• No results found

A thesis submitted in fulfilment of the requirements for the degree

N/A
N/A
Protected

Academic year: 2023

Share "A thesis submitted in fulfilment of the requirements for the degree "

Copied!
110
0
0

Loading.... (view fulltext now)

Full text

This research project aimed to investigate the effects of particle size distribution (PSD) on the shrink-core model (SKM) for determining the control regime of the ammonium thiosulphate leaching of gold from waste mobile phone PCBs. These PSD models were used to estimate the central values ​​(median and mean magnitudes) and covariances of the PSDs.

Background to Research Problem

Over the years, hydrometallurgical treatment of electronic waste has replaced pyrometallurgical methods due to its lower environmental footprint, ease of operation and relatively higher metal extraction efficiency (Cui & Anderson, 2016; Hanafi et al., 2012; Grosse et al. , 2003). Thiosulfate has been praised for its higher selectivity, lower toxicity and corrosiveness (Hanafi et al., 2012; Grosse et al., 2003).

Table 1-1: E-waste processed in Gauteng (Widmer & Lombard, 2005; ATE, 2012)
Table 1-1: E-waste processed in Gauteng (Widmer & Lombard, 2005; ATE, 2012)

Statement of Research Problem

The leaching process is a series of diffusion stages between the liquid and solid phase as well as kinetic stages. Previous research has established that neglecting the PSD can lead to erroneous conclusions regarding the control regime of the leaching process when modeling with the SCM is performed (Prosser, 1996; Gbor & Jia, 2004).

Research Questions

However, there is still a gap in the body of knowledge regarding the impact of incorporating the PSD and PD into the shrinking core model (SCM) to describe the thiosulfate leaching of gold from waste mobile phone printed circuit boards.

Research Aim and Objectives

Research Hypothesis

Scope of this Study

Thesis Outline

Electronic Waste

Mobile Phone Printed Circuit Boards (PCBs)

2006) state that mobile phone and computer printed circuit boards differ in that computer PCBs are not multiplayer. A mobile phone consists of a polymer fraction, a printed circuit board (PCB), crystal liquid display (LCD), battery, keypad and an antenna (Kasper et al., 2011).

Figure 2-3: A typical printed circuit board (Veit et al., 2006)
Figure 2-3: A typical printed circuit board (Veit et al., 2006)

Characterisation of Mobile Phone PCBs

General E-Waste Processing Scheme

The pre-treatment or refining phase is of crucial importance because it frees the PCB particles from the unwanted non-metallic fraction. It is performed by mechanical-physical separation phases based on physical, magnetic and electrical properties of the PCB constituents (J. Guo et al., 2015; X. Guo et al., 2015).

Metal Extraction by Pyrometallurgy

Copper smelter feed includes domestic concentrates, imported concentrates and spent secondary waste. Potential loss of precious and base metals from slag caused by the presence of ceramic components in e-waste. iv) Incomplete separation of metals is currently achievable, which makes the pyrometallurgical technique more suitable as a pre-treatment operation.

Figure 2-7: The Noranda smelting process (Veldhuizen & Sippel, 1994; Biswas & Davenport, 1994)
Figure 2-7: The Noranda smelting process (Veldhuizen & Sippel, 1994; Biswas & Davenport, 1994)

Metal Extraction by Hydrometallurgy

  • Aqua Regia
  • Cyanide Leaching
  • Halide Leaching
  • Thiourea Leaching
  • Thiosulphate Leaching
  • Metal Extraction by Biohydrometallurgy
  • Comparison of Current Leaching Technologies

However, several studies have reported adverse environmental effects due to the significant amount of cyanide-containing wastewater produced (Zhang et al., 2012). According to La Brooy et al. 1994), the use of halide systems for dissolving gold predates cyanidation. The use of thiourea ((𝑁𝐻2)2𝐶𝑆) as a gold leaching agent has shown promise to extract gold from ores (Veglio et al., 2003).

It is also used in the extraction of other base metals such as nickel and zinc (Rohwerder et al., 2003). However, many of the investigations in this field have focused on copper and gold as metals of interest (Baniasadi et al., 2019).

Figure  2-9  indicates  possible  gold  dissolution  by  using  non-cyanide  lixiviants
Figure 2-9 indicates possible gold dissolution by using non-cyanide lixiviants

Ammonium Thiosulphate Leaching of Gold from E-waste

Leaching Mechanism

According to Isildar et al., (2017), the presence of ammonium ions and cupric ions in the leach system promotes the complex formation of precious metals such as gold and silver. Cupric ion acts as a catalyst in the dissolution reaction and NH4+ as a stabilizer of the system, thereby accelerating the anodic dissolution. According to Molleman & Dreisinger (2002), the chemistry of the ammoniacal thiosulfate system for gold mining involves many interrelated chemical equilibria that are not yet fully understood.

This complexity can be attributed to the presence of the three essential components involved in the leaching system, namely ammonia, thiosulphate and copper (Aylmore, 2016). Previous studies have reported that this leaching process is strictly dependent on pH in the range of 9-10, and operates at ambient temperature to further enhance the thermodynamic stability of the Cu(II)-ammonia-thiosulfate system (Perez & Galaviz, 1987 Zipperian et al. al., 1988; Molleman & Dreisinger, 2002).

Figure 2-10: Electrochemical-catalytic mechanism model for gold leaching with ammonium  thiosulphate (adapted from Aylmore & Muir, 2001)
Figure 2-10: Electrochemical-catalytic mechanism model for gold leaching with ammonium thiosulphate (adapted from Aylmore & Muir, 2001)

Factors Influencing the Ammonium Thiosulphate Leaching of Gold

  • Thermodynamics
  • Temperature
  • Thiosulphate Concentration
  • Ammonia Concentration
  • Cupric Ion Concentration
  • Particle Size

Higher pH conditions can cause sulfide formation, which, in turn, reduces gold extraction efficiency (Zipperian et al., 1988). According to Hung et al. 2011), due to the activation energy exceeding 25 kJ.mol-1, the reaction is chemically controlled. At higher pH values>9, however, gold dissolution becomes independent of ammonia concentration and dependent on thiosulfate concentration, suggesting that the complex 𝐴𝑢(𝑆2𝑂3)23− is more dominant than 𝐴𝑢(𝑁𝝐3)2+ ., 2010).

According to Bell et al. 1995), the most important thiosulphate leaching parameters include temperature, oxygen partial pressure and ammonia concentration. According to Tripathi et al. 2012), the presence of copper ions promotes the dissolution of gold in the thiosulphate solution.

Figure 2-11 shows that the cupric amine complex is most stable within a pH range of 8.5 to 9.5
Figure 2-11 shows that the cupric amine complex is most stable within a pH range of 8.5 to 9.5

Previous Studies on the Thiosulphate Leaching of Gold from E-waste

Thus it can be argued that a significant increase in gold dissolution can be achieved by reducing the PCB particle size. Furthermore, Cu layers cover most of the surface of a typical PCB, which means that Au will leach out with more strain from full-size PCBs (Li & Miller, 2007; Wei et al investigated the effect of particle size on acid thioureamination of gold from milled central processing units (CPUs) with particle sizes smaller than 3 and 0.1 mm The gold extracts obtained were 18.2 and 82%, respectively, showing improvement significantly with smaller particle sizes as a result of improved contact area between the particles and the solution.

Shrinking-Core Model for the Predictive Analysis of Thiosulphate Leaching

The shrinking core model outlined in this section applies only to particles of constant size. The SCM conversion time expressions for non-shrinking particles are given in Table 2-5 and graphically in Figure 2-16. The variable (1 – X) describes the shrinkage of the unconverted core and is directly linked to the conversion.

As the reaction progresses, a gradual increase in conversion is accompanied by a decrease in the unreacted core until all the solid reactant is converted and the remaining particles are converted to ash. The SCM is expressed in terms of unreacted core shrinkage and fractional time for complete conversion, which helps determine the rate-limiting rate of the reactive system (Levenspiel, 1999; Othusitse & Muzenda, 2015).

Figure 2-14: Illustration of unreacted core shrinking as the reaction takes place from the outer  layer (Levenspiel, 1999)
Figure 2-14: Illustration of unreacted core shrinking as the reaction takes place from the outer layer (Levenspiel, 1999)

Previous Work on the Shrinking-Core Model in Hydrometallurgy

Particle Size Distribution and Shrinking-Core Model

Equation (25) can be used for chemical reaction control and liquid film diffusion control since 𝑓(𝐷, 𝑡) can be expressed algebraically for these mechanisms. Equation (25) can be solved numerically using available third-party software packages and libraries. Gbor & Jia (2004) showed that when the covariance or CV coefficient of variation of the particle size distribution can be used to assess whether PSD can be included and excluded from the shrinkage kernel model when determining the limiting rate step or control regime of a drainage process.

The authors indicated that for CV values ​​less than 0.3, the linear relationship between 1 − (1 − 𝑋)1 3⁄ and 𝑡 is maintained for chemical reaction control and PSD can be excluded from SCM. For systems with higher CV values ​​(>0.7), ignoring the PSD will erroneously switch the leaching mechanism from liquid film diffusion to ash diffusion.

Table 2-6: Common particle size distribution functions (Herbst, 1979; Gbor & Jia, 2004; Crundwell et  al., 2013)
Table 2-6: Common particle size distribution functions (Herbst, 1979; Gbor & Jia, 2004; Crundwell et al., 2013)

Research Design

PCB Size Reduction

Unused mobile phones of various makes and models were collected from various local mobile phone outlets.

Aqua Regia Leaching

Ammonium Thiosulphate Leaching

PCBs were added and the reaction mixture was adjusted to 500 mL with deionized water. The temperature of the mixture was adjusted to 30°C and the pH of the solution was adjusted and maintained at 9.5 with the aid of ammonia solution. The leaching process was allowed to proceed for 3 h, and leachate samples were collected every 30 min for metal analysis by AAS.

Table 3-2: Fixed experimental parameters (Tripathi et al., 2012)
Table 3-2: Fixed experimental parameters (Tripathi et al., 2012)

Description of Experimental Apparatus

Description of Materials

  • Ammonium Thiosulphate
  • Copper Sulphate
  • Ammonia Solution
  • Nitric Acid
  • Hydrochloric Acid

It is a blue crystalline salt and regulates the redox potential of the thiosulphate leaching system. It was therefore used in a fume hood with the appropriate protective equipment (respiratory mask, laboratory gloves, safety glasses and shoes). Nitric acid is a strong acid that was used in combination with hydrochloric acid to form aqua regia.

Table 3-4: Copper sulphate specifications
Table 3-4: Copper sulphate specifications

Personal Protection Equipment

Particle Size Distributions

Not only were the GGS and RR models used to estimate the mean magnitude of the PSDs, but they were also used to determine covariances based on the best model fit to the data. It is worth noting that the covariance was used to assess the requirement to incorporate the PSD in the SCM characterization of the leaching process which is further discussed in section 4.4. However, an examination of Figure 4-3 indicates that RR model better fits PSD 2 to 4 due to the closeness of predicted and actual median (D50) magnitudes compared to the GGS model.

These results are in agreement with the literature as the GGS and RR models have been reported to characterize the size distribution of large and small particles, respectively (Gbor & Jia, 2004; Ahmed & Ahmed, 2008). These results are discussed in connection with gold mining and the shrinking core model in the rest of the chapter.

Figure 4-2: Fitting of PSD data to (a) Gates-Gaudin-Schuhmann (GGS) model and (b) Rosin- Rosin-Rammler (RR) model
Figure 4-2: Fitting of PSD data to (a) Gates-Gaudin-Schuhmann (GGS) model and (b) Rosin- Rosin-Rammler (RR) model

Gold Extraction

Reducing pulp density from 120 to 40 g/L improved gold dissolution from average values ​​of 35 to 80. Increasing pulp density resulted in excess PCB solids, making thiosulfate the limiting reagent. Particle size distribution had a significant effect on gold extraction at the lowest pulp density of 40 g/L.

The gold extraction was found to fluctuate between 40 and 50% at 80 g/l pulp density and between 35 and 40% at 120 g/l pulp density, with no clear behavioral pattern of gold extraction with PSD variation. Therefore, based on the experimental results presented in Figure 4-4, it can be concluded that the effect of the pulp density variation was more pronounced than the particle size distribution variation in terms of gold extraction.

Statistical Analysis of Leaching Results

Overall, the estimated marginal means of gold extraction (Figure 4-5) also support the higher significance of the pulp density variation compared to PSD due to the higher variation in gold extraction obtained at different pulp densities compared to different PSDs.

Particle Size Distribution, Shrinking-Core Model and Rate-limiting Mechanism . 56

Particle Size Distribution in SCM Characterisation

SCM fitting to Thiosulphate Leaching Data

At higher pulp densities (80 and 120 g/L), the unreacted core size reductions did not reach satisfactory levels, regardless of the PSD level. None of the predictive mechanisms of the SCM fit the kinetic data (see the model fitting plots in Figure C-1 and Figure C-2 of Appendix C). In addition, they hampered the application of the shrink core model to describe, predict and provide insight into the rate-limiting mechanism of the leaching process.

The possible explanation for the SCM not fitting the leaching data, for larger particles with low pulp density and all particle sizes with higher pulp density, is that the leaching kinetics under these conditions may be influenced by any of the following factors (which were not considered in this research): (i) SCM assumes that a single general parameter, the effective diffusivity (including all mechanical resistances to leaching) can adequately describe the transfer of mass within a reacting solid particle, which can lead to difficulties in implementing SCM for data mining (Liddell, 2005); (ii) the pseudo-steady state assumption (PSSA) of SCM as presented by Levenspiel (1999) suggests that there is a discrete interface between the reacted and unreacted solids sections, but this assumption does not apply to all drainage systems (Liddell, 2005); iii) the effects of mechanical agitation-induced particle disruption and segregation on leaching kinetics (Velardo et al., 2002). The control regime can be further isolated by determining the activation energy of the flushing process (Gbor & Jia, 2004; Crundwell et al., 2013; Khezri et al., 2020).

Figure 4-6: Fitting of SCM to leaching experimental data for 40 g/L pulp density R² = 0.1075
Figure 4-6: Fitting of SCM to leaching experimental data for 40 g/L pulp density R² = 0.1075

Conclusion

Recommendation for Future Research

Precious metal recovery from waste printed circuit boards using cyanide and non-cyanide lixiviants - a review. Recovery of precious metals from waste printed circuit boards using thiosulphate leaching and ion exchange resin. Recovery of metallic values ​​from waste printed circuit boards using an alkali fusion-leaching-separation process.

Extraction of gold, silver, copper and niobium from printed circuits using the leaching column technique. A new process for extracting precious metals from scrap printed circuit boards: using gold concentrate as a fluxing material. Leaching of gold from waste mobile phone printed circuit boards (PCBs) with ammonium thiosulphate.

Bioleaching of copper from waste printed circuit boards by a bacterial consortium enriched with acid mine drainage.

Table A-1: Particle size distribution on frequency basis  Particle Size
Table A-1: Particle size distribution on frequency basis Particle Size

Figure

Table 2-2: The general weight composition (wt. %) of mobile phone PCBs
Figure 2-5: General e-waste processing scheme (Hanafi et al., 2012; Syed, 2012; Lu & Xu, 2016)
Figure 2-6: A typical beneficiation scheme for e-waste treatment (Wang & Xu, 2015)
Figure 2-7: The Noranda smelting process (Veldhuizen & Sippel, 1994; Biswas & Davenport, 1994)
+7

References

Related documents

CONSTITUTIONAL COURT OF SOUTH AFRICA CASE CCT 88/14 WCHC NO: 5244/13 In the matter between: CHEVRON SA PTY LIMITED Applicant and DENNIS EDWIN WILSON t/a WILSON’S TRANSPORT