2018 Volume 59 Issue 3 Pages 443-449
In this work, vacuum distillation experiments of Zn-Al-Fe alloy were performed. A method for predicting the infinite dilute activities in the Zn-Fe system was examined. The activities of components of the Zn-Al-Fe system were calculated based on the Wilson equation. Vacuum distillation of Zn-Al-Fe alloy was discussed based on the experimental investigations of the distillation temperature, the holding time and the thickness of the raw materials mixture by response surface methodology. The results showed that there were remarkable influences of the distillation temperature, holding time and little influence of the thickness of the raw materials mixture on vacuum distillation. The zinc in Zn-Al-Fe alloy were effectively recovered at 800℃–850℃ for 60 min–75 min. This work aims to investigate the vacuum distillation of Zn-Al-Fe alloy to optimize the process for high zinc recovery. After vacuum distillation, the direct yield and volatilization rate for zinc were 99.8% and 99.79% when the distillation temperature was 800℃, and the soaking time was 75 min. The calculations and experimental results demonstrated that this method can guide the vacuum distillation of Zn-Al-Fe alloy based on the Wilson equation.
Hot dip galvanizing is the most common and effective process for steel corrosion prevention. Hot dip galvanized steel sheet has excellent corrosion resistance, low cost, high quality surface and other important properties, and is widely used in the light industry, home appliances, and automobile and construction industries. With the rapid development of the automobile industry, the demand for hot dip galvanized steel sheet has seen continued growth. Increasing numbers of large iron and steel enterprises have taken on the research and development of hot dip galvanized steel sheet as an important development strategy1).
Hot galvanizing slag consists of aluminum and iron, which dissolves in liquid zinc under high temperature conditions. Essentially, hot galvanizing slag consists of intermetallic compounds of Fe-Zn, Fe-Al and Zn-Al-Fe systems. Under the process conditions of continuous hot galvanizing, liquid zinc will always be iron saturated. Adding zinc ingot will absorb large amounts of heat and transfer a large amount of aluminum to the zinc pot because of segregation, reducing the solubility of iron and aluminum in the liquid zinc. Therefore, the separation of intermetallic compounds of Fe-Zn system, Fe-Al system and Zn-Al-Fe system by the liquid zinc is unavoidable2).
Satisfying the requirements of industry has made the extraction of zinc and aluminum from primary mineral resources difficult. Hot dip galvanizing slag contains a large amount of zinc and aluminum, which will cause significant waste and pollution to the environment if the hot dip galvanizing slag is directly abandoned. Therefore, it is necessary to recover the hot galvanizing slag.
There are many problems associated with the traditional method of alloy separation, including long process times, high energy consumption, low yield and poor economic benefit3). Vacuum distillation is a new type of metallurgy method, which is performed in a closed container that is clean and highly efficient. Vacuum distillation has been widely used in the separation of alloys and the refining of crude metals. Kunming University of Science and Technology has been engaged in small scale, expanded and industrialized experiments of alloy separation for many years and has achieved good results in those experiments4,5). However, there are fewer studies on the theoretical basis, thus further work is needed.
Activity is an important parameter in the metallurgical process that can guide the experiment6). However, it is very difficult to experimentally measure activity because of the high temperature detection. Therefore, it is very important to predict the thermodynamic properties of the solution by using a theoretical or semi empirical mathematical model. Thus far, there have been many types of solution model and geometric model methods, including Regular Solution Model7), Sub-Regular Solution Model8), Quasi-Regular Solution Model9), Molecular Interaction Volume Model(MIVM)10), Wilson Equation11), new generation solution model (ZHOU model)12), etc. However, each of these models has some applicable conditions and limitations. The Wilson Equation can predict the thermodynamic properties of multicomponent alloy systems only by the infinite dilute activity coefficients, $\gamma_{i}^{\infty}$ and $\gamma_{j}^{\infty}$. The author predicted infinite dilute activity coefficients ($\gamma _{Zn - Fe}^{\infty}$) by a few activity coefficients of two components at a certain temperature. In addition, the Wilson Equation is suitable for the liquid phase completely miscible system as well as the single phase liquid region of the liquid phase partially miscible system. Therefore, the author calculated any two components of the activity coefficients of Zn-Al-Fe alloy ($\gamma_{Zn - Al}^{\infty}$, $\gamma_{Al - Fe}^{\infty}$, $\gamma_{Zn - Fe}^{\infty}$) by the Wilson Equation and generated a vapor-liquid phase diagram to analyze the separation of zinc, aluminum, and iron.
The infinite dilution activity coefficient is a numerical characteristic of a real solution and an important piece of information in solution theory that represents the largest non-ideal condition of the solution as well as the behavior of solute molecules completely surrounded by solvent molecules. In the actual metallurgical production process, the measure of infinite dilute activity coefficients is both time and cost intensive, which is meaningless among a large quantity of alloy systems and high temperature production processes. At this time, the prediction model is a good method.
Ying13) presents a method to calculate the infinite dilute activity coefficient as follows:
| \[ In \gamma_{1} = \int\limits_{0}^{1} \alpha_{2} dx_{2} \] | (1) |
| \[ In \gamma_{2} = \int\limits_{0}^{1} \alpha_{1} dx_{1} \] | (2) |
Using eqs. (1) and (2), the infinite dilute activity coefficient can be easily calculated as long as the corresponding function of the two elements can be obtained. Through a few experimental measurements of activity coefficients of two components at a certain temperature, function α of the two elements can be calculated. Then, it will calculate an arbitrary temperature and contents of the two elements by using the Wilson Equation.
2.2 Saturated vapor pressureThe principle of vacuum distillation to separate two components of a metal is the difference between their saturated vapor pressures. The greater the saturated vapor pressure, the easier the metal will volatilize. At the same time, the component with the smaller saturated vapor pressure will remain in the liquid phase. Thus, the greater the difference in saturated vapor pressures, the greater the separation efficiency of the two components14).
The saturated vapor pressures can be calculated by the following formula:
| \[ \lg P = AT^{-1} + B \lg T + CT + D \] | (3) |
| Element | A | B | C | D |
|---|---|---|---|---|
| Zinc | −6220 | −1.255 | 0 | 14.465 |
| Aluminum | −16380 | −1.0 | 0 | 14.44 |
| Iron | −19710 | −1.27 | 0 | 15.39 |
The saturated vapor pressures of zinc, aluminum and iron at different temperatures are calculated separately in Table 2 below.
| P/Pa T/℃ |
450 | 500 | 550 | 600 | 650 | 700 | 800 | 900 | 100 |
|---|---|---|---|---|---|---|---|---|---|
| $\lg p_{zn}^{*}$ | −1.52089 | −0.10721 | 1.044481 | 2.000087 | 2.805178 | 3.492259 | 4.601622 | 5.45687 | 6.135 |
| $\lg p_{A{\rm l}}^{*}$ | −24.6132 | −21.019 | −18.0822 | −15.6382 | −13.5729 | −11.8051 | −8.93809 | −6.71424 | −4.94 |
| $\lg p_{F{\rm e}}^{*}$ | −31.7796 | −27.4577 | −23.9266 | −20.9883 | −18.5055 | −16.3804 | −12.9344 | −10.2619 | −8.13 |
According to the data in Table 2, the relation curves of saturated vapor pressure and temperature of zinc, aluminum and iron are shown in Fig. 1.
The change curve of saturated vapor pressure and temperature of zinc, aluminum and iron.
Under the control of certain temperature and vacuum conditions, the greater the saturation vapor pressure, the more volatile the component. The smaller the saturation vapor pressure, the more the component remains in the liquid phase. From Fig. 1, the saturated vapor pressure of each component increases with increasing temperature. The saturated vapor pressure of zinc is far greater than those of aluminum and iron, which suggests that under suitable conditions, zinc can be very easily volatilized into the gas phase. Therefore, zinc can be easily separated.
2.3 The Wilson EquationIn 1964, Wilson put forward the Wilson Equation for describing the excess Gibbs free energy of a non-electrolyte solution, which was based on the Flory-Huggins Equation.
There are strong and weak interactions between the molecules in solution, thus the local concentration of each component is not equal to the average concentration of any molecule in the solution. Based on the above assumptions, Wilson first brought forward the concept of local molecular fraction, which is the ratio of the number of coordination molecule, j, around the central molecule, i, to the total number of central molecules and coordination molecules around the central molecule. This coordination number is confined to the first coordination layer. Based on the Maxwell-Boltzmann distribution, the correlations between the local molecular fraction and the Boltzmann factor can be obtained by the analogy of the odds ratio, xij/xjj and xji/xii.
The excess Gibbs free energy of the multicomponent solution can be expressed as:
| \[ \frac{{G_{\rm m}}^{\rm E}}{RT} = - \sum_{i=1}^{k} x_{i} \ln \left( \sum_{j=1}^{k} x_{j} A_{ji} \right) \] | (4) |
Equation (4) is the Wilson Equation, where Aij is an adjusting parameter (Aii = Ajj).
For a binary alloy system, i and j, the activity coefficients of components i and j can be described as follows:
| \[ In \gamma_{i} = - In(x_{i} + x_{j} A_{ij}) + x_{j} \left( \frac{A_{ij}}{x_{i} + x_{j} A_{ij}} - \frac{A_{ji}}{x_{j} + x_{i} A_{ji}} \right) \] | (5) |
| \[ In \gamma_{j} = - In(x_{j} + x_{i} A_{ji}) - x_{i} \left( \frac{A_{ij}}{x_{i} + x_{j} A_{ij}} - \frac{A_{ji}}{x_{j} + x_{i} A_{ji}} \right) \] | (6) |
The adjusting parameters Aij and Aji can be described as follows:
| \[ A_{ij} = \frac{\upsilon_{j}}{\upsilon_{i}} \exp [-(\lambda_{ij} - \lambda_{ii})/RT] \] | (7) |
| \[ A_{ji} = \frac{\upsilon_{i}}{\upsilon_{j}} \exp [-(\lambda_{ji} - \lambda_{jj})/RT] \] | (8) |
Then, activity coefficients can be calculated using the Newton-Raphson method.
The Zn-Al-Fe alloy sample used for the experiments was obtained from a factory in China. The material and sample information are summarized in Table 3. The experiment was performed in the vacuum furnace in Fig. 2 at the National Engineering Laboratory of vacuum metallurgy, Kunming University of Science and Technology. The working pressure range for the vacuum furnace is 1 Pa to atmospheric pressure, and the working temperature range is from ambient temperature to approximately 1800℃. The vacuum furnace adopts a silicon-controlled transformer to control the temperature and measures the residual vapor pressure by a standard McLeod vacuum gauge. The temperature detection system consists of Pt-100 probes, which connect to a digital temperature meter (ANTHONE LU-900M) and predicts temperatures to 0.01℃. The residual vapor pressure can be measured by placing the sample in a graphite crucible in the constant temperature zone of the vacuum furnace after accurate weighing and drying treatment. The silicon-controlled transformer was regulated when the vacuum degree of the vacuum furnace met the experimental requirements. The vacuum furnace was heated to the preset temperature, and then the heat preservation treatment was performed in the vacuum furnace. After the treatment, the power was turned off. When the temperature was below 100℃, turned on the furnace to remove residues and condensate, which were weighed and tested. The direct yield and volatilization rate were calculated by the following formulas:
| \[ {\rm The\ direct\ yield} = \left( \frac{x''_{i} \times m''}{x_{i} \times m} \right) \times 100 \] | (9) |
| \[ {\rm Volatilization\ rate} = \left( \frac{x_{i} \times m - x_{i}^{\prime} \times m'}{x_{i} \times m} \right) \times 100 \] | (10) |
| Sample | xZn | xAl | xFe | xPb | xCu | xSn |
|---|---|---|---|---|---|---|
| raw material(mass%) | 36.52 | 60.23 | 1.98 | 0.54 | 0.16 | 0.12 |
Schematic diagram of the internal structure of the vertical vacuum furnace: 1- furnace cover; 2- vacuum furnace body; 3- furnace chassis; 4- electrode; 5- condensation plate; 6- observation hole; 7- thermal insulation sleeve; 8- heater; 9- crucible.
According to the characteristics and properties of hot galvanizing slag combined with industrial tests, this research adopts the experimental design software Design-Expert 8.0.6 experimental design (BBD) to study the influencing factors, including the interaction in the distillation process, distillation temperature and time and the quality of the raw materials, on the refining of Zn-Al-Fe alloy. The distillation temperature was 600℃ to 1000℃, the soaking time was 15 min to 75 min, and the material quantity was 40 g to 80 g, for performing the vacuum distillation experiment by the response surface method. The experimental design results are shown in Table 4.
| Factors | Code | Coding level | |||
|---|---|---|---|---|---|
| −1 | 0 | 1 | |||
| Distillation temperature, T/℃ | X1 | 600 | 800 | 1000 | |
| Feeding materials, m/g | X2 | 40 | 60 | 80 | |
| Soaking time, t/min | X3 | 15 | 45 | 75 | |
The values of the element content and element yield were selected as the response surface. In this paper, we studied the repeatability and random error, and provide additional freedom for error estimation. The experiment was repeated 5 times to ensure the accuracy of the experimental results.
The relationship between the factors and the results is shown in eq. (11):
| \[ Y = \beta_{0} + \sum_{i=1}^{k} \beta_{i} X_{i} + \sum_{i=1}^{k} \beta_{ii} X_{i}^{2} + \sum_{i}^{i<j} \sum\nolimits_{j} \beta_{ij} X_{i} X_{j} \] | (11) |
The graphite crucible was used in the experiment, so the graphite was doped when the material was collected. However, the chemical analysis method was used in the experiment, so nonmetals like graphite would not have an impact on the experimental results.
4.1 Infinite dilute activity predictionThe author calculated infinite dilute activity coefficients of a variety of binary alloy systems (Table 5) using eqs. (1) and (2) with a few activity coefficients of two components at a certain temperature15), where $\gamma_{i - \exp}^{\infty}$ and $\gamma_{j - \exp}^{\infty}$ indicate the experimental values of infinite dilute activity coefficients of component i and j, respectively; $\gamma_{i - {\rm Re}}^{\infty}$ and $\gamma_{j - {\rm Re}}^{\infty}$ indicate the calculated values of infinite dilute activity coefficients of component i and j, respectively; and $S_{i - {\rm Re}}$ and $S_{j - {\rm Re}}$ indicate the relative deviation of infinite dilute activity coefficients of component i and j, respectively.
| i-j | T/℃ | $\gamma_{\rm i - \exp}^{\infty}$ | $\gamma_{\rm j - \exp}^{\infty}$ | $\gamma_{\rm i - Re}^{\infty}$ | $\gamma_{\rm j - Re}^{\infty}$ | Si-Re | Sj-Re |
|---|---|---|---|---|---|---|---|
| Sn-Zn | 477 | 4.578 | 1.956 | 4.54512 | 1.96359 | 0.71822 | 0.38804 |
| Cd-Zn | 527 | 4.154 | 3.304 | 4.16273 | 3.30033 | 0.21016 | 0.11108 |
| Sn-Sb | 632 | 0.411 | 0.411 | 0.40895 | 0.40895 | 0.49878 | 0.49878 |
| Bi-Tl | 477 | 0.027 | 0.238 | 0.02795 | 0.23259 | 3.51852 | 2.27311 |
| Al-Cu | 1000 | 0.002 | 0.042 | 0.00211 | 0.04361 | 5.50000 | 3.83333 |
| Bi-Pb | 427 | 0.490 | 0.467 | 0.48505 | 0.46480 | 1.01020 | 0.47109 |
| Au-Ni | 877 | 10.21 | 5.149 | 10.11677 | 5.14077 | 0.91312 | 0.15984 |
| Cu-Fe | 1600 | 9.512 | 10.570 | 9.52975 | 10.49823 | 0.18661 | 0.67900 |
| Fe-Mn | 1177 | 1.765 | 1.543 | 1.78120 | 1.55727 | 0.91785 | 0.92482 |
| Al-Fe | 1600 | 0.058 | 0.027 | 0.05799 | 0.02667 | 0.01724 | 1.22222 |
| Al-In | 900 | 6.604 | 10.102 | 6.84927 | 10.08231 | 3.71396 | 0.19491 |
| Cu-Sn | 1127 | 0.317 | 0.007 | 0.30564 | 0.00729 | 3.58360 | 4.14286 |
The average relative deviation of the infinite dilute activity coefficients of components i and j are so low that it can forecast the infinite dilute activity coefficients of the Zn-Fe system. Partial activity coefficients and predicted values of infinite dilute activity coefficients of the Zn-Fe system are shown in Table 6.
| xZn | γZn | α | $\gamma_{\rm Fe}^{\infty}$(Zn-Fe) |
|---|---|---|---|
| 0.1 | 3.576 | 1.57314 | 2.4704 |
| 0.2 | 2.455 | 1.40332 | |
| 0.3 | 1.829 | 1.23218 | |
| 0.4 | 1.467 | 1.06450 | |
| 0.42 | 1.415 | 1.03189 |
The Wilson parameters for Zn-Al, Al-Fe and Zn-Fe binary systems are listed in Table 7. For the Zn-Al-Fe ternary system, the activity coefficients of any component can be calculate by using eqs. (4) and (5), which mostly require only binary parameters for the three constituent binaries. With the Wilson Equation, using binary data only, Zn-Al-Fe ternary system parameters can be calculated by substituting the corresponding $\gamma_{\rm Zn}$, $\gamma_{\rm A}$, $\gamma_{\rm Fe}$, T, P, ${\rm P}^{*}_{\rm Zn}$, ${\rm P}^{*}_{\rm Al}$, and ${\rm P}^{*}_{\rm Fe}$ at different temperatures with simple thermodynamic formula. Binary system infinite activity coefficients are the most important parameter in the Wilson Equation. By predicting the infinite activity coefficient of Zn-Fe system and inquiring other parameters, such as the infinite activity coefficients of Zn-Al and Al-Fe systems, the vapor composition and the equilibrium temperature can be predicted on the basis of liquid-residue composition of the Zn-Al-Fe ternary system by the Wilson Equation, as shown in Fig. 3.
| i-j | T/℃ | Aij | Aji |
|---|---|---|---|
| Zn-Al | 600 | 0.4117 | 0.7639 |
| 800 | 0.4808 | 0.8115 | |
| 1000 | 0.5349 | 0.8457 | |
| Al-Fe | 600 | 4.4528 | 26.8474 |
| 800 | 3.6618 | 13.3844 | |
| 1000 | 3.2047 | 8.2977 | |
| Zn-Fe | 600 | 0.8327 | 0.1293 |
| 800 | 0.9276 | 0.1759 | |
| 1000 | 0.9998 | 0.2169 |
Phase diagram of some experimental data and calculations of Zn-Al-Fe alloy for samples at 10 Pa.
Point a is the component of raw materials, points b through e are some experimental data, which are shown in Table 8. Additionally, points b through e are liquid phase points, whereas the points that collect in the lower right corner are the vapor phase points. From the above-calculated results, point b is clearly close to phase equilibrium, which can be broken by extending the distillation temperature and soaking time. Although this method could inevitably introduce certain errors during calculation and experiment, points d and e have shown satisfying experimental results for vacuum distillation.
| Point | Temperature, T/℃ |
Time, t/min | xZn | xAl | xFe |
|---|---|---|---|---|---|
| b | 600 | 15 | 0.3276 | 0.6239 | 0.0207 |
| c | 600 | 75 | 0.1165 | 0.7782 | 0.0582 |
| d | 800 | 75 | 0.0012 | 0.9523 | 0.0187 |
| e | 1000 | 45 | 0.0025 | 0.9436 | 0.0493 |
The experimental results of the vacuum distillation with contents of zinc, aluminum, and iron in residuum (liquid phase) and volatiles (vapor phase) are listed in Table 9.
| Sample | Temperature, T/℃ |
Pressure, P/Pa |
Time, t/min |
Weight, m/g |
xZn | xAl | xFe | yZn | yAl | yFe | Direct yield(%) |
Volatilization rate(%) |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1# | 800 | 10 pa | 45 | 60 | 0.0173 | 0.9041 | 0.0466 | 0.9726 | 0.0003 | 0.0001 | 92.175 | 96.88 |
| 2# | 800 | 75 | 80 | 0.0006 | 0.9221 | 0.0602 | 0.9696 | 0.0086 | 0.0004 | 94.476 | 99.89 | |
| 3# | 600 | 75 | 60 | 0.1165 | 0.7782 | 0.0582 | 0.9754 | 0.0001 | <0.0001 | 67.278 | 76.59 | |
| 4# | 800 | 45 | 60 | 0.0025 | 0.9051 | 0.0436 | 0.9790 | <0.0001 | <0.0001 | 92.788 | 99.54 | |
| 5# | 800 | 45 | 60 | 0.0148 | 0.9057 | 0.0489 | 0.9754 | 0.0005 | <0.0001 | 92.264 | 96.98 | |
| 6# | 1000 | 15 | 60 | 0.0042 | 0.9310 | 0.0035 | 0.9647 | 0.0050 | 0.0009 | 95.227 | 99.28 | |
| 7# | 800 | 15 | 40 | 0.0125 | 0.9260 | 0.0424 | 0.9819 | 0.0001 | <0.0001 | 90.325 | 97.74 | |
| 8# | 1000 | 45 | 40 | 0.0022 | 0.9317 | 0.0482 | 0.9295 | <0.0001 | <0.0001 | 98.2 | 99.61 | |
| 9# | 600 | 15 | 60 | 0.3276 | 0.6239 | 0.0207 | 0.9570 | <0.0001 | <0.0001 | 13.947 | 15.99 | |
| 10# | 800 | 45 | 60 | 0.0093 | 0.8972 | 0.0457 | 0.9766 | <0.0001 | <0.0001 | 95.183 | 98.32 | |
| 11# | 600 | 45 | 40 | 0.1067 | 0.7778 | 0.0274 | 0.9950 | <0.0001 | <0.0001 | 69.845 | 78.28 | |
| 12# | 800 | 75 | 40 | 0.0012 | 0.9523 | 0.0187 | 0.9776 | 0.0070 | 0.0002 | 99.8 | 99.79 | |
| 13# | 1000 | 45 | 80 | 0.0025 | 0.9436 | 0.0493 | 0.9589 | 0.0064 | 0.0002 | 99.5 | 99.56 | |
| 14# | 800 | 15 | 80 | 0.0205 | 0.9284 | 0.0457 | 0.9872 | 0.0003 | 0.0001 | 91.604 | 96.28 | |
| 15# | 800 | 45 | 60 | 0.0031 | 0.9111 | 0.0316 | 0.9702 | 0.0036 | <0.0001 | 90.944 | 99.43 | |
| 16# | 600 | 45 | 80 | 0.3236 | 0.6340 | 0.0187 | 0.9776 | <0.0001 | <0.0001 | 17.876 | 17.01 | |
| 17# | 1000 | 75 | 60 | 0.0087 | 0.9122 | 0.0472 | 0.9784 | 0.0062 | 0.0004 | 94.57 | 98.4 |
As shown in Fig. 4, the Zn-Al-Fe alloy was under a distillation temperature 600℃ to 1000℃, soaking time of 15 to 75 min, and feeding materials of 60 g. The effect of these factors on the direct yield of zinc was examined.
The effect of factors on the direct yield of zinc.
As shown in Fig. 4, the saturated vapor pressure of zinc increases gradually with increasing temperature, which made the direct yield of zinc gradually increase. With the increased of temperature, the direct yield of zinc increased sharply. When the temperature was 600℃ and the soaking time was 75 min, the direct yield of zinc was approximately 67.278%. When the temperature was 800℃ and the soaking time was 75 min, the direct yield of zinc was approximately 99.8%. This was because the large amount of zinc volatilized in gas phase. With further increased of temperature, the direct yield of zinc fell slowly, and when the temperature was 1000℃ and the soaking time was 75 min, the direct yields of tin was 94.57%. This is because when the temperature is higher than 800℃, other impurities also evaporated into the gas phase. Prolonged the distillation time can also improve the direct yield of zinc. When the distillation temperature and soaking time were above 800℃ and 70 min, respectively, the direct yield of zinc was no longer rising.
4.3.2 The effect of factors on the volatilization rate of zincAs shown in Fig. 5, the volatilization rate of zinc increased gradually with increase soaking time and distillation temperature because more zinc evaporated with increasing temperature and time, which gradually increases the volatilization rate of zinc. The volatilization rate of elemental zinc increased from 15.99% to 99.79% under the experimental conditions. When the temperature was 800℃ and the soaking time was 75 min, the volatilization rate of zinc was approximately 99.79%. With further increased of temperature, the direct yield of zinc was flat or even falling, and when the temperature was 1000℃ and the soaking time was 75 min, the volatilization rate of zinc was 98.4%. This is because some of the zinc returned to the liquid phase during the long cooling time. When the distillation temperature and soaking time were above 800℃ and 70 min, respectively, the volatilization rate of zinc stopped increasing which was same as the direct yield.
The effect of factors on the volatilization rate of zinc.
As shown in Fig. 6, the content of aluminum increased gradually with increasing soaking time and distillation temperature. With the increased of temperature and the extension of time, the zinc in the Zn-Al-Fe alloy essentially evaporated, leaving most of the aluminum and iron, thus the content of aluminum increased. The content of aluminum correlates strongly with the distillation temperature. When the distillation temperature and soaking time were above 800℃ and 70 min, respectively, the aluminum content in the residue no longer increased. When the temperature was 800℃ and the soaking time was 75 min, the aluminum content in the residue was approximately 95.23 mass%. The aluminum from these residues can be recycled in other ways.
The effect of factors on the content of aluminum in liquid phase.
As shown in Fig. 7, the content of iron increased gradually with increasing soaking time and distillation temperature. However, the iron content changed only slightly from 2.07 mass% to 6.02 mass% because of the low content of iron in the raw materials despite the volatilization of zinc.
The effect of factors on the content of iron in liquid phase.
According to the theoretical analysis, such as pure material boiling point and saturated vapor pressure of pure substances, arsenic volatile went into the gas phase in the vacuum distillation process, and aluminum was enriched in the liquid phase. The prediction of infinite dilute activity coefficients of a binary system can make the distillation process more efficient. The Wilson Equation was used to calculate the experimental phase equilibrium data. The phase equilibrium can be broken by extending the distillation temperature and soaking time. The purification experiment showed that the zinc in Zn-Al-Fe alloy were effectively recovered at 800℃–850℃ for 60 min–75 min. These results indicated that the calculation method of phase equilibrium was reliable for the process of vacuum distillation of alloys by comparing the experimental data and the phase equilibrium diagrams. This calculation could help control the conditions of the vacuum distillation of Zn-Al-Fe alloy to achieve separation as well as provide an efficient and convenient method to guide the process of vacuum metallurgy.
This work has been founded by the Fund of Yunnan science and technology plan project under Grant No. 2013FZ012, the National Natural Science Foundation of China Youth Foud of Study on mechanism of vapor condensation from complex gas of As-Pb No. 51504115, the Cultivating Plan Program for the Leader in Science and Technology of Yunnan Province under Grant No. 2014HA003.
