Modeling copper leaching from non-pulverized printed circuit boards at high concentrations of bioregenerated ferric sulfate

Minerals Engineering 217 (2024) 108913 Available online 16 August 2024 0892-6875/© 2024 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). Modeling copper leaching from non-pulverized printed circuit boards at high concentrations of bioregenerated ferric sulfate Pablo Ramírez a , * , Nieves Iglesias-Gonz´ alez a , Antonio D. Dorado a , b a Chemical Engineering Department, University of Seville, Calle Profesor García Gonz´ alez, Facultad de Química, 41012 Sevilla. Spain b Department of Mining, Industrial and ICT Engineering, Universitat Polit` ecnica de Catalunya, Avinguda de les Bases de Manresa 61-73, 08242 Manresa, Spain ARTICLE INFO Keywords: Non-pulverized PCBs Biohydrometallurgy Copper Bioreactor Modeling Leaching kinetics ABSTRACT This work studies the leaching of copper contained in waste printed circuit boards (PCB) with ferric sulfate, with the aim of improving the results found in the literature for an industrial application of the process. For this purpose, temperatures and concentrations of ferric ion higher than those of previous works (up to 60 ◦C and 40 g/L) are used. Furthermore, ferric sulfate can be continuously regenerated by ferro-oxidizing bacteria immobilized in a bioreactor, avoiding physical contact between waste and microorganisms to optimize independently operation conditions. A kinetic model was developed for the interpretation of the experimental results validated at different conditions. The model was based on a shrinking core model limited by mass transfer inside the PCB where the geometry and structure of the PCB has been considered and an excellent fit of the model to the experimental data was obtained. The connection with the bioreactor for biooxidation of the generated ferrous ion allows maintaining high concentrations of ferric when operating at high solid concentrations (5–10 %). In all the cases studied, more than 99 % of the copper contained in the non-pulverized PCB parts was dissolved in less than 24 h, with an average copper extraction rate of 0.6 g/L⋅h higher than those found in previous works. 1. Introduction Printed circuit boards (PCBs) are components of electronic waste with high content of base and precious metals, generally at higher concentrations than those found in natural ores. Due to the current technological development, there is a continuous growth of e-waste and, consequently, of PCBs, which makes it necessary to improve the knowledge and investigate the development of new processes to return the valuable metals contained to the value chain, that is, to generate circular economy processes to allow metals to give a second life and minimize the environmental impact. On the one hand, waste is treated, and on the other, valuable elements are recovered and it is minimized the dependence of exploitation from nature (Kumari and Samadder, 2023). Among the metals used for the production of PCBs, copper stands out for its abundance, between 10 and 30 % (Oguchi et al., 2011). Moreover, copper has been classified as critical raw material in the last European Commission Report due to it is considered a strategical raw material (European Commission, 2023). The most common process used to recover copper from waste PCBs is based on grinding and mechanical/ magnetic separation of the components (Mir and Dhawan, 2022) resulting in ground copper with low purity for its use. Subsequently, the copper recovered by this way is conducted to a pyrometallurgical treatment to obtain high purity copper. Taking into account that smelting is an energy consuming, the reliance on fossil fuels can be partially offset by taking advantage of the plastic content of PCBs as a fuel, as well as a reducing agent in smelting. However, this treatment involves a large environmental burden (Rao et al., 2020). In recent years, numerous studies have been carried out to offer an alternative for the recovery of copper from PCBs that does not imply the use of pyrometallurgical processes, particularly processes in which copper is obtained in solution by hydrometallurgical and biohydrometallurgical methods. Biohydrometallurgical methods arouse the most interest for using processes with low consumption of reagents and at low temperatures, meaning an alternative of great interest from an environmental point of view (Baniasadi et al., 2019; Ilyas et al., 2021). However, currently, the productivity obtained in these processes is still limited. The proposed processes can be divided into all-in-one processes in which the pulverized or chunked PCBs are introduced into a bioreactor together with the bacteria that regenerate the ferric ion, the leaching agent, or two-stage processes in which the regeneration of the ferric occurs in a bioreactor separated from the PCB residue (Iglesias-Gonz´ alez * Corresponding author. E-mail address: [email protected] (P. Ramírez). Contents lists available at ScienceDirect Minerals Engineering journal homepage: www.elsevier.com/locate/mineng https://doi.org/10.1016/j.mineng.2024.108913 Received 24 April 2024; Received in revised form 16 July 2024; Accepted 6 August 2024 Minerals Engineering 217 (2024) 108913 2 et al., 2021; Shoda et al., 2020). Carrying out the process in two independent stages has the great advantage of being able to optimize each of them individually, since the best conditions for the leaching reaction do not coincide with the best conditions for the bioregeneration of the leaching agent. The two-step bioleaching process offers several advantages for extracting metals from electronic wastes (e-wastes): continuous regeneration of Fe(III), higher waste concentration utilization, and increased metal yields (Cui and Zhang, 2008). This approach mitigates toxicity concerns, enhancing metal extraction during bioleaching (Benzal et al., 2020), and prevents adverse effects in high-pulp-density ewaste scenarios (Shah et al., 2015). The two-step process minimizes direct contact between biomass and metal-containing waste, enabling potential biomass recycling and avoiding microbial biomass contamination of waste materials. Moreover, it allows for the optimized formation of the leaching agent in the absence of waste material (Brandl et al., 2001). It is therefore very interesting to study each of the stages separately, that is, to study, on the one hand, the kinetics of the leaching of copper with ferric sulfate (reaction (1)), and, on the other, the biooxidation of the ferrous ion contained in the leaching liquors (reaction (2)) for, subsequently, study the connection of both. 2Fe3++Cu→2Fe2++Cu2+(r.1) 2Fe2++1/2O2+2H+→2Fe3++H2O (r.2) The studies found in the literature on the leaching of printed circuit boards with ferric sulfate are summarized in Table 3. The concentration of leaching agent used in these studies range from 6 g/L of Fe(III) (Benzal et al 2020) to 40 g/L (Iglesias-Gonz´ alez et al, 2024). In previous studies it has been shown that the leaching rate of copper increases with the concentration of ferric ion used as a leaching agent (Iglesias- Gonz´ alez et al., 2021; Van Yken et al., 2020; Vardanyan et al.,2022), this implies the use of high efficiency bioreactors for the biooxidation of the generated ferrous ion like the one developed by Mazuelos et al., 2023 at extreme conditions of iron concentration and acidity (up to 57 g/L of iron and pH up to 0.44). Most authors use both commercial and biogenic ferric sulfate, without finding a different behavior due to the source of the reagent. The working pH was between 1 and 1.8, except for the study carried out by van Yken et al.,2020 at pH 0.6. A low pH seems to favor the reaction, which may be related to the existence at low pH values of the most electroactive ferric iron species (FeHSO 4 2+ and Fe 3+ ), which are responsible for leaching (Crundwell, 1987). Reported studies have been performed mainly in Erlenmeyer flasks with orbital stirring, stirred tank reactors (STR) with and without air supply or in some cases by means of a percolation column or by a reactor equipped with a device to hold multiple WPCB plates in a vertical position (Shoda et al., 2020). The working temperature was usually around 30 ◦C, with some exceptions at 40–45 ◦C. The effect of temperature on the reaction rate for ferric leaching of copper has been studied by Iglesias-Gonz´ alez et al. (2021) finding an activation energy value of 16.9 kJ/mol, characteristic of processes limited by mass transfer. Recently, the improvement of copper leaching from PCB by working at 60 ◦C have been shown (Iglesias-Gonz´ alez et al., 2024). Under the hypothesis that copper leaching is limited by mass transfer, an increase in temperature would increase the diffusion of the leaching agent and would improve the leaching kinetics. It is known that bioleaching processes are limited by the temperature tolerated by mesophilic microorganisms; however, separating the chemical process (reaction (1)) from the biological one (reaction (2)) in different reactors would allow moderately high temperatures in the chemical reactor (50–80 ◦C), without altering the bacteria fixed in a bioreactor at the optimal growth temperature. It is worth noting, that process productivity can be enhanced by operating at high pulp density as long as the kinetics of copper extraction are not affected. Nevertheless, most of the previous work has been carried out with pulp densities below 5 % (Sethurajan and van Hullebusch, 2019; Iglesias-Gonz´ alez et al., 2021; Van Yken et al., 2020; Vardanyan et al.,2022; Benzal et al., 2020), whereas few ones have been done at higher pulp densities, but showing a much slower kinetics than the works at lower pulp densities (Khatri et al., 2018; Thacker et al., 2021; Iglesias-Gonzalez et al., 2022; Vardanyan et al., 2023). In order to be able to work with high pulp densities and maintain the kinetics of the ferric leaching of copper from PCB and therefore increasing the productivity of the process, the ferric iron has to be renewed by the action of the ferrous oxidizing bacteria. One of the main goals of the present work is to show that, working in a process configuration where the ferric leaching of copper from PCB is carried out in a separate reactor from the bioreactor for ferric iron regeneration, an increase in productivity can be achieved by working at solid concentrations as high as 10 % (w/w) keeping the kinetics of the copper dissolution. Particle sizes range from the order of μ m in pulverized samples to cut pieces in the range from mm to surfaces up to several cm 2 . The size of printed circuit boards (PCBs) plays a crucial role in the ferric leaching process of copper, primarily because it affects mass transfer limitations (Iglesias-Gonz´ alez et al., 2021). In a study by Shah et al. (2015), the impact of particle size on metal extraction percentages from e-waste was investigated. It was observed that as the particle size decreased the percentage of metal extraction increased. Moreover, working at higher particle sizes, enable a more efficient subsequent solid–liquid separation than with dust, and reduce the cost of mechanical pretreatment associated with grinding. Similarly, Sodha et al (2020) and Thacker et al (2022) reported recoveries superior to 90 % of Cu, Zn and Ni with nonpulverized waste cell phone printed circuit boards using NaOH pretreatment to remove the protective epoxy layer. Regarding copper recovery, most authors give values higher than 85 %, with reaction times that range from 1 to 25 days. The kinetics of ferric leaching of copper is usually modeled by using the shrinking core model (Wang et al., 2018) in which the mass transfer from the bulk of the solution to the surface of copper particles is the rate-limiting step. However, in this work we are dealing with PCB pieces of rectangular shape where the copper sheets that are consumed in the course of the reaction are sandwiched between plastic sheets as schematically shown in Fig. 1. Fig. 1. (A) Simplified scheme of a PCB piece. Copper sheet is shown in orange and at the beginning of the experiment had four faces in contact with the ferric iron solution. (B) Simplified scheme of a PCB piece once the reaction takes place. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.) P. Ramírez et al. Minerals Engineering 217 (2024) 108913 3 Therefore, to correctly analyze the experimental data, a kinetics model that takes into account the geometry of the PCB pieces used was developed. The main goals of the present work are: to study and develop a kinetics model of the ferric leaching of copper from non-pulverized PCB where copper is sandwiched between non-metallic sheets; to use a twostage configuration to regenerate the ferric in a bioreactor capable to operate at high total iron concentration and low pH values; to show that productivities higher than obtained so far can be achieved with nonpulverized PCB. 2. Materials and methods 2.1. Sample, lixiviant and inoculum The PCB sample object of this study was supplied by electronic waste manager company (RECILEC S.A, Spain). It was first cut with a laboratory cutting machine (Struers Labotom), then manually cut into square parallelepiped shape, and sieved between 0.84 and 3.2 mm. The average copper grade analyzed was 15 ±2 %. Other metals such as Zn, Ni, Fe and Au were also measured showing a low concentration: 271, 41, 384, 12.5 mg/kg, respectively. The inoculum was a bacterial consortium originally isolated from Rio Tinto Mines drainage waters consisted mainly of Acidithiobacillus ferrooxidans and Leptospirillum ferrooxidans, and some heterotrophic bacteria (mostly related with Acidophilium genus) as determined by DNA extraction, PCR amplification, sequencing and phylogenetic analysis (Mazuelos et al., 2012). 2.2. Experimental 2.2.1. Study of the kinetics of Cu leaching The experimental device used to study the kinetics of copper leaching consists of a 100 mL glass reactor with two necks for sampling and for temperature probe, and of a hot plate magnetic stirrer with thermoregulation. One hundred grams of biogenic ferric sulfate at pH =1 (adjusted with concentrated sulfuric acid) was placed in the reactor, and when the working temperature was reached the corresponding amount of PCB was added, at this moment, the test started. Periodically, the concentration of dissolved copper was analyzed to obtain the kinetic curves. ORP and pH were also monitored during the test. At the end of the test, the content of the reactor was filtered, Fe(II) and Fe(III) concentrations were determined in the liquid, and the solid residue was digested with aqua regia (nitric acid and hydrochloric acid in a molar ratio of 1:3) to quantify the copper that remains without leaching. Tests were carried out at temperatures of 30 ±1, 45 ±1 and 60 ± 1◦C, with 9, 20, 30 and 40 g/L of Fe(III), and with PCB concentration of 2.5 %. 2.2.2. Study of ferric leaching of Cu with biological regeneration of leaching agent In this study, the reactor (described in section 2.2.1.) was connected to a flooded packed bed bioreactor (FPBB) with continuous recycling of the liquid between them. The reactor outlet stream was pumped by a peristaltic pump at the bottom of the bioreactor whose outlet stream was returned by overflow to the reactor. The conditions for leaching in the reactor were 60 ◦C, 40 g/L of Fe (III) and PCB concentrations of 5 and 10 % (w/w). The FPBB was placed in a thermostatic chamber at 31 ◦C and consisted of a 2.7 cm diameter glass column filled with 5 mm glass spheres as biomass support. Cells were attached to the spheres following the procedure described by Mazuelos et al. (2000). Air was fed through the bottom of the FPBB at a flow rate of 0.5 L/min controlled by a rotameter. The useful bioreactor volume was 170 mL. In order to avoid precipitation of ferric iron due to the increase in pH associated with the consumption of protons from the biooxidation reaction of ferrous iron, a programmable dispenser of a concentrated solution of sulfuric acid (5 M) was used. The measured final pH in both experiments was of 1.05 therefore preventing ferric precipitation. 2.2.3. Analysis and control The PCB samples and leach residues were subjected to acid digestion with aqua regia (nitric acid and hydrochloric acid in a molar ratio of 1:3) to determine the copper grade. Copper was measured by atomic absorption spectrophotometry (PinAAcle 500, Perkin Elmer). Ferric iron concentration was determined by the sulfosalicylic acid method (Karamanev et al., 2002) using a spectrophotometer (UV–Vis C-7100, Peak Instruments Inc) and ferrous iron concentration was determined by automatic titration with K 2 Cr 2 O 7 (TitraLab AT 1000 Series, Hatch). 2.3. Preparation of samples for observation under the scanning electron microscope The morphology and composition of the PCB before and after leaching was studied by scanning electron microscopy (SEM), energy dispersive X-ray spectroscopy (EDS) and mapping with a high-resolution scanning electron microscope (Teneo SEM, FIE) in CITIUS (Centro de Investigaci´ on, Tecnología e Innovaci´ on de la Universidad de Sevilla). PCB sample was cut into pieces with a square geometry of 3 mm length and the cross section of the samples was examined. They were placed in a mold supported by a pin so that they were resting on the bottom on one side and then covered with epoxy resin (EpoThin 2, Buehler) and allowed to cold cure for 8 h. Then, the samples were carefully roughened with 400, 800, and 1200 grit SiC paper. Finally, carbon was deposited to make them conductive. 2.4. Model development Copper was embedded in the PCB pieces in the form of thin sheets (Fig. 1A). The copper dissolution rate was modeled considering that the copper dissolution proceeds from the outside of the PCB to the inside of it as the copper foil dissolves as shown in Fig. 1B. It was considered that Fig. 2. Ferric iron concentration profile as a function of the distance in the direction of one of the sides of the copper sheet base at two instant times (t 2 > t 1 ). In the figure is shown the following characteristic lengths: l 0 is the initial length of the side of the copper sheet; l t is the length of the side of the copper sheet at a time t. P. Ramírez et al. Minerals Engineering 217 (2024) 108913 4 copper dissolution rate was mass transfer limited as has been recently shown for copper sheets in contact with ferric sulfate solutions (Iglesias- Gonz´ alez et al., 2021). Redox reaction between ferric iron and copper (reaction (1)) was supposed to be instantaneous and, therefore, at the plane of the copper in contact with the solution the ferric iron concentration was equal to zero. Fig. 2 shows a hypothetical concentration profile of ferric iron as a function of the distance for the direction of one side of the base of the copper sheet. At the plane of the copper sheet the ferric iron concentration is zero and a linear increase of concentration is assumed until reaching the plane of the PCB piece in contact with the bulk solution. It is assumed that the resistance to mass transfer in the liquid film is much lesser than the resistance to internal mass transfer. The reaction rate of ferric ion on the copper surface is equal to the mass transfer rate through the PCB pore that can be expressed as the product of a mass transfer coefficient and a driving force given by the concentration gradient of ferric iron inside the PCB piece: ( − rFe3+) = k([Fe3+]b−[Fe3+]s)(1) where kis the mass transfer coefficient in m/s, [Fe 3+ ] b is the ferric iron bulk concentration in mol/m 3 and [Fe 3+ ] s is the ferric iron concentration in contact with the inner copper surface which is taken as 0. According to the stoichiometry given by reaction (1), the reaction rate of ferric ion on the surface is related to the reaction rate of copper on the surface as follows: (rCu2+) = ( − rFe3+) = 1 2k[Fe3+]b(2) Multiplying Eq. (2) by the area of reaction (A r ) the reaction rate of copper ion production is obtained: dNCu2+ dt = (rCu2+)Ar=1 2k[Fe3+]bAr(3) The copper area that reacts at each moment is equal to the area exposed to the ferric ion solution. This area is the sum of the four sides of the copper sheet that are in contact with the ferric ion solution. The area of each of these faces is the product of the length of the base (l) and the width of the copper sheet (e). Therefore, Eq. (3) can be written as follows: dNCu2+ dt =1 2k[Fe3+]b4⋅e⋅l(4) Moreover, the moles of Cu 2+ in solution are related to copper conversion (X Cu ) according to the following equation: NCu2+=mCu,0XCu AW(Cu)(5) where m Cu,0 is the total mass of copper in the PCB piece and AW(Cu) is the atomic weight of copper. Therefore, the reaction rate of copper ion production can be expressed as follows: dNCu2+ dt =mCu,0 AW(Cu) dXCu dt (6) Copper conversion (X Cu ) is the ratio between the mass of dissolved copper divided by the total mass of copper: XCu =mCu,0−mCu mCu,0 (7) The mass of copper is the product of the density of copper and the volume of the copper sheet, which is given by the product of the area of the base of the sheet (l 2 ) and the width (e). Hence, Eq. (7) can be expressed as: XCu =mCu,0−mCu mCu,0 =1− ρ Cul2e ρ Cul2 0e=1−l2 l2 0 (8) and, the length of the base of the copper sheet at each moment (l) is related to conversion according to the following expression: l=l0(1−XCu) 1 / 2 (9) Substituting Eqs. (6) and (9) into Eq. (4) gives: dXCu dt =2AW(Cu) ρ Cul0 k[Fe3+]b(1−XCu) 1 / 2 (10) Furthermore, the internal mass transfer coefficient, k, depends on the physical properties of the fluid, the geometry and the linear flow velocity and through the dimensionless Sherwood number: Sh =kL D(11) where Lis the characteristic length of the diffusion process and Dis the diffusivity of ferric iron. In this case, Lcan be defined as the total length of the pore trough ferric iron diffuses: L= (l0−l)(12) which turn out to be function of time as long as the copper sheet is consumed. Therefore, the mass transfer coefficient can be expressed as: k=Sh⋅D (l0−l)(13) Eq. (10) can be rewritten as: dXCu dt =2AW(Cu) ρ Cul0 Sh⋅D (l0−l)[Fe3+]b(1−XCu) 1 / 2 (14) and substituting lwith the Eq. (9) it can be also expressed as: dXCu dt =kʹʹ[Fe3+]b (1−XCu) 1 / 2 ⎛ ⎝1− (1−XCu) 1 / 2 ⎞ ⎠ (15) being kʹʹ =2AW(Cu)ShD ρ Cul2 0. In batch experiments, the ferric iron bulk concentrations diminish as long as the reaction takes place. The ferric iron concentration at each time can be related to the copper conversion according to the following expression: [Fe3+]b=[Fe3+]0(1−2[Cu2+]max [Fe3+]0 XCu )(16) where [Fe 3+ ] 0 is the initial ferric iron concentration, [Cu 2+ ] max is the maximum copper ion concentration once all the copper in the PCB has been extracted. Finally, Eq. (13) can be rewritten as: dXCu dt =kʹʹ[Fe3+]0(1−2[Cu2+]max [Fe3+]0 XCu )(1−XCu) 1 / 2 ⎛ ⎝1− (1−XCu) 1 / 2 ⎞ ⎠ (17) This expression is valid as long as X Cu >0. In order to obtain the integrated expression of Eq. (15), a new variable D =(1-X Cu ) 1/2 is used. Eq. (15) expressed as function of variable D was: dD dt = − 1 2kʹʹ[Fe3+]0(1−a(1−D2)) 1 (1−D)(18) P. Ramírez et al. Minerals Engineering 217 (2024) 108913 5 with a =2[Cu2+]max [Fe3+]0 By separating variables and integrating between t 0 >0 and t and D 0 and D we arrive at the following expression: A much simpler expression can be obtained if the ferric iron concentration remains constant during the copper leaching process, that is the case when ferric iron is regenerated by the biooxidation of the ferrous iron produced. In this case, the following expression is obtaining by integration with initial conditions t =t 0 and X =X 0 : (1−XCu) − 2(1−XCu) 1 / 2 − (1−X0) + 2(1−X0) 1 / 2 =kʹʹ[Fe3+]0(t−t0) (20) Therefore, a linear plot is obtained when plotting the left-hand side (l.h.s) of Eqs. (19) and (20) as a function of (t-t 0 ) with a slope equal to k’’⋅[Fe 3+ ] 0 . 3. Results and discussion 3.1. Cu leaching in batch mode First, experiments in batch mode and without biological regeneration of ferric ions were carried out at different temperatures in the range 30–60 ◦C and with initial ferric ion concentration ranging from 9 to 40 g/L. All the experiments were carried out at 2.5 % of solid/liquid ratio. Fig. 3 shows the scanning electron microscope image of the cross section of the PCB sample before (left) and after 1 h of leaching (right). Below each microphotograph the respective mapping result is presented. A sandwich-type structure can be clearly observed, composed of two external copper sheets and a network of fibers inside. In the leached sample image (right) it can be seen how the copper disappears below the Fig. 3. SEM photomicrographs and mapping of the sample before (left) and after leaching (right) and the working conditions. Fig. 4. Plot of the copper conversion (X Cu ) as a function of time for the batch experiments: experimental values (symbols) and theoretical model from Eq. (19) with the values of k ″ given in Table 1 (lines). the initial ferric concentrations and temperature in each test are shown in the legend. 1 a[ln(a(D2−1)+1 a(D2 0−1)+1)+2 a 1−a √(tan−1( a 1−a √D0)−tan−1( a 1−a √D))]= =kʹʹ[Fe3+]0(t−t0) (19) P. Ramírez et al. Minerals Engineering 217 (2024) 108913 6 mask due to the reaction with ferric as this ion accesses the surface of the copper. Fibers are composed of Ca, Al, Si and O. Fig. 4 shows the experimental values of the conversion versus time. Along with the experimental values, the values of X Cu as a function of time obtained from Eq. (19) have also been represented with the values of k ″ given in Table 1. As can be observed, experimental data fits satisfactorily to model predictions in all the tested conditions. These results indicate that copper dissolution from PCB waste is mainly governed by a mass transfer process, and for the same particle size, experiments carried out at higher temperatures and ferric ion concentrations show higher dissolution rates. In order to explore in more detail, the effect of temperature the natural logarithm of the kinetic constant (k ″ ) as a function of the inverse of the absolute temperature was plotted. Fig. 5 shows the Arrhenius plot from where the activation energy can be determined from the slope of the linear regression of the experimental data a value of E a of 20 kJ/mol was obtained, which is a value that supports the fact that the copper leaching from PCB pieces is a process limited by mass transfer. 3.2. Cu leaching coupled with biological regeneration of ferric iron In order to increase productivity, it was interesting to study the copper dissolution kinetics for experiments at higher pulp densities. Nevertheless, increasing pulp density would lead to a depletion of ferric ion concentration, and therefore the copper dissolution will no longer take place. Because of its low cost and circularity, the regeneration of ferric iron carried out by microorganisms is an interesting alternative to overcoming this issue. Nevertheless, the bacterial growth is severely affected by temperature. As shown in the previous section, faster leaching kinetics can be achieved working at higher temperatures than the optimum temperature for mesophilic bacteria growth. As a result, a two stage operation where the biological action is separated from the copper leaching was used. This experimental setup it is increasingly accepted as a promising alternative for the biological leaching of copper and other metals from PCBs (Benzal et al., 2020; Iglesias-Gonz´ alez et al., 2021; Van Yken et al., 2023). The biological reactor used was a flood packed bed bioreactor (FPBB) as described in the experimental section. The goal was to be able to regenerate the ferric iron consumed in the chemical leaching reactor while operating at high total iron concentration (up to 40 g/L) and maintaining the pH below 1.5 in order to avoid a massive ferric iron precipitation, as recently reported in a similar bioreactor (Mazuelos et al., 2023). Before being connected to the reactor, the bioreactor operated in continuous recirculation with a 1L tank with culture medium and an initial ferrous iron concentration of 25 g/L and a total iron concentration of 40 g/L and a feed flow rate of 12 mL/h for 95 h with the objective of promoting the adhesion of the bacteria to the glass spheres inside the column. Once a bacterial biofilm was formed inside the bioreactor, the feed flow rate was increased from 12 to 108 mL/h. When the connection to the reactor was set, the biooxidation rate of the bioreactor achieved a value of 2.1 g/L⋅h. In Fig. 6 is shown the scheme of the experimental setup used. Fig. 7 shows the experiments carried out at 60 ◦C and initial ferric ion concentration of 40 g/L, at 2.5 % solid/liquid ratio and at 5 % and 10 % solid/liquid ratio with ferric iron regeneration. For the experiments with the biological reactor connected with the PCB ferric leaching reactor the Table 1 Experimental conditions of each test given by the values of [Fe 3+ ] 0 , T and a for each experiment with the value of the fitting parameter k’’ and its coefficient of determination. [Fe 3+ ] 0 (g/L) T (◦C) a=2[Cu2+]max [Fe3+]0 k“⋅10 6 (L/mol⋅s) R 2 40 30 0.191 9 0.996 40 60 0.199 21 0.978 20 30 0.334 12 0.994 20 60 0.413 29 0.974 30 45 0.284 18 0.960 9 30 0.821 17 0.984 Fig. 5. Plot of the natural logarithm of k ″ as a function of the inverse of the absolute temperature (Arrhenius plot). The dashed line is the linear regression of the experimental data from where an activation energy, Ea, of 20 kJ/mol has been obtained. Fig. 6. Experimental setup for the experiments with biological regeneration of ferric iron. P. Ramírez et al. Minerals Engineering 217 (2024) 108913 7 ferric ion concentration remains practically constant over the entire experiment. The fitting parameters of the model are shown in Table 2. In the three scenarios tested at different solid/liquid ratios, the copper was totally extracted in less than 1500 min. Nevertheless, as shown in Fig. 5 the copper dissolution kinetics was affected by mass transfer when increasing the proportion of solid in the process. In Table 2 is shown the values of k ″ for the three experiments. A decrease in the kinetics constant of about 42 % from the value at 2.5 % S/L compared to the value at 10 % S/L was observed. This result is likely due to inefficient mixing leading to lower Sherwood number and therefore to a lower ass transfer coefficient as expressed in equation (11). Table 3 summarizes the experimental conditions and the most relevant results of the studies on the leaching of copper from PCBs with ferric sulfate found in the most recent literature, as well as the conditions and results of the present work. The percentage of Cu leached and the average Cu leaching rate (g L −1 h −1 of copper leached) are compared. The highest rates are achieved at a high solid/liquid ratio, which requires ferric iron concentrations of at least 20 g/L. This is observed in the present work ([Fe(III] =40 g/L, T=60 ◦C) in which productivity increases proportionally to the solid–liquid ratio. In comparison with previous works, a positive effect of temperature and ferric ion concentration on daily productivity is observed, despite the fact that the copper grade in the current case is low (15 %) and that the sample has not been pulverized. 4. Conclusions The ferric leaching of PCBs cut in a rectangular shape with a typical sandwich-type structure in which the copper is interspersed in layers of non-metallic material was studied depending on the temperature and ferric iron concentration. A kinetic model was developed to interpret the experimental results. The proposed model considers that the kinetics of ferric leaching is limited by the internal mass transfer from ferrous in the bulk solution to the copper surface. It is considered that the thickness of the boundary layer is equal to the distance between the bulk solution to the surface of the copper within the structure. The experimental data fit perfectly with the proposed model in which the kinetics with respect to the leaching agent is of order one, that is, it increases linearly with the concentration of ferric in solution. Likewise, an Arrhenius dependence of the kinetics with the temperature was observed with an activation energy of 20 kJ/mol consistent with a process limited by internal mass transfer. The possibility of regenerating the leaching agent using ferrooxidizing microorganisms in a flooded fixed-bed bioreactor was studied. The main goal was to separate the chemical and biological stages, seeking the use of high temperatures in the chemical reactor without affecting the mesophilic microorganisms. Before connecting both reactors, the biooxidation rate was determined, obtaining a value of 2.1 g/ L⋅h. It has been shown that through this configuration it is possible to work at higher solid/liquid ratios. Experiments were carried out with solid/liquid ratios of 5 % and 10 %. In both cases, 99 % of the copper was leached in less than 24 h. However, increasing the solid/ratio from 5 to 10 % caused a decrease in the kinetics of the process, possibly due to poor mixing. Finally, it should be noted that a much higher leaching rate (up to 0.62 g/L⋅h of leached copper) than previously reported for other Fig. 7. Experimental values of copper dissolution from PCB pieces as a function of reaction time (squares –2.5 % S/L; circles –5 % S/L; triangles –10 % S/L) and fitting to the proposed model (lines). All the experiments were carried out at 60 ◦C and with an initial ferric ion concentration of 40 g/L. Table 2 Parameters fitting of experiments conducted at different solid/liquid ratio (between 2.5 and 10 %) at 60 ◦C and 40 g/L of ferric concentration. S/L ratio (%) [Fe 3+ ] 0 (g/L) T (◦C) K ″ ⋅10 6 (L/mol⋅s)* R 2 2.5 40 60 19 0.992 5 40 60 14 0.999 10 40 60 11 0.992 *Obtained from Eq. (20), i.e., assuming constant ferric iron concentration. Table 3 Comparison of recent results of ferric batch bioleaching of PCBs. Fe(III) (g/L) Ferric source pH T (◦C) S/L % Cu grade % Particle size ( μ m) Reactor time (days) Cu leached % leaching rate (g/L⋅h) Authors 16.7 Chem 1.25 30 10 27.5 200–250 Flask 5 99 0.227 Khatri et al. (2018) 11.17 Chem −20 0.1 16.6 0–500 Flask 1 95 0.007 Sethurajan and van Hullebusch (2019) 20 Bio & chem 0.6 25 1 58.7 365–750 Flask 2 87 0.106 Van Yken, et al. (2020) 8.38 Bio 1.75 30 0.75 39 200–1000 Flask 2 95 0.058 Benzal et al. (2020) 20 Bio 1.25 30 2.5 27 5000–10000 STR 2 90 0.127 Iglesias-Gonz´ alez et al. (2021) 9 Bio 1.8 30 0.35 43.5 750 Flask 10.4 74.6 0.005 Santaolalla et al. (2021) 20 Bio 1.25 30 13.6 30 20,000 Percolator 25 90 0.061 Iglesias-Gonzalez et al. (2022) 15.5 Bio 1 40 1 21.5 <2000 STR 1 87 0.078 Vardanyan et al.(2022) 20 Bio 1.25–1.64 45 1 22.2 365–750 STR 2 97 0.045 Van Yken et al. (2023) 20 Bio 1 40 10 24.6 <2000 STR 2 95 0.487 Vardanyan et al.(2023) 40 Bio 1 60 1 35 800–2000 STR 9 (hours) 99 0.398 Iglesias-Gonz´ alez et al. (2024) 40 Bio 1 60 2.5 15 840–3200 STR 1 99 0.155 Present work 40 Bio 1 60 5 15 840–3200 STR 1 99 0.310 Present work 40 Bio 1 60 10 15 840–3200 STR 1 99 0.619 Present work P. Ramírez et al. Minerals Engineering 217 (2024) 108913 8 researchers has been obtained. Furthermore, this productivity has been achieved with non-pulverized PCB samples. CRediT authorship contribution statement Pablo Ramírez: Writing –review &editing, Writing –original draft, Validation, Supervision, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Nieves Iglesias-Gonz´ alez: Writing – review &editing, Writing –original draft, Validation, Supervision, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Antonio D. Dorado: Writing –review &editing, Writing – original draft, Validation, Supervision, Methodology, Formal analysis, Data curation, Conceptualization. Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Data availability Data will be made available on request. Acknowledgements We thank Consuelo Cerrillos Gonz´ alez for her technical support in the scanning electron microscopy study. References Baniasadi, M., Vakilchap, F., Bahaloo-Horeh, N., Mousavi, S.M., Farnaud, S., 2019. Advances in bioleaching as a sustainable method for metal recovery from e-waste: a review. J. Ind. Eng. Chem. 76, 75–90. https://doi.org/10.1016/j.jiec.2019.03.047. Benzal, E., Sol´ e, M., Lao, C., Gamisans, X., Dorado, A.D., 2020. Elemental Copper Recovery from e-Wastes Mediated with a Two-Step Bioleaching Process. Waste Biomass Valor 11, 5457–5465 (2020). doi: 10.1007/s12649-020-01040-2. Brandl, H., Bosshard, R., Wegmann, M., 2001. Computer-munching microbes: metal leaching from electronic scrap by bacteria and fungi. Hydrometall. 59 (2001), 319–326. https://doi.org/10.1016/S0304-386X(00)00188-2. Crundwell, F.K., 1987. Kinetics and mechanism of the oxidative dissolution of a zinc sulphide concentrate in ferric sulphate solutions. Hydrometall. 19 (2), 227–242. https://doi.org/10.1016/0304-386X(87)90007-7. Cui, J., Zhang, L., 2008. Metallurgical recovery of metals from electronic waste: a review. J. Hazard. Mater. 158, 228–256. https://doi.org/10.1016/j.jhazmat.2008.02.001. European Comission, 2023. European Comission, Study on the Critical Raw Materials for the EU 2023 –Final Report. https://op.europa.eu/en/publication-detail/-/publica tion/57318397-fdd4-11ed-a05c-01aa75ed71a1. Iglesias-Gonz´ alez, N., Carranza, F., Mazuelos, A., Romero, R., Lorenzo-Tallafigo, J., Romero-García, A., Ramírez, P., 2021. The BRISA process as a path for efficient copper recovery from waste PCBs. Hydrometall. 205, 105750 https://doi.org/ 10.1016/j.hydromet.2021.105750. Iglesias-Gonz´ alez, N., Dorado, A.D., Ramírez, P., Mazuelos, A., 2024. A high productivity bioprocess for obtaining metallic copper from printed circuit boards (PCBs). Miner. Eng. 205, 108459 https://doi.org/10.1016/j.mineng.2023.108459. Iglesias-Gonzalez, N., Ramirez, P., Lorenzo-Tallafigo, J., Romero-García, A., Mazuelos, A., Romero, R., Carranza, F., 2022. Copper recovery from unground printed circuit board by biogenic ferric at high solid/liquid ratio. Miner. Eng. 180 https://doi.org/10.1016/j.mineng.2022.107471. Ilyas, S., Srivastava, R.R., Kim, H., Das, S., Singh, V.K., 2021. Circular bioeconomy and environmental benignness through microbial recycling of e-waste: a case study on copper and gold restoration. Waste Manag. 121, 175–185. https://doi.org/10.1016/ j.wasman.2020.12.013. Karamanev, D.G., Nikolov, L.N., Mamatarkova, V., 2002. Rapid simultaneous quantitative determination of ferric and ferrous ions in drainage waters and similar solutions. Miner. Eng. 15 (5), 341–346. https://doi.org/10.1016/S0892-6875(02) 00026-2. Khatri, B.R., Sodha, A.B., Shah, M.B., Tipre, D.R., Dave, S.R., 2018. Chemical and microbial leaching of base metals from obsolete cell-phone printed circuit boards. Sust. Environ. Res. 28(6), 333–339. https://doi.org/10.1016/j.serj.2018.10.002. Kumari, R., Samadder, S.R., 2023. Evaluation of the recycling potential of obsolete mobile phones through secondary material resources identification: a comprehensive characterization study. J. Environ. Manage. 345, 118550 https://doi. org/10.1016/j.jenvman.2023.118550. Mazuelos, A., Carranza, F., Palencia, I., Romero, R., 2000. High efficiency reactor for the biooxidation of ferrous iron. Hydrometall. 58, 269–275. https://doi.org/10.1016/ S0304-386X(00)00141-9. Mazuelos, A., Moreno, J.M., Carranza, F., Palomino, C., Torres, A., Villalobo, E., 2012. Biotic factor does not limit operational pH in packed-bed bioreactor for ferrous iron biooxidation. J. Ind. Microbiol. Biotechnol. 39, 1851–1858. https://doi.org/ 10.1007/s10295-012-1187-9. Mazuelos, A., Moreno-Perez, M., Perdigones, B., Ramirez, P., Iglesias-Gonzalez, N., 2023. Ferrous iron biooxidation in a flooded packed-bed bioreactor at extreme conditions of iron concentration and acidity. Miner. Eng. 204, 108408 https://doi.org/ 10.1016/j.mineng.2023.108408. Mir, S., Dhawan, N., 2022. A comprehensive review on the recycling of discarded printed circuit boards for resource recovery. Resour. Conserv. Recycl. 178, 106027 https:// doi.org/10.1016/j.resconrec.2021.106027. Oguchi, M., Murakami, S., Sakanakura, H., Kida, A., Kameya, T., 2011. A preliminary categorization of end-of-life electrical and electronic equipment as secondary metal resources. Waste Manag. 31 (9–10), 2150–2160. https://doi.org/10.1016/j. wasman.2011.05.009. Rao, M.D., Singh, K.K., Morrison, C.A., Love, J.B., 2020. Challenges and opportunities in the recovery of gold from electronic waste. RSC Adv. 10 (8), 4300–4309. https://doi. org/10.1039/C9RA07607G. Santaolalla, A., Lens, P., Barona, A., Rojo, N., Ocio, A., Gallastegui, G., 2021. Metal extraction and recovery from mobile phone PCBs by a combination of bioleaching and precipitation processes. Minerals 11 (9), 1004. https://doi.org/10.3390/ min11091004. Sethurajan, M., van Hullebusch, E.D., 2019. Leaching and selective recovery of Cu from printed circuit boards. Metals (Basel). 9, 1–12. https://doi.org/10.3390/ met9101034. Shah M.B., Tipre D.R., Mamta S.P., Dave S.R., 2015. Development of two-step process for enhanced biorecovery of Cu-Zn-Ni from computer printed circuit boards. J. Biosci. Bioeng. 120, 167–173. doi: 10.1016/j.jbiosc.2014.12.013. Sodha, A.B., Tipre, D.R., Dave, S.R., 2020. Optimisation of biohydrometallurgical batch reactor process for copper extraction and recovery from non-pulverized waste printed circuit boards. Hydrometall. 191 (1), 105170 https://doi.org/10.1016/j. hydromet.2019.105170. Thacker, S.C., Nayak, N.S., Devayani, R., Tipre, D.R., Dave, S.R., 2022. Multi-metal mining from waste cell phone printed circuit boards using lixiviant poprduced by a consortium of acidophilic iron oxidizers. Environ. Eng. Sci. 39, 287–295. https://doi. org/10.1089/ees.2020.0389. Van Yken, J., Cheng, K.Y., Boxall, N.J., Nikoloski, A.N., Moheimani, N., Valix, M., Sahajwalla, V., Kaksonen, A.H., 2020. Potential of metals leaching from printed circuit boards with biological and chemical lixiviants. Hydrometall. 196, 105433 https://doi.org/10.1016/j.hydromet.2020.105433. Van Yken, J.; Cheng, K.Y.; Boxall, N.J.; Nikoloski, A.N.; Moheimani, N.; Valix, M.; Kaksonen, A.H., 2023. An integrated biohydrometallurgical approach for the extraction of base metals from printed circuit boards. Hydrometallurgy 216, 105998. doi: 10.1016/j.hydromet.2022.105998. Vardanyan, A., Vardanyan, N., Aˆ atach, M., Malavasi, P., Gaydardzhiev, S., 2022. Bioassisted leaching of non-ferrous metals from waste printed circuit boards—importance of process parameters. Metals 12(12), 2092. https://doi.org/ 10.3390/met12122092. Vardanyan, A., Vardanyan, N., Gaydardzhiev, S., 2023. Biological extraction of Cu and Ni from printed circuit boards via redoxolysis with concomitant material characterization. Hydrometall. 221, 106145 https://doi.org/10.1016/j. hydromet.2023.106145. Wang, L., Li, Q., Li, Y., Sun, X., Li, J., Shen, J., Wang, L., 2018. A novel approach for recovery of metals from waste printed circuit boards and simultaneous removal of iron from steel pickling waste liquor by two-step hydrometallurgical method. Waste Manag. 71, 411–419. https://doi.org/10.1016/j.wasman.2017.10.002. P. Ramírez et al.