Cu Evaporation from Liquid Iron Alloy in Stream

Abstract

The accumulation of copper in steel scrap is becoming an increasingly problematic issue in the steelmaking industry. Accordingly, the present study was undertaken to investigate the removal of copper from a liquid Fe–Cu alloy via tapping under vacuum. Furthermore, the impact of surface-active components sulfur and oxygen was examined. For this purpose, four Fe–0.5 wt% Cu alloys with varying oxygen and sulfur contents were melted and subsequently poured at a pressure of 100 Pa. The findings indicate that alloys with low oxygen and sulfur content exhibited enhanced copper evaporation. Additionally, the evaporation of other tramp metals, including manganese, phosphorus, and tin, was observed, and the influence of sulfur and oxygen on this process was discussed. Furthermore, the vacuum treatment conditions for copper evaporation in industrial settings were explored.

1. Introduction

The quantity of CO₂ emissions from fossil fuels and industry increased from 23 to 37.5 billion metric tons between 1990 and 2022 [1]. While the steelmaking sector is responsible for approximately 7% of global emissions, it is important to note that the blast furnace–basic oxygen furnace route, responsible for about 70.8% of the world steel production, with 2.33 million tons of CO₂ emissions per ton of crude steel, is the largest emitter of CO₂ globally [2,3]. In contrast, an electric arc furnace (EAF), which utilizes steel scrap as raw material, produces the lowest carbon emissions with 0.66 million tons of CO₂ [4]. Consequently, the scrap melting process is favorable for reducing CO₂ emissions in the steelmaking sector.

Steel scrap is a source of iron and nonferrous metals, and its use in steelmaking will increase due to its environmental benefits. Furthermore, the global stock of steel in use is projected to reach 100 billion tons by 2050 [5]. However, an insufficient sorting process of steel scrap has led to an increase in the content of copper, which is undesirable, with the annual rise in its concentration being a concern [6]. The primary challenges associated with copper are attributed to its lower melting point (1084.85 °C) in comparison to iron (1537.8 °C) and the unlimited solubility of copper in liquid iron [7]. According to the iron-copper phase diagram, no intermetallic phase is observed between iron and copper [8]. Consequently, during the hot rolling process, copper exists in the liquid phase and penetrates into the austenite grain boundaries, contributing to the development of hot fracture [9]. Furthermore, according to the Ellingham diagram, iron exhibits a stronger affinity to oxygen than copper, which precludes the formation of copper oxides in the presence of iron [10].

There are various methods to remove copper, such as transferring copper to different metals such as lead, which is insoluble in iron [11,12], increasing the sulfur concentration in steel to 1.5 wt% S to further remove sulfur and copper using sodium compounds [13], treating steel with Na₂S- or Na₂SO₄-based fluxes [14]. However, due to the complexity of these processes and the need for additional costs, these methods have not been used in the industry.

Copper vaporization is one of the promising methods. This method is based on the difference between the vapor pressure of iron and that of copper or its compound. Therefore, it can be easily integrated into the industry since vacuum treatment is an integral part of the production of quality steels. The existing vacuum treatment processes are shown in Figure 1. Various authors have studied Cu evaporation from the liquid melt [15,16,17,18,19]. Savov and Janke [15] studied the evaporation of copper and tin from iron-based melts at reduced pressure and found that copper evaporation is higher at lower pressures. The same result was obtained by Labaj [20] who studied the effect of pressure on the process of copper removal from liquid iron-based alloys by evaporation. The effect of sulfur and oxygen on the evaporation of copper has also been studied by various researchers. Chen et al. [18] studied the evaporation of copper from liquid iron under reduced pressure and found that the evaporation rate was increased by adding sulfur to the melt. Ohno [21] studied the kinetics of copper evaporation from iron alloys under reduced pressure and concluded that sulfur and oxygen hardly affect the evaporation rate of copper at 1873 K. On the contrary, Jung and Kang [17] found out that sulfur in the liquid iron can both increase and decrease the copper evaporation rate depending on the concentration of dissolved sulfur in the melt. Wei et al. [22] studied the decopperization of molten steel with ZnAl₂O₄ and found that high oxygen content significantly affected the copper evaporation due to the surfactant behavior of oxygen. In addition, oxygen reduces the activity coefficient of copper in the molten iron [23,24].

As mentioned above, copper evaporation can be easily integrated into the steel making industry. Vacuum degassing processes such as VD/VOD and RH/RH-OB are the most common in the industry. However, their technical characteristics should be considered for the evaporation of copper from liquid melt. It is known that the reaction surface area [26,27] and the content of elements such as sulfur [16] and oxygen [22,28], which are surface-active substances, can affect the evaporation of copper from the melt surface. Furthermore, the presence of liquid slag on the melt will impede copper evaporation due to the isolation of the melt surface area. Consequently, the decarburization process in RH/RH-OB, where oxygen content is high, and vacuum treatment in VD/VOD, where slag remains on the surface, will result in ineffective copper removal. To eliminate these issues, a method for copper evaporation from the liquid metal stream during vacuum tapping was utilized in this study. For this purpose, four Fe-Cu alloys with different oxygen and sulfur concentrations were tapped under vacuum. The effect of oxygen and sulfur concentration on the evaporation of copper from the liquid metal stream during tapping under vacuum was investigated.

2. Materials and Methods

A vacuum induction melting furnace (VIM-12; ALD Vacuum Technologies GmbH, Hanau, Germany) was used for the experiments in this work. The detailed description of the furnace can be found elsewhere [15]. To produce Fe–0.5 wt% Cu alloy with varying oxygen and sulfur contents, approximately 25 kg of pure Armco iron (chemical composition presented in

Table 1. Chemical composition of pure Armco iron in ppm; Fe is in balance.

C	Mn	P	S	Cu	N	Sn	Si	Al	Cr	Mo	Ni	Co
70	450	30	30	90	46	20	30	30	260	30	210	30

) was used as the initial material.

The melt was alloyed with pure copper, FeS (36.47 wt% S), and analytically pure Fe₂O₃ to achieve the required oxygen and sulfur content, as well as electrolytic aluminum to deoxidize the melt. The objective was to produce four alloys with different sulfur and oxygen contents:

• Alloy 1—high oxygen and low sulfur;
• Alloy 2—high in both oxygen and sulfur;
• Alloy 3—low oxygen and low sulfur;
• Alloy 4—low oxygen content and high sulfur.

The starting materials were charged and melted in an alumina crucible under a protective argon atmosphere at the system pressure of 16 kPa. After melting, the first sample was taken by immersion of a steel cup into the melt, and the chemical composition was analyzed. Then, if additional alloying was required, alloying elements were added through a vacuum lock without breaking the vacuum in the chamber. After homogenization at constant input power, the final sample was taken. The temperature was controlled throughout the process using a type B thermocouple. The chemical composition of the samples before tapping is shown in

Table 2. Chemical composition of Fe-Cu alloy before tapping in ppm.

Alloy	Cuinitial	Sinitial	Oinitial
1	5030	41	584
2	5250	380	624
3	5230	25	51
4	5160	244	51

. It should be noted that the chemical composition presented in Table 2 includes only the element of interest excluding other metals (C, Si, Mn, P, Cr, Mo, Ni, Al, Sn, B, Zr, As, N). Then, the vacuum pumps were turned on, the pressure was reduced to 100 Pa, and the temperature was controlled at 1600 °C. After reaching 100 Pa, the melt was poured into a water-cooled Cu mold installed in the vacuum chamber. The duration of tapping was 140 s for each melt to obtain a thin jet of melt for proper vacuum treatment. After tapping, the samples were cooled under a protective atmosphere and analyzed.

The chemical composition of Fe-Cu alloys before and after the experiments was analyzed utilizing a ICP-OES-Spectrometer 5100 VDV Dual View (Agilent Technologies, Santa Clara, CA, USA), and G4 Icarus (Bruker AXS, Karlsruhe, Germany) was used to determine sulfur and G8 Galileo (Bruker AXS, Karlsruhe, Germany) was used to determine oxygen content in the samples.

3. Results

3.1. The Evaporation Mechanism of Copper from the Melt

The evaporation of copper from liquid melt treated under vacuum can be described by the following steps:

1.

Convective transport of copper from the bulk of the liquid alloy to the diffusion boundary layer in the melt,

2.

Diffusion of copper through the liquid phase interface,

3.

Evaporation of copper at the gas–metal interface,

4.

Diffusion of copper vapor through the gas phase interface,

5.

Transport of the vapor through the bulk to the site of condensation,

6.

Condensation.

However, considering the fact that steps 1, 5, and 6 occur very fast, steps 2, 3, and 4 are the main determining steps for the evaporation process. According to Fischer et al. [29], the apparent evaporation rate constant can be expressed by Equation (1):

$$k_{Cu}={\displaystyle \frac{1}{\frac{1}{k_{Cu}^L+k_{Cu}^V+k_{Cu}^G}}}$$

(1)

where $k_{Cu}^L{,k}_{Cu}^V\mathrm{a}\mathrm{n}\mathrm{d}{k}_{Cu}^G$ denote the liquid phase mass transfer coefficient (m·s⁻¹), rate coefficient of the vaporization reaction (m·s⁻¹), and gas phase mass transfer coefficient (m·s⁻¹). Therefore, in order to increase the evaporation rate, it is necessary to understand the factors that affect it.

The liquid phase mass transfer of copper to the free surface in liquid metal is mainly controlled by diffusion and can be described by Equation (2), proposed by Machlin [30]:

$$k_{Cu}^L=2(C-C_S)\sqrt{{\displaystyle \frac{2·D_{Cu}·ʋ}{\pi ·r·h^2}}}$$

(2)

where C is the concentration of copper on the melt, C_S is the copper concentration at the melt surface, D_Cu is the diffusion coefficient of copper in liquid iron (m²·s⁻¹), ʋ is the average radial velocity of metal at the free surface (m·s⁻¹), r is the radius of the liquid metal surface, and h is the melt height.

The gas phase mass transfer coefficient $k_{Cu}^G$ of copper evaporating in the argon atmosphere can be described by Equation (3) [29]:

$$k_{Cu}^G={\displaystyle \frac{D_{Cu-Ar}}{\delta }}·{\displaystyle \frac{{\gamma }_{Cu}^0·P_{Cu}^0}{\rho }}·{\displaystyle \frac{M_{Fe}}{RT}}$$

(3)

where $D_{Cu-Ar}$ is the interdiffusivity of copper vapor and Ar gas (m²·s⁻¹), δ is the thickness of the gas phase diffusion boundary level (m), ${\gamma }_{Cu}^0$ is the Raoultian activity coefficient at infinite dilution, $P_{Cu}^0$ is the vapor pressure of pure liquid copper (Pa), $\rho$ is the density of liquid iron (g·cm⁻³), $M_{Fe}$ is the atomic mass of Fe (55.85 g·mol⁻¹), R is the gas constant (8,314,510 Pa·cm³·mol⁻¹·K⁻¹), and T is the temperature (K).

According to Richardson [31], at a constant gas temperature, $D_{Cu-Ar}$ is linearly dependent on the reciprocal of the pressure in the reaction chamber 1/P:

$$D_{Cu-Ar}={\displaystyle \frac{C_2}{P}}$$

(4)

where C₂ is a constant independent of the chamber pressure (Pa·m²·s⁻¹).

Evaporation at the gas–metal interface can be expressed by Equation (5) using Hertz–Knudsen–Langmuir’s equation for the evaporation of pure liquid metals in a vacuum [29]:

$$k_{Cu}^V={10}^{-3}·{\displaystyle \frac{\alpha ·{\gamma }_{Cu}^0·P_{Cu}^0}{\rho }}·\sqrt{{\displaystyle \frac{M_{Fe}^2}{2\pi ·RT·M_{Cu}}}}$$

(5)

where α is the surface evaporation coefficient (is 1 for liquid metals), R is the gas constant of 8314 Pa·L·mol⁻¹·K⁻¹, and M_Cu is the atomic weight of copper (63.546 g·mol⁻¹).

3.2. Results of Experiments

Table 3. Chemical composition of Fe-Cu alloy after tapping in ppm.

Alloy	Cuend	Send	Oend	ΔCu	ΔS	ΔO
1	4980	38	572	50 (0.99) *	3	12
2	5220	306	593	30 (0.53)	74	31
3	5020	23	22	210 (4.02)	2	29
4	5100	234	38	60 (1.16)	10	13

* Values in brackets are the Cu removal in wt%.

shows the chemical composition of alloys 1–4 after tapping at a pressure of 100 Pa. The evaporation process took place during the tapping of approximately 25 kg of liquid Fe-Cu alloy into the water-cooled copper mold. The tapping time was approximately 140 s. Table 3 shows that Cu, S, and O evaporated from the melt during tapping. The highest copper evaporation was obtained from alloy 3. Convsersely, the worst results were obtained with alloy 2. Figure 2 shows the effect of sulfur and oxygen content on copper vaporization. It can be clearly seen that as the sulfur and oxygen content increased, the evaporation of copper from the melt decreased.

3.3. Effect of Sulfur and Oxygen on the Vaporization

As evidenced in the previous chapter, sulfur and oxygen had a pronounced effect on copper evaporation from the melt. Both sulfur and oxygen are well known for their capacity to act as surface-active agents, which accumulate on the surface of the liquid iron alloy. Even a low content of oxygen and sulfur can significantly influence the surface tension of liquid iron [28,32]. Consequently, the surface-active elements can decelerate the evaporation of copper from the melt by blocking the surface, as illustrated in Figure 3. This phenomenon has also been interpreted by Langmuir’s ideal adsorption, which represents the degree of surface coverage [23,33].

The influence of sulfur and oxygen on the evaporation process has been corroborated by numerous researchers. Jung and Kang [17] conducted a study on Fe-Cu alloys with varying carbon and sulfur contents to investigate the evaporation mechanism of copper. Their findings indicated that the overall evaporation of copper at low sulfur contents (less than 0.03 wt%) is slower than that observed in the absence of sulfur, due to the blockage of the surface by sulfur molecules. However, high sulfur content (more than 0.25 wt%) was beneficial for copper evaporation through the formation of CuS_(g). This effect was similarly reported by Chen et al. [18]. Wei et al. [22] reported that high oxygen content greatly prevented the copper evaporation from Fe-Cu alloys due to the surface-active nature of oxygen in the liquid iron. Furthermore, oxygen reduces the activity coefficient of copper [23]. Consequently, it is essential to minimize the sulfur and oxygen content in the liquid Fe-Cu alloy in order to facilitate optimal copper evaporation.

3.4. Evaporation of Other Component from the Liquid Melt

As previously stated in Section 2, the liquid Fe-Cu alloy contained not only copper, sulfur, and oxygen, but also other impurities in the melt, which were not listed in the chemical compositions before and after the vacuum treatment. However, a comparison of the chemical compositions before and after tapping revealed that the behavior of some impurities was worthy of attention. The content of these impurities before and after tapping is listed in

Table 4. Content of other impurities in Fe-Cu alloy before and after tapping in ppm.

Alloy	Pinitial/Pend	ΔP, wt%	Mninitial/Mnend	ΔMn, wt%	Sninitial/Snend	ΔSn, wt%
1	127/92	27.56	171/153	10.53	41/32	21.95
2	156/120	23.08	132/118	10.61	49/41	16.33
3	132/120	9.09	146/104	28.77	45/30	33.33
4	141/130	7.80	133/121	9.02	43/38	11.63

.

As illustrated in Figure 4, not only did copper evaporate from the melt, but also phosphorus, manganese, and tin. The evaporation of tin from the iron-based alloy is also a topic worthy of further investigation, as tin enters the furnace with steel scrap and cannot be oxidized or reduced using a conventional refining method. Consequently, evaporation represents a promising method for the removal of tin from the liquid melt, a topic that has been studied by various authors, including [15,16,34,35]. In the present experiments, the highest tin evaporation was observed in alloy 3, where approximately one-third of tin was evaporated from the melt during the slow tapping of the steel melt. In contrast, in other experiments, a lower amount of tin was evaporated, as in the case of copper due to the presence of surface-active elements such as oxygen and sulfur. A similar effect was observed in the case of manganese evaporation from the liquid melt. This suggests that tin and manganese evaporate from the melt as Sn_g and Mn_g, and the presence of sulfur did not lead to the formation of SnS and MnS, which could increase the evaporation rate of the elements from the melt. Conversely, the evaporation rate of phophorus from the liquid melt was approximately three times higher in the case of alloys 1 and 2 than in alloys 3 and 4. This may be attributed to the higher oxygen content in the melt, which may have led to the formation of gaseous PO_x. In their study of the vaporization of melts in the SnO-ZnO-P₂O₅ system, Shilov et al. observed the evaporation of PO and PO₂ as gaseous species [36]. Consequently, the evaporation procedure can be employed to reduce the content of other impurities in the melt. These results are summarized in

Table 5. Optimal conditions for the evaporation of trump elements in dependence of oxygen and sulfur content in the steel.

	Unkilled/ Desulfurized	Unkilled/ Non-Desulfurized	Killed/ Desulfurized	Killed/ Non-Desulfurized
Cu	− *	−	+ **	+ ***
P	+	+	−	−
Mn	−	−	+	−
Sn	−	−	+	−

* conditions are not optimal; ** conditions are optimal; *** conditions are optimal only at [S] > 0.25 wt% according to [16].

.

4. Discussion of Results

A number of researchers [15,18,20,26,37] have conducted studies on the evaporation of copper from a liquid iron alloy under reduced pressure. The objective of these studies was to investigate a method for decopperization that can be integrated on an industrial scale. All experiments in the past were conducted in different types of VIM furnaces, with the melt being treated in a crucible under reduced pressure up to 0.06 Pa. Thereby, it was assumed that this method can be integrated into the industry in the form of VD/VOD. However, the decopperization in VD/VOD would entail a prolonged period of treatment, higher heat lost from the melt in the ladle, and higher refractory wear, which could result in additional operational costs. Furthermore, in VD/VOD, the evaporation process is constrained by the limited surface area of the metal exposed to the vacuum, as it is covered by slag. The findings of the aforementioned studies [15,20] indicate that at pressures above 10 Pa, the evaporation of copper is predominantly influenced by gas phase mass transfer, rather than liquid mass transfer or vaporization at the gas–metal interface. However, this is only applicable to the experimental conditions of their works, where the melt was in the crucible and was treated for a long period without slag.

In order to facilitate the evaporation of copper from the melt in VD/VOD, it is necessary to remove additional slag in order to expose the metal surface to the vacuum. This will result in an increased temperature loss of the melt and a prolonged treatment time due to the limited surface area available for evaporation, which is constrained by the diameter of the ladle. In the case of the RH/RH-OB method, copper evaporation is possible when the melt has undergone decarburization; however, the limiting factor will be the small portion of the melt presented to the vacuum. The optimal approach is the pouring stream treatment illustrated in Figure 1, as the steel has already been killed and desulfurized. However, in vacuum ingot casting, there is a risk due to the low casting temperature, which increases the likelihood of the early solidification of the steel. Additionally, as evidenced in the work of Savov and Janke [15], low temperatures are insufficient for copper evaporation.

Therefore, the experimental procedure presented in this work involved a short-time treatment of the melt during tapping under reduced pressure. The main advantages of this approach are as follows:

• The increased surface layer of Fe-Cu alloy presented for copper evaporation, benefiting the rate coefficient of the vaporization reaction;
• The decreased height of the liquid melt benefits the liquid phase mass transfer;
• An appropriate pressure of the system for operational conditions in steelmaking,
• The absence of a time-consuming treatment.

This was achieved through the slow tapping of approximately 25 kg of liquid melt at 100 Pa. However, as previously discussed, this method is highly sensitive to the presence of surface-active elements, such as oxygen and sulfur, which can harm the evaporation of copper from the surface layer. Therefore, it is crucial to control their content in the melt for the proper vacuum treatment.

The integration of this process on an industrial scale can be implemented by steelmakers who (1) deoxidize the melt just before tapping from the melting unit, which means that the oxygen content is low enough for proper evaporation, and (2) produce steel grades with low sulfur content. As one example, steelmakers may recover valuable metals from the liquid slag into the metal by adding reducing agents just before tapping by producing steel grades with low sulfur content. As reported by Nadif et al. [38], the sulfur content in the converter vessel just before tapping can be as low as 40 ppm at ArcelorMittal Flat Carbon Western Europe when producing low-sulfur steel grades. Therefore, considering these conditions, this method can be well adapted for copper evaporation just by adding a vacuum treatment of the stream during tapping from the vessel to the ladle.

Based on the results obtained for alloy 2, where 210 ppm of copper was evaporated from deoxidized and desulfurized melt, it can be assumed that, on average, approximately 21 kg or 0.021 wt% of copper can be evaporated per 100 t of melt during the tapping of crude steel. Therefore, this method has the potential to solve the problem of copper accumulation in metal scrap.

5. Conclusions

This study proposes copper evaporation from liquid Fe-Cu alloy that can be integrated on an industrial scale. The method suggests vacuum treatment of the jet of liquid iron alloy during its tapping. The findings demonstrate that during the tapping of approximately 25 kg of liquid melt for 140 s, 210 ppm of copper was evaporated. This indicates the significant potential of this method for further solving the issue of copper accumulation in metal scrap. Furthermore, the impact of sulfur and oxygen on copper evaporation was investigated. It was demonstrated that surface-active elements, such as sulfur and oxygen, can significantly impede copper evaporation from liquid melt by blocking the surface. Furthermore, the evaporation of phosphorus, manganese, and tin from the liquid melt was observed. It was found that the surface-active sulfur and oxygen elements inhibited the evaporation of manganese and tin from the melt. However, the presence of oxygen was beneficial for phosphorus evaporation.