ISIJ International
Online ISSN : 1347-5460
Print ISSN : 0915-1559
ISSN-L : 0915-1559
Note
Conversion of Calcium Phosphide to Calcium Phosphate in Reducing Dephosphorization Slags by Oxygen Injection
Jae Hong ShinJoo Hyun Park
Author information
JOURNAL OPEN ACCESS FULL-TEXT HTML

2013 Volume 53 Issue 12 Pages 2266-2268

Details
Abstract

We tried to confirm the conversion of Ca3P2 to calcium phosphate by oxygen injection into the molten Reducing Dephosphorization (RDP) slag as well as to find the stoichiometry of calcium phosphate compound by the treatment. It was experimentally confirmed that the Ca3P2 was converted to beta-3CaO·P2O5 and 4CaO·P2O5 phases by oxygen injection treatment of the molten RDP slag by XRD analysis and thermodynamic estimation.

Study

High manganese steels (HMnS, 10 to 25 mass% Mn) are of interest due to their good mechanical properties including superior strength and good ductility.1,2,3) Accordingly, the demand for ultra low phosphorous manganese alloys such as FeMn and SiMn alloys has recently increased. However, the conventional oxidizing dephosphorization technique is not applicable to manganese alloys because silicon or manganese will be oxidized before phosphorous under oxidizing conditions.4,5,6,7,8) Therefore, the dephosphorization should be carried out under a strongly reducing atmosphere to produce a low phosphorus manganese alloys.

The mechanism of reducing dephosphorization (RDP) using lime-based highly basic fluxes has been reported by several authors,4,5,6,7,8,9,10,11) and recently reviewed and experimentally confirmed by the present authors in previous study.12) The equilibrium reaction of RDP by using lime-based flux is as follows:   

3( CaO ) +2[ P ]=( C a 3 P 2 ) +3[ O ] (1)

However, because the calcium phosphide (Ca3P2), which is a reaction product of RDP mechanism, is very active when it is exposed to moisture, the hazardous phosphine gas (PH3) evolves.5,9) The equilibrium reaction of the emission of phosphine gas is as follows:   

C a 3 P 2 ( s ) +3 H 2 O( g ) =3CaO( s ) +2P H 3 ( g ) (2)
  
C a 3 P 2 ( s ) +6 H 2 O( g ) =3Ca ( OH ) 2 ( s ) +2P H 3 ( g ) (3)

Phosphine gas is hazardous to environment as well as to human being.13,14) Inhalation is the most likely route of exposure to phosphine gas. Symptoms are non-specific and include irritation of the respiratory tract, headaches, dizziness, abdominal pain, sickness, and vomiting. Severe phosphine poisoning can cause convulsions, damage to the lungs, heart, liver and kidney, and death.13,14) Therefore, it is very important to suppress the evolution of phosphine gas by Eqs. (2) and/or (3) after reducing refining operations.

On the line of the above background, we reported in our previous work that the basicity (CaO/SiO2 ratio) of the slag after RDP process should be controlled to be lower than 1.35 to minimize the emission of phosphine gas during cooling of RDP slag even under highly wet conditions.15) When the basicity of RDP slag is greater than about 1.35, the lime and dicalcium silicate compounds precipitated during solidification, resulting in an increase in the emission rate of PH3 gas due to a dramatic increase in the reaction area occurred from a disintegration of slags. However, when the basicity of RDP slag is lower than about 1.35, the fluorite, cuspidine, and wollastonite compounds precipitated during cooling, resulting in a negligible amount of PH3 evolution because the reaction between Ca3P2 and H2O was confined to the surface of bulk slag.15)

Even though the above results are very useful to restrain the emission of phosphine gas from RDP slag during cooling, the problem due to phosphine emission at room temperature is still remained. Actually, the safety treatment of RDP slag by oxygen blowing was originally reported by Arato et al.16) and Katayama et al.17) They primarily investigated the dephosphorization efficiency of molten stainless steel by applying calcium alloy and CaC2–CaF2 flux, respectively, and proposed that oxygen blowing was effective for reducing PH3 emission from RDP slag. They claimed that some phosphorus in the RDP slag was vaporized during oxygen blowing treatment.

The above results unambiguously provided very useful technological information for making RDP slag harmless. Nevertheless, there was no experimental confirmation not only whether calcium phosphide (Ca3P2) was converted to calcium phosphate or not but also which kind of phosphate compound was formed, e.g. 4CaO·P2O5 (monoclinic), 3CaO·P2O5 (β-rhombohedral, α-monoclinic, α’-hexagonal), 2CaO·P2O5 (γ-phase, β-tetragonal, α-monoclinic), CaO·P2O5 (monoclinic), CaO·2P2O5 (monoclinic) and 2CaO·3P2O5 (monoclinic).18)

Therefore, in the present study, we tried to confirm the conversion of Ca3P2 to calcium phosphate by oxygen injection into the molten RDP slag as well as to find the stoichiometry of calcium phosphate compound by the treatment. The RDP slag (300 g), which was supplied from an industrial SiMn refining process, was initially melted in the graphite crucible (OD: 56 mm, ID: 50 mm, HT: 96 mm) at 1773 K using a high frequency induction furnace with a graphite heater under a purified Ar-3%H2 atmosphere. The composition of the slag is listed in Table 1.

Table 1. The composition of RDP slag supplied from industry (mass%).
SiO2CaOCaF2MnOPOthers
17.548.728.42.30.222.9

After the slag was melted, the power of the furnace was switched off, after which the slag was slowly cooled down. At this time, the gas was switched from Ar–H2 gas mixture to purified O2 gas, and then a stainless steel lance (OD: 6 mm, ID: 4 mm) for injecting O2 gas (flow rate=1.0 l/min) was inserted into the molten slag and kept 5 mm above the bottom of the crucible. Strong agitation of the molten slag by O2 gas injection resulted in fast attainment of the oxidation of RDP slag during cooling. The schematic diagram of the experimental apparatus is shown in Fig. 1. After cooling, the crystalline phases of the solidified slags were identified using an X-ray diffraction analysis (RAD-3C; Rigaku, Tokyo, Japan).

Fig. 1.

Schematic diagram of the experimental apparatus.

The experimental results of XRD analyses for the as-received RDP slag and oxygen-treated slag are shown in Figs. 2 and 3, respectively. Both of the slags were mainly composed of calcium fluoride (CaF2), wollastonite (CaSiO3), cuspidine (Ca4Si2F2O7), dicalcium silicate (Ca2SiO4), iron oxide (Fe2O3) and manganese oxide (MnO) based on the reference peak from a JCPDS.19,20) The RDP slag was mainly CaO–CaF2 system coupled with small amounts of SiO2, Fe2O3, and MnO, those were formed during SiMn refining process.15) The cuspidine phase is believed to crystallize during solidification of the RDP slag.15)

Fig. 2.

XRD pattern for the as-received reducing dephosphorization (RDP) slag from Mn smelting company.

Fig. 3.

XRD pattern for the oxygen-treated RDP slag in the present experiments.

It is very interesting in Figs. 2 and 3 that the calcium phosphide (Ca3P2) was identified in the as-received RDP slag,21) whereas the β-tricalcium phosphate (β-3CaO·P2O5) which is the most stable phase among the tricalcium phosphate compounds at room temperature and tetracalcium phosphate (4CaO·P2O5) were detected at the expense of Ca3P2 phase in the oxygen-treated slag. The Ca3P2 in molten RDP slag was oxidized to 3CaO·P2O5 phase at high temperature by oxygen injection. Also, the tetracalcium phosphate is believed to form by the oxidation of Ca3P2 and the reaction with excess CaO in RDP slag as shown in Eqs. (4) and (5), respectively.   

C a 3 P 2 ( s ) +4 O 2 ( g ) =3CaO P 2 O 5 ( s ) (4)
  
C a 3 P 2 ( s ) +4 O 2 ( g ) +CaO( s ) =4CaO P 2 O 5 ( s ) (5)

The standard Gibbs free energy changes of these reactions, which were calculated using FactSageTM6.3 software, are shown in Fig. 4.22,23) The Gibbs free energy changes have significantly negative values even in room temperatures. However, because a difference in the Gibbs free energy changes of conversion reactions (Eqs. (4) and (5)) is relatively small, it is not easy to confirm which one is more dominant than the other. Therefore, 3CaO·P2O5 and 4CaO·P2O5 compounds are commonly detected after safety (oxygen) treatment of RDP slag as shown in Fig. 3.

Fig. 4.

Gibbs free energy changes of the phosphide-to-phosphate conversion reactions. Calculated using FactSageTM6.3.

Consequently, it was experimentally confirmed that the Ca3P2 was converted to β-3CaO·P2O5 and 4CaO·P2O5 phases by oxygen injection treatment of the molten RDP slag by XRD analysis and thermodynamic estimation.

Acknowledgment

The authors express their sincere appreciation to Prof. Dong Joon MIN, Department of Materials Science and Engineering, Yonsei University, Seoul, Korea, and the anonymous reviewers for their helpful discussion on this topic.

References
 
© 2013 by The Iron and Steel Institute of Japan
feedback
Top