Phosphorus Distribution Ratio between Multi Phase CaO–FeO_t–SiO₂–P₂O₅ (6%–13%) Slags with MP Near Hot Metal Temperature and C-saturated Molten Iron at 1573 K

Abstract

Aimed at the pretreatment for hot metal with varied phosphorus contents, by applying the indirect approach, the distribution ratio of phosphorus between multi phase slags with 6% to 13% P₂O₅, whose melting points were close to the hot metal temperature, and carbon-saturated molten iron at 1573 K, , were measured. The results demonstrated was determined by the composition of multi phase slags and that at different levels of the P₂O₅ content, the combination of the (FeO) content and the mass ratio of solid particles to molten slag with which reached the maximum was different. In correspondence with high-P content hot metal, under the condition of about 13% P₂O₅, a decrease in (%FeO) and a simultaneous increase in solid particles/molten slag mass ratio was benefit to the increase of , and the maximum reached 130. However, when P₂O₅ content was about 10% and 6%, could only reach 107.98 and 60.7 at most, respectively. The results of microcosmic measurements showed that in quenched multi phase slags, there exist two kinds of precipitates with different phosphorus contents, i.e. the low-P phase (6C₂S-mC₃P, m=0.6294~1.3400) and the high-P phase (C₂S-nC₃P, n=1.1064~1.4303), and also 2CaO·SiO₂ phase with no phosphorus and CaO–SiO₂–Fe_tO slag.

1. Introduction

At the present, that is practically used in the industrial production is basically all the homogeneous flux of hot metal pretreatment dephosphorization. It is also usual to add fluorspar to promote melting. However, adding fluorspar will increase slag volume and pollute the environment, and the final slag cannot be used as phosphate fertilizer as well.¹⁾ Therefore, it is urgent to develop high efficient environmental conscious dephosphorization fluxes with less or no fluorine.

A multi phase dephosphorization flux contains molten slag and solid particles at the hot metal temperature. For different compositions of the molten slag, the solid particles can be CaO, 2CaO·SiO₂, 3CaO·P₂O₅, 4CaO·P₂O₅, 2CaO·SiO₂-3CaO·P₂O₅ solid solution, 3CaO·P₂O₅·CaF₂ and (Fe, Mg)O. It is noticed early that the phosphorus in BOF slag largely presents in dicalcium silicate, producing a silicophosphate solid solution.²⁾ In industrial production of the MURC process with high [Si] hot metal treated, Nippon Steel Corporation, Japan, found out that the generated 2CaO·SiO₂ particles could dissolve P₂O₅, which was favorable to the dephosphorization reaction.³⁾

Uchida et al. suggested that it was possible to substantially reduce slag volume by applying multi phase dephosphorizing fluxes based on the Condensed Phase Law.⁴⁾ This is because in this case the number of phases is just equal to the number of components, and if temperature is moreover kept constant the freedom of the system will become zero, in other words, no matter how to change the concentration of FeO and P₂O₅, their activities will all maintain unchanged, and thus it is no longer necessary as using homogeneous fluxes to add extra FeO or CaO in later stages of the dephosphorization process.

There have been extensive theoretical studies on multi-phase fluxes of hot metal dephosphorization. For instance, Hasegawa and Iwase⁵⁾ and Jeoungkiu IM et al.⁶⁾ respectively measured the activity of Fe_xO and P₂O₅ in the 3 three-phase regions of CaO–P₂O₅–Fe_xO system. Tsukihashi et al.^7,8,9,10,11) systematically investigated the formation reaction of phosphate compound in multi phase flux and the mass transfer processes between CaO/2CaO·SiO₂ particles and molten slag. In 1982 at the earliest, Ito et al.¹²⁾ measured the phosphorus distribution between 2CaO·SiO₂ and CaO–SiO₂–FeO–Fe₂O₃ molten slag. Suito and Inoue¹³⁾ applied EMPA-EDX techniques to the measurement of the distribution of phosphorus between CaO/2CaO·SiO₂ grains and molten slag, and determined the phosphorus distribution between multi phase slag and Fe-0.5%P alloy at 1560°C¹⁴⁾ and that between multi phase slag and Fe-5%P alloy at 1400°C.¹⁵⁾

In the Western Hubei Province of China, there exists a 1.7 billion reserve of high phosphorus iron ores, it is desired to enlarge their dosage in the blast furnace. But phosphorus contents in the produced hot metal will be very high. This paper is intended to apply multi phase fluxes with melting points near hot metal temperature to phosphorus removal of the hot metal, especially the high-P hot metal, to measure the distribution ratio of phosphorus between multi phase slags and C-saturated molten iron at 1573 K, and to provide the basis for determining reasonable recipes of dephosphorization fluxes.

2. Experimental

2.1. Experimental Apparatus

Tests were carried out in a high temperature tubular electric resistance furnace, refer to Fig. 1. The furnace tube was made of high purity sintered alumina. Heating elements of lanthanum chromite were used. It is allowed to place three crucibles in the furnace tube at one time, between them was positioned a Pt-30%Rh/Pt-6%Rh thermocouple for temperature measurement. The temperature was controlled by a microcomputer PID controller of P909 type, the error was ±1 K.

Fig. 1.

Set-up for phosphorus equilibrium experiment.

In the test, high purity argon gas of 99.999 vol.% was purged for protecting, the flow rate was 1 to 2 liters per minute. Before entering the furnace, it passed through a desiccant for moisture removal, and then through in order copper flakes and magnesium foils at 823 K for residual oxygen removal. The equilibrium oxygen partial pressure was calculated by using FactSage 6.1 thermodynamic software: the p_O2 that is determined by reaction 4Cu(s)+O₂(g)=2Cu₂O(s) is equal to 1.745×10⁻⁹ Pa, and the p_O2 that is determined by reaction 2Mg(s)+O₂(g)=2MgO(s) is equal to 8.361×10⁻⁶¹ Pa.

In order to make it sure that no air penetrates into the furnace tube, a slight positive pressure was maintained throughout the experimental process.

2.2. Experimental Materials

The specimens were made up of FeO, SiO₂, Ca₃(PO₄)₂, CaO and so on, weighing 30 grams each. CaO was prepared by decomposing pure CaCO₃ at 1273 K for one hour. FeO was prepared by decomposing analytically pure oxalate (FeC₂O₄·2H₂O) at 1123 K for one hour in a high purity argon.¹⁶⁾ The other agents were all of analytical purity and phosphorus was added in the form of Ca₃(PO₄)₂. Before weighing, CaO and Ca₃(PO₄)₂ should be heated at 1273 K to remove H₂O and CO₂.

The composition of multi phase dephosphorization fluxes were determined in accordance with a requirement that their melting points are near the hot metal temperature, this is because that this kind of fluxes completely melt at 1673 K and precipitate 2CaO–SiO₂ fine particles as temperature decreases to 1573 K, as a result the preparation and addition of fine materials (CaO or 2CaO·SiO₂) can be omitted. Moreover, among all sorts of multi phase fluxes with a same composition of molten slag, the mass ratio of solid particles to molten slag, abbreviated as SP/MS mass ratio hereinafter, of these fluxes is the smallest, and therefore their kinetic condition for the phosphorus removal reaction is better, the requirement for stirring energy is lower, and hence it can be not only expect to achieve higher practical phosphorus removal degrees but also to further reduce the production cost.

P₂O₅ content in the dephosphorization final slag is increased with increasing the initial phosphorus content in hot metal. According to the literature information,^17,18) P₂O₅ content in the final slag for low-P, medium-P and high-P hot metal pretreatment is about at a level of 6%, 10% and 13%, respectively. By taking the four points of A, B, C and D on the 1400°C liquidus contour of CaO–SiO₂–Fe_tO phase diagram as base slags, refer to Fig. 2, and adding three different amounts of P₂O₅, totally twelve specimens of three groups for the phosphorus distribution experiment were prepared. Their compositions are summarized in Table 1. In the order of A→B→C→D, the FeO content was gradually increased while SP/MS mass ratio was gradually decreased. The SP/MS mass ratio was calculated based on the liquidus contour in the CaO–SiO₂–FeO ternary system according to the Lever Principle.

Fig. 2.

Chemical composition of slag samples for P distribution experiment and the phosphorus distribution ratio of slags with 10% P₂O₅.

Table 1. Composition of slag samples for P distribution experiment at 1573 K.

Samples	A1-0	A2-0	A3-0	B1-0	B2-0	B3-0	C1-0	C2-0	C3-0	D1-0	D2-0	D3-0
Solid, %	16.22	16.22	16.22	12.24	12.24	12.24	8.33	8.33	8.33	8.16	8.16	8.16
Liquid, %	83.78	83.78	83.78	87.76	87.76	87.76	91.67	91.67	91.67	91.84	91.84	91.84
SP/MS, -	0.1936	0.1936	0.1936	0.1395	0.1395	0.1395	0.0909	0.0909	0.0909	0.0889	0.0889	0.0889
(FeO)0, %	18.80	18.00	17.40	28.20	27.00	26.10	37.60	36.00	34.80	47.00	45.00	43.50
(P2O5)0, %	6	10	13	6	10	13	6	10	13	6	10	13

Note: SP/MS is the mass ratio between solid particles and molten slag, -

The crucibles and iron foils were made of high purity iron and their chemical compositions are given in Table 2.

Table 2. Chemical composition of pure iron crucibles and iron foils (mass%).

	C	Si	P	Mn	S
Crucibles	0.008	0.0020	0.006	0.017	0.005
Iron foils	0.070	0.0032	0.021	0.036	0.001

A direct measurement of phosphorus distribution between high FeO slag and high C liquid iron is not able to be carried out because the composition of slag and iron will change due to the reaction between FeO and C. For this reason, it is considered to conduct measurement indirectly by applying an approach which makes slag reach equilibrium with other appropriate metal phases. Iwasaki et al.¹⁹⁾ proposed at the earliest an indirect approach, i.e. firstly to measure phosphorus distribution ratio between slag and solid iron and then to convert it into that between slag and C-saturated liquid iron. This approach has been accepted and improved by some researchers.^6,20,21)

The indirect approach utilized in present study requires two equilibrations for each experiment, i.e. the equilibration between thoroughly mixed chemicals and the crucible material (Reaction: (Fe₂O₃)+Fe(s)⇌(FeO)) and the phosphorus equilibrium between the slag, which was acquired with the first equilibration, and iron foils. In a continuous type operation, the system is in a stable state, with the temperatures, concentrations and volumes in the reactor all keeping constant. In this study, the crucibles were under Ar flowing atmosphere of 10^–61 Pa oxygen partial pressure, which was thermodynamically estimated, and the system could reach equilibrium only if the reaction time is long enough. Werme and Lundh²⁰⁾ specially designed three series of tests and demonstrated that the reduction of Fe₂O₃ by Fe in the first equilibration is completed within 8 hours and the phosphorus distribution approaches equilibrium in approximately 10 hours.

The first equilibration was carried out by heating the crucible containing a dephosphorizing flux for one hour at 1673 K, then lowering the temperature to 1573 K and holding the crucible at this temperature for 11 hours. The increased temperature during the first hour permitted the slag having a composition just inside the two phase region to melt uniformly and precipitate upon cooling to 1573 K. Having been equilibrated for 12 hours, the crucibles were removed from the furnace and immediately quenched in water to prevent oxidation.

After the first equilibration, the slag was removed from the crucible, crushed and placed in a second iron crucible together with some 1 gram of iron foil (thickness 0.1 mm). In the second equilibration, the crucible was held in the furnace at 1573 K for 12 hours. After the experiments, the crucible was water quenched to prevent oxidation of the materials and to facilitate microcosmic examinations of the slag structure after preparation. Special care was taken in cleaning slag adhered on foils by polishing and a subsequent ultrasonic cleaning in an acetone solution.

After the second equilibration experiment, the chemical composition of slags was measured by an X-ray fluorescence spectrometer (XRF), while phosphorus dissolved in iron was analyzed by the phosphovanadomolybdate yellow photometry.

For quenched multi-phase slags, the microscopic morphology was observed and the concentration of different elements in different precipitates were analyzed by the FESEM (Nova 400 Nano, FEI)/BSED-EDS method while the mineralogical composition was examined by the XRD method (D/max 2500 type, 18 kW and Cu target, a product of Rigaku, with a scanning rage of 5 to 80 degrees and scanning speed of 4 degrees per minute).

3 Results and Discusstions

3.1. Phosphorus Distribution Ratio between Slag and C-saturated Molten Iron

The phosphorus distribution ratio between slag and solid iron is defined as follows:   

$$L_{\text{P}}^{\text{Fe(}\alpha ,\gamma \text{)}}=\frac{(\%\text{P})}{{[\%\text{P}]}^{\text{Fe(}\alpha ,\gamma \text{)}}}$$(1)

The phosphorus distribution ratio between slag and C-saturated molten iron is defined as follows:   

$$L_{\text{P}}^{\text{Fe}-\text{C}}=\frac{(\%\text{P})}{{[\%\text{P}]}^{\text{Fe}-\text{C}}}$$(2)

In Eqs. (1) and (2), (%P) represents the bulk concentration of phosphorus in multi-phase slag. Through a series of thermodynamic conversion, the following two correlations are obtained:⁶⁾   

$$\frac{{[\%\text{P}]}^{\text{Fe}(\alpha )}}{{\text{[\%P]}}^{\text{Fe-C}}}=0.779\mathrm{\hspace{0.17em} }\mathrm{\hspace{0.17em} }([\mathrm{P}]\mathrm{\hspace{0.17em} }\mathrm{i}\mathrm{n}\mathrm{\hspace{0.17em} }\mathrm{s}\mathrm{o}\mathrm{l}\mathrm{i}\mathrm{d}\mathrm{\hspace{0.17em} }\mathrm{i}\mathrm{r}\mathrm{o}\mathrm{n}>0.41\%)$$(3)

  

$$\frac{{[\%\text{P}]}^{\text{Fe}(\gamma )}}{{\text{[\%P]}}^{\text{Fe-C}}}=0.413\mathrm{\hspace{0.17em} }\mathrm{\hspace{0.17em} }([\mathrm{P}]\mathrm{\hspace{0.17em} }\mathrm{i}\mathrm{n}\mathrm{\hspace{0.17em} }\mathrm{s}\mathrm{o}\mathrm{l}\mathrm{i}\mathrm{d}\mathrm{\hspace{0.17em} }\mathrm{i}\mathrm{r}\mathrm{o}\mathrm{n}<0.41\%)$$(4)

It can be learned from the above correlations that, based on whether the P content in solid iron exceeds 0.41 pct., $L_{\text{P}}^{\text{Fe}-\text{C}}$ is equal either to $L_{\text{P}}^{\text{Fe}(\alpha )}$ times by 0.779 or to $L_{\text{P}}^{\text{Fe}(\gamma )}$ times by 0.413. The calculated $L_{\text{P}}^{\text{Fe}-\text{C}}$ and the chemical composition of slag and iron are listed in Table 3. In this table, (Fe_tO)=(FeO)+0.9×(Fe₂O₃).

Table 3. The phosphorus distribution ratios and chemical compostion of slags and iron foils.

No.	Composition of slag and metal, mass%							$L_{\text{P}}^{\text{Fe}(\gamma )}$	$L_{\text{P}}^{\text{Fe}-\text{C}}$
	(FeO)	(Fe2O3)	(CaO)	(SiO2)	(P2O5)	(FetO)	[P]
A1	15.7	1.96	47.1	29.78	5.74	17.46	0.0640	39.15	16.17
A2	14.79	2.31	48.41	26.59	9.46	16.87	0.0178	232.06	95.84
A3	16.09	2.25	46.92	22.99	12.76	18.12	0.0177	314.77	130.00
B1	24.83	2.01	36.09	29.42	5.80	26.64	0.0496	51.07	21.09
B2	23.74	0.91	38.86	24.50	10.24	24.56	0.0171	261.45	107.98
B3	25.11	0.90	36.63	21.86	13.28	25.92	0.0279	207.82	85.83
C1	32.88	3.19	32.33	23.38	5.97	35.75	0.0181	144.02	59.48
C2	34.11	2.42	32.00	20.47	9.33	36.29	0.0221	184.33	76.13
C3	34.93	2.43	31.52	16.85	12.69	37.12	0.0351	157.85	65.19
D1	40.94	4.24	30.50	18.43	6.16	44.76	0.0183	146.97	60.70
D2	41.75	4.70	27.21	15.31	9.06	45.98	0.0253	156.34	64.57
D3	43.39	5.00	28.03	11.94	12.85	47.89	0.0337	166.49	68.76

3.2. Effect of the Solid Phase Concentration on Phosphorus Distribution Ratio between CaO-SiO₂-Fe_tO-10%P₂O₅ Slag and C-saturated Molten Iron

The phosphorus distribution ratio of multi phase slags with about 10% P₂O₅ were marked in Fig. 2, and also in this figure the phosphorus distribution ratio of homogeneous and multi phase slags with the same level of P₂O₅ content reported by Werme and Lundh²⁰⁾ were marked. It can be seen from Fig. 2 that at about a same Fe_tO content, $L_{\text{P}}^{\text{Fe}-\text{C}}$ of slags A2, B2, C2 and D2 are bigger than that of the homogeneous liquid slags in the region on the right side of the 1400°C liquidus contour but smaller than that of the multi phase slags with a higher SP/MS mass ratio in the region on the left side of the 1400°C liquidus contour. This demonstrated that the introduction of solid phases promotes the dephosphorization capacity of pretreatment fluxes and that a higher solid phase concentration is beneficial to the improvement of $L_{\text{P}}^{\text{Fe}-\text{C}}$.

3.3. Relationship between $L_{\text{P}}^{\text{Fe}-\text{C}}$ and Slag Composition for Multi Phase Slag with Various P₂O₅ Contents

The P₂O₅ content in Nos. 1, 2 and 3 groups of slag samples for phosphorus distribution experiment is 6%, 10% and 13%, respectively, corresponding respectively to the low-P, medium-P and high-P hot metal pretreatment, see Table 1. The measured phosphorus distribution ratios are plotted against SP/MS mass ratio (solid lines and legends) and also against FeO content (dash lines and blank legends), resulting in Fig. 3. In this figure, the cubic legends represent the slag samples of 6% P₂O₅, the circle legends represent the slag samples of 10% P₂O₅, and the upper triangle legends represent the slag samples of 13% P₂O₅.

Fig. 3.

Relationship between $L_{\text{P}}^{\text{Fe}-\text{C}}$ and the SP/MS mass ratio or (%FeO) of slag samples.

It can be seen from Fig. 3 that for the slag samples of 6% P₂O₅, $L_{\text{P}}^{\text{Fe}-\text{C}}$ shows a clear tendency of increasing with the decrease in SP/MS mass ratio or with the increase in (%FeO), the increasing speed of $L_{\text{P}}^{\text{Fe}-\text{C}}$ was very fast particularly in the SP/MS mass ratio range of 0.1395 to 0.0889 or in the (%FeO) range of 28.20 to 37.60, and that reached a maximum of 60.70 when SP/MS mass ratio was decreased to 0.0889 and (%FeO) was increased to 47.

It can also be seen from Fig. 3 that for the slag samples of 13% P₂O₅, there is a turning point of 0.0909 for SP/MS mass ratio: when smaller than this value $L_{\text{P}}^{\text{Fe}-\text{C}}$ decreased with the increase in SP/MS mass ratio but when bigger than this value $L_{\text{P}}^{\text{Fe}-\text{C}}$ increased with the increase in SP/MS mass ratio. And for these slags, similarly, there is also a turning point of 34.80% for (%FeO): when smaller than this value, $L_{\text{P}}^{\text{Fe}-\text{C}}$ rapidly decreased with the increase in (%FeO), but when bigger than this value, $L_{\text{P}}^{\text{Fe}-\text{C}}$ increased in a small extent with the increase in (%FeO). $L_{\text{P}}^{\text{Fe}-\text{C}}$ reached a maximum of 130 when SP/MS mass ratio was equal to 0.0889 and (%FeO) was equal to 47.

Figure 3 also shows that for the slag samples of 10% P₂O₅, $L_{\text{P}}^{\text{Fe}-\text{C}}$ reached a maximum of 107.98 when SP/MS mass ratio was equal to 0.1395 and (%FeO) was equal to 27.00, and that $L_{\text{P}}^{\text{Fe}-\text{C}}$ will be reduced when whether SP/MS mass ratio was lower or higher than 0.1395 or when whether (%FeO) was lower or higher than 27.00.

Different from the homogeneous dephosphorizing fluxes having only molten slag, the multi phase fluxes consist of FeO-containing liquid slag and solid 2CaO·SiO₂. The role of FeO is to oxidize P in hot metal (Reaction: 2[P]+8(FeO)=(3FeO·P₂O₅)+5[Fe]) while that of 2CaO·SiO₂ is to absorb generated P₂O₅ and form 2CaO·SiO₂-3CaO·P₂O₅ solid solution, leading to the decrease of the activity of P₂O₅ in molten slag and the promotion of oxidation of phosphorus in iron.^9,14) For phosphorus removal from high-P hot metal, the amount of generated P₂O₅ is large and thus the increase of added 2CaO·SiO₂ amount by enlarging SP/MS mass ratio can increase the amount of absorbed P₂O₅ amount, leading to a higher (%P₂O₅) and thus a higher $L_{\text{P}}^{\text{Fe}-\text{C}}$. Comparatively, for phosphorus removal from low-P hot metal, an increase in (%FeO) is benefit to the increase of oxygen partial pressure in the system, promoting the [P] oxidation, and also due to less P₂O₅ generated, the amount of 2CaO·SiO₂ required is smaller, as a consequence, the reasonable recipe of multi phase fluxes in this case should be a higher (%FeO) and a smaller SP/MS mass ratio.

The above analysis demonstrates that the higher the hot metal phosphorus content, the more beneficial the application of multi phase fluxes to the pretreatment dephosphorization: at a same flux consumption, the maximum $L_{\text{P}}^{\text{Fe}-\text{C}}$ corresponding to high-P hot metal can reach 130 (slag A3), the maximum $L_{\text{P}}^{\text{Fe}-\text{C}}$ corresponding to medium-P hot metal can reach 107.98 (slag B2), but the maximum $L_{\text{P}}^{\text{Fe}-\text{C}}$ corresponding to low-P hot metal can only reach 60.70 (slag D1), less than a half of the maximum $L_{\text{P}}^{\text{Fe}-\text{C}}$ for high-P hot metal. Higher phosphorus distribution ratios imply that it is possible to reduce the flux consumption, and it is thus not only to decrease the production cost, but also beneficial to the decrease of temperature drop during the phosphorus removal process.

3.4. Microstructure of Quenched Multi Phase Slags

Figure 4 shows the FESEM morphology of four quenched multi phase slags containing about 13% P₂O₅, Table 4 lists the elemental composition of all twelve quenched multi phase slags and the type of all phases, and Fig. 5 shows the XRD analysis results of selected multi phase slags. These microcosmic measurements were summarized and elaborated, and it is revealed that phosphorus distributes very unevenly in the multi phase slags, the P₂O₅ content is partial to three concentration ranges of 13.07% to 14.05%, 2.97% to 5.33% and 0%. The phases containing no phosphorus are 2CaO·SiO₂ and CaO–SiO₂–Fe_tO slag.

Fig. 4.

Morphology of quenched multi phase slags containing about 13% P₂O₅ (×1000). (a)- Slag A3, (b)- Slag B3, (c)- Slag C3, (d)- Slag D3.

Table 4. Elemental composition of quenched multi phase slags analyzed by EDS method and the estimated type of phases.

Slag	Point	Elemental composition, mass%					Phase
		O	Si	P	Ca	Fe
A1	1	37.46	14.81	2.97	38.23	6.53	6C2S+mC3P
	2	35.19	18.65	0	24.28	21.88	slag
B1	3	37.37	13.10	4.53	35.97	9.03	6C2S+mC3P
	4	35.28	17.82	0	21.71	25.18	slag
	5	39.20	7.00	13.07	39.38	1.34	C2S+nC3P
C1	6	39.90	12.37	4.64	36.31	6.78	6C2S+mC3P
	7	40.26	6.92	13.46	37.90	1.47	C2S+nC3P
	8	20.59	0.25	0	0.64	78.52	wustite
	9	36.53	16.43	0	21.65	25.39	slag
	10	43.23	11.82	6.00	33.07	5.88	6C2S+mC3P
D1	11	41.33	12.03	4.96	35.61	6.06	6C2S+mC3P
	12	21.47	0	0	0	78.53	wustite
	13	40.36	12.46	4.87	35.76	6.54	6C2S+mC3P
	14	34.37	16.37	0	20.70	28.56	slag
	15	38.35	7.11	13.32	39.73	1.50	C2S+nC3P
A2	16	41.551	13.10	4.42	34.13	6.84	6C2S+mC3P
	17	37.05	16.24	0	21.59	25.12	slag
B2	18	39.94	6.62	13.37	38.02	2.05	C2S+nC3P
	19	0	0	0	0	100	M.Fe
	20	35.26	16.38		19.74	28.61	slag
	21	39.19	12.80	5.33	35.60	7.07	6C2S+mC3P
C2	22	18.79	0	0	0.54	80.67	wustite
	23	36.69	18.07	0	22.83	22.41	slag
	24	39.27	6.64	13.10	38.73	2.26	C2S+nC3P
	25	0	0	0	0	100	M.Fe
D2	26	37.58	18.01	0	23.07	21.35	slag
	27	42.03	6.78	13.49	37.71	0	C2S+nC3P
	28	23.09	0	0	0	76.91	wustite
A3	29	38.70	13.25	5.29	36.15	6.61	6C2S+mC3P
	30	38.41	6.92	13.66	39.45	1.56	C2S+xC3P
	31	33.11	14.30	0	19.66	32.94	slag
B3	32	34.15	16.46	0	20.52	28.88	slag
	33	39.70	7.19	13.07	38.19	1.86	C2S+nC3P
	34	1.84	0.11	0	0.59	97.45	M.Fe
C3	35	0	0	0	0	100	M.Fe
	36	21.37	0	0	0	78.63	wustite
	37	35.35	17.87	0	16.28	30.50	slag
	38	38.20	6.47	14.05	38.74	2.53	C2S+nC3P
D3	39	21.56	0	0	0.65	77.78	wustite
	40	38.23	6.84	13.37	37.75	3.80	C2S+nC3P
	41	36.13	17.90	0	15.20	30.77	slag

Note: Oxygen results are of semi-quantitative analysis, and the type of phases are presumed according to Si, P, Ca and Fe concentrations. 6C₂S+mC₃P (m=0.6294~1.3400) represents the low-P phase while C₂S+nC₃P (n=1.1064~1.4303) represents the high-P phase.

Fig. 5.

XRD spectrum of some quenched multi phase slags.

Also, the existence of wustite and metallic iron phases were observed. Wustite probably originates from the precipitation of FeO from molten slag upon cooling. Some wustite contained a little Ca (0.54% to 0.64%) and Si (0.25%). The shape of wustite was usually dendrite crystals, but when the FeO content was higher there possibly were wustite precipitates of grain shape. The precipitation of bright white metallic iron is probably because that when FeO content in molten slag is higher the disproportionation reaction (4FeO(s)=Fe₃O₄(s)+Fe(s)) has taken place for a part of FeO during quenching.¹⁶⁾ This phenomenon was also observed by PAHLEVANI et al.²²⁾ The metallic iron in point 34 also contained 0.11% Si and 0.55% Ca.

The presence of Ca₂FeO₅ (srebrodolskite) was also observed. It may originate from the CaO·FeO layers around CaO particles, which would absorb CaO from the molten slag and be oxidized by water upon quenching.

The chemical formula of phospho-calcium silicate is Ca₅(PO₄)₂SiO₄ or 2CaO·SiO₂-3CaO·P₂O₅ (C₂S+C₃P), and its phosphorus content is equal to 12.86%. The phosphorus contents of the high-P phase range from 13.07% to 14.05% and it is, therefore, to express its chemical formula as C₂S+C₃P, n=1.1064 to 1.4303. The chemical formula of mineral Ca₁₅(PO₄)₂(SiO₄)₆ can also be written as 6C₂S+C₃P and its phosphorus content is equal to 4.62%. The phosphorus contents of the low-P phase range from 2.97% to 5.33% and it is thus to express its chemical formula as 6C₂S+mC₃P, m=0.6294 to 1.3400. Fix et al.²³⁾ reported that at steelmaking temperature 3CaO·P₂O₅ and 2CaO·SiO₂ can form silicophosphate solid solution ((C₂S+C₃P)_ss) in almost the whole concentration range. Comparatively in this work, however, it is revealed that at hot metal temperature silicophosphate solid solution obviously has two concentration regions of low P and high P.

It is also found that (C₂S+C₃P)_ss commonly contained a little Fe (1.47% to 7.07%), and that the Fe content in the low-P phase was higher while the Fe content in the high-P phase was lower. The reason is possibly because that there is a CaO–FeO layer on the surface of 2CaO·SiO₂ particles which absorbed P₂O₅ from molten slag.⁷⁾

4. Conclusions

By using the indirect approach, the distribution ratio of phosphorus between CaO–FeO_t–SiO₂–P₂O₅ (6%–13%) multi phase slags with melting points near hot metal temperature and C-saturated molten iron at 1573 K, $L_{\text{P}}^{\text{Fe}-\text{C}}$, were measured, and the phase composition of quenched multi phase slags were studied with various microcosmic measuring techniques, leading to the following conclusions:

(1) Different from the homogeneous fluxes having only liquid slag, among the factors influencing the dephosphorization ability of multi-phase fluxes, represented by $L_{\text{P}}^{\text{Fe}-\text{C}}$, there is not only the content of ferrous iron that oxides phosphorus in iron, but also the amount of 2CaO·SiO₂ particles that absorb generated P₂O₅.

(2) The slag composition with which $L_{\text{P}}^{\text{Fe}-\text{C}}$ reaches a maximum changes along with the level of P₂O₅ content in slag. When the P₂O₅ content was at 13% level (corresponding to high-P hot metal pretreatment), increasing the SP/MS mass ratio and decreasing the (%FeO) are beneficial to the increase of $L_{\text{P}}^{\text{Fe}-\text{C}}$. In comparison with its application to medium-P and low-P hot metal pretreatment, it is possible to obtain larger $L_{\text{P}}^{\text{Fe}-\text{C}}$ for the application of a multi phase slag to high-P hot metal pretreatment, implying a potential of the decrease in slag volume and temperature drop during the process.

(3) In the (C₂S+C₃P)_ss precipitated from quenched multi phase slags of hot metal pretreatment, 2CaO·SiO₂ and 3CaO·P₂O₅ do not dissolve into each other in almost entire concentration range, but there exist apparently two kinds of minerals with different phosphorus contents, i.e. the low-P phase 6C₂S+mC₃P (m=0.6294~1.3400) and the high-P phase C₂S+nC₃P (n=1.1064~1.4303). Nurse et al. reported some of the equilibrated phases in the system Ca₂SiO₄–Ca₃P₂O₈ at high temperatures did not survive quenching, and in this study, high-temperature microscopy and X-ray analysis were not applied for the phase analysis. It is therefore unlikely that the two kinds of minerals with different phosphorus contents are equilibrated phases at 1573 K. It is speculated that they are precipitated during quenching.

Acknowledgments

Thanks are going to the Ministry of Science and Technology, China, for financial support to this research. The authors also want to thank Professor Yan JIN and Professor Wei XIONG for their invaluable discussions and proposals throughout the work.

References

• 1)   J.  Diao,  B.  Xie,  Y.  Wang and  X.  Guo: ISIJ Int., 50 (2010), 768.
• 2)   H.  Suito,  Y.  Hayashida and  Y.  Takahashi: Tetsu-to-Hagané, 63 (1977), 1252.
• 3)   S.  Kitamura,  K.  Yonezawa,  Y.  Ogawa and  N.  Sasaki: Ironmaking Steelmaking, 29 (2002), 121.
• 4)   Y.  Uchida,  M.  Iwase,  A.  Mclean,  K.  Takahashi,  Y.  Kikuchi and  K.  Wakimoto: Proc. Asia Steel Int. Conf. 2000, The Chinese Society for Metals, Metallurgical Industry Press, Beijing, (2000), 61.
• 5)   M.  Hasegawa and  M.  Iwase: Proc. Sec. Baosteel Biennial Academic Conf., Baosteel, Metallurgical Industry Press, Beijing, (2006), 210.
• 6)   J.  Im,  K.  Morita and  N.  Sano: ISIJ Int., 36 (1996), 517.
• 7)   T.  Hamano,  S.  Fukagai and  F.  Tsukihashi: ISIJ Int., 46 (2006), 490.
• 8)   S.  Fukagai,  T.  Hamano and  F.  Tsukihashi: ISIJ Int., 47 (2007), 187.
• 9)   X.  Yang,  H.  Matsuura and  F.  Tsukihashi: ISIJ Int., 49 (2009), 1298.
• 10)   X.  Yang,  H.  Matsuura and  F.  Tsukihashi: ISIJ Int., 50 (2010), 702.
• 11)   X.  Gao,  X.  Yang,  H.  Matsuura and  F.  Tsukihashi: Proc. Asia Steel 2012, The Chinese Society for Metals, Metallurgical Industry Press, Beijing, (2012). 1.
• 12)   K.  Ito,  M.  Yanagisawa and  N.  Sano: Tetsu-to-Hagané, 68 (1982), 342.
• 13)   R.  Inoue and  H.  Suito: ISIJ Int., 46 (2006), 174.
• 14)   H.  Suito and  R.  Inoue: ISIJ Int., 46 (2006), 180.
• 15)   R.  Inoue and  H.  Suito: ISIJ Int., 46 (2006), 188.
• 16)   Y.  Zhou,  Y.  Gao,  X.  Ma,  X.  Zheng,  M.  Wang and  B.  Wang: J. Wuhan Univ. Sci. Technol. (Mater. Sci. Ed.), 36 (2013), 383.
• 17)   T.  Moriya,  Y.  Matsumura,  K.  Katohgi,  M.  Sato and  A.  Mori: Tetsu-to-Hagané, 72 (1986), S1001.
• 18)   A.  Werme: Proc. 67th Steelmaking Conf., AIME, Warrendale, PA, (1984), 309.
• 19)   K.  Iwasaki,  N.  Sano and  Y.  Matsushita: Tetsu-to-Hagané, 67 (1981), 536.
• 20)   A.  Werme and  P.  Lundh: Scand. J. Metall., 16 (1987), 33.
• 21)   J.  Zhou,  X.  Bi and  F.  Yang: Ironmaking Steelmaking, 41 (2014), 298.
• 22)   F.  Pahlevani,  S.  Kitamura,  H.  Shibata and  N.  Maruoka: ISIJ Int., 50 (2010), 822.
• 23)   W.  Fix,  H.  Heymann and  R.  Heinke: J. Am. Ceram. Soc., 52 (1969), 346.