The Distribution Behavior of Elements during the Top-Blowing Smelting Process of Electronic Waste

Abstract

In this work, the local equilibrium modeling method of a non-equilibrium multi-phase reaction system in the top-blowing melting process of electronic waste was studied. The automatic judgment mechanism of phase transformation and the improvement of the trace component solving algorithm were explored to build the mathematical model of the element migration and transformation. Secondly, to determine the distribution mechanism of various elements in top-blowing smelting of electronic waste, the thermodynamic digital simulation system was developed according to the software platform of metallurgical process calculation. On this basis, combined with the industrial production practice, the coupling simulation experiment was carried out to investigate the influence of oxygen:feed ratio, oxygen concentration, amount of additive iron powder and CaO:SiO₂ ratio of the slag on the smelting process. In addition, the direct yields of metals in the slag were Cu 90.69 wt%, Au 98.57 wt%, Ag 94.84 wt%, and Pd 97.87 wt% under the optimum conditions. Finally, the simulated values were consistent with industrial data, which can provide theoretical guidance for the industrial production practice of the top-blowing smelting of electronic waste.

1. Introduction

The twenty-first century is an era of information and technology, which has greatly promoted the progress and development of society. Especially, the renewal cycle of electronic products has been shortened, resulting in a large accumulation of obsolete products due to the dramatic changes to the electronics industry [1,2,3]. More importantly, electronic waste will cause new pollution problems and serious effects on human health and the ecological environment if it was not disposed properly, which contains a lot of heavy metals and other hazardous substances [4,5,6,7]. On the other hand, the electronic waste is essentially a valuable urban mine which contains a large number of valuable metals, such as copper, lead, tin, nickel, antimony, gold, etc. [8,9]. Thus, recycling and utilizing these valuable metals can not only obtain good economic benefits but also produce far-reaching ecological benefits to achieve economic and environmental benefits in a win-win situation.

At present, the harmless treatment of electronic waste has become a hot spot in the world. Many scholars have developed new methods for efficient and environmentally friendly recycling of valuable metals from electronic waste, which can be differentiated into the mechanical process, hydrometallurgical process, biometallurgical process and pyrometallurgical process [10,11]. Among them, the mechanical treatment method is used to separate and enrich valuable components by gravity separation, magnetic separation, electric separation, eddy current separation and flotation based on the differences in density, specific gravity, conductivity, magnetism and toughness among heavy metals, precious metals, silicon and resin in the electronic waste [12]. Generally, it is a pretreatment process for the recycling of electronic waste, because the purity of the separated metal phases cannot meet the requirements of industrial production. In contrast, the hydrometallurgical process is used to dissolve the crushed electronic waste particles in acid or alkali at first, after separation and deep purification to remove impurities, the valuable metals are concentrated and enriched by solvent extraction, adsorption or ion exchange, and then recovered according to electrowinning, chemical reduction or crystallization [13]. It has the advantages of high purity of products, great controllability and flexibility, but it also has shortcomings, such as low leaching rate of metals, long process flow, large consumption of leaching agent and difficulty in wastewater treatment [14]. In addition, the biometallurgy technology is used to oxidize the valuable metals into solution based on the catalytic oxidation of microorganisms under acidic condition, which has been successfully industrialized for leaching of the sulfide ore. However, as the valuable metals in the electronic waste mainly exist in the form of a simple substance or alloy, it is only at the research stage in terms of the limited strains, small scale and long leaching time [15].

Currently, many scholars believe that pyrometallurgical technology is still the most feasible method for the efficient treatment of electronic waste [16], which has the characteristics of a large processing capacity and strong adaptability to raw materials. It is burned in the furnace or molten pool to remove refractory oxides such as plastics and glass fibers so that the impurity metal oxides form the slag phase and separate from the molten metal phase. In the past 20 years, pyrometallurgical technology has successively developed incineration dissolution, high-temperature oxidation smelting, pyrolysis method, electric arc furnace sintering and other processes [17]. At present, the outstanding electronic waste processing companies adopt the pyrometallurgical processes, including Kaldor furnaces of Boliden, ISA furnaces of Germany and Belgium and Ausmelt furnaces of Japan and South Korea. In addition, the Nerin Recycling Technology (NRT) also belongs to the pyrometallurgical technology of electronic waste, which the smelting furnace independently developed by China is a high-temperature-intensified top-blowing smelting furnace with prominent prospects for extension and application.

However, how to increase the direct yield of valuable metals, reduce the emission of toxic and harmful gases and realize the high-efficiency and clean production of electronic waste is the current technical problem of the high-temperature top-blowing molten pool smelting.

According to the kinetics research of top-blowing smelting [18,19], the reaction between the electronic waste and gas has the typical characteristics of “uniform mixing and rapid reaction”, which is an obvious high-temperature non-equilibrium multiphase reaction system. The physical and chemical behavior of electronic waste is directly related to the melting, decomposition, oxidation, slagging, organic cracking, etc., and effects the migration and transformation behavior of various components in the product and the degree of organic cracking reaction, thus determining the technical parameters, including the direct yield of valuable metals and the toxic equivalent of flue gas [20]. Many scholars have done a lot of research on the release of dioxins during the high-temperature smelting of electronic waste. To reduce the production of dioxin, Sakai et al. [21] studied the low-temperature comprehensive heterogeneous catalysis in the post-combustion zone, and successfully converted HBr to Br₂ without adding S. In addition, Bientinesi et al. [22] proposed a method of using lye to absorb and recover Br₂ from the absorption liquid based on the proportion of bromide in the smoke and dust in the process of mixed combustion and staged vaporization of electronic waste. Based on the high-temperature combustion experiments, Ming [23] found that the organic bromine can be converted into inorganic bromine at high enough temperature and air excess coefficient. Besides, 99.9% of the bromine will enter the flue gas to form HBr and Br₂ when the air excess factor up to 1.3 at 1200 °C. Thus far, the industry mainly avoids the production of dioxins by destroying the precursors of polychlorinated diphenyl (PCDD) and polychlorinated diphenyl furan (PCDF). For solving this problem, the Umicore of Belgian, one of the most technologically advanced recycling companies of electronic waste in the world, achieves the effective separation of the organic phase and the metal phase based on the ultra-high temperature processing technology of the ISA furnace.

Therefore, to increase the recovery rate of valuable metals and reduce the emission of toxic gases, it is necessary to investigate the phase migration and chemical evolution during the top-blowing smelting of electronic waste. After building the mathematical model and simulation system for the transformation of electronic waste via the principles of material balance, heat balance and chemical balance, the micro-physicochemical behavior of electronic waste in the top-blowing smelting was studied combined with the production practice of the industry. In this study, the multi-factor coupling experiments were carried out, and the effects of oxygen:feed ratio, oxygen concentration, amount of additive iron powder and CaO:SiO₂ ratio of the slag on the distribution behavior of elements in the top-blowing smelting process were investigated. It can provide great theoretical and practical significance for promoting the transformation and upgrading of industrial technology for green and efficient recycling of electronic waste.

2. Thermodynamic Mathematical Model and Simulation System

2.1. The Top-Blowing Smelting Process of Electronic Waste

The top-blowing smelting of electronic waste mainly enriches the rare and precious metals in the crude copper to form a metal phase, which can realize the recovery of valuable metals after separation from the non-metal phases, as shown in Figure 1. It can be seen that the crushed electronic waste, slagging flux, industrial oxygen, air and additives (iron powder) were placed into the high-temperature furnace together. Among them, refractory oxides such as plastics and glass fibers could be combusted and form the flue gas and smoke dust phases according to the waste heat boiler, while the impurity metals were oxidized and form the slag phases with fluxing. In particular, the metal elements, such as copper, gold, silver, platinum and palladium, form the metal phases which were mutually insoluble and layered with the slag due to different densities.

The distribution behavior of each element in the top-blowing smelting process of electronic waste is an important part of the entire process and has a significant impact on the optimization of smelting and separation process parameters. In addition, the component migration and chemical evolution of electronic waste in various products are directly related to the processes including melting, decomposition, oxidation and slagging, which in turn affect parameters of the direct yield of metals and dusting rate. NRT smelting technology is a typical top-blowing smelting process and has the advantages of low cost, small investment, simple operation, large-scale production capacity, environmentally friendly, etc. It is expected to become a new integrated treatment technology of electronic waste [24].

In this work, the cutout of the copper-clad laminate was selected as the raw material, and provided by a non-ferrous smeltery from Jiangxi Province in China, as displayed in

Table 1. Chemical compositions of the cutout of the copper-clad laminate.

Components	Cu	Au (g/t)	Ag (g/t)	Pd	Pb	Fe	SiO2	CaO	Al2O3	B2O3	H2O
(wt%)	19.43	0.20	0.50	0.48	0.038	0.19	16.91	9.59	6.84	16.89	5.00
Organic	C2H3Cl	C2H4		CH3NO2		C7H8O2	C12H6Cl4		C12H8Br2	C15H16O2
(wt%)	1.20	2.39		0.72		1.67	0.72		2.87	14.34

. It was mainly composed of copper, silicon dioxide, boron trioxide, aluminum oxide and calcium oxide, the contents of which were 19.43, 16.91,16.89, 6.84 and 9.59 wt%, respectively. However, the ratio of Fe:SiO₂ was only 0.011 in terms of the low content of iron, which was difficult to obtain the slag with suitable melting-point and viscosity. To effectively solve this problem, the additives of iron powders and calcium oxide were added to the slag with mass fractions of 95 and 86 wt%, respectively, according to our previous work [25]. Moreover, the amount of organic matter in the copper-clad laminate was approximately 23.9 wt%.

2.2. Improvement of the Thermodynamic Mathematical Model

As the top-blowing smelting with a certain scale treatment of electronic waste is a stabilization process, it can be assumed that the entire process is divided into n regions with the same reaction time. Firstly, the reaction gas and electronic waste can be considered to reach or close to the chemical equilibrium state in terms of rapid reaction and uniform mixing during the top-blowing melting process. In the first region, after reaching the chemical equilibrium state, the raw material (m_Raw) with a mass of ∆m reacts with oxygen by a volume of V_las-phase and consumes ∆V gas. Similarly, at the last region, the remaining material of ∆M will reach chemical equilibrium with the reaction gas by a volume of [V_Total − (n − 1) ∆V]. Finally, the thermodynamic model of the top-blowing smelting process can be obtained by building the multi-phase equilibrium model of each region and using the phase transfer between the regions, as shown in Figure 2.

However, the top-blowing smelting of electronic waste is a multi-phase transformation process, during which it is difficult to calculate the trace components of the new phase or the disappearing phase. In addition, the typical free energy minimization algorithms are prone to iterative divergence and negative overflow, resulting in model calculations that cannot be performed normally [18]. Hence, an effective judgment mechanism is investigated to automatically determine the phase change, and an efficient algorithm can be constructed to realize the effective solution of trace components. That is, the element potential algorithm is introduced to avoid negative overflow during the solution process, so as to solve the problems of phase transition and trace component calculation in the top-blowing smelting multi-phase reaction system, which can provide an effective approach for the analysis of the chemical evolution principle of the electronic waste.

In addition, the top-blowing smelting of electronic waste is an oxygen-rich top-blowing submerged smelting technology based on the reactions among the combustible components, metals, non-metals and oxygen in materials, which has the characteristics of uniform mixing of the solid-gas phases and rapidly reaction. Consequently, the top-blowing smelting process can be regarded as a multi-phase reaction system in which the electronic waste and reaction gas continuously reach a local equilibrium. After introducing the local equilibrium assumption of the non-equilibrium system, the time discretization in milliseconds will be carried out. Simultaneously, the smelting process is divided into some time infinitesimal locales, and the multiphase equilibrium mathematical model can be constructed based on the principle of minimum Gibbs free energy. Secondly, based on the transfer relationship between the electronic waste and gas in each micro-element localization, the local models are correlated to establish a mathematical model for the migration and transformation of electronic waste during the top-blowing smelting. Finally, combing with the Visual C#.net, Delphi and SQL Server database, the corresponding digital simulation system can be developed by the method of object-oriented programming technology.

For the local area of i, the total Gibbs free energy can be expressed according to Equation (1). Based on the principle of minimum Gibbs free energy, the phase equilibrium composition of the i local area is obtained after solving the minimum value.

$$G_i={\displaystyle \sum _{p=1}^p{\displaystyle \sum _{c=1}^{Cp}[x_{pc}{G}_{pc}^{\theta }}}+RT\ln ({\gamma }_{pc}{x}_{pc}/{\displaystyle \sum _{k=1}^{Cp}{x}_{pk}}),$$

(1)

where P denotes the number of phases in the i area; C_p is the component fraction in the p phase; x_pc denotes the mole fraction of c component in the p phase; γ_pc is the activity coefficient of c component in the p phase; $G_{pc}^{\theta }$ denotes the standard Gibbs free energy of formation of component c in the p phase; R is the universal constant of gas; T denotes the reaction temperature.

To avoid the phase transition of P and the small amount of newly formed phase components (that is, x_pc tends to 0) during the melting process. By implanting the phase transformation judgment mechanism and introducing the element potential algorithm, the problem of multi-phase transition in the top-blowing smelting process is solved. In view of the mechanism of phase transformation judgment, the value of P is initially determined according to the top-blowing smelting process of electronic scrap. For instance, the value of P is five when the crude copper, slag, magnetite and flue gases reach equilibrium. Secondly, the disappearance or formation of phase can be judged by calculating the amount of phase formation or the activity of component to determine the p value. The magnetite phase is about to disappear if the generated amount of magnetite approaches zero and the value of P should be reduced by one. On the contrary, the activity of Fe₃O₄ in the slag phase approaches one which means that the magnetite phase is about to be formed and the value of P should be increased by one. Thus, the information of the corresponding phase can be deleted or added from the system according to the change of the P value.

At the beginning of each phase transition, the number of moles x_pc of the new phase or the disappearing phase will tend to zero. At this time, the iterative solution process of the conventional algorithm will overflow with negative numbers and cannot proceed smoothly, and the element potential is introduced in this work. In addition, the number of moles of components in different phases can be related to the number of atoms of the elements by the Lagrange factor, which becomes the exponential function of the standard Gibbs free energy of formation, the number of element atoms and the corresponding Lagrange factor, as shown in the Equation (2).

$$x_{p_c}=\exp [-G_{p_c}^{\theta }/(RT)+{\displaystyle \sum _{e=1}^E{A}_{p_{ce}}{\lambda }_e/(RT)}]/{\gamma }_{p_c}$$

(2)

where A_pce denotes the number of e atoms in the c component of the p phase; λ_e is the Lagrange factor of the e atom.

2.3. Establishment of the Simulation System

The phases involved in the top-blowing smelting process of electronic waste mainly included the crude copper, slag, magnetite, flue gases, etc. The electronic waste, temperature and reaction atmosphere were constantly changing during the smelting process within 3 to 5 h. In addition, the entire reaction system continuously underwent phase transformation and migration, and gradually completes complex reactions such as oxidation, combustion, slagging, and copper production, which were gradually completed in the whole reaction system and had obvious characteristics of non-equilibrium heterogeneous chemical reactions. For this kind of complex system, it is difficult to characterize the actual situation in laboratory experiments. Many researchers try to directly carry out the experimental exploration in industrial furnaces. However, there are many difficulties in carrying out a large number of experiments in the industry due to the production practice requires relatively stable process conditions. In addition, it is not easy to investigate the impact of various process parameter changes on the physicochemical behavior of electronic waste.

Therefore, according to the above-mentioned thermodynamic mathematical model and the software platform of metallurgical process calculation [19], combined with the object-oriented programming method for studying the migration and transformation of digital models [26], a thermodynamic digital model simulation system for the top-blowing smelting process of electronic waste was developed, as illustrated in Figure 3. More importantly, this system can combine different metallurgical units in an orderly manner, providing an effective means for effectively calculating the material and heat balance of electronic waste, and studying the distribution behavior of various elements during the smelting process.

3. Results and Discussion

3.1. Analysis of Coupling Simulation

The thermodynamic mathematical models can be verified and optimized based on obtaining process parameters and reaction characteristic data with the production practice of the NRT smelting, and further multi-factor coupling simulation experiments are carried out. In this work, the influences of oxygen:feed ratio, oxygen concentration, amount of additive iron powder and CaO:SiO₂ ratio of the slag on the yield of each phase and the recovery rate of copper, gold, silver and platinum were investigated. In addition, the phase transformation, component migration and chemical evolution of electronic waste during the top-blowing smelting are discussed in detail.

3.1.1. Influence of Oxygen:Feed Ratio

The influences of various oxygen:feed ratios (that is, 315, 320, 325, 330, 335, 340, and 345 Nm³/t) on the top-blowing smelting of electronic waste were investigated, as displayed in Figure 4. Especially, the simulation conditions control are as follows: The feed rate of scrap copper-clad laminate with 3.13 t/h, amount of additive iron powder by 120 kg/h, CaO:SiO₂ ratio of 1.22, the oxygen concentration of 21%, smelting time of 4 h, and the temperature of the crude copper and slag is 1250 and 1280 °C, respectively.

It can be seen from Figure 4a that the oxygen:feed ratio has a significant impact on the production rate of crude copper and slag, but has almost no impact on the smoke and dust. That is, with the increase of oxygen:feed ratio, the yields of crude copper and slag continued to decrease and increased significantly, respectively. At the same time, the contents of FeO and Fe₃O₄ in the slag slowly decrease and increase, respectively, while the Cu content increases rapidly as seen in Figure 4b, which means that the FeO is further oxidized to Fe₃O₄, and a large amount of Cu is oxidized into the slag phase. Figure 4c,d, respectively show the influence of different oxygen:feed ratios on the distribution rate of the valuable metals of copper, gold, silver, palladium in the blister copper phase and the slag phase during the smelting process. With the increase of oxygen:feed ratio, the distribution ratio of copper in the blister copper phase decreases continuously, while the slag phase shows an opposite trend. This is because more oxygen enters the furnace, which increases the degree of copper oxidation. On the other hand, the distribution ratios of precious metals, such as gold, silver, palladium, etc., in the blister copper and slag phase slowly decrease and increase, respectively. This indicates that the increase of oxygen:feed ratio will accelerate the oxidation loss of precious metals, and the changing trend is less than that of copper, mainly because the noble metals have a low affinity to oxygen.

In addition, the effect of different oxygen:feed ratios on the temperature of flue gas during the smelting process is shown in Figure 4e. It can be found that the flue gas temperature drops from 1325 to 1278 °C when the oxygen:feed ratio increases from 315 to 345 Nm³/t, that is, the increase of oxygen:feed ratio increases the amount of nitrogen brought into the furnace, thereby increasing the heat removal from the flue gas. The flue gas temperature will decrease according to the calculation of heat balance under the condition of the constant temperature of crude copper and slag. Moreover, the relationship between the oxygen potential (Log PO₂) and oxygen:feed ratio can be seen in Figure 4f, and the oxygen potential increases rapidly as increasing the oxygen:feed ratio. Especially, the Log PO₂−9.77 corresponds to the oxygen:feed ratio of 315 Nm³/t.

Thus, to have a high direct yield of copper, gold, silver, and palladium, and ensure the complete combustion of non-metallic organics, it is appropriate to select the oxygen:feed ratio at about 315 Nm³/t.

3.1.2. Influence of Oxygen Concentration

The influences of various oxygen concentrations (that is, 21%, 25%, 30%, 35%, 40%, 45%, and 50%) on the top-blowing smelting of electronic waste were investigated, as shown in Figure 5. In addition, the simulation conditions control are as follows: The feed rate of scrap copper-clad laminate with 3.13 t/h, amount of additive iron powder by 120 kg/h, CaO:SiO₂ ratio of 1.22, the oxygen:feed ratio of 315 Nm³/t, smelting time of 4 h, and the temperature of the crude copper and slag is 1250 and 1280 °C, respectively.

It can be seen that the oxygen concentration has no significant effect on the production rate of blister copper, slag, and smoke dust. However, with the increase of oxygen concentration, the content of copper and silver in the slag showed a decreasing trend, as displayed in Figure 4b. At the same time, the influence of oxygen concentration on the distribution ratio of different valuable metals in the copper phase and slag phase can be seen in Figure 5c,d, respectively. In addition, the increase of oxygen concentration has no obvious effect on the distribution ratio of copper, gold, silver and palladium in different phases. As can be seen in Figure 5e, with the increase of oxygen concentration from 21% to 50%, the amount of flue gas dropped significantly, from 6700 to about 3200 Nm³. Moreover, the amount of N₂ brought into the furnace has dropped significantly, which leads to a continuous decrease in the amount of flue gas. In addition, the flue gas temperature also increases rapidly with the increase of oxygen concentration.

3.1.3. Influence of Amount of Additive Iron Powder

The influences of various amounts of additive iron powder (that is, 80, 85, 90, 95, 100, 105, 110, 115 and 120 kg/h) on the top-blowing smelting of electronic waste were investigated, as shown in Figure 6. The simulation conditions control are as follows: The feed rate of scrap copper-clad laminate with 3.13 t/h, oxygen concentration of 21%, CaO:SiO₂ ratio of 1.22, the oxygen:feed ratio of 315 Nm³/t, smelting time of 4 h and the temperature of the crude copper and slag is 1250 and 1280 °C, respectively.

It can be found that with the increasing amount of additive iron powder, the production rates of the blister copper, slag and smoke dust phases remain stable. However, the concentrations of Cu, FeO and Fe₃O₄ in the slag and the amount of iron oxidized have been increased at the same time. In addition, the effect of the amount of additive iron powder on the distribution rate of valuable metals in different phases can be illustrated in Figure 6c,d. With increasing the amount of additive iron powder, the distribution rate of copper in the blister copper and slag phases slowly decreases and increases, respectively, while having little effect on precious metals such as gold, silver and palladium. Simultaneously, the temperature of the flue gas during the smelting of electronic scrap increased slightly since the increase of heat released by iron oxidation.

3.1.4. Influence of CaO:SiO₂ Ratio

The influences of various CaO:SiO₂ ratios (that is, 1.1, 1.15, 1.2, 1.25, 1.3, 1.35 and 1.4) on the top-blowing smelting of electronic waste were investigated, as shown in Figure 7. In addition, the simulation control conditions are as follows: The feed rate of scrap copper-clad laminate with 3.13 t/h, amount of additive iron powder by 120 kg/h, the oxygen:feed ratio of 315 Nm³/t, the oxygen concentration of 21%, smelting time of 4 h and the temperature of the crude copper and slag is 1250 and 1280 °C, respectively. It can be found from Figure 7a that the production rate of crude copper slowly decreased with the increase of CaO:SiO₂ ratio in the slag, while the yield of the slag phase continued to increase. This is mainly because more flux of limestone is added to improve CaO:SiO₂ ratio in the system. On the other hand, the amount of FeO decreased and Fe₃O₄ increased slightly in the slag by the increase of CaO:SiO₂ ratio, and the copper content of the slag increased rapidly as shown in Figure 7b. As shown in Figure 7c,d, the degree of copper oxidation is increased in terms of the increase of oxygen entering the furnace. Then the distribution ratio of copper in the blister copper phase continues to decrease, while the opposite trend is presented in the slag phase. In addition, the distribution ratio of precious metals including gold, silver and palladium in the blister copper and slag phase slowly decreased and increased respectively, which means that the increase of CaO:SiO₂ rate is unfavorable for the recovery of valuable metals. Figure 7e shows that, as the CaO:SiO₂ rate increases, the temperature of flue gas continues to drop because the decomposition of limestone consumes a lot of heat.

In addition, according to the previous research results [25], to obtain a slag type with good viscosity and melting point, it is necessary to add an appropriate flux to increase the content of CaO in the slag. At the same time, to ensure a higher direct yield of valuable metals, the CaO:SiO₂ ratio in the slag should be about 1.20 in this work.

3.2. Distribution Mechanism of Elements in the Top-Blowing Smelting

According to the above simulation results, the optimal conditions for top-blowing smelting of scrap copper clad laminates are as follows: Processing capacity of 3.13 t/h, amount of additive iron powder of 120 kg/h, smelting time 4 h, the CaO:SiO₂ ratio of 1.20, oxygen:feed ratio of 315 Nm³/t, the oxygen concentration of 21%, smelting time of 4 h and the temperature of crude copper and slag is 1250 and 1280 °C, respectively. In addition, the temperature of smoke and flue gas is automatically calculated via the heat balance.

Combining with the industrial production practice of the NRT smelting, the typical top-blowing smelting of electronic waste, the verification of the simulation system was investigated in this study. In addition the calculation of the material balance, contents of products and heat balance were carried out, respectively.

The input–output material balance with the NRT smelting of electronic waste is shown in

Table 2. The input–output material balance of the NRT smelting.

No.	Input				Output
	Phase	°C	kg/h	Nm3/h	Phase	°C	kg/h	Nm3/h
1	Scrap of copper-clad laminate	25	3130.00		Crude copper	1250	566.07
2	Iron powder	25	120.00		Slag	1280	2052.88
3	Limestone	25	718.32		Smoke	1327	8970.25	6766.85
4	Arenaceous quartz	25	0.00		Dust	1327	162.50
5	Industrial oxygen	25	0.00	0.00
6	Air	25	7783.39	6046.96
Total			11,751.70	6046.96	Total		11,751.70	6766.85

. After inputting the corresponding materials according to the above-mentioned optimal process parameters, the crude copper of 566.07 kg/h, the slag of 2052.88 kg/h, the smoke of 8970.25 kg/h and the dust of 162.50 kg/h are produced. In addition, the temperature of smoke and dust can be reached at 1327 °C.

Table 3. The product contents of the NRT smelting.

No.	Crude Copper		Slag		Smoke			Dust
	Phase	wt%	Phase	wt%	Phase	wt%	v%	Phase	wt%
1	Cu	97.42	FeO	5.96	CO2	25.27	17.06	Cu	16.98
2	Au	1.04 × 10−4	SiO2	25.46	O2	0.12	0.11	Pb	0.05
3	Ag	2.50 × 10−4	Cu2O	1.59	N2	66.62	70.66	Fe	13.21
4	Pb	1.29 × 10−4	CaO	30.55	H2O	7.07	11.66	SiO2	23.49
5	Pd	2.58	Fe3O4	0.33	Pb	0.00	0.00	CaO	13.32
6			PbO	0.05	PbO	0.00	0.00	Al2O3	9.49
7			Au	0.00	HCl	0.38	0.31	B2O3	23.46
8			Ag	0.00	Cl2	0.00	0.00
9			Pd	0.02	Br2	0.01	0.00
10			Ag2O	0.00	HBr	0.53	0.19
11			Al2O3	9.67
12			B2O3	23.90
13			Other	2.47
	Total	100.00	Total	100.00	Total	100.00	100.00	Total	100.00

shows the contents of products obtained from the NRT smelting of electronic waste based on the thermodynamic simulation system. It can be seen that the crude copper is mainly composed of Cu, Au, Ag and Pd, meanwhile, a small amount of copper and lead are present in the dust. In addition, 5.96, 25.46, 30.55, 9.67 and 23.87 wt% of the FeO, SiO₂, CaO, Al₂O₃ and B₂O₃ existed in the slag phase. Moreover, the smoke phase consisted of CO₂ and N₂ with the corresponding contents of 25.27 and 66.62 wt%, respectively. Especially, other chemical compounds such as small molecular alkanes, benzene compounds, bromine volatile, coke and glass fiber may exist in the gas and dust phases according to the study of a predecessor [17].

The input-output heat balance with the NRT smelting of electronic waste can be seen in

Table 4. The input-output heat balance of the NRT smelting.

No.	Heat-Income					Heat-Output
	Type	Materials	T/°C	MJ/h	%	Materials	T/°C	MJ/h	%
1	Physical heat	Scrap of copper-clad laminate	25	0	0	Copper	1250	354.04	1.62
		Iron powder	25	0	0	Slag	1280	3601.73	16.43
		Limestone	25	0	0	Smoke	1327	14,366.36	65.55
		Arenaceous quartz	25	0	0	Dust	1327	243.78	1.11
		Industrial oxygen	25	0	0
		Air	25	0	0
		Total		0	0	Total		18,565.91	84.71
2	Chemical heat	-	25	21,916.25	100		25
3	Exchanged heat	Top cooling water	25			Top cooling water	30	1044.34	4.77
4	Natural Cooling	-					208	2306	10.52
	Total	-		21,916.25	100			21,916.25	100

. The heat in the furnace is ensured by exothermic reactions, where the amount of heat income was 21,689.46 MJ/h. On the other hand, the heat output consists of physical heat, chemical heat, exchanged heat and natural heat dissipation. Among them, the smoke heat played an important role in the heat-output stage, which released approximately 14,366.36 MJ/h of the heat and the total physical heat can be up to 18,565.91 MJ/h. Besides, the exchanged heat and natural heat dissipation were 1044.34 and 2306.00 MJ/h, respectively.

Moreover, under the optimal process conditions, the mass percentages of copper, gold, silver and palladium in the slag of the NRT smelting are 1.41%, 4.15 × 10⁻⁷%, 3.75 × 10⁻⁶%, and 0.02%, respectively. In addition the direct yields were 90.69, 98.57, 94.84 and 97.87 wt%, respectively. Therefore, the distribution mechanism of elements in the top-blowing smelting of electronic waste can be revealed.

3.3. Verifications with Industrial Results

The chemical components of the samples obtained from the top-blowing melting process were analyzed via direct-reading inductively coupled plasma-optical emission spectrometer (ICP-OES, IRIS Intrepid II XSP, Thermo Fisher Scientific Co., Waltham, MA, USA) and X-ray Fluorescence Spectrometry (XRF, AXIOS-MAX, Almelo, The Netherlands), respectively. More importantly, the verification of the simulation system can be carried out according to industrial production. Then the phase composition, feed rate and other parameters of the optimization model will directly guide the production practice.

Figure 8 shows the verifications of the thermodynamic mathematical model for the simulation system between the industrial results and simulated results according to the NRT smelting production practice. It can be seen that the industrial results of various elements in the crude copper are consistent with the simulation values. In addition, the predicted values of each phase in the slag are in good agreement with the industrial data. More notably, the part of industrial values is slightly higher than the simulation results due to the mechanical inclusion. Moreover, photographs of the crude copper and slag phases were obtained from the non-Ferrous metallurgical enterprises. Thus, the simulation system based on the thermodynamic mathematical model is of great significance for the industrial production practice of the top-blowing smelting process of electronic waste.

4. Conclusions

• After investigating the automatic judgment mechanism of phase transformation and the improvement of the micro-component solution algorithm, the mathematical model of element migration and transformation in the top-blowing smelting process of electronic waste was verified and optimized, and the corresponding simulation system was also established;
• The multi-factor coupling simulation experiment was carried out to investigate the influences of oxygen:feed ratio, oxygen concentration, amount of additive iron powder and CaO:SiO₂ ratio of the slag on the production rate of different phases, the distribution ratio of valuable metals and the temperature of the flue gases. In addition, the conditions for the top-blowing smelting were clarified, and the direct yields of copper, gold, silver and palladium can reach 90.69, 98.57, 94.84 and 97.87 wt%, respectively.
• The simulation results are in good agreement with the industrial data in this study, which can provide theoretical guidance for the industrial production practice of the top-blowing smelting process of electronic waste.