In-Line Co-Processing of Stainless Steel Pickling Sludge Using Argon Oxygen Decarburization Slag Bath: Behavior and Mechanism

By inergency On Feb 26, 2024

[ad_1]

1. Introduction

China’s burgeoning stainless steel industry has led to the substantial generation of stainless steel pickling sludge (SSPS), estimated at 750,000 tons annually, or 2.5 wt%~5.0 wt% of the nation’s stainless steel output [1]. Predominantly, domestic manufacturers use H₂SO₄ or hybrid acids for surface treatment, producing high-sulfur sludge through lime neutralization, which contains S (5 wt%~9 wt%), Cr (3 wt%~5 wt%), and Fe (15 wt%~22 wt%) [2,3]. Classified as hazardous waste (HW17) by the “National Dangerous Waste Catalog (2016)” for its toxicity and corrosiveness [4], SSPS management is now tightly regulated under the “Control Law of Solid Waste Environmental Pollution” [5,6]. Therefore, managing SSPS within the plant and throughout the production line is of paramount importance and demands immediate focus.

However, in the disposal process of SSPS, the method most stainless steel companies adopt is landfilling or stacking. This undoubtedly increases the risk of heavy metal leaching and also leads to the wastage of land resources [1,2,3]. The best way to dispose of such hazardous solid waste as sludge is to make it harmless. The existing research on sound disposal mainly focuses on the separation and recovery of heavy metal Cr and physicochemical solidification treatment. High-temperature carbothermal reduction process is widely considered as a complete detoxification way to treat stainless steel pickling sludge. Li et al. [7] investigated the thermodynamics of sludge with carbon desulfurization to avoid S into the metal to recover valuable metals. Wang et al. [8] used sludge as raw material for carbothermal reduction to produce low-sulfur low-carbon Fe-Cr-Ni alloy, and the process also removed the S element first. Wu et al. [9] studied the effect of temperature, C/O ratio, and reduction time of low-temperature direct reduction–magnetic separation on the reduction of Cr in stainless steel sludge. Liu et al. [10] proposed a semi-molten method for sludge reduction recovery after studying the growth and migration trend of low-melting-point Fe-Cr-C droplets. The described reductive separation and recovery methodology is distinguished by two principal characteristics: firstly, the sulfur element is removed to avoid its incorporation into the metal. Secondly, a reduction technique is utilized to prepare a low-melting-point alloy, enabling metal recovery. However, this disposal strategy may result in the release of sulfur-rich gases, representing a grave environmental hazard. Furthermore, should the alloy be synthesized via direct reduction, it is likely to yield a metal product with an excessively high sulfur concentration for practical application. Zhao et al. [11] and Yang et al. [12] prepared glass ceramics with pickling sludge as a doping material and then solidified Cr. Pan et al. [13] and others used sludge and gypsum to prepare glass ceramics, first removing the sulfur in the mixture and then preparing glass ceramics. In the context of solidification and stabilization treatment, while a heightened level of safety performance is achieved, this approach results in the underutilization of Cr resources, coupled with the detrimental environmental impact due to the release of sulfur. Moreover, the intricacy of the treatment process results in suboptimal efficiency when tasked with managing sludge volumes of significant scale. Different from SSPS, the Cr concentration in argon oxygen decarburization (AOD) slag is relatively low, exempting it from classification as hazardous solid waste.

Presently, the utilization of AOD slag is notably limited [14,15,16]. Post-cooling, the slag exhibits high γ-C₂S content, leading to volume instability. Additionally, the slag contains Cr concentrations ranging from 0.1 to 3 wt%, presenting significant leaching risks of the heavy metal Cr(VI) [17,18]. These factors substantially limit the direct application of AOD slag in supplementary cementitious materials. Certain studies [19,20,21] have demonstrated the carbonation and solidification treatment of AOD slag, converting γ-C₂S into calcium carbonate, thereby reducing expansion, enhancing cementitious properties, and lowering leaching toxicity. This offers significant theoretical support for the application of AOD slag as a supplementary cementitious material. While carbonation is a feasible approach, an unresolved issue identified in these studies is the presence of heavy metal Cr. Direct immobilization of Cr in cementitious materials not only wastes metal resources but also poses potential environmental hazards. Furthermore, SSPS and AOD slag are two waste products simultaneously produced in stainless steel plants. Their joint disposal process may hold greater research and application potential, an aspect yet unreported.

This study introduces a novel process utilizing high-temperature molten AOD slag baths for treating carbon-containing pickling sludge briquettes (C-SSPS). The method aims to separate and recover valuable metals such as Cr and Fe from SSPS and AOD slag, while also stabilizing sulfur in the slag to reduce SO₂ emissions. The process further achieves detoxification of the mixed AOD slag and SSPS, enabling the use of Cr-free AOD slag in cementitious materials. Additionally, the study delves into the reduction mechanisms within the slag bath treatment, providing significant insights for industrial production.

3. Results and Discussion

A.: Thermodynamic model of pre-reduction and slag-bath smelting

The slag-bath co-processing process proposed and studied in this paper involves two processes: (1) low-temperature pre-reduction of C-SSPS and (2) subsequent high-temperature slag-bath smelting of the briquettes. The selective reduction of SSPS is based on the reaction of metal oxides and sulfates [23,24,25,26,27]. These reactions describe the thermodynamic properties of the carbothermal reduction and mineral phase formation of single species contained in sludge.

C(s) + 1/4Fe₃O₄(s) = CO(g) + 3/4Fe(s)

(6)

C(s) + Fe₃O₄(s) = CO(g) + 3FeO(s)

(7)

C(s) + 1/3Fe₂O₃(s) = CO(g) + 2/3Fe(s)

(8)

C(s) + NiO(s) = CO(g) + Ni(s)

(9)

C(s) + 1/3Cr₂O₃(s) = CO(g) + 2/3Cr(s)

(10)

C(s) + 1/2CaSO₄(s) = CO₂(g) + 1/2CaS(s)

(11)

C(s) + 1/3CaSO₄(s) = 1/3CO₂(g) + 2/3CO(g) + 1/3CaS(s)

(12)

C(s) + 2CaSO₄(s) = CO₂(g) + 2SO₂(g) + 2CaO(s)

(13)

Fe₂O₃(s) + Cr₂O₃(s) = FeCr₂O₄(s)

(14)

C(s) + 1/4FeCr₂O₄(s) = 1/4Fe(s) + 1/2Cr(s) + CO(g)

(15)

C(s) + 1/3MgCr₂O₄(s) = 1/3MgO(s)+ 2/3Cr(s) + CO(g)

(16)

MgO(s) + FeCr₂O₄(s) = MgCr₂O₄(s) + FeO(s)

(17)

Al₂O₃(s) + MgFe₂O₄ (s) = MgAl₂O₄ (s) + Fe₂O₃(s)

(18)

MgO(s) + Al₂O₃(s) = MgAl₂O₄(s)

(19)

It should be noted that the carbothermal reduction products of CaSO₄ are complex in the low-temperature range, the desulfurization product is CaO, and the sulfur fixation product is CaS. The ΔG curve of the main phase carbothermal reaction during SSPS calcination is shown in Figure 2a, while Figure 2b shows the ΔG curve of phase formation during slag bath reduction. As shown in Figure 2a, the carbothermal reduction of nickel oxide is the most favorable among the metal oxides, followed by iron oxides, with the reduction temperatures of different iron oxides being relatively close to one another. In comparison to the aforementioned categories, Cr oxides exhibit the least favorable carbothermal reduction, with their ΔG remaining positive even at a temperature of 1000 °C [9,25]. There are two ways to form solid-phase CaS from CaSO₄, as shown in Equations (11) and (12). Low carbon content is more favorable than high carbon content to form CaS, which releases part of CO. In contrast, the formation of solid-phase CaO and SO₂ by CaSO₄ is not thermodynamically favorable, and the minimum temperature for the reaction to occur needs to be higher than 800 °C, which has been confirmed in many studies [4,9,23]. It can be seen from Figure 2b that at the considered slag bath reaction temperature, FeCr₂O₄ and MgCr₂O₄ are conducive to carbothermal reduction to form alloys, and the reaction of FeCr₂O₄ combined with MgO to form MgCr₂O₄ mineral phase also exists. In Equations (16) and (17), involving Al₂O₃, the ΔG for the conversion of MgFe₂O₄ to MgAl₂O₄ exhibits a more negative value compared to the ΔG for the direct formation of MgAl₂O₄ from MgO, indicating a thermodynamically more favorable pathway.

Based on the basic thermodynamic model, during the calcination process, the iron–nickel alloy can be formed, and the sulfur-fixed product CaS can be formed without causing SO₂ to escape, which is the expected result. At the same time, some FeCr₂O₄ mineral phases are inevitably formed. The calcined products enter the slag bath smelting stage. After the briquettes contact the slag pool, the MgO or Al₂O₃ in the slag participates in the reactions of Equations (15) and (16). However, after considering the kinetic process, the product formed by the combination of FeCr₂O₄ and MgO can be in the form of Mg(Fe,Cr)₂O₄ complex, that is, MgFe₂O₄ and MgCr₂O₄ participate in related reactions together. Therefore, MgCr₂O₄ containing Cr in the slag is finally reduced to metal and MgO, while MgFe₂O₄ will be converted to MgAl₂O₄, that is, MgAl₂O₄ still exists in the slag at the end of the reduction.

B.: Pre-reduction stage of C-SSPS

Figure 3 illustrates the change law for the weight loss rate and gas products of C-SSPS particles as analyzed by the TG-MS online analysis system. The findings from TG and DTG analyses indicated a total weight loss rate of 41.9 wt% for the particles, characterized by five stages with peaks of varying intensities. Integrating temperature and MS analysis, the initial phase demonstrated that around 100 °C, the adsorbed water began to release from the particles. The rate of water release increased with the rise in temperature, reaching a peak at 605 °C, suggesting that bound water was the dominant form. The disposal of moisture is also very important in pyrometallurgical processes and may cause fumes or an increase in hydrogen content in the raw material. The findings of Zhao et al. [2] indicated that the temperature for the water of crystallization in SSPS needed to exceed 550 °C to ensure its rapid and complete elimination.

When the drying temperature reached around 200 °C in the second stage, CO₂ gas began to be emitted. Equation (11) illustrates the corresponding reaction. The reaction continued to speed up as a result of the rising temperature and minor rise in the resultant gas. As indicated in Equation (12), a significant weight loss peak of DTG resulted from the simultaneous production of a large amount of CO and CO₂ at temperatures above 358 °C. This was brought on by the excess C in the particles, which in some areas had led to a C/S molar ratio of 3. At this temperature, only the conversion of calcium sulfate resulted in the production of carbon monoxide; the metal oxides in SSPS had not yet reached the reaction temperature. In the third and fourth stages, the formation rate of CO₂ significantly exceeded that of CO, indicating that Equation (11) predominated over Equation (12). The gas phase of SO₂ was not discovered until the fourth stage’s temperature reached 827 °C. It is possible that a small amount of calcium sulfate had not fully reacted due to the rapid consumption of the reducing agent graphite, resulting in the C/S molar ratio being less than 0.5 as given in Equation (13). The progressive drop in SO₂ level during the continuous isothermal process showed that the reaction requirements of Equation (13) were difficult to achieve. In contrast to the strategy of Wang et al. [8] and Zhao et al. [23], the SO₂ escape situation is strictly considered in this study to avoid increasing the cost of flue gas desulfurization.

The amount of CO₂ produced grew dramatically in the last stage, or about 50 min after isothermal, along with the creation of CO gas, although no SO₂ gas was formed. In conclusion, the results of the online TG-MS analysis demonstrated that the gas phase sulfur development could be fully prevented by limiting the calcination temperature below 800 °C.

2.

Mineral phase formation of the product

The TG-MS data intuitively suggest that the pre-reduction process can control sulfide release, but more research is needed to understand the formation of sulfur-bearing minerals. As shown in Figure 4 and Table 3, combined with XRD patterns and SEM-EDS analysis, it was confirmed that following the pre-reduction of C-SSPS at 800 °C, the predominant mineral phases within the sludge consisted of unreacted graphite and the sulfur-fixed product, CaS. The primary microstructures observed included flake graphite, large particles of FeCr₂O₄-CaSO₄-CaS-FeS composite phase, small-sized CaSO₄ particles, and transformed CaS phase. Notably, CaS was predominantly distributed around the carbon particles. This indicated that the substantial quantity of CaSO₄ initially present in the SSPS converted to CaS due to the influence of temperature and carbon, whereas the Fe and Cr oxides also resulted in the formation of FeCr₂O₄. This was in agreement with the results of the thermodynamic and correlation studies described above [9,10,23].

C.: Slag bath reduction stage

Mass spectrometry was used to study the release pattern of sulfide during slag bath smelting. Using the air sample as a reference, the ion current intensity value was used to monitor the waste gas collected in the C-SSPS slag bath reduction experiment. The SO₂ concentration in the exhaust gas between 0 and 10 min matched the air sample values (Figure 5), indicating no additional SO₂ production. Most of the existing forms of sulfur had been converted, with no occurrence of further desulfurization reactions. Trace amounts of SO₂ were detectable within the 10 to 20 min interval. The reaction progressed as the briquettes dissolved into the slag pool, dispersing minimal unreacted sulfate that subsequently decomposed and escaped. At the slag bath stage, unreacted CaSO4 in the briquettes attained its decomposition temperature [4], yet a delay in SO₂ release was observed, suggesting that the briquettes and initial pre-reduction facilitated sulfur fixation product integration into the slag. Comparative analysis of waste gas and air samples revealed CO production during the initial stages of the slag bath treatment, primarily attributed to the high-temperature reduction of metal oxides like Fe and Cr [8,9]. The SO₂ concentration in the air is estimated to be 0.05 mg/m³, that is, in the pretreatment process and the initial stage of slag bath reduction, even if the gas does not undergo desulfurization treatment, the sulfur element concentration will not exceed the industrial limit of 0.25 mg/m³ [3].

The high-temperature AOD slag bath reduction resulted in high metal recovery rates and sulfur retention in the residue. As shown in Figure 5, the recovery rate of the Fe-Cr-based alloy in the system exceeded 95%, with Fe and Cr contents in the metal phase being 59 wt% and 25 wt%, respectively. The metal phase also contained necessary carbon diffusion elements, partially reduced Mn, and trace amounts of Ni. The sulfur content of the impurity elements in the metal was 0.034%. The Cr content in the slag was only 0.15 wt%, while the sulfur content was 0.692 wt%, with over 94.29% of the sulfur remaining as sulfides in the slag phase. For elemental sulfur, only 5.71% was distributed in the gas phase, and 0.21% was in the recovered metal. To achieve better results, further optimization of the experimental plan is necessary. By adding 27 wt% CaO to adjust the chemical composition of SSPS, the sulfur retention rate in the solid phase exceeded 99.7%. According to Equation (11), when a certain amount of CaO is added, the reaction shifts to the left, resulting in less SO₂ generation and an increase in Ca²⁺ combining with S²⁻. Studies [8,25] have shown that adjusting the alkalinity of slag can not only improve the desulfurization effect, but also promote the recovery of Cr.

2.

Phase transformation at the interface between briquettes and slag

There are two significant processes in the gradual dissolution of the briquettes into the slag: the internal changes within the briquettes and the interface reaction with the slag. As illustrated in Figure 6a and Table 4, after 5 min in the slag bath, numerous fine C-Fe-Cr-Ni metal particles began to form within the C-SSPS, alongside a considerable amount of flake graphite and irregularly shaped CaS particles. After 10 min in the slag bath, larger metal particles had formed, settling to the bottom layer due to gravity. However, the initially smaller metal particles (region 12) continued to aggregate and grow (region 13), ultimately existing as Fe-Cr-based metal particles. It can be seen that Fe-Cr metal formation occurs through two distinct pathways: direct reduction to form Fe-Cr and reductive carburization to form Fe-Cr-C. The formation of Fe-Cr primarily results from the carbothermal reduction of spinel. The pathway to Fe-Cr-C alloy formation begins with the reductive carburization of Fe-containing oxides, leading to the production of Fe-C alloys with lower melting points. This process sets the stage for the complex formation of Fe-Cr-C. Specifically, during the reduction of Cr-containing oxides, Cr₇C₃ intermediates are formed, which impede the rate of Cr reduction. However, these Cr₇C₃ intermediates dissolve when interacting with pre-formed high-iron alloys, eventually forming Fe-Cr-C alloys [28,29]. Concurrently, with the incorporation of the low-melting-point phase CaF₂-Ca₂SiO₄ in the briquettes, the sulfur-fixing product CaS became interconnected.

Figure 6b illustrates the interface reaction between the briquettes and slag. Five minutes into the slag bath treatment, the interface between C-SSPS and the slag pool could be distinctly categorized into three zones. In zone I, close to the briquettes, the main phases were FeCr₂O₄, CaS, and Fe-Cr-Ni-C metal particles from the briquettes. Zone II, the middle diffusion layer, predominantly featured fine Mg(Fe,Cr)₂O₄ particles and the Ca₃SiO₅ phase in the slag. Zone III, the slag layer distant from the briquettes, mainly contained the Ca₂MgSi₂O₇ phase, consistent with the original slag pool. This suggested that CaS and FeCr₂O₄ in the briquette gradually migrated into the slag. The metal particles, having aggregated and grown, began to diffuse and settle into the slag. The AOD slag contained some MgO. In the MgO diffusion zone, the strong binding ability between MgO and Cr₂O₃ resulted in the formation of the MgCr₂O₄ spinel phase, with the presence of the composite Mg(Fe,Cr)₂O₄ phase not ruled out.

After 10 min in the slag bath, the near-briquette region IV had comprised Fe-Cr-based metal particles and the bulk CaS phase. The intermediate product layer V was characterized by Mg(Al,Cr)₂O₄. In comparison to the situation at the 5-minute mark, the metal particles had grown larger. With the diminishing concentration gradient of MgO at the interface between the briquette and the molten slag, Al³⁺ ions had more readily reacted with MgO, displacing Fe³⁺ ions to form MgAl₂O₄. This reaction also led to the formation of the Mg(Al,Cr)₂O₄ phase. These observations were in agreement with the thermodynamic calculations as shown in Equations (16)–(19). After 18 min in the slag bath, the briquettes had completely dissolved in the AOD liquid slag, the slag phase was uniformly distributed, and the metal phase had aggregated and grown. During the dissolution, the products of sulfur fixation remained in the stabilized forms of CaS and MnS.

D.: Slag Bath Reduction Mechanism and Model Establishment

Figure 7 presents a detailed model of the reaction mechanism, divided into two parts: the reduction behavior inside the briquette and the briquette’s dissolution behavior in the slag. The sulfur fixation mechanism in the briquette and the evolution rules of metal particle formation, aggregation, and growth were revealed.

[ad_2]