Home Physical Sciences Thermodynamic Analysis on the Minimum of Oxygen Content in the Deoxidation Equilibrium Curve in Liquid Iron
Article Open Access

Thermodynamic Analysis on the Minimum of Oxygen Content in the Deoxidation Equilibrium Curve in Liquid Iron

  • Pei-Wei Han EMAIL logo , Pei-Xian Chen and Shao-Jun Chu
Published/Copyright: June 12, 2015

Abstract

The minimum of oxygen content in the deoxidation equilibrium in liquid iron was thermodynamically analyzed in the present paper. Two criteria were developed to determine the existence of the minimum. The first criterion was 0xγMM+yγOMmin(x/4.606[%M]ex2,(xeMM+yeOM)2/3.474x) with xeMM+yeOM<0, or xγMM+yγOM<0. And the second criterion was (xeMO+yeOO)+y/2.303[%O]ex>0. The criteria in terms of first-order activity interaction parameters were the special case of present thermodynamic analysis with neglecting the second-order activity interaction parameters. They were not fit for the case of xeMM+yeOM>0, in which case the criteria in terms of second-order activity interaction parameters should be taken into account to determine the existence of the minimum. The value 0.11 of eSiSi was smaller based on the existence of the minimum for the Fe-O-Si system. It was guaranteed that the minimum value of oxygen content on the deoxidation equilibrium curve existed at silicon content 20 mass%, when the value 0.32 of eSiSi was chosen, and the second-order activity interaction coefficients γSiSi and γOSi satisfied the condition γSiSi+2γOSi=1.54×103.

Introduction

The deoxidation equilibrium curve can be obtained depending on the calculation of the deoxidation equilibrium in liquid steel by using thermodynamic data for different elements, as shown in Figure 1 [1, 2]. According to Figure 1, there are minima for the Fe-O-Al, Fe-O-C and Fe-O-Cr systems, but no minimum exists in the deoxidation equilibrium curve of Fe-O-Si system.

Figure 1: Deoxidation equilibrium curve of different elements in iron melt at 1,873 K.
Figure 1:

Deoxidation equilibrium curve of different elements in iron melt at 1,873 K.

Silicon–oxygen equilibrium in liquid iron was experimented by Gokcen and Chipman in the early 1950s [3]. Their results showed that there was no minimum for the Fe-O-Si system, even the silicon content up to 15 mass%. However, it was reported recently by Shibaev et al. [4] that the minimum occurred at 20 mass% Si, who measured the solubility of oxygen in iron–silicon melts in equilibrium with silica at 1,873 K, as shown in Figure 2. The criteria for a minimum developed by Hone et al. [1] and Wei [2] are not fit for the condition that the deoxidizer content is fairly high. The aim of the present study is to correct the criteria for the existence of the minimum of oxygen content in the deoxidation equilibrium curve.

Figure 2: Oxygen solubility in Fe–Si melts in equilibrium with silica at 1,873 K.
Figure 2:

Oxygen solubility in Fe–Si melts in equilibrium with silica at 1,873 K.

Thermodynamic analysis

The reaction between deoxidizer and oxygen can be represented as follows:

(1)x[M]+y[O]=MxOy
(2)m=aMxaOy

where x and y are stoichiometric coefficients, [M] and [O] are metallic deoxidizer and oxygen in liquid iron, respectively, m is the reciprocal of equilibrium constant, aM and aO are the activities of [M] and [O]. By replacing aM and aO with fM[%M] and fO[%O], respectively, eq. (3) is obtained.

(3)lgm=xlgfM+xlg[%M]+ylgfO+ylog[%O]

The dependence of fM and fO on composition can be expressed as eqs (4) and (5) [5].

(4)lgfM=eMM[%M]+eMO[%O]+γMM[%M]2+γMO[%O]2+γMM,O[%M][%O]
(5)lgfO=eOO[%O]+eOM[%M]+γOO[%O]2+γOM[%M]2+γOM,O[%M][%O]

Neglecting the higher order terms of [%O] and the second-order terms γMM,O[%M][%O] and γOM,O[%M][%O], eqs (4) and (5) reduce to eqs (6) and (7), respectively.

(6)lgfM=eMM[%M]+eMO[%O]+γMM[%M]2
(7)lgfO=eOO[%O]+eOM[%M]+γOM[%M]2

Substitution of eqs (6) and (7) into eq. (3) yields

(8)lgm=[%M](xeMM+yeOM)+[%O](xeMO+yeOO)+[%M]2(xγMM+yγOM)+xlog[%M]+ylog[%O]

If there is a minimum, the [%O] versus [%M] curve should meet the following two conditions:

(9)d[%O]d[%M]=0
(10)d2[%O]d[%M]2>0

Equation (11) is obtained by differentiating eq. (8) and further rearranging:

(11)d[%O]d[%M]=xeMM+yeOM+2[%M](xγMM+yγOM)+x2.303[%M]xeMO+yeOO+y2.303[%O]=0

With the first derivative zero, the second derivative of eq. (8) is

(12)d2[%O]d[%M]2=2(xγMM+γrOM)+x2.303[%M]2xeMO+yeOO+y2.303[%O]>0

According to eq. (11), eq. (13) is obtained.

(13)xeMM+yeOM+2[%M](xγMM+yγOM)+x2.303[%M]=0

If xγMM+yγOM=0, eq. (13) is a linear equation, to ensure the existence of a positive real solution for it, the following condition should be satisfied:

(14)xeMM+yeOM<0

However, eq. (13) is a quadratic equation if xγMM+yγOM0. Equation (15) meets the condition that the quadratic equation has real solutions. And the corresponding real solutions are given in eq. (16).

(15)(xeMM+yeOM)23.474x(xγMM+yγOM)
(16)[%M]=(xeMM+yeOM)±(xeMM+yeOM)23.474x(xγMM+yγOM)4(xγMM+yγOM)

For the case of xγMM+yγOM<0, there is always a positive real solution for eq. (13) no matter whether xeMM+yeOM is positive or negative. And eq. (15) is always satisfied if xγMM+yγOM<0.

But when it comes to the case of xγMM+yγOM>0, eq. (14) should be satisfied to ensure the existence of positive real solutions for eq. (13). In this case, there are two positive real solutions. This is similar to the work of St. Pierre [6], but in which the constraints were not discussed such as xγMM+yγOM>0 and the following eqs (18) and (19).

Equation (12) is needed to determine whether the positive real solution of eq. (13) is the minimum in the deoxidation equilibrium curve. Let [%M]ex be the positive real solution of eq. (13) and [%O]ex be the corresponding oxygen content obtained by substitution of [%M]ex into eq. (8). Hence, eq. (12) can be written as

(17)2(xγMM+yγOM)+x2.303[%M]ex2xeMO+yeOO+y2.303[%O]ex>0

For the case of xγMM+yγOM0, since the denominator term of eq. (11) is positive, the denominator term of eq. (17) follows that

(18)(xeMO+yeOO)+y/2.303[%O]ex>0

For the case of xγMM+yγOM0, it is reasonable that the smaller solution (xeMM+yeOM)(xeMM+yeOM)23.474x(xγMM+yγOM)4(xγMM+yγOM) of eq. (13) should be the minimum, since the oxygen in liquid iron decreases initially as the content of deoxidizer increases. As a consequence, the denominator term of eq. (11) should be positive. It means that it is also positive for the denominator term of eq. (17). Therefore the denominator term of eq. (11) should be greater than zero and satisfy the following equation:

(19)xγMM+yγOM<x/4.606[%M]ex2

From the thermodynamic analysis above, two criteria are developed for determining the existence of the minimum in the deoxidation equilibrium curve.

The first criterion is 0xγMM+yγOMmin(x/4.606[%M]ex2,(xeMM+yeOM)2/3.474x) with xeMM+yeOM<0, or xγMM+yγOM<0.

And the second criterion is (xeMO+yeOO)+y/2.303[%O]ex>0 where min(A,B) is the minimum between A and B. It should be noted that there is always a minimum no matter whether xeMM+yeOM is positive or negative, in the case of xγMM+yγOM<0 and (xeMO+yeOO)+y/2.303[%O]ex>0.

Discussion

The limitation of the criteria for the minimum in terms of first-order activity interaction parameters

Since the amount of deoxidizer addition is not large in steelmaking practice in general, the first-order activity interaction parameters are adopted to determine the existence of the minimum in the deoxidation equilibrium curve [1, 2, 7, 8]. In other words, γMM and γOM are both regarded as zero. The first criterion for the minimum is xeMM+yeOM<0 in this condition, which is consistent with eq. (14). It can be noted that this criterion is the special case of xγMM+yγOM=0 in Section 2.

In the work of Wei [2], the second criterion is given as

(20)xeMO+yeOO<0

It is considered that eq. (20) is questionable. Of course, xeMO+yeOO>0 satisfies eq. (18). On the other hand, it is not guaranteed that the value of (xeMO+yeOO) is smaller than that of y/2.303[%O]ex in the case of xeMO+yeOO<0. Thus, the second criterion is eq. (18) rather than eq. (20). [%O]ex is obtained by eqs (21) and (22) in terms of first-order criterion parameters. Then, the existence of the minimum is confirmed by substitution of [%O]ex into eq. (18).

(21)[%M]ex=x2.303(xeMM+yeOM)
(22)lgm=[%M]ex(xeMM+yeOM)+[%O]ex(xeMO+yeOO)+xlog[%M]ex+ylog[%O]ex

It is worth noting that only the first criterion is not sufficient for the existence of the minimum. It is not persuasive enough to confirm the existence of a minimum in the work of St. Pierre and Blackburn [7] and the work of Feldman and Kirkaldy [8].

The value of activity interaction parameter can be judged by the existence of the minimum in the deoxidation equilibrium curve. It can be concluded that the value −0.13 of eOC is incorrect [2], since there is no minimum for the Fe-O-C system with eOC=0.13. This is the contribution of the criteria in terms of the first-order activity interaction parameters.

However, there are limitations of criteria for the minimum in terms of first-order activity interaction parameters.

First, it is only suitable to the range that lgfMM and lgfOM have a linear relationship with [%M] due to the neglect of second-order interaction parameters. As [%M] increases, it is also lager for the degree of lgfMM and lgfOM deviating from the linear relationship with [%M]. The influence of Ti elements on the oxygen solubility in liquid iron at 1,873 K is shown in Figure 3 [8], in which the full line is the theoretical solubility curves calculated in terms of first-order activity interaction parameters, and the dots are experimental data from the work of Fruehan [9]. It can be seen from Figure 3 that the position of the minimum is accurately predicted by the criteria in terms of the first-order activity interaction parameters for some elements. But the deviation between the calculated value and the experimental value is larger as [%M] increases.

Figure 3: Influence of Ti element on the oxygen solubility in liquid iron at 1,873 K (reproduced from Ref. [8]).
Figure 3:

Influence of Ti element on the oxygen solubility in liquid iron at 1,873 K (reproduced from Ref. [8]).

Second, it is concluded that there is no minimum in the case of xeMM+yeOM>0 in Refs. [1] and [2]. Thus the minimum is considered to be inexistence for the Fe-O-Si system. According to Figure 2, there is a minimum in the deoxidation equilibrium curve at high silicon content. Thus, even if xeMM+yeOM>0, there is still a minimum when it satisfies the condition of xγMM+yγOM<0 and (xeMO+yeOO)+y/2.303[%O]ex>0.

The deoxidation equilibrium curve of silicon in liquid iron

There is a disagreement in the literature on the existence of the minimum in the deoxidation equilibrium curve for Fe-O-Si system. With the criteria in terms of first-order activity interaction parameters, there is a minimum with the data from Refs. [10] and [11], but not with data from Refs. [1] and [2]. The data used on calculation and the results are shown in Tables 1 and 2, respectively. The values of [%Si]ex and [%O]ex in Table 2 are obtained by eqs (21) and (22), respectively.

Table 1:

Data on calculation of silicon deoxidation equilibrium curve at 1,873 K.

xyeSiSieOSieSiOeOOlg mNote
120.32−0.14−0.25−0.20−4.66Refs [1, 2]
0.11−0.131−0.23−0.20−4.64Refs [10, 11]
Table 2:

Verification on the minimum of Fe-O-Si system.

eSiSi+2eOSieSiO+2eOO[]/mass%[]/ppmy/2.303[%O]exThe minimum
0.04−0.65No*
−0.152−0.632.86471.85 × 102Yes**

The difference of results in Table 2 is mainly caused by the value of eSiSi. There is a big disagreement about activity interaction parameters obtained by different research, due to the experimental methods, experimental conditions, the high temperature control, etc.

There is a minimum at silicon content 2.86 mass% for the Fe-O-Si system with data from Refs [10] and [11]. It differs from the experimental results as shown in Figure 2. According to Figure 2, the minimum of oxygen content is 1.4 ppm at the silicon content 20 mass% [4].

Depending on the above thermodynamic analysis, the first criterion for a minimum is derived from eq. (13). Substitution of [%Si]ex=20 and eSiSi+2eOSi=0.152 into eq. (13) yields

(23)γSiSi+2γOSi=3.26×103

Then application of the second criterion, substitution of eSiO+2eOO=0.63, [%Si]ex=20 and [%O]ex=1.4×104 into eq. (17) yields

(24)γSiSi+2γOSi<5.43×104

Obviously, the value of eq. (23) is out of the range of eq. (24). It can be concluded that the value 0.11 of eSiSi is smaller.

The data of eSiSi=0.32 and eOSi=0.14 from Refs. [1] and [2] and the effect of second-order activity interaction parameters are taken into account. Substitution of [%Si]ex=20 and eSiSi+2eOSi=0.04 into eq. (13) yields

(25)γSiSi+2γOSi=1.54×103

The second criterion is also satisfied by substituting [%O]ex=1.4×104 and eSiO+2eOO=0.65 into eq. (18). It means that it determines the existence of the minimum at silicon content 20 mass% theoretically.

It is interesting to note that the values of γSiSi and γOSi are −0.0021 and 0, respectively, in Ref. [10]. Their sum is similar to the calculated result as shown in eq. (25) in the present paper.

Conclusions

Two criteria are developed for determining the existence of the minimum in the deoxidation equilibrium curve in liquid iron in the present paper. The first criterion is 0xγMM+yγOMmin(x/4.606[%M]ex2,(xeMM+yeOM)2/3.474x) with xeMM+yeOM<0 or xγMM+yγOM<0. And the second criterion is (xeMO+yeOO)+y/2.303[%O]ex>0.

The criteria in terms of first-order activity interaction parameters are the special case of present thermodynamic analysis with neglecting the second-order activity interaction parameters. They are not fit for the case of xeMM+yeOM>0. In this case, the criteria in terms of second-order activity interaction parameters are taken into account to determine the existence of the minimum.

The value 0.11 of eSiSi is smaller based on the existence of the minimum in the deoxidation equilibrium curve for the Fe-O-Si system. It is guaranteed that the minimum value of oxygen content on the deoxidation equilibrium curve existed at silicon content 20 mass%, when the value 0.32 of eSiSi is chosen, and the second-order activity interaction coefficients γSiSi and γOSi satisfy the condition γSiSi+2γOSi=1.54×103.

Funding statement: Funding: The authors gratefully acknowledge research funding from the National Natural Science Foundation of China (No. 51274030).

List of symbols

γii

The second-order self-interaction parameters of element i in liquid iron on 1 mass% solution standard

γij

The second-order interaction parameter of element j on element i in liquid iron on 1 mass% solution standard

γii,j

The cross-product second-order interaction coefficient of elements i and j on element i in liquid iron on 1 mass% solution standard

ai

Activity of element i in liquid iron

eii

The first-order self-interaction parameters of element i in liquid iron on 1 mass% solution standard

eij

The first-order interaction parameter of element j on element i in liquid iron on 1 mass% solution standard

fi

Activity coefficient of element i on 1 mass% solution standard

fij

The influence of element j on fi on 1 mass% solution standard

[i]

Elements i in liquid iron

[%i]

Weight percentage of element i in liquid iron

m

The reciprocal of equilibrium constant

min(A,B)

The minimum between A and B

MxOy

Deoxidizing product

[%M]ex

The positive real solution of eq. (13)

[%O]ex

The corresponding oxygen content obtained by substitution of [%M]ex into eq. (8)

x

Stoichiometric coefficient in deoxidizing product

y

Stoichiometric coefficient in deoxidizing product

References

[1] M.Hone, S.Houot and M.Rigaud, Can. Metall. Q., 13 (1974) 619623.Search in Google Scholar

[2] S.K.Wei, Thermodynamics of Metallurgical Process, Shanghai Scientific and Technical Publishers, Shanghai (1980), pp. 288291.Search in Google Scholar

[3] N.A.Gokcen and J.Chipman, Trans. TMS-AIME, 194 (1952) 171181.Search in Google Scholar

[4] S.S.Shibaev, P.V.Krasovskii and K.V.Grigorovitch, ISIJ Int., 45 (2005) 12431247.Search in Google Scholar

[5] C.H.P.Lupis and J.F.Elliott, Acta Metall., 14 (1966) 529538.Search in Google Scholar

[6] G.R.St. Pierre, Metall. Trans. B, 8 (1977) 215217.10.1007/BF02657649Search in Google Scholar

[7] G.R.St. Pierre and R.D.Blackburn, Trans. TMS-AIME, 242 (1968) 24.Search in Google Scholar

[8] S.E.Feldman and J.S.Kirkaldy, Can. Metall. Q., 13 (1974) 625630.10.1179/cmq.1974.13.4.625Search in Google Scholar

[9] R.J.Fruehan, Metall. Trans., 1 (1970) 34033410.10.1007/BF03037871Search in Google Scholar

[10] G.K.Sigworth and J.F.Elliott, Met. Sci., 8 (1974) 298310.10.1179/msc.1974.8.1.298Search in Google Scholar

[11] J.X.Chen, Steelmaking Common Chart and Data Manual, Metallurgical Industry Press, Beijing (1984), pp. 513515.Search in Google Scholar

Received: 2014-11-23
Accepted: 2015-3-12
Published Online: 2015-6-12
Published in Print: 2016-4-1

©2016 by De Gruyter

This article is distributed under the terms of the Creative Commons Attribution Non-Commercial License, which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.

Downloaded on 12.12.2025 from https://www.degruyterbrill.com/document/doi/10.1515/htmp-2014-0217/html
Scroll to top button