamic modeling and optimization of Al-Te system.

Studies of the Al-Te system phase diagram showed the existence of aluminum telluride in solution [14].

The congruent nature of the compound evaporation obtained [15] as a result of Al₂Te₃ dissociation at melting point 1168 K–9.97 × 10⁻³ kPa indicates the possibility of considering the Al-Te system as two particular Al-Al₂Te₃ and Al₂Te₃-Te systems [17].

The study results showed that the thermodynamics of the tellurium chalcogenide compound with aluminum impurity present as tellurium sesquihalcogenide at high temperature and low pressure were insufficiently studied. The data obtained are required to create a physical and chemical basis for the distillation processes, as well as to solve technical problems of chalcogenide separation from aluminum impurities and to obtain top-quality tellurium.

Thermodynamic studies of the Al-Al₂Te₃ and Al₂Te₃-Te systems are partially described in our two works [2,3,4].

The purpose of this study is to perform fundamental research on the issue of technological recommendations. The data obtained enable determining the presence or absence of the possibility of separating the studied system and refining tellurium from aluminum and its compounds.

2. Experimental Part

The purification process for the secondary aluminum raw material from tellurium can be solved by distillation method. The possibility to separate components can be judged based on phase diagrams with the liquid and vapor coexistence fields. These diagrams can be constructed based on the vapor pressure values of the components. When the vapor pressure values of tellurium and aluminum are compared, the boiling point method (isothermal version) is considered to be the most acceptable way to determine the pressure of the most volatile component, tellurium.

Therefore, preliminary experiments were performed to determine the onset of Al₂Te₃ intense evaporation and the possibility to determine the saturated vapor pressure values of the compound by the boiling point method [22]. The boiling point method (isothermal variant) was used to determine the values of saturated aluminum telluride vapor in the Al-Al₂Te₃ system and the total vapor pressure in the Al₂Te₃-Te system.

The procedure intended to determine the vapor pressure from the boiling points was as follows. A sample of an alloy with a certain composition was transferred into a quartz crucible (Figure 1), suspended on a spring scales in a quartz retort filled with argon.

The quartz retort was put down into a shaft-type furnace. The crucible with a sample was transferred to the isothermal zone of the furnace. When the inert gas was evacuated from the retort, the mass loss (Δm) and the corresponding pressure (P) (isothermal version of the method) were recorded in the system. The pressure was considered equal to the vapor pressure when a sharp increase in the evaporation rate was observed. The vapor pressure was determined with the help of the joint solution of equations describing the dependence P = f(Δm) before and after the observed break in the curve. The sample temperature was maintained with an accuracy of ±5 °C; the pressure was measured with an aneroid barometer with an accuracy of ±67 Pa (±0.5 mm Hg).

An isobaric variant of the method was used to raise the alloy temperature at constant pressure. The temperature was considered equal to the boiling point when a sharp increase in the evaporation rate was observed.

The comparability of the saturated vapor pressures of Al₂Te₃ and tellurium was also established. In this regard, the vapor phase composition was previously determined by a static method for a Al₂Te₃-Te particular system.

Method intended to determine the steam composition by the static method. The vapor composition was determined using a vessel made of quartz glass with the design shown in Figure 2.

The vessel included two ampoules: a small one for the transfer of a telluride sample, and a large one for the vapor phase in equilibrium with the alloy, both were connected with a capillary. The time to reach equilibrium was pre-set. The experimental procedure was as follows. A weighed amount of aluminum telluride was transferred into the ampoule for the alloy through the filling offshoot. The vessel was washed with argon three times, evacuated to a pressure of 1 Pa, and then the offshoot was sealed off. The vessel prepared in this way was transferred into the isothermal zone of a furnace heated to the test temperature and maintained for 24 h. The vessel was removed after the exposure time and quenched in water. The vapor phase ampoule was cut off. The steam condensate was dissolved in nitric acid. The number of elements in the solution was determined by the chemical method with the help of an atomic absorption spectrophotometer. In this case, the amount of aluminum in the condensate was assigned to the Al₂Te₃ compound. The molar ratio of the elements in the vapor was calculated based on the data on the mass amount of sulfur and selenium in the condensate.

Preparation of alloys. Synthesized alloys were used as an object of research. Their composition is specified in

Table 1. Composition of Al-Te system alloys.

Alloy Numbers	Alloy Composition, Mass %		Alloy Composition, at.
	Te	Al	Te	Al
1	40.21	57.79	12.45	87.55
2	59.54	60.46	23.73	76.27
3	72.52	27.48	35.82	64.18
4	80.47	19.53	46.56	53.44
5	87.65	12.35	60.00	40.00
6	90.13	9.87	65.88	34.12
7	92.74	7.26	72.98	27.02
8	95.88	4.12	83.11	16.89
9	97.91	2.09	90.83	9.17

.

The alloys were prepared from tellurium (99.99 wt %) and aluminum (99.99 wt %) in quartz ampoules. The furnace was heated slowly 100 °C per hour to the delamination area, and the alloy was kept for 12 h followed by quenching in water.

Our work [2] presents data on the vapor pressure of aluminum telluride and aluminum and the values of saturated vapor pressure over liquid and crystalline aluminum sesquitelluride confirming the correctness of the experimental data.

3. Results

The primary factors for the separation of the system into components are liquid–vapor equilibrium and the coexistence of field boundaries in the liquid alloy and vapor: boiling point and composition of the vapor. The data on them suggest the possibility of separating the system into its components by evaporation of one or more alloy components.

The liquid-vapor phase transition boundaries were determined based on the partial pressures of the saturated vapor components of the system.

Boundaries of liquid-vapor phase transitions: boiling point and the corresponding vapor composition of the Al-Al₂Te₃ system ($0\le x_{A{l}_{2}T{e}_3}\le 1$) at atmospheric pressure and 6.67 kPa are specified in

Table 2. Liquid-vapor phase transition boundaries of Al-Al₂Te₃ system ($0\le x_{A{l}_{2}T{e}_3}\le 1$ ) at atmospheric and low pressures.

Al2Te3 Content in the Alloy, mol.	At Pressure, Pa:
	101,325		6670
	Boiling Point, °C	Al2Te3 Content in the Steam, Mole Fraction	Boiling Point, °C	Al2Te3 Content in the Steam, Mole Fraction
1	1352	1	911	1
0.8	1407	0.9999	927	~1
0.6	1441	0.9997	937	~1
0.4	1459	0.9996	941	~1
0.2	1496	0.9994	959	~1
0.1	1591	0.9983	1010	~1
5 × 10−2	1749	0.9923	1089	~1
1 × 10−2	2252	0.7412	1373	0.9978
5 × 10−3	2385	0.4769	1544	0.9831
3 × 10−3	-	-	1683	0.9325
1 × 10−3	2493	0.1156	1905	0.5726
5 × 10−4	-	-	1967	0.3315
1 × 10−4	-	-	2014	0.0739
0	2520	0	2026	0

.

The pressure value of 6.67 kPa is due to the fact that the boiling point of the Al₂Te₃ compound becomes lower than its melting point at lower pressure.

Liquid-vapor phase transition boundaries of Al-Al₂Te₃ system ($0\le x_{Te}\le 1$) at atmospheric and 6.67 kPa are specified in

Table 3. Liquid-vapor phase transition boundaries of Al-Al₂Te₃ system ($0\le x_{Te}\le 1$ ) at atmospheric and low pressures.

Te Content in the Alloy, Mole Fraction	At Pressure, Pa:
	101,325		6670
	Boiling Point, °C	Te Content in the Steam, Mole Fraction	Boiling Point, °C	Te Content in the Steam, Mole Fraction
1	989	1	745	1
0.8	1017	0.9708	774	0.9488
0.6	1049	0.9152	795	0.7858
0.4	1085	0.8446	804	0.6955
0.2	1145	0.7166	811	0.6536
0.1	1205	0.5597	830	0.5605
5 × 10−2	1258	0.3905	855	0.4188
1 × 10−2	1327	0.1143	895	0.1336
0	1352	0	911	0

.

The known state diagram made by the authors [23] is supplemented by us [3] with the melt-vapor coexistence fields for the Al-Te system (Figure 3) at atmospheric pressure and in a vacuum of 6.67 kPa (highlighted by dashed line).

Thus, when the pressure decreases from atmospheric pressure (101,235 Pa) to high vacuum (0.01 Pa), the phase transition temperature change is 5.6 × 10⁻³ K according to our calculations based on the results of the study [24]. The decrease in the temperature of condensed phase transformations at low pressure (6.67 kPa) is not taken into account in the diagram.

The following conclusion can be made from the consideration of the position of the liquid and vapor coexistence field boundaries at atmospheric pressure and in vacuum. Aluminum can be quite completely purified from Al₂Te₃ and Te by distillation in a vacuum in one operation at temperatures above 1273 K. Tellurium will be in a complete vapor state under these conditions—above the boiling line in the Al₂Te₃-Te system.

It is difficult to refine tellurium from aluminum impurities due to the presence of a joint transition of tellurium and aluminum telluride in the condensate which requires the removal of aluminum at the preliminary stage. Only 1.32 × 10⁻³ atm. % Al will be present in the condensate during distillation of aluminum telluride from an alloy containing 0.5 atm. % Te at 6.67 kPa in the vapor phase. The technological rarefaction limit makes 6.67 kPa (50 mm Hg) during the distillation process in forevacuum, and aluminium telluride crystallization from a cube residue is possible below this limit

The partial and integral thermodynamic characteristics of mixing and evaporation were determined based on the thermodynamic activity values of Al-AlTe₃ system and AlTe₃-Te partial pressure of saturated vapor [3].

For isobaric-isothermal conditions, the activity of each component ($a_i$), of the constituent system is related to the partial Gibbs free energy of mixing ($\Delta {\overline{G}}_i$) with the expression: $\Delta {\overline{G}}_i=RT\ln a_i$, used to determine the partial change in the mixing entropy of the component by differentiation ($\Delta {\overline{S}}_i$): ${\left(\frac{\partial \Delta \overline{G}}{\partial T}\right)}_P=-\Delta {\overline{S}}_i$, and further the mixing enthalpies ($\Delta {\overline{H}}_i$): $\Delta {\overline{H}}_i=\Delta {\overline{G}}_i+T\Delta {\overline{S}}_i$.

The change in the Al-Al₂Te₃ system activity is shown in Figure 4, and the change in the Al₂Te₃-Te system activity is given in our work [3].

It can be seen that the particular Al-Al₂Te₃ system is characterized with a strong positive deviation from the law of ideal solutions that is characteristic of stratifying melts.

The partial and integral mixing functions of the system components are speccified in

Table 4. Changes in partial and integral mixing entropies of the Al-Te system.

Composition of the Alloy, atm. %		$\mathbf{\Delta }{\overline{\mathit{S}}}_{\mathit{T}\mathit{e}}^{\mathit{m}\mathit{i}\mathit{x}}$ J/(mole·K)	$\mathbf{\Delta }{\overline{\mathit{S}}}_{\mathit{A}{\mathit{l}}_2\mathit{S}{\mathit{e}}_3}^{\mathit{m}\mathit{i}\mathit{x}}$ J/(mole·K)	$\mathbf{\Delta }{\overline{\mathit{S}}}_{\mathit{A}\mathit{l}}^{\mathit{m}\mathit{i}\mathit{x}}$ J/(mole·K)	$\mathbf{\Delta }{\mathit{S}}_{\mathit{A}\mathit{l}-\mathit{S}\mathit{e}}^{\mathit{m}\mathit{i}\mathit{x}}$ J/(mole·K)
Te	Al
0	100	-	-	0 ± 0	0 ± 0
10	90	-	9.10 ± 0.53	0.86 ± 0.05	2.24 ± 0.13
20	80	-	8.00 ± 0.47	1.20 ± 0.07	3.46 ± 0.20
30	70	-	7.24 ± 0.42	1.77 ± 0.10	4.51 ± 0.26
40	60	-	5.88 ± 0.34	3.78 ± 0.22	5.18 ± 0.30
50	50	-	3.62 ± 0.21	11.17 ± 0.65	4.87 ± 0.28
60	40	-	0 ± 0.0	-	0 ± 0.0
70	30	15.84 ± 1.38	4.62 ± 0.40	-	7.43 ± 0.65
80	20	−2.66 ± 0.23	15.30 ± 1.34	-	6.32 ± 0.55
90	10	−4.75 ± 0.41	16.59 ± 1.45	-	0.59 ± 0.05
100	0	0 ± 0	-	-	0 ± 0

and

Table 5. Change of partial and integral mixing enthalpy of Al-Te system.

Composition of the Alloy, atm. %		$\mathbf{\Delta }{\overline{\mathit{H}}}_{\mathit{T}\mathit{e}}^{\mathit{m}\mathit{i}\mathit{x}}$ J/(mole·K)	$\mathbf{\Delta }{\overline{\mathit{H}}}_{\mathit{A}{\mathit{l}}_2\mathit{S}{\mathit{e}}_3}^{\mathit{m}\mathit{i}\mathit{x}}$ J/(mole·K)	$\mathbf{\Delta }{\overline{\mathit{H}}}_{\mathit{A}\mathit{l}}^{\mathit{m}\mathit{i}\mathit{x}}$ J/(mole·K)	$\mathbf{\Delta }{\mathit{H}}_{\mathit{A}\mathit{l}-\mathit{T}\mathit{e}}^{\mathit{m}\mathit{i}\mathit{x}}$ J/(mole·K)
Te	Al
0	100	-	-	0 ± 0	0 ± 0
10	90	-	6.37 ± 0.37	−0.25 ± 0.01	0.86 ± 0.05
20	80	-	7.05 ± 0.41	−0.44 ± 0.03	2.06 ± 0.12
30	70	-	6.44 ± 0.38	0.05 ± 0.00	3.24 ± 0.19
40	60	-	5.14 ± 0.30	1.95 ± 0.11	4.08 ± 0.24
50	50	-	3.20 ± 0.19	8.29 ± 0.48	4.05 ± 0.24
60	40	-	0 ± 0.0	-	0 ± 0.0
70	30	6.12 ± 0.53	3.59 ± 0.31	-	4.22 ± 0.37
80	20	−12.20 ± 1.06	14.03 ± 1.22	-	0.91 ± 0.08
90	10	−9.99 ± 0.87	7.01 ± 0.61	-	−5.74 ± 0.50
100	0	0 ± 0	-	-	0 ± 0

.

The formation of alloys in the Te-Al system is accompanied by an increase in system disorder in the range up to eutectic concentrations of Te, reaching a maximum of 7.43 J/(mol × K) at ~74 atm. Insignificant ordering is observed in alloys above these concentrations, and the minimum negative value of the integral mixing entropy is −1.3 J/(mol × K).

The mixing enthalpy is variable and insignificant in value. Aluminum telluride-based alloys form with heat absorption with a maximum of +4.22 kJ/mol at ~70 atm. % Te, and this process is exothermic for tellurium-based melts −6.35 kJ/mol at ~94 atm. % Te and is positive over the entire range of melt concentrations, therefore, the formation of solutions goes with heat absorption-endothermically.

Some of the illustrative materials for thermodynamic functions (partial and integral mixing functions and evaporation of the Al-Te system) are presented in our two works [2,3].

Similar works for phase equilibria have been performed for many years by the authors [25,26,27].

4. Conclusions

The classical method of boiling points is presented based on the analysis of the determination methods. It is designed to determine the boiling point and the sum of the partial pressures of the elements. The experimental part for the determination of the total vapor pressure in the tellurium–aluminum system was performed by the determination of the boiling point of the substance based on the vapor pressure value in the unit with a vertical location of the reactor and continuous weighing of the sample. The composition of the vapor phase was determined by the static method. The standardized measurement systems for the parameters and reliable methods of mathematical processing of the results provide uniformity and the required accuracy of the measurements that are used for the reliability of the experimental research.

Some conclusions and recommendations for technology intended to refine elements by distillation in a vacuum can be made based on the conducted research, the constructed state diagram, and thermodynamic constants.

Al refining from an impurity Te, which will be in the form of Al₂Te₃ in the melt, will not constitute any technological difficulties; the vapor phase, in this case, will be represented practically with the Al₂Te₃ compound.

Te refining from Al impurity by Te distillation from the melt will be accompanied by Al₂Te₃ accumulation in the cube residue; the vapor phase will be represented by Te and Al₂Te₃ in comparable quantities.

The coexistence field of liquid solutions and the vapor phase in the Al₂Te₃-Te system has a small temperature interval on the state diagram, both at atmospheric and reduced pressure, and, therefore, a large number of condensate re-evaporation cycles will be required to separate the Al-Te alloys into their constituent elements, which are typical for the rectification process.

Distillation refining of aluminum from tellurium impurity does not imply technological difficulties, except for the rarefaction value equal to 6.67 kPa (50 mm Hg). The latter would involve crystallization of aluminum telluride from the melt. It would result in an even greater increase in distillation-condensation cycles with a transition to rectification. In this regard, the purification of tellurium from aluminum by distillation of chalcogen in a vacuum is difficult for the above reasons from a technological point of view.