Finds documents with both search terms in any word order, permitting "n" words as a maximum distance between them. Best choose between 15 and 30 (e.g. NEAR(recruit, professionals, 20)).
Finds documents with the search term in word versions or composites. The asterisk * marks whether you wish them BEFORE, BEHIND, or BEFORE and BEHIND the search term (e.g. lightweight*, *lightweight, *lightweight*).
Activate our intelligent search to find suitable subject content or patents.
Select sections of text to find matching patents with Artificial Intelligence.
powered by
Select sections of text to find additional relevant content using AI-assisted search.
powered by
(Link opens in a new window)
Abstract
The article delves into the dissolution behaviors of recycled cement paste (RCP) and lime in EAF slag under static conditions, aiming to explore their potential as fluxes in the steelmaking process. The study compares the dissolution rates, interfacial microstructures, and mechanisms of RCP and lime, revealing that RCP dissolves faster than lime. The authors conduct experiments at 1400°C and 1500°C, observing that RCP can be completely dissolved in 60 seconds at 1500°C. The dissolution process of RCP is characterized by direct dissolution, while lime undergoes indirect dissolution with the formation of a C2S product layer. The article also presents a dynamic dissolution model based on the experimental data, which can predict the dissolution process of lime and RCP in EAF slag. This research is significant for understanding the potential of RCP as a sustainable flux in steelmaking, contributing to waste reduction and resource efficiency.
AI Generated
This summary of the content was generated with the help of AI.
Abstract
Recycled cement paste (RCP) is a CaO-rich resource that is recycled from waste concrete and waste cement. For reducing environmental impact and promoting comprehensive resource utilization, RCP could partially replace lime as flux in electric arc furnace (EAF) steelmaking process. In this study, the dissolution behaviors of RCP and lime in EAF slag at 1400 °C and 1500 °C were investigated using static dissolution experiments. The measured dissolution rate constant of lime at 1400 °C and 1500 °C is 3.12×10−5 and 8.42×10−5 m/s, respectively, while that for RCP is 1.37×10−4 and 2.91×10−4 m/s, respectively. For the lime dissolution, a dense dicalcium silicate (C2S) product layer was generated at the dissolution interface, with the diffusion of Ca2+ in the C2S layer being the limiting step for the lime dissolution. However, for the RCP dissolution, no product layer was observed at the dissolution interface. Instead, a dissolution intermediate layer forms due to slag penetration. During the dissolution of RCP, Ca2+ and SiO44− diffuse from RCP layer to slag, while Fe2+ and AlO45− diffuse from slag to RCP layer accumulating in the dissolution intermediate layer. Furthermore, a dissolution model was developed based on current experimental results, which helps to predict dynamic dissolution process of lime and RCP in EAF slag. In general, the dissolution rate of RCP in EAF slag was much higher than that of lime, partially adding RCP as flux in EAF could accelerate the liquid slag formation in EAF steelmaking process.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Introduction
In Europe, 180 million tons of concrete demolished waste (CDW) are produced every year, and burying the CDW can cause significant environmental pollution.[1] CDW can be recycled for environmental impact reduction and comprehensive resource utilization. Recycled cement paste (RCP), a CaO-rich resource obtained through the separation and recycling from CDW,[2‐5] has the potential to partially replace lime as a flux in the electric arc furnace (EAF) steelmaking process. In the EAF steelmaking process, rapid slag formation is important in accelerating refining reactions, resulting in a shortened refining duration and energy savings, and the slag formation is heavily influenced by the dissolution of flux into the slag.[6‐9]
Over the last few decades, numerous investigations have been conducted to study the dissolution behaviors of fluxes in steel slag.[10‐19] Lesiak et al.[10] investigated the effect of calcination condition of dolomite-based materials dissolution in EAF slag, and found that the amount of the material dissolved decreases with deceasing its porosity. Fruehan et al.[15] studied the dissolution of magnesite and dolomite in EAF slag using both the dipping test and rotating cylinder test, and found that during the dissolution of dolomite, CaO dissolved away first, and then the MgO particles entered the solution. The dissolution of lime in steel slag is a very common topic for steelmaking. The factors influencing dissolution could be summarized as: (1) chemical composition of solid and liquid, (2) temperature, (3) particle size, and (4) forced convection. As well documented by Li et al.,[20] the dissolution experiments, based on the experimental techniques, could be divided into five types: (1) under static conditions, (2) rotating rod/disc method, (3) under forced convection, (4) direct observation and (5) sampling from industry.
Advertisement
Considering the dissolution mechanism, the dissolution could be divided into direct dissolution and indirect dissolution. Direct dissolution, is characterized by the absence of an intermediate phase and involves only physical diffusion without accompanying chemical reactions. Direct dissolution processes such as the dissolution of Al2O3[21,22] and SiO2[23,24] into CaO–Fe2O3-based slag, dissolution of Al2O3 into CaO–Al2O3–SiO2 slag.[25] Dissolution of solid oxide into slag may involve the formation of an intermediate reaction product, and this type of dissolution is called indirect dissolution. As for indirect dissolution, Sandhage and Yurek[26] observed that Spinal (MgAl2O4) formed at the Al2O3/CaO–SiO2–MgO–Al2O3 interface. Yu et al.[27] and Yang et al.[28] found the MgFe2O4 and CaTiO3 produced at the MgO/CaO–Fe2O3 and TiO2/CaO–Fe2O3 interface respectively. The product layer of CaAl4O7 could be found near the sapphire/CaO–SiO2–Al2O3 interface, reported by Oishi et al.[29] As for the lime dissolution into CaO–SiO2–FeO based slag (generally basic oxygen steelmaking slag and EAF slag), reports[10‐14] confirmed the formation of C2S (dicalcium silicate) phase during the dissolution process, and the C2S crystal distribution depends on the experimental conditions. The reaction of CaO and SiO2 to produce C2S is considered inevitable in these conditions. Typically, the C2S phase demonstrates high thermodynamic stability in CaO–SiO2–MgO–Al2O3–FexO slag systems, particularly under basicity ranging from 1.5 to 3.0 at high temperatures.
As mentioned above, RCP has the potential to partially replace lime as a flux in EAF for resource utilization. The dissolution process, dissolution mechanism and dissolution rate of RCP have never been investigated in the EAF slag formation. Thus, the present study aims to explore the dissolution behaviors of RCP and lime (for comparison) in simulated EAF slag at 1400 °C and 1500 °C under static conditions. The dissolution process, dissolution interfacial microstructure, and dissolution mechanisms were thoroughly investigated. Additionally, a dynamic dissolution model of lime and RCP was developed based on the present experimental data. This model assists in determining the dissolution rate and predicting the dissolution process of lime and RCP in EAF slag.
Experimental
Materials and Slag Preparation
In this study, the initial EAF slag was synthesized by chemical reagents and its chemical composition is given in Table I. The chemical reagents of CaO, SiO2, MgO, Al2O3 and FeO with a purity of 99.5 wt pct (Alfa Aesar) were used to prepare the slag samples. The chemical reagent powders were mixed using a ball mill (Retsch, PM100) at a mass ratio of grinding ball to material of 5:1 for 30 minutes. Subsequently, the mixtures were pressed into tablets and then pre-melted in a high-purity argon atmosphere at 1500 °C for 30 minutes for homogenization.
Table I
Chemical Composition of Samples (Wt Pct) Used in This Study (R2 = CaO/SiO2, Binary Basicity)
Items
CaO
SiO2
MgO
Al2O3
FeO
Fe2O3
K2O
Na2O
TiO2
R2
Slag
40.00
30.00
6.00
4.00
20.00
—
—
—
—
1.33
RCP
68.09
19.46
3.61
0.74
—
3.61
0.49
0.12
0.25
3.50
Lime
99.50
—
—
—
—
—
—
—
—
—
The RCP sample was produced from commercial Portland cement by well mixing the Portland cement and water at a water to cement mass ratio of 0.6, then curing the sample at room temperature for one month in a sealed condition to avoid evaporation. After curing, these cement paste bricks were heated up around 450 °C to 500 °C in a muffle furnace to produce RCP. The chemical and phase compositions of RCP were examined by XRF (PANalytical Epsilon 3) and XRD (PANalytical Empyrean, Co target), as shown in Table I and Figure 1, respectively. RCP is mainly composed of CaO and SiO2 with a binary basicity R2 of 3.5 (R2 = CaO/SiO2, wt pct). For the phases composition, main phases such as C2S (dicalcium silicate), C3S (tricalcium silicate), CaO, and a small amount of Ca(OH)2 and SiO2 were detected in the RCP samples. During the heating process (450 °C to 500 °C), the calcium silicate hydrate phase (C–S–H) dehydrated to C3S and C2S. Then the C3S phase would decompose into C2S and CaO because C3S is not thermodynamically stable at the roasting temperature. In addition, some of Ca(OH)2 was also detected in the RCP samples due to the moisture absorption of CaO.
Fig. 1
XRD patterns of RCP sample and standard components
In order to minimize the influences of porosity and density of solid samples on dissolution, the sintered dense solid tablets of CaO and RCP were prepared for dissolution experiments. The CaO and RCP powders were loaded into a cylinder shape mould with a diameter of 5 mm under 1 ton pressure for 1 minutes, then the pressed tablets were roasted in a muffle furnace in air for 1 hour at 1400 °C and 1000 °C, respectively. Subsequently, the dense CaO and RCP tablets were polished to be flat for dissolution experiment, and their weight and dimension (e.g., height before dissolution h0) were measured, as shown in Table II.
Table II
Experimental Conditions and Measured Parameters for Samples
No.
Temperature (°C)
Dissolution Time (s)
Slag Weight (mg)
Solid Weight (m0, mg)
Average Diameter (D0, mm)
Height Before Dissolution (h0, mm)
Solid Density (g/cm3)
Height After Dissolution (h1, mm)
Height Decrease (∆h, mm)
Weight Loss (∆m, mg)
Weight Loss Fraction
L1
1400
30
143.434
44.113
4.46
1.06
2.6638
1.00
0.06
4.779
0.1067
L2
1400
60
148.267
39.31
4.52
0.93
2.6342
0.84
0.09
6.316
0.1673
L3
1400
90
145.068
43.635
4.55
1.01
2.6571
0.90
0.11
8.702
0.1967
L4
1400
120
143.341
41.201
4.46
1
2.6372
0.87
0.13
9.115
0.2285
L5
1500
30
153.261
46.34
4.43
1.13
2.6606
0.97
0.16
12.104
0.2647
L6
1500
60
150.005
41.544
4.44
1.02
2.6306
0.78
0.24
16.048
0.3917
L7
1500
90
149.94
45.675
4.46
1.11
2.6339
0.82
0.29
20.065
0.4409
L8
1500
120
148.05
44.256
4.51
1.05
2.6384
0.74
0.31
21.075
0.4784
R1
1400
30
155.272
38.261
3.99
1.23
2.4878
1.01
0.22
13.556
0.3536
R2
1400
60
146.394
37.108
3.99
1.22
2.4326
0.84
0.38
20.353
0.5460
R3
1400
90
144.291
37.347
3.95
1.23
2.4778
0.78
0.45
23.227
0.6244
R4
1400
120
152.063
36.149
3.96
1.17
2.5086
0.67
0.50
24.650
0.6800
R5
1500
30
146.063
35.101
3.96
1.15
2.4782
0.58
0.57
25.620
0.7421
R6
1500
60
150.327
34.525
3.95
1.13
2.4933
—
—
34.525
1
R7
1500
90
144.886
32.345
3.94
1.08
2.4564
—
—
32.345
1
R8
1500
120
141.998
32.151
3.93
1.06
2.5004
—
—
32.151
1
Dissolution Experiments
A confocal laser scanning microscope furnace (CLSM) was employed for the dissolution experiments, and the schematic diagram of the apparatus is shown in Figure 2. The chamber is ellipsoid in shape, with the samples and heating element positioned at the upper and lower focal points of the ellipsoid. The heating principle involves the reflection and focusing of thermal radiation through the gold coating on the chamber. To minimize the dissolution time errors, the heating and cooling rates were set as quickly as possible at 400 and −1200 K/min, respectively. The dissolution start time (i.e., zero dissolution time) was defined as the moment when samples reached the desired temperature.
Fig. 2
Schematic of the confocal laser scanning microscope furnace and dissolution experiment
The dissolution experiments were carried out at 1400 °C and 1500 °C in argon, with different dissolution times ranging from 30 to 120 seconds. Experimental conditions and measured dissolution parameters, such as the geometry of solid part, are summarized in Table II. The samples labelled from L1 to L8 represent the lime dissolution samples, while the samples labelled R1 to R8 are the RCP dissolution samples.
After dissolution, all samples with crucible were mounted in resin, cut and polished along the cross-section, and subsequently coated with Au/Pd at around 20 nm (Bio-Rad SC650 Sputter Coater Lab). A digital optical microscope (OM, Keyence VHX7000), as well as a scanning electron microscope (SEM, Zeiss, Sigma) equipped with energy dispersive X-ray spectroscopy (EDS, Oxford, Ultim Extreme) were employed for microstructural observation.
Results and Discussion
Dissolution Behaviors
The typical optical images of the cross-sections of samples R2 and L2 are shown in Figure 3, and the thickness of solid parts after dissolution was measured using the digital optical microscope. In order to minimize the measurement error, the thickness (height) was measured across the sample for 10 times, and the obtained average thickness of the solid part was defined as the height after dissolution (h1), as shown in Table II. Meanwhile, the density of RCP ranges from 2.43 to 2.51 g/cm3, while that of lime ranges from 2.63 to 2.66 g/m3. The density of liquid slag is 2.83 g/cm3 and can be determined using Eq. [1], where ω(CaO) represents the mass percent of component (CaO, SiO2, etc.) in the liquid slag.[24]
Fig. 3
Optical microscope images of cross-sections of samples, (a) through (d): L2 (lime, 1400 °C, 60 s), and (e) through (h): R2 (RCP, 1400 °C, 60 s)
During the dissolution process, the solid samples of RCP and lime tend to float in the liquid slag due to their lower densities compared to the liquid slag. Assuming that the shape of the solid parts in slag during the dissolution remains cylindrical, and the dissolution of the solid sample in liquid slag primarily occurred through the bottom and sides of the cylindrical samples. Consequently, the reduction in specimen height (Δh) and specimen radius (Δr) can be considered as the same value, that is Δh = Δr. Based on above assumption, the volume, weight, weight loss and its fraction were calculated and also summarized in Table II, while the weight loss fraction of samples is depicted in Figure 4.
Fig. 4
Measured weight loss fractions of RCP and lime at different temperature and in different dissolution time
Measured results show that the weight loss fractions of RCP were higher than those of lime, revealing that the dissolution rate of RCP is greater. The dissolution of RCP at 1500 °C was observed to complete at 60 seconds, which is evidenced by the fact that no residual solid RCP part could be observed in samples R6, R7 and R8. Further discussions on the dissolution rate of RCP and lime are described in Section III–D.
Interfacial Microstructure
Figure 5 displays the SEM image and EDS results illustrating the microstructure of the lime–slag interface. After dissolution, a layered structure formed at the dissolution interface. The presence of the FexO was detected in the lime layer, indicating that some slag penetrated into the lime layer during dissolution. Spectra 3 and 4 reveal that the bright area is primarily composed of Ca, Fe, and O, suggesting that this phase corresponds to a calcium ferrite-based slag phase. As seen in spectra 5, 6 and EDS mapping, a product layer of C2S, with a thickness of approximately 20 μm, is clearly visible at the forefront of the dissolution interface layer. The observed interfacial microstructure during the dissolution of lime into EAF slag aligns with previously reported findings.[10‐14] Those studies also noted the accumulation of a C2S product layer at the dissolution interface of lime into steelmaking slag (i.e., CaO–SiO2–FeO based slag). Deng and Du[13] investigated the dissolution of lime in steelmaking slag (CaO–SiO2–FeO) under forced convection and concluded that the removal of the C2S dense product layer by liquid forced mobility could accelerate the lime dissolution process. They also identified some tricalcium silicate (C3S) crystals generated at the dissolution interface and within the lime layer. The dissolution of lime into slag at high temperatures depends on Ca2+ diffusion; thus, the diffusion of Ca2+ through the C2S product layer into slag is the limiting step for lime dissolution under static conditions. In slag layer, massive crystals of C2S precipitated along with a small amount of fine iron oxide crystals. No tricalcium silicate (C3S) phases were found at the dissolution interface in this study. The reaction of C2S + CaO=C3S is the main reaction for C3S formation.[30] The formation of C3S requires certain conditions, such as a specific temperature (>1300 °C) and sufficient reaction time. However, the maximum dissolution duration in the present experiments is 120 seconds, which may be insufficient for the C3S formation.
Fig. 5
SEM image and EDS results for dissolution interface of sample L2 (lime, 1400 °C, 60 s)
As for the dissolution of RCP in EAF slag, the SEM images and EDS results depicting the interfacial microstructure of RCP dissolution are shown in Figure 6. In comparison with the dissolution of lime, no dense C2S product layer was observed in the samples, and some dendritic C2S crystals were found in the slag. An obvious layer between the slag and RCP can be observed and it is defined as dissolution intermediate layer. The thickness of dissolution intermediate layer measured from the image is about 80 μm, and spectra 4, 5 and 6 show its chemical composition.
Fig. 6
SEM image and EDS results for dissolution interface of sample R2 (RCP, 1400 °C, 60 s)
The SEM image and EDS results for the cross-section of sample R2, focusing on the dissolution intermediate layer, are shown in Figure 7. Some C2S crystals and fine iron oxides could be detected in residual slag. As shown in spectrum 3 for the slag layer, the Al2O3 content in slag is 7.9 wt pct, which is higher compared with the original slag, indicating that a small amount of Al2O3 dissolved from alumina crucible into slag. In the dissolution intermediate layer, no significant product is formed, and the content of CaO, SiO2, MgO, and FeOx in the intermediate layer lies between that of the RCP and the slag layer. Namely the intermediate layer is composed of liquid slag and RCP, some liquid slag penetrated into RCP to form this layer. The Al2O3 content in the dissolution intermediate layer is higher than that in the slag bulk, indicating that AlO45− diffuses through the slag bulk and accumulates in the interlayer.
Fig. 7
SEM image and EDS results for cross-section of sample R2 near the dissolution intermediate layer
The SEM images and EDS results for the cross-section of sample R2, focusing on the undissolved RCP layer, are shown in Figure 8. As marked in Figures 8(c) and (d), the brighter net-like area is liquid phase. Figure 8(c) presents the microstructure at the interface between RCP and intermediate layer, with the liquid phase content gradually decreasing from the bottom (intermediate layer) to top (RCP). Figures 8(d) and (e) show the microstructure in the RCP layer, and also some liquid phase could be detected. But the liquid phase content in the RCP layer is much lower than that in the intermediate layer. EDS results show that the liquid phase is composed of primarily CaO, FexO and Al2O3 with minor SiO2 and MgO contents. Cracks (in black in the figures) can be detected with high CaO content in the EDS mapping, which could be formed in samples due to the strong moisture absorbability.
Fig. 8
SEM image and EDS results for solid part of dissolution of RCP (sample R2), (a) solid part for sample R2, and (b) through (e) are specific areas in (a)
Figure 9 presents the thermodynamic calculation for phase composition and liquid slag composition in RCP at 1400 °C and 1500 °C. The calculated liquid slag phase proportion in RCP was 10.65 wt pct at 1400 °C and 16.63 wt pct at 1500 °C, respectively, indicating that the RCP consists of both solid and liquid phases at the experimental dissolution temperatures. The liquid phase compositions of RCP are shown in Figures 9(c) and (d), which are in agreement with the report from Winter,[30] who pointed out that the liquid phase in cement clinker is mainly composed of oxides of calcium, iron and aluminium, with some silicon and other minor elements. The existence of liquid slag phase in RCP at the experimental temperatures is beneficial for the slag penetration, and therefore could be the main reason leading to faster dissolution of RCP compared to lime.
Fig. 9
Thermodynamic calculation of RCP phase composition, (a) phase composition of RCP at 1400 °C, (b) phase composition of RCP at 1500 °C, (c) liquid slag composition in RCP at 1400 °C, and (d) liquid slag composition in RCP at 1500 °C
As mentioned above, the dissolution of lime into EAF slag is an indirect dissolution, whereas the RCP dissolution is of direct dissolution. An intermediate C2S product layer gathered at the lime/slag interface, however, there is no product layer observed in the RCP/slag interface, instead, a slag penetration layer was found in RCP samples. For better understanding the dissolution mechanisms of lime and RCP in EAF slag, the average thicknesses of the C2S layer in lime dissolution samples and the slag penetration layer in RCP dissolution samples were measured and shown in Figure 10.
Fig. 10
Measured average thickness of intermediate layers. (a) Thickness of C2S product layer in lime samples, and (b) Thickness of slag penetration layer in RCP samples
For the lime dissolution, the measured average thickness of the C2S product layer at 30 seconds was around 32 and 21 μm at 1400 °C and 1500 °C, respectively. Then the thickness decreased to constant at around 20 and 17 μm at 1400 °C and 1500 °C, respectively. The indirect dissolution of lime could be regarded as two process: (A) C2S layer formation and (B) C2S layer dissolution. The formation rate of the C2S layer is symbolised as RA1, and the dissolution rate of the C2S layer is symbolled as RB1. The variation in thickness of the C2S layer could be considered as the competing effect of these two processes. At the dissolution time of 60 seconds or more, the thickness of the C2S product layer kept almost constant, that means the rate of the C2S layer formation and the rate of the C2S layer dissolution reached dynamic balance (RA1 ≈ RB1). However, the thicknesses of the C2S layer at 30 seconds was greater than that at 60 seconds at both 1400 °C and 1500 °C, indicating that the C2S formation rate was higher than the C2S dissolution rate at 30 seconds (RA1 > RB1). It is worth noting that the thickness of the C2S layer at 1400 °C was higher than that at 1500 °C, indicating that the net difference between RA and RB at 1400 °C was larger than that at 1500 °C.
The measured average thickness of the penetration layer in RCP samples (i.e., the dissolution intermediate layer) at 1400 °C was found to increase with dissolution time and then kept at around 80 μm. Meanwhile, the dissolution of RCP also could be regarded as the following two processes: (A) slag penetration and (B) intermediate layer dissolution, while the rates for these two processes could be symbolised as RA2 and RB2. The penetration depth gradually increased with increasing the dissolution time, indicating that the slag penetration rate was greater than the rate of the intermediate layer dissolution (RA2 > RB2). In addition, there was a significant increase in the penetration depth as raising temperature, contributed by factors in the descending order of (1) liquid phase portion in RCP and (2) slag viscosity. As illustrated before, no C2S product layer was formed in RCP samples, indicating that the driving force for C2S layer formation was insufficient. It is worth noting that the XRD pattern (Figure 1) and thermodynamic calculation (Figure 9) reveal that the large amounts of C2S were already present in RCP, leading to the insufficient driving force for the C2S product layer formation.
Furthermore, a schematic diagram of the dissolution mechanisms for both lime and RCP under static conditions is shown in Figure 11. Where the X-axis in Figure 11(a) represents dissolution distance, Y-axis is the CaO concentration, and the blue line represents the CaO concentration at the dissolution interface. Two-way diffusion does exist in both lime and RCP dissolution. Regarding lime dissolution, Fe2+ diffuses from the slag bulk toward the lime layer, while Ca2+ diffuses from the lime layer toward the slag bulk. Subsequently, CaO reacts with SiO2 to produce a dense layer of C2S, and C2S could not be well removed under static conditions. The diffusion rate of Fe2+ is higher than that of Ca2+, while the diffusion rate of SiO44− is the lowest.[20] In the lime dissolution, CaO reacts with SiO2 to produce C2S at the lime–C2S interface, while C2S dissolves into slag at the C2S–slag interface. A dynamic balance between C2S formation and C2S dissolution was established at the dissolution interface, and the diffusion of Ca2+ in C2S product layer is the limiting step for lime dissolution. As for RCP dissolution, slag penetrated into the RCP layer to form a dissolution intermediate layer, both Ca2+ and SiO44− diffuse from the RCP layer to slag bulk, while Fe2+ and AlO45− diffuse from slag bulk to the dissolution interface. During RCP dissolution, no dense C2S layer was observed at the dissolution interface, and RCP consists of both liquid and solid phases at the experimental temperatures. The presence of a liquid phase sounding solid phase facilitates the liquid EAF slag penetration into the RCP layer resulting in its faster dissolution than lime.
Fig. 11
Schematic diagrams of dissolution mechanisms of lime and RCP in slag, (a) dissolution of lime, and (b) dissolution of RCP
Due to the relatively high CaO content both in lime and RCP, the concentration of CaO in slag would increase during dissolution process. The saturation concentration of CaO in slag is a key parameter for dissolution, and it was calculated by FactSage8.3 using Equilibrium mode (FactPS and FToxid databases). The following compositions of 100 g slag (40 g CaO + 30 g SiO2 + 20 g FeO + 6 g MgO + 4 g Al2O3, as described in Table I) plus xg of CaO (i.e., varying amount of CaO) were selected for calculation, and the activity of Fe was set as 1.0 to prevent the oxidation of FeO in the slag. The calculated phase equilibrium diagram for current slag at 1400 °C and 1500 °C are shown in Figure 12, where the blue line represents the precipitated CaO phase. The CaO concentration when CaO begins to precipitate is considered to be the saturation concentration of CaO in the slag. As the additional CaO content increases, the liquid slag proportion would decrease, and CaO would participate when the addition of CaO is 9.95 g at 1400 °C and 15.85 g at 1500 °C. Thus, the saturation concentrations of CaO in present slag at 1400 °C and 1500 °C are 0.4543 and 0.4821, respectively.
Fig. 12
Thermodynamic equilibrium calculation for slag with different CaO content at different temperatures, (1) 1400 °C, and (2) 1500 °C
As mentioned earlier, assuming that the dissolution of solid samples in slag occurs primarily through the bottom and sides of the cylindrical sample in present experiments. The schematic for dissolution of cylindrical sample is shown in Figure 13(a), where the blue and red parts represent solid sample before and after dissolution. The dissolution area Adiss (m2) could be calculated in Eq. [2], where Abotm (m2), Aside (m2), R (m) and H (m) are the area of bottom, area of side, radius, and height of cylindrical sample, respectively.
Fig. 13
(a) Geometry for dissolution of cylindrical samples in present experiments, (b) Geometry for dissolution of sphere powders, and (c) Calculation flowsheet for the present model
In the dissolution process, the dissolved mass Δmdiss (kg) in a time step Δt (seconds) can be calculated by Eq. [3]. The dissolution rate Rd [kg/(m2·s)], defined by Eq. [4], is a parameter of how fast the dissolution occurs. Here, k0 is a constant, which is relative to temperature and the characteristics of the solid phase, γ (m2/s) is the kinematic viscosity of liquid slag, D (m2/s) is the diffusion coefficient of solute in slag. Csat (kg/m3) and Cbulk (kg/m3) represent the saturated mass concentration of solute in slag and mass concentration of solute in the entire melts.
The diffusion coefficient D could be calculated using Stokes–Einstein–Sutherland equation[31] (Eq. [5]), where η, r, kB, and T refer to liquid viscosity (Pa·s), the radius of spread particle (m) (radius of Ca2+ is 0.099 nm), Boltzmann constant (1.380649×10−23 J/K), and degree Kelvin (K). It is thought that the chemical composition and temperature of liquid affects the mass transfer coefficient by affecting both viscosity and diffusivity of CaO (or Ca2+) in the liquid slag.
$$D=\frac{{k}_{B}T}{6\pi r\eta }.$$
(5)
If the liquid slag composition and temperature remain relatively stable, the density, viscosity, and kinematic viscosity of the slag can be considered constant. Thus, the Eq. [5] could be expressed as Eq. [6], where k′ (m/s) is the dissolution rate constant. The difference between Csat and Cbulk is the dissolution driving force, and the dissolution rate Rd is proportional to the dissolution driving force. In present model, the dissolution rate constant k′ is used as the fitting parameter.
Assuming the density of the solid ρsolid (kg/m3) does not change during the dissolution process. The change of solid part volume ΔV (m3) could be calculated through geometric relations, as shown in Eq. [7].
$$\text{Mass of dissolution:}\, {\Delta m}_{diss}^{i+1}={R}_{d}^{i+1}\cdot {A}_{diss}^{i+1}\cdot \Delta \text{t}.$$
(19)
The mass solid, the mass of slag, as well as some parameters for dissolution could be expressed as Eqs. [8] through [19]. In these equations, the superscript i represents the parameters for time step i, while superscript i+1 represents the next calculation time step. Figure 13(c) shows the calculation flowsheet for the present model. The iterative method and mathematic calculation software of MATLAB_R2022a were used to solve the above explicit equations. There are three conditions for calculation: (1) \({m}_{solid}^{i}>0\) and \(\left({C}_{sat}-{C}_{bulk}^{i}\right)>0\), it means that there is undissolved solid part remaining in slag and the dissolution driving force is greater than zero, it will continue to calculate until it turns into the following two conditions. (2) \({m}_{solid}^{i}\le 0\) and \(\left({C}_{sat}-{C}_{bulk}^{i}\right)>0\), it represents the solid phase is completely dissolved into slag and the concentration of solute (CaO) in slag does not reach the saturation concentration; (3) \({m}_{solid}^{i}>0\) and \(\left({C}_{sat}-{C}_{bulk}^{i}\right)\le 0\), that means there is still some undissolved solid part in slag, but the concentration of solute (CaO) reaches saturation concentration, the dissolution driving force is zero. The calculation process will stop in conditions (2) and (3). In current dissolution model, the parameter of the dissolution rate constant k′ in Eq. [6] is employed as a curve-fitting parameter. The dissolution rate constant of lime and RCP at temperatures derived from experimental results is shown in Table III, it can be seen that dissolution rate constant of RCP is much higher than that of lime. Figure 14 illustrates the experimental results and fitting curves for dissolution weight loss fraction of lime and RCP at 1400 °C and 1500 °C. It is evident that the present model provides a good fit for the dissolution of RCP and lime.
Table III
Dissolution Rate Constant Values Derived Experimentally for Lime and RCP at Temperatures
Samples
k′ (m/s) at Temperatures (°C)
1400
1500
Lime
3.12 × 10−5
8.42 × 10−5
RCP
1.37 × 10−4
2.91 × 10−4
Fig. 14
Experimental results and model fitting curves of dissolution weight loss fraction of lime and RCP at 1400 °C and 1500 °C
In practice, the method for charging powder materials into the EAF often involves injection, which results in high material utilization efficiency.[32] Assuming the shape of powder particles is spherical, the schematic of the geometry for the dissolution of spherical-shaped samples is depicted in Figure 13(b). Equations from Eqs. [11] through [17] could be replaced by Eqs. [20] through [22] due to the differing geometric relations and dissolution interface area for spherical shape. Under the spherical powder shape assumption, the calculated dynamic dissolution process of lime powder in EAF slag at different temperatures and in different solid–liquid mass ratio (β) is shown in Figure 15(a). It can be seen that the dissolution rate of lime would increase with the temperature increasing and decrease with the solid–liquid mass ratio. In case 1 at 600 seconds, the mass of solid part and dissolution driving force is greater than 0 (msolid > 0 and Csat−Cbulk > 0), the lime will continuously dissolve into slag. In case 2 at 400 seconds, the mass of solid is greater than 0 but the dissolution driving force is near 0 (msolid > 0 and Csat−Cbulk = 0). That means the concentration of CaO in liquid slag reaches saturation concentration, lime cannot dissolve into slag and some undissolved lime will still remain in slag. In case 3 at 200 seconds, the dissolution driving force is greater than 0 but the mass of the solid is 0 (msolid = 0 and Csat−Cbulk > 0), that means the lime has already completely dissolved into slag. The concentration of CaO in slag is easier to reach saturation concentration when the mass ratio of solid to liquid increases.
Fig. 15
Calculated dynamic dissolution process of RCP and lime from present model. (a) Dissolution of lime in different solid–liquid mass ratios, (b) Dissolution of lime and RCP in different particle sizes and temperatures, (c) Relationship between complete dissolution time and particle size, and (d) Dissolution process of RCP in different CaO content
As mentioned above, the mass ratio of solid to liquid is a crucial factor for dissolution. Fixing the mass ratio of solid to liquid at 0.1, the calculated dynamic dissolution process of lime and RCP in different particle sizes and at various temperatures is depicted in Figure 15(b). Lime dissolves very slowly at 1400 °C, and it is not fully dissolved into the slag within 300 seconds. However, the particle size of RCP decreases faster than that of lime, and it can be fully dissolved into slag within 140 seconds and 40 seconds at 1400 °C and 1500 °C, respectively. In practice, complete dissolution time of flux in slag is important. Fixing the solid–liquid mass ratio at 0.1, Figure 15(c) presents the calculated relationship between the complete dissolution time and particles of lime and RCP at different temperatures within the particle size range from 0 to 2.5×10−3 m. The parts above the curve represents the region where complete dissolution is achievable. For example, RCP with a particle size less than 1.5×10−3 m can be fully dissolved into the slag at 1400 °C within 150 seconds. In general, the particle size of lime used in EAF injection falls within the range of 0.01 to 1.5 mm.[6,32] The complete dissolution duration increases with an increase in particle size among all samples. Compared to the rapid increase in the complete dissolution duration for lime with its particle size, the complete dissolution duration of RCP increases gradually (i.e., at a much slower pace) with increasing RCP particle size.
As discussed earlier, the dissolution process of both RCP and lime reaches its endpoint when the CaO concentration in the slag reaches saturation concentration. The CaO content in RCP may vary depending on the specific resources and recycling methods employed in practice. Assuming a solid–liquid ratio is 0.2, Figure 15(d) illustrates the calculated dynamic dissolution process of RCP with different CaO content in the slag. It can be observed that the dissolution rate of RCP decreases with the increasing CaO content. In the case of the dissolution of RCP with 75 pct CaO content, the particle size of RCP decreases slowly after 200 seconds, and the dissolution rate continuously decreases during the dissolution process. The dissolution rate of RCP in slag increases with the decreasing CaO content in RCP. However, in practice, there is a desire to obtain a higher CaO content in RCP, which helps reduce the flux addition amount and slag weight.
Conclusions
The dissolution behaviors of recycled cement paste (RCP) and lime in the simulated initial EAF slag were investigated through dissolution experiments under static conditions. The dissolution rates, interfacial microstructure and dissolution mechanisms were fully discussed, leading to the development of a dissolution kinetics model based on the present experimental data. The following conclusions have been obtained.
(1)
The RCP could be completely dissolved into slag in 60 seconds at 1500 °C, and the dissolution rate of RCP is much higher than that of lime. The dissolution rate constant of lime at 1400 °C and 1500 °C is 3.12×10−5 and 8.42×10−5 m/s, respectively, while that of RCP is 1.37×10−4 and 2.91 × 10−4 m/s, respectively. Partially adding RCP as flux in EAF could accelerate the liquid slag formation process in EAF steelmaking process.
(2)
For lime dissolution, a dense product layer of dicalcium silicate is generated at the dissolution interface, with the diffusion of Ca2+ in the C2S layer being identified as the limiting step for lime dissolution. Calcium ferrite could be observed at the interface between the lime and C2S layer.
(3)
During the RCP dissolution, no apparent product layer of C2S was observed at the dissolution interface. Instead, a dissolution intermediate layer forms due to slag penetration. Liquid phase, containing primarily CaO and FeO, was detected in the undissolved RCP layer, indicating RCP was in solid–liquid two phases region at dissolution temperatures.
(4)
During the dissolution of RCP, Ca2+ and SiO44− diffuse from the RCP layer to slag, while Fe2+ and AlO45− diffuse from the slag to RCP layer, accumulating in the dissolution intermediate layer.
(5)
A dissolution model was developed based on the current experimental results, which helps to predict dynamic dissolution process of lime and RCP in EAF slag.
Acknowledgments
The authors gratefully acknowledge the financial support from EPSRC Grant No. EP/W026104/1.
Conflict of interest
On behalf of all authors, the corresponding author states that there is no conflict of interest.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
D. Gastaldi, F. Canonico, L. Capelli, L. Buzzi, E. Boccaleri, and S. Irico: Cem. Concr. Compos., 2015, vol. 61, pp. 29–35.CrossRef
2.
A. Bordya, A. Younsib, S. Aggoun, and B. Fiorioa: Constr. Build. Mater., 2017, vol. 132, pp. 1–8.CrossRef
3.
X. Fang, D. Xuan, B. Zhan, W. Li, and C.S. Poon: J. Clean. Prod., 2021, vol. 290, p. 125192.CrossRef
4.
J. Wang, M. Mu, and Y. Liu: Constr. Build. Mater., 2018, vol. 190, pp. 1124–32.CrossRef
5.
S. Zhutovsky and A. Shishkin: Constr. Build. Mater., 2021, vol. 280, p. 122510.CrossRef
6.
D. Mombelli, G. Dall’Osto, G. Villa, C. Mapelli, S. Barella, A. Gruttadauria, L. Angelini, C. Senes, M. Bersani, P. Frittella, R. Moreschi, R. Marras, and G. Bruletti: Steel Res. Int., 2021, vol. 92, p. 2100083.CrossRef
7.
Y. Sung, S. Lee, K. Han, J. Koo, S. Lee, D. Jang, C. Oh, and B. Jang: Processes, 2020, vol. 8(10), p. 1202.CrossRef
8.
S. Timoshenko, M. Gybinskyi, Y. Yaroshenko, and T. Vvedenska: KnE Eng., 2018, vol. 3(5), pp. 35–49.CrossRef
9.
R. Chairaksa-Fujimoto, Y. Inoue, N. Umeda, S. Itoh, and T. Nagasaka: Int. J. Miner. Metall. Mater., 2017, vol. 22, pp. 788–97.CrossRef
10.
S. Lesiak, E. Cheremisina, J. Rieger, J. Schenk, F. Firsbach, W. Johnson, T. Chopin, and M. Nispel: Steel Res. Int., 2022, vol. 93, p. 2100675.CrossRef
11.
S. Amini, M. Brungs, and O. Ostrovski: ISIJ Int., 2007, vol. 47(1), pp. 32–37.CrossRef
12.
J. Martinsson, B. Glaser, and S. Du: Metall. Mater. Trans. B, 2018, vol. 48B, pp. 3164–70.CrossRef
13.
T. Deng and S. Du: Metall. Mater. Trans. B, 2012, vol. 43B, pp. 678–86.
14.
Z.S. Li, M. Whitwood, S. Millman, and J. Boggelen: Ironmak. Steelmak., 2014, vol. 4, pp. 112–21.CrossRef
15.
R.J. Fruehan, Y. Li, and L. Brabie: Dissolution of magnesite and dolomite in simulated EAF slags. In ISSTech-Conference Proceedings, Iron and Steel Society, 2003, pp. 799–812.
16.
S. Kitamura: ISIJ Int., 2017, vol. 57(10), pp. 1670–76.CrossRef
17.
E. Cheremisina, J. Schenk, L. Nocke, A. Paul, and G. Wimmer: 48 Seminário de Aciaria, Fundição e Metalurgia de Não-Ferrosos, 2017, São Paulo, SP, Brazil, pp. 74–84.
18.
E. Cheremisina, J. Schenk, L. Nocke, A. Paul, and G. Wimmer: Metall. Mater. Trans. B, 2019, vol. 50B, pp. 1269–76.CrossRef
19.
L. Vieira, H. Oliveira, V. Telles, E. Junca, E. Vieira, and J. Oliveira: J. Mater. Res. Technol., 2020, vol. 9(6), pp. 14878–86.CrossRef
20.
Z.S. Li, J. Li, S. Spooner, and S. Seetharaman: Steel Res. Int., 2022, vol. 93, p. 2100167.CrossRef
21.
B. Yu, X.W. Lv, S.L. Xiang, and J. Xu: Metall. Mater. Trans. B, 2016, vol. 47B, pp. 2063–73.CrossRef
22.
M.R. Yang, X.W. Lv, R.R. Wei, J. Xu, and C.G. Bai: Metall. Mater. Trans. B, 2018, vol. 49B, pp. 1331–45.CrossRef
23.
B. Yu, X.W. Lv, S.L. Xiang, and C.G. Bai: ISIJ Int., 2015, vol. 55(3), pp. 483–90.CrossRef
24.
S.L. Xiang, X.W. Lv, B. Yu, J. Xu, and J.Q. Yin: Metall. Mater. Trans. B, 2014, vol. 45B, pp. 2106–17.CrossRef
25.
W. Cho and P. Fan: ISIJ Int., 2004, vol. 44, pp. 229–34.CrossRef
26.
K.H. Sandhage and G.J. Yurek: J. Am. Ceram. Soc., 1988, vol. 71(6), pp. 478–89.CrossRef
27.
B. Yu, X.W. Lv, S.L. Xiang, C.G. Bai, and J.Q. Yin: ISIJ Int., 2015, vol. 55(8), pp. 1558–64.CrossRef
28.
M.R. Yang, X.W. Lv, R.R. Weim, and C.G Bai: Metall. Mater. Trans. B, 2018, vol. 49B, pp. 2667–80.CrossRef
29.
Y. Oishi, Y.A.R. Cooper Jr., and W.D. Kingery: J. Am. Ceram. Soc., 1965, vol. 48(2), pp. 88–95.CrossRef
30.
N.B. Winter: Understanding Cement, Microanalysis Consultants, St Ives, 2012.
31.
K.R. Harris: J. Chem. Phys., 2009, vol. 131(5), p. 054503.CrossRefPubMed
32.
H. Bayat: Optimization of Lime Injection in EAF Systems by Using Pneumatic Sidewall Injectors, Master of Science Thesis. Åbo Akademi University, Turku, Finland, 2022.
