2013 年 53 巻 7 号 p. 1120-1124
The oxygen solubility in the zirconium-containing iron-nickel melts has been experimentally studied at 1873 K using the Fe–40% Ni alloy as an example. It was shown that zirconium considerably decreases the oxygen solubility in this melt. The equilibrium constant of interaction of zirconium and oxygen dissolved in the Fe–40% Ni melt (logK(1)(Fe–40%Ni) = –6.221), the Gibbs energy of this reaction ( = 223078 J/mol), and the interaction parameters characterizing these solutions ( = –7.16; = –1.25; = 2.16) were determined. The equilibrium constants of interaction of zirconium and oxygen dissolved in the Fe–20% Ni and Fe–30% Ni melts, the Gibbs energy of this reaction, and the interaction parameters characterizing these solutions were calculated for 1873 K. With an increase in the nickel content in melt the deoxidation ability of zirconium decreases. This is due to the fact that, when the nickel content in melt increases, the bond strengths of zirconium with the melt base ( = 4.59·10–4; = 1.40·10–10) rise to a considerably greater degree than the bond strengths of oxygen weaken ( = 0.0105; = 0.357).
The Fe–Ni alloys are widely used in modern technologies. The presence of oxygen in these alloys worsens their serviceability. Physicochemical properties of oxygen solutions in melts of iron and nickel are studied in detail. Thermodynamic parameters characterizing these solutions are given in publications.1,2) However, since the additivity of properties of oxygen solutions in the Fe–Ni melts with respect to pure iron and nickel is not observed, their thermodynamic parameters in these melts should be studied. This allows one to optimize the production of these alloys. Upon production of iron-nickel melts, zirconium is used as a modifying agent. This element has the higher affinity to oxygen in comparison with iron and nickel. If zirconium is added into nondeoxidized melt, its considerable portion can be oxidized and lost. For this reason, the study of thermodynamics of oxygen solutions in the zirconium-containing iron-nickel melts is of both scientific and commercial importance.
The oxygen solubility in these melts was not studied earlier. Our investigations of thermodynamics of oxygen solutions in the iron-nickel melts containing chromium, vanadium, silicon, carbon, titanium, and aluminum showed the oxygen solubility in these alloys to be able differ appreciably from its solubility in pure iron and nickel due to a change in the bond strengths of element-deoxidizer and oxygen with the melt base.3)
The interaction of zirconium with oxygen can be described as
(1) |
(1a) |
The thermodynamics of oxygen solutions in the zirconium-containing iron melt is studied by many authors.4,5,6,7,8,9,10,11,12,13,14,15,16) However, the results obtained are not always in satisfactory agreement with one another. This is exemplified1) by the temperature dependence of the equilibrium constant of reaction (1) (Fig. 1). Based on a detailed analysis of these studies, the results of paper7) are recommended as most reliable data. In that investigation, the following parameters characterizing reaction (1) and zirconium solutions in pure iron were obtained:
logK(1)(Fe) = –57000/T + 21.8,
Relations between log K(1) (log
The authors7) determined the dependence of oxygen solubility in liquid iron on the zirconium content only for four very low zirconium concentrations in a range of 7·10–5 ÷ 2·10–3% Zr (Fig. 2), which hampers the extrapolation of these data to the wider concentration range. Recently, the dependence of oxygen solubility in liquid iron on the zirconium content was studied again.17) In this case, the zirconium concentration range studied is appreciably wider (2·10–3 ÷ 0.65% Zr) (Fig. 2). This figure shows that the results of papers7,17) are complementary. Figure 2 also demonstrates the data15,16) given in the reference book,1) which are close to the recommended values but the zirconium deoxidation abilities determined in these studies are somewhat lower than those in papers.7,17)
Dependence of the equilibrium oxygen concentration in pure iron and in the Fe–40% Ni melt on the zirconium and aluminum contents.
The oxygen concentration in equilibrium with a given zirconium content can be calculated according to the equation
(2) |
(2a) |
(3) |
Since, as was shown above, the results7,17) are mutually well-complementary, it is reasonable to process these experimental data together using Eq. (3). This allows one to determine the thermodynamic parameters of the Fe–Zr–O system for the wider range of zirconium concentrations. The data were processed by the method of regression analysis using a Quattro Pro program. The following coefficients were obtained for members of the equation (coefficient of determination R2 = 0.72):
(4) |
In Fig. 2 curve 1 corresponds to Eq. (4).
In Eq. (3)
A = ½logK(1); B = –½
С = –½
Taking into account the values of coefficients in Eq. (4), the fact that
Figure 2 also shows the dependence of the oxygen solubility on the aluminum content in iron melts according to data1) (curve 2). As is shown, up to about 0.006% zirconium is characterized by the higher deoxidation ability in comparison with aluminum; at the higher zirconium content, aluminum is the more strong deoxidizer.
The deoxidation of iron-nickel melts with zirconium was experimentally studied by the example of the Fe–40% Ni alloy. These experiments have been carried out in an induction furnace fed by a 10 kV·A high-frequency (400 kHz) generator. The scheme of experimental apparatus is shown in Fig. 3.
Schematic diagram of the experimental apparatus.
As initial materials, carbonyl iron (99.95%), electrolytic nickel (99.95%), and iodide zirconium (99.99%) were used. Total charge was 100 g. This charge containing iron and nickel in the ratio corresponding to the Fe–40% Ni alloy was placed into a ZrO2 crucible, which, in turn, was located in an external protective Al2O3 crucible. The charge was melted under Ar–Н2 atmosphere. Hydrogen and argon were preliminarily purified from oxygen impurities, water vapor, sulphides, organic compounds, and mechanical and other impurities. We used high-pure argon (99.992%), where the volume percent of oxygen was ≤ 0.0007% according to the Russian Standard GOST 10157-79. When was passed through a gas purification system of experimental unit, the volume fraction of oxygen in argon was less than 10–5%. Argon and hydrogen consumptions were 150 and 50 ml/min, respectively. When metal was completely melted, the supply of hydrogen was stopped and the melt was held under argon atmosphere (150 ml/min, 1873 K). Zirconium was introduced into the melt without breaking the furnace air-tightness; then this melt was held at the given temperature under Ar atmosphere until the equilibrium was stated. The temperature was measured by the Pt–6%Rh/Pt–30%Rh thermocouple. As was shown in previous experiments, where samples were taken every five minutes and the zirconium and oxygen concentrations were determined, the equilibrium is stated in this system for 18–20 min after the addition of zirconium.19) To be sure that the equilibrium is stated this time was increased to about 30 min. When the equilibrium was achieved the melt was sampled to be analyzed. The oxygen concentration in melt was determined with an accuracy of ±5·10–5% using a Leco TC-600 Gas Analyzer. Zirconium and nickel were analyzed with an accuracy of ±0.001% using an Ultima 2 Horiba Jobin Yvon ICP optical emission spectrometer.
The experimental results are shown in the Table 1 and in Fig. 2. Comparison of the experimental results with published data on the deoxidation ability of zirconium in iron melt (Fig. 2) shows that the deoxidation ability of zirconium in the Fe–40% Ni melt is somewhat lower as compared with pure iron.
[Ni], % | [Zr], % | [O], % | logK(1) | K(1) |
---|---|---|---|---|
40.5 | 0.030 | 0.00427 | –6.304 | 4.97·10–7 |
40.2 | 0.045 | 0.00439 | –6.110 | 7.77·10–7 |
40.6 | 0.047 | 0.00341 | –6.304 | 4.97·10–7 |
40.2 | 0.064 | 0.00368 | –6.111 | 7.74·10–7 |
40.5 | 0.091 | 0.00335 | –6.047 | 8.98·10–7 |
40.5 | 0.110 | 0.00193 | –6.439 | 3.64·10–7 |
38.2 | 0.310 | 0.00149 | –6.280 | 5.25·10–7 |
38.8 | 0.760 | 0.00128 | –6.175 | 6.68·10–7 |
38.9 | 1.390 | 0.00113 | –6.236 | 5.81·10–7 |
The experimental data were processed by the method of regression analysis using a Quattro Pro program according to Eq. (3). The following coefficients were obtained (coefficient of determination R2 = 0.81):
(5) |
Taking into account the values of coefficients in Eq. (5) and
Figure 4 shows the obtained dependence of the oxygen solubility in the Fe–40% Ni melt at 1873 K on the zirconium content in comparison with similar data on the oxygen solubility in this melt depending on the concentrations of chromium,20) manganese,21) vanadium,22) silicon,21) carbon,23) titanium,24) and aluminum.25) As is seen, the strongest deoxidation ability is characteristic of aluminum, it is lower but sufficiently high for zirconium. In the order of decreasing of this value other elements can be arranged as titanium, carbon, silicon, manganese, vanadium, chromium. The horizontal line in Fig. 4 shows the oxygen solubility in the Fe–40% Ni melt.26)
Oxygen solubility in the Fe–40% Ni melt depending on the content of element-deoxidizer (R) at 1873 K.
Using the parameters obtained for the melts of iron and Fe–40% Ni, one can calculate the equilibrium oxygen contents depending on the zirconium concentrations in the Fe–20% Ni and Fe–30% Ni melts. Reaction (1) can be imagined as a sum of reactions
18) (6) |
(7) |
(8) |
MFe–Ni = MFe XFe + MNi XNi .
The Gibbs energy for reaction (1) can be calculated as
whence it follows that
The
Parameter | Ni, % | |||
---|---|---|---|---|
0 | 20 | 30 | 40 | |
|
299726 | 265118 | 244291 | 223078 |
|
745195 | 745195 | 745195 | 745195 |
|
–199057 | –240820 | –268872 | –299463 |
|
–246412 | –239257 | –232032 | –222654 |
logK(1) | –8.359 | –7.402 | –6.820 | –6.221 |
|
4.59·10–4 | 3.11·10–5 | 5.11·10–6 | 7.13·10–7 |
|
0.010527) | 0.0131 | 0.0164 | 0.0221 |
|
2.25 | 2.21 | 2.18 | 2.16 |
|
–3.54 | –2.42 | –1.84 | –1.25 |
|
–20.21 | –13.81 | –10.52 | –7.16 |
|
–0.171) | –0.137 | –0.123 | –0.104 |
As for iron and nickel, the published data27) on the activity coefficients at infinite dilution
(9) |
In the case of zirconium, this equation for the Fe–40% Ni melt takes form
(9a) |
Using
The Fe–Ni melts are characterized by slight deviations from ideality.30) The
(10) |
With the knowledge of the interaction parameters for iron and и Fe–40% Ni melts, one can calculate those both for pure nickel and for Fe–20% Ni and Fe–30% Ni melts (Table 2).
The equilibrium oxygen concentrations in the Fe–20% Ni and Fe–30% Ni melts at different zirconium contents were calculated according to Eq. (2a). The oxygen concentrations depending on the zirconium contents in these melts are given below:
(11) |
(12) |
The dependences of the equilibrium oxygen concentration on the zirconium and nickel contents in melt at 1873 K, which were calculated by Eqs. (11) and (12), are shown in Fig. 5. The deoxidation ability of zirconium in the Fe–20% Ni and Fe–30% Ni melts is lower in comparison with iron but it is higher than in the Fe–40% Ni melt.
Dependence of the equilibrium oxygen concentration in the Fe–Ni melts on the zirconium and nickel contents at 1873 K.
With an increase in the nickel content in melt, the deoxidation ability of zirconium decreases. This is due to the fact that, as the nickel content increases, the bond strength of zirconium with the alloy base rises to a considerably more degree (
The curves of oxygen solubility in the melts pass through a minimum. To determine the zirconium concentrations corresponding to the minimum oxygen contents ([%O]min), the equation31)
(13) |
(13a) |
The zirconium concentrations calculated by Eq. (13a) and the oxygen contents corresponding to these concentrations are given below:
Ni, % | 0 | 20 | 30 | 40 |
---|---|---|---|---|
[%Zr]* | 0.09 | 0.17 | 0.29 | 1.28 |
[%O]min | 0.00037 | 0.00081 | 0.00120 | 0.00114 |
(1) The solubility of oxygen in the zirconium-containing iron-nickel alloys was experimentally studied at 1873 K as exemplified by the Fe–40% Ni alloy. It was shown that zirconium considerably decreases the oxygen solubility in this melt. The equilibrium constant of interaction of zirconium with oxygen dissolved in the Fe–40% Ni melt (logK(1)(Fe–40%Ni) = –6.221), the Gibbs energy of this reaction (
(2) The equilibrium constants of interaction of zirconium and oxygen dissolved in the Fe–20% Ni and Fe–30% Ni melts (logK(1)(Fe–20%Ni) = –7.402; logK(1)(Fe–30%Ni) = –6.820), the Gibbs energy of this reaction (
(3) As the nickel content in melt increases, the deoxidation ability of zirconium decreases. This is due to the fact that, when the nickel content in melt increases, the bond strengths of zirconium with the melt base (