The following article is a component of the October 1997 (vol. 49, no. 10) JOM
and is presented as JOM-e. Such articles appear exclusively
on the web and do not have print equivalents.

Research Summary

The Solubility of Silica in Calcium Ferrite Slags

C.M. Acuña and A. Yazawa

JOM-e Logo

In copper smelting processes, lime flux offers several advantages over silica; however, copper concentrates contain levels of silica that may impair the benefits of lime (i.e., the formation of dicalcium silicate). To fully utilize the benefits of lime fluxing, the solubility of silica in the final slag must be investigated. In this study, the solubility of silica in calcium-ferrite slags was investigated under different atmospheric conditions between 1,473-2,573 K.


Calcium-ferrite slags are widely studied as an alternative to traditional iron-silicate slags in copper smelting. The interest in these slags arises from their high capacity to absorb acidic impurities (e.g., arsenic and antimony) while presenting a wide, homogeneous liquid area when mixed with iron oxides. These two aspects partly solve troublesome magnetite precipitation during the smelting and converting of copper concentrates and reduce bath contamination in the electrolytic refineries.

Several studies regarding the kinetic aspects of calcium-iron silicates1 and the effect of CO/CO2 mixtures and sulfur content on calciowustite have been reported.2,3 In the systems CaO-FeO-Fe2O34-6 and CaO-FeOx-SiO2,7 most of the data are for iron-making temperatures, but few studies relate to nonferrous processing temperatures. Takeda et al.8 provided basic information on the activity of components at 1,473 K and 1,573 K for the CaO-FeO-Fe2O3 system, while the liquidus isotherms were investigated by Schumman et al.9 Furthermore, the phase relations of the CaO-Fe2O3-SiO2 and CaO-FeO-SiO2 systems under air atmosphere and in equilibrium with gamma iron have been reported.10-12

copper concentrates contain significant levels of siliceous components. To successfully use lime flux, the dissolution of silica in calcium-ferrite slags must be determined. By adding different levels of silica to a CaO-FeOx slag, the liquidus isotherm lines at 1,473-1,573 K under iron saturation and air and other atmospheres were investigated.


Figure 1
Figure 1. The effect of oxygen potential on the solubility of silica in calcium-ferrite slags melted at different temperatures.
A slag containing 18 percent CaO was prepared by mixing p.a. grade reagents (i.e., iron, calcium oxide, and hematite) and smelting in an iron crucible under N2 atmosphere. CaO or Fe-Fe2O3 mixtures were added to adjust the required CaO or FeOx level for different compositions. The experiments were carried out under gamma-iron saturation at 1,473 K, 1,523 K, and 1,573 K; experiments under other atmospheres, controlled by carbon-monoxide and carbon-dioxide mixtures, were done at 1,523 K and 1,573 K.

Equilibrium smelting tests were carried out at a fixed initial composition, temperature, and atmosphere with slags containing several silica contents. Once equilibrium was attained, the samples were rapidly solidified, and their upper and lower halves analyzed for silica. Any compositional difference was attributed to the presence of a primary solid, which was later confirmed by microscopic observations. The solubility limit for silica was estimated as the value where no difference in composition between the upper and lower halves of the sample was determined, as proposed by Takeda et al.8 The samples were chemically analyzed by standard methods for calcium, silica, and ferrous and total iron when no second phases were observed via optical microscopy.


At low oxygen potentials, the primary solid in the slags can be regarded mainly as dicalcium silicate because of its low free energy and dissolution of silica.10 This primary solid, realized via optical microscopic observations and checked by electron probe microanalysis, is highly favored as compared to any other compound such as fayalite or monocalcium silicate; however, as the partial pressure of oxygen increases, other compounds tend to form, and the appearance or precipitation of dicalcium silicate is jeopardized. Furthermore, high oxygen potentials involve an increase in the dissolution of silica. The precipitation of second or solid phases requires more stringent conditions (i.e., higher silica, iron oxides, and/or lime contents).

Figure 2
Figure 2. The effect of lime content on the oxidation degree (Fe+3/Fe+2) of the slag for samples melted at different temperatures and oxygen potentials (logPO2)—(a) 1,573 K, (b) 1,523 K, (c) 1,473 K. Dashed lines represent estimated behavior.
The effect of the oxygen potential upon the solubility of silica at various temperatures is shown in Figure 1. Generally, a continuous decrease in solubility with increasing oxygen potential is observed. At higher equilibrium temperatures, the solubility of silica in the slag is higher.

The effect of the oxygen potential upon the oxidation degree of the slag can be realized, at a first stage, by considering the equilibrium wustite-hematite, according to

FeO(s) + 1/4 O2(g) = FeO1.5(s)(1)

By replacing the appropriate thermodynamic data,14 the following is obtained.

log(aFeO1.5/aFeO) = 1/4 logPO2 - (7441.14/T-2.57)(2)

which, in terms of the ferric-to-ferrous ratio (assuming Henrian behavior), can be written as:

log(Fe+3/Fe+2) = log(YFeO/YFeO1.5) + 1/4 logPO2 - (7441.14/T - 2.57)(3)

The effect of temperature upon the oxidation degree can be explained by Equation 2. A decrease in temperature results in the second term on the right side of the equation becoming larger. If the oxygen potential and the ratio of activity coefficients in Equation 3 are assumed constants, the result is a decrease in ferric-to-ferrous ratio, as observed from Equation 3 in agreement with the data presented in Figure 2.

In Figure 2b, the estimated Fe+3/Fe+2 ratios corresponding to oxygen potentials of 10-8.59 and 10-10.80 at 1,523 K are presented as thick dashed lines. This range of oxygen potentials was selected because the slags show the wider range in calcium content. A comparison of the data in Figure 2a and Figure 2b shows that a decrease in temperature results in a lower Fe+3/Fe+2 ratio at a constant oxygen potential.

Although calcium ferrites are the main compounds formed, the following equilibria may also be considered.

2 CaO(s) + Fe2O3(s) = Ca2Fe2O5(s)(4)

CaO(s) + Fe2O3(s) = CaFe2O4(s)(5)

CaO(s) + 2 Fe2O3(s) = CaFe4O7(s)(6)

At low oxygen potentials under the coexistence of lime, wustite, and calcium ferrite, Reaction 5 may be written as

CaO(s) + 2 FeO(s) + 1/2 O2(g) = CaFe2O4(s)(7)

Table I. Equilibrium Constants for Calcium Ferrites and Dicalcium Silicate Formation
Reaction Temperature (K) Equilibrium Constant (Keq)
CaO(s) + 2 FeO(s) + 1/2 O2(g) = CaFe2O4(s) 1,573 2.24x104
CaO(s) + 2 FeO(s) + 1/2 O2(g) = CaFe2O4(s) 1,573 4.93x104
2 CaO(s) + Si2(s) = Ca2SiO4(s) 1,573 2.87x104
2 CaO(s) + Si2(s) = Ca2SiO4(s) 1,573 3.94x104

The equilibrium constants for calcium ferrites and dicalcium silicate formation are given in Table I;13 the tendency to form monocalcium ferrite is as strong as the tendency to form dicalcium silicate. It has been reported that reaction products such as calcium ferrites decrease the tendency to form dicalcium silicate,14 and the interaction between lime and iron oxides becomes stronger with increasing oxygen partial pressures.15 In fact, under these conditions, monocalcium ferrite (CaFe2O4) was always detected.16

Figure 3
Figure 3. The effect of lime on the solubility of silica in calcium-ferrite slags melted under iron saturation.
In the case of an acidic slag system (i.e., iron-silicate melts), increasing the silica content along isopotential lines results in a decrease of the ferric-to-ferrous iron ratio. However, in the situation of a basic slag system (i.e., iron, lime melts), the opposite behavior is observed. Therefore, if acidic silica replaces a basic oxide such as lime, the concentration of oxygen ions increases and, consequently, the slag will gradually become more basic. Under this condition, a stronger tendency to form ferric ions should be realized and the ferric-to-ferrous ratio should increase.


On the extreme side of low and high oxygen potentials, the system can be considered as the pseudoternaries FeO-CaO-SiO2 and Fe2O3-CaO-SiO2, respectively. From the data of Allen et al.17 for the system FeO-CaO-SiO2 with 15-35% CaO, the solubility of silica along the isotherms 1,473 K and 1,573 K is estimated as 1.0% and 1.5%, respectively. These values are in agreement with the present results (Figure 3). Minimum solubility is observed at 18-20% CaO, regardless of temperature. Above or below this range, an increase in solubility occurs.

Figure 4
Figure 4. The effect of lime on the solubility of silica in calcium-ferrite slags melted at 1,573 K under different oxygen potentials (logPO2).
In the ternary Fe2O3-CaO-SiO2 system with 30-40% CaO, the solubility of silica is practically constant at around 5%; however, it must be remembered that the data were obtained under air atmosphere.17 Although there is a tendency toward increasing silica solubility with increasing oxygen potential, the present results are nearly half those presented by Allen et al.

As the oxygen potential increases, the nose in the FeOx-CaO-SiO2 is wider and shifted toward the CaO corner.17 Therefore, as the slag dissolves, the change in silica solubility with lime content becomes steeper. Nevertheless, there is experimental evidence for the opposite arguments: the position of the nose depends on the oxygen potential and is extended away from the CaO corner toward lower silica contents.16 In view of this situation, further research on the effect of oxygen potential and temperature upon the solubility of silica at low lime contents is advisable.

Increasing the equilibrium temperature results in higher solubilities of silica. Furthermore, a decrease in temperature stretches the range at which calcium oxide melted. Results obtained at 1,573 K for equilibrium tests under different oxygen potentials are presented in Figure 4, which shows similar behavior to the test under iron saturation, with the exceptions of the narrower range of lime at which the silica level is constant and the faster change in solubility toward low lime contents.

Figure 5
Figure 5. A comparison of the experimental results and Takeda's Equation 8 (denoted by dashed lines)—(a) 1,573 K, (b) 1,523 K, (c) 1,473 K.
In experiments run at an oxygen potential of PO2 = 10-0.68 and 1,573 K, the solidus lines indicate around 5.2% SiO2 and 2.8% SiO2 at 19% CaO and 28% CaO, respectively (Figure 4). In comparison to experiments run at an oxygen isopotential of PO2 = 10-3.44, 3.7% SiO2, and 2.4% SiO2 at 16% CaO and 25% CaO, respectively, the results may provide an explanation for the difference in solubility between Allen's results and those in the present study. In fact, Allen et al. do not report the dissolution of silica along isopotentials lines other than iron saturation and air atmosphere.

An approach to realize the effect of lime upon the solubility of silica is provided by Takeda et al.,8 who relate the oxidation degree of the slag with its lime content. Although their analysis is based on the plain CaO-FeOx system, it can be extrapolated to CaO-FeOx-SiO2 due to the low solubility of silica. Takeda et al. relate the Fe+3/Fe+2 ratio with the CaO content in the slag under iron saturation up to oxygen potentials of -4 through

log(Fe+3/Fe+2) = 0.170·logPO2 + 0.018·(%CaO) + 5500/T - 2.52(8)

With this equation, it is possible to estimate the effect of oxygen potential, temperature, and lime content upon the ferric-to-ferrous iron ratio. The present results and those evaluated through Equation 8 are presented in Figure 5. At the lowest oxygen potentials, silica decreases the ferric-to-ferrous ratio in agreement with the behavior shown in the FeO-Fe2O3-SiO2 system; however, as the oxygen potential increases, so does the ferric-to-ferrous ratio. These results suggest that there is a compromise between lime and silica in respect to ferric and ferrous iron. For example, in Figure 5a and Figure 5b, the results at the lowest oxygen potentials when compared to Takeda's show lower ferric-to-ferrous ratios. However, at higher oxygen potentials, the opposite behavior is observed, except at 1,473 K.


The experimental work was done at the Research Institute of Mineral Dressing and Metallurgy, Tohoku University. Support from the Ministry of Education of Japan in a grant to C. Acuna is deeply acknowledged.


1. S.K. El-Rahaiby et al., Met. Trans., 17B (1968), p. 307.
2. F. Nakibogly, D.H. St. John, and P.C. Hayes, Met. Trans.,, 17B (1986), p. 375.
3. D.H. St. John, F. Nakibogly, and P.C. Hayes, Met. Trans.,, 17B (1986), p. 383.
4. R.W. Gurry and L.S. Darken, J. Am. Chem. Soc., 72 (1950), p. 3906.
5. E.T. Turkdogan, Trans. AIME, 221 (1961), p. 1090.
6. T. Arato and M. Tokuda, Bull. Res. Inst. Min. Dress. Metallurgy, Tohoku University, 35 (1979), p. 160.
7. M. Timucin and A.E. Morris, Met. Trans.,, 1 (1970), p. 3193.
8. Y. Takeda, S. Nakazawa, and A. Yazawa, Can. Met. Quart., 19 (1980), p. 297.
9. S. Schummann and G. Kraune, Arch. Eisenhuttenwes, 47 (1980), p. 435.
10. E.F. Osborn and A. Muan, "Phase Diagrams for Ceramists," J. Am. Chem. Soc., (1964), p. 456.
11. E.F. Osborn and A. Muan, "Phase Diagrams for Ceramists," J. Am. Chem. Soc., (1964), p. 586.
12. A. Muan and E.F. Osborn, Phase Equilibria Among Oxides in Steelmaking (New York: Addison-Wesley, 1965), p. 113.
13. K. Itakagi, Non-Ferrous Extractive Metallurgy, ed. A. Yazawa (Sendai, Japan: Japanese Institute of Metals, 1980), p. 315.
14. R. DasGupta, Second Int. Symp. Metall. Slags and Fluxes, ed. H.A. Fine and D.R. Gaskill (Warrendale, PA: TMS-AIME, 1984), p. 649.
15. S. Hara, T. Araki, and H. Ogino, Second Int. Symp. Metall. Slags and Fluxes, ed. H.A. Fine and D.R. Gaskill (Warrendale, PA: TMS, 1984), p. 521.
16. C. Acuna, "Equilibria Between Ferrite Slags and Copper Matte," Dr.Eng. thesis, Tohoku University, Japan (1987).
17. W.C. Allen and R.B. Snow, J. Am. Chem. Soc., 42 (1959), p. 413.

Cesar M. Acuña earned his DIC in materials science at Imperial College of Science and Technology, United Kingdom, in 1977 and his Dr.Eng. in extractive metallurgy at Tohoku University, Japan, in 1987. He is currently section head of R&D at the Chuquicamata smelter of Codelco Chile.

Akira Yazawa earned his Dr.Eng. in extractive metallurgy at Tohoku University, Japan, in 1953. He is currently emeritus professor at Tohoku University and is an international consultant.

For more information, contact C.M. Acuña, Codelco Chile, Chuquicamata Smelter R&D/Surgerencia Fundición de Concentrado, Chuquicamata—Chile; telephone (055) 325437; fax (055) 323488; e-mail acuna@chuq.codelco.cl.

Copyright held by The Minerals, Metals & Materials Society, 1997

Direct questions about this or any other JOM page to jom@tms.org.

Search TMS Document Center Subscriptions Other Hypertext Articles JOM TMS OnLine