Ferrous iron biooxidation in a flooded packed-bed bioreactor at extreme conditions of iron concentration and acidity

Minerals Engineering 204 (2023) 108408 Available online 6 October 2023 0892-6875/© 2023 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/bync-nd/4.0/). Ferrous iron biooxidation in a flooded packed-bed bioreactor at extreme conditions of iron concentration and acidity Alfonso Mazuelos a , * , Martin Moreno-Perez a , Blanca Perdigones a , Pablo Ramirez a , Nieves Iglesias-Gonzalez a a Chemical Engineering Department, Faculty of Chemistry, University of Seville, C/ Profesor Garcia Gonzalez, Edificio de Quimica, Planta 1, 41012 Seville, Spain ARTICLE INFO Keywords: Ferrous iron biooxidation Acidithiobacillus ferrooxidans Leptospirillum ferrooxidans Flooded packed-bed bioreactor Low pH High iron concentration ABSTRACT The utility of ferrous iron biooxidation is to regenerate Fe(III) at required rates and conditions in specific hydrometallurgical contexts, such as metal extraction from ores/concentrates and mining and electronic waste. In these applications, considerable kinetics improvements can be achieved by increasing [Fe(III)], but pH must be decreased to avoid precipitation of this oxidant. Information about continuous biooxidation operation is limited to [Fe] <20 g/L and 2.3 >pH >1, therefore, it is interesting to test wider ranges for these parameters. A 1L flooded packed-bed bioreactor (30 cm in height), inoculated with Acidithiobacillus ferrooxidans and Leptospirillum ferrooxidans, was operated 60 days in continuous mode without interruptions at [Fe] up to 57 g/L and pH up to 0.44. Total Fe(II) conversion was achieved when operating at [Fe] =57 g/L and pH =1.2. The maximum biooxidation rate reached was 3.5 g/L⋅h for [Fe] =40 g/L and 1 h of hydraulic retention time. Biooxidation rate decreases by 32 % when pH decreases from 1.2 to 0.44. Nevertheless, the biofilm remained stable at this low pH and steady state was achieved. When comparing the relative decreases in biooxidation rate and oxygen solubility, the drop of efficiency can be explained by aeration limitations and salting out effect. 1. Introduction The term “ferrous iron biooxidation” refers to the rapid oxidation of Fe(II) to Fe(III) with oxygen catalyzed by microorganisms (Reaction 1). Fe(II) + H++1 4O2→ microorganisms Fe(III) + 1 2H2O(Reaction 1) The microorganisms capable of catalyzing this reaction are known as iron-oxidizing microorganisms (Nemati et al., 1998; Norris et al., 2000; Johnson, 2008). Among them, the species of the genera Acidithiobacillus and Leptospirillum stand out. The applications of Fe(II) biooxidation are depuration and valorization of acid mine drainage (AMD) (Sandstr¨ om and Mattsson, 2001; Song et al., 2022), desulfurization of combustible gases (Malhotra et al., 2002; Lin et al., 2013), valorization of copper slags (Carranza et al., 2009; Kaksonen et al., 2017), valorization of e-wastes (Hubau et al., 2020; Iglesias-Gonz´ alez et al., 2021; Iglesias-Gonzalez et al., 2022) and metal extraction from ores and concentrates. All of them can be considered clean technologies because they are based on circularity and sustainability. With exception of the first one, they are inspired by indirect bioleaching with separation of effects chemical and biological (Carranza et al., 1993). Indirect bioleaching consists of two simultaneous stages (Nemati et al., 1998; Schippers & Sand, 1999; Mishra et al., 2023): - A chemical leaching stage, in which the oxidizing agent Fe(III) takes electrons from the metal sulfide (MeS), oxidizing it to the metal (Me) and S◦, and modifying its oxidation state to Fe(II) (Reaction 2). MeS +Fe(III)→Me(II) + So+Fe(II)(Reaction 2) Abbreviations: a, Ratio C X0 /Y XS ; AMD, Acid Mine Drainage; ε , Gas hold-up; EW, Electrowinning; [Fe(II)] 0 ino , Initial ferrous iron concentration of the inoculum; FPBB, Flooded Packed-Bed Bioreactor; Me, Metal; q s , Specific rate of substrate; C X , Biomass concentration; C X0 , Initial biomass concentration; C Xf ino , Final biomass concentration of the inoculum; μ max , Specific growth rate; ORP, Oxidation Reduction Potential; S, Oxygen solubility; S 0 , Oxygen solubility in water; SX, Solvent Extraction; V ino , Volume of the inoculum; V cul , Volume of the culture; Y XS , Yield of biomass from substrate. * Corresponding author. E-mail address: [email protected] (A. Mazuelos). Contents lists available at ScienceDirect Minerals Engineering journal homepage: www.elsevier.com/locate/mineng https://doi.org/10.1016/j.mineng.2023.108408 Received 16 June 2023; Received in revised form 31 July 2023; Accepted 23 September 2023 Minerals Engineering 204 (2023) 108408 2 - A biological stage, the biooxidation of Fe(II) regenerating the oxidizing agent Fe(III) (Reaction 1). It is possible to accelerate indirect bioleaching by carrying out the chemical leaching and biological Fe(II) oxidation stages separately, allowing the optimization of each independently (Carranza et al., 1993). With the exception of depuration of acid mine drainage (AMD), in the applications mentioned above a sulfide, a metal sulfide, H 2 S or an elemental metal is oxidized by O 2 , with the redox couple Fe(II))/Fe(III) acting as an intermediate in electron transfer, and microorganisms catalyzing the regeneration of the oxidizing agent Fe(III) (Fig. 1). In all these processes, the chemical leaching state can be accelerated by increasing the concentration of Fe(III). Along with this increase in Fe (III) concentration, sufficient acidic conditions are required to prevent the precipitation of this oxidizing reagent. In the Pourbaix diagram (Fig. 2) it can be seen that a pH lower than 1.5 is required for a Fe(III) concentration of 1 M (56 g/L) (Pourbaix, 1974). In order to achieve an accurate control in the operation of continuous bioreactors, this thermodynamic interdependence between acidity and iron concentration must be assumed since it conditions critical operational aspects such as microorganism performance, oxygen supply and physicochemical stability of biofilms. As far as the authors know, biooxidation carried out at iron concentrations higher than 25 g/L has only been tested in batch Fig. 1. Conceptual engineering of metal extraction from ores and concentrates, slags and e-waste, and desulfurization of gases. Fig. 2. Iron Pourbaix diagram (Pourbaix, 1974). A. Mazuelos et al. Minerals Engineering 204 (2023) 108408 3 experiments. Kawabe et al. (2003) informed that biooxidation rate drops by more than 80 % in the presence of Fe(III) at a concentration of 25 g/L when strains of Acidithiobacillus ferrooxidans are used as inoculum. Pyrite bioleaching performance with A. ferrooxidans is considerably depressed when Fe(III) concentration is 35 g/L, being slightly affected by Fe(II) concentration up to 27 g/L (Battaglia et al., 1994). Nevertheless, tolerance of A. ferrooxidans to Fe ions can be increased to 36 g/L of total iron by an adaptation process based on subculturing (Saavedra et al., 2020). In fact, there is no mechanism of inhibition caused by Fe (III) described in the existing literature (Amouric et al., 2009; Moinier et al., 2017; Ponce et al., 2012). Rawlings et al. (1999) reported that Leptospirillum species are more resistant to Fe(III) than Acidithiobacillus species. In continuous operation, only liquors with an iron concentration lower than 25 g/L have been biooxidized (Mousavi et al., 2007; Frias et al., 2008). pH has a great influence in the metabolism of the iron-oxidizing microorganisms and their ability to oxidize Fe(II). A. ferrooxidans typically grows at a pH >1.5 with an optimum pH between 1.8 and 2.5 whereas L. ferrooxidans typically grows at pH >1 and its optimum pH is between 1.3 and 2 (Torma, 1977; Karamanev & Nikolov, 1988; Smith et al., 1988; Breed & Hansford, 1999; Oren, 2010). Using a mixed culture of these microorganisms allows continuous Fe(II) biooxidation in a range of pH between 1 and 2.3 (Mazuelos et al., 2010, 2012). Fe(II) biooxidation is a heterogeneous reaction, in which microorganisms, Fe(II) and oxygen are in different phases. Therefore, in addition to the biochemical reaction itself, the transport phenomena of matter must be considered. As a result, the transport of oxygen from the gas to the liquid significantly influences the overall kinetics of the biooxidation reaction, and aeration is usually a key design factor (Savic et al., 1998; Mazuelos et al., 2002). Oxygen solubility in aqueous solutions depends on ion concentration due to salting out effect (Mazuelos et al., 2017). When the iron concentration is increased and pH is decreased (increasing SO 4 2- , HSO 4 - concentrations), a decrease in oxygen solubility must be assumed and, as a consequence, a lower oxygen concentration in the liquid medium is available for the bacteria. Understanding the structure that allows the microorganisms to attach themselves to surfaces is essential to operate continuous bioreactors based on biofilms. Jarosite provides iron-oxidizing microorganisms with a porous structure in which they can attach (Karamanev, 1991). Jarosite formation depends on the Fe(III) concentration and pH (Dutrizac, 1983; Kaksonen et al., 2014). pH has been mentioned as a key design factor (Pogliani & Donati, 2000; Kinnunen & Puhakka, 2004; van der Meer et al., 2007; Kahrizi et al., 2009; Mazuelos et al., 2010) because it affects the biofilm stability. pH must be low enough to avoid precipitation of these iron compounds but not so low as to compromise the structure of the biofilm due to the dissolution of the inorganic matrix. At a low pH (<1), an excessive dissolution of these precipitates occurs, resulting in the destabilization of the biofilm and in a cell wash-out. Precipitation of Fe(III) compounds can lead to an accumulation of precipitates, causing clogging of the channels meant for the liquid and the air, hindering nutrients diffusion to the microorganisms (Curutchet et al., 1992; Mazuelos et al., 2010). The utility of biooxidation is to regenerate Fe(III) at demanded rates and conditions in the above mentioned industrial contexts. On a commercial scale, only continuous operation is of interest. Continuous biooxidation can be carried out using bioreactors designed specifically for this purpose. In the last two decades, several types of reactors have been studied such as stirred tank reactors operating in both batch and continuous mode (Ojumu et al., 2008; Candy et al., 2009), airlifts (Kaksonen et al., 2014; He et al., 2022) and packed-bed bioreactors (Mazuelos et al., 2000; Chowdhury & Ojumu, 2014; Abbasi et al., 2021). Of the bioreactors found in the literature, those that have shown the highest efficiency are the flooded packed-bed bioreactors (FPBBs), which can obtain biooxidation rates higher than 3.5 kg/h⋅m 3 (Mazuelos et al., 2000). This bioreactor has been successfully tested at pilot plant scale, integrated in both hydrometallurgical and environmental processes (Frias et al., 2008; Avila et al., 2011). In FPBBs, the inlet liquor and air are continuously fed at the bottom, where an efficient mixing must be promoted between the gas and liquid phases. Both fluids ascend through the packed-bed occupying all the hollow volume between the particles on which the cells are supported. The liquid overflows from the top of the bioreactor (Mazuelos et al., 1999). The advantages associated with FPBBs are: •Ease of operation as no mechanical elements are required and low associated energy costs, as mechanical agitation systems are not required. •Stability of the biofilm. The cells are fixed on immobile particles. •Mean residence time and residence time distribution for the liquid can be controlled during operation from the inlet flow. •Gas-liquid interfacial area can be controlled during operation from gas injection conditions. •High gas–liquid interfacial areas can be achieved, due to the retention of the gas phase in the packed-bed. The objective of this work is to study continuous Fe(II) biooxidation at wider ranges of iron concentration and pH than those reported in literature ([Fe] <25 g/L and 2.5 >pH >1). For this purpose, a 1L FPBB was operated in continuous, thus testing the flexibility of this design regarding feed composition at extreme operational conditions. Control and robustness are also checked by planning more than 60 days of continuous operation without interruptions and reaching consecutives steady states for each operational condition tested. 2. Materials and methods 2.1. Microorganisms The culture used as inoculum for the present study was originally obtained from the Rio Tinto Mine in Huelva, Spain. It consists mainly of the mesophilic acidophilic iron oxidizing species Acidithiobacillus ferrooxidans and Leptospirillum ferrooxidans (Mazuelos et al., 2012). This culture has been regularly maintained on a modified Silverman and Lundgren 9 K nutrient medium (Silverman & Lundgren, 1959) at pH 1.25 (adjusted with concentrated H 2 SO 4 ) and 31 ◦C. 2.2. Batch tests A set of preliminary batch Fe(II) biooxidation tests have been carried out in stirred Erlenmeyer flasks at 180 rpm and 31 ◦C. These cultures were monitored by oxidation–reduction potential (ORP) probes (reference to Ag/AgCl) (Atlas Scientific ®) connected to a PC with an Arduino UNO controller. Influence of Fe(III) concentration and pH was studied in order to observe the tolerance and the adaptation of the culture to these parameters and its impact on biooxidation kinetics. All the tests were performed with 100 mL of 9 K nutrient medium, whose Fe(III) and Fe(II) concentrations and pH were modified to the set values for each case, and 10 mL of inoculum. Specific growth rate ( μ max ) was calculated from ln[Fe(III)] vs time curves. The rate of Fe(II) consumption is given by: d[Fe(II)] dt = − qs⋅CX(1) •where C X is the biomass concentration and assuming exponential growth can be calculated as follows: CX=CX0⋅e μ max⋅t (2) where C X0 is the initial biomass concentration. A. Mazuelos et al. Minerals Engineering 204 (2023) 108408 4 •and q s is the specific rate of substrate uptake and, when neglecting substrate uptake for maintenance, can be expressed as: qs= μ max YXS (3) where Y XS is the yield of biomass from substrate and μ max is the specific growth rate. By substituting Eqs. 2 and 3 into Eq. 1 and integrating with initial condition: t =0 and [Fe(II)] =[Fe(II)] 0 , the following expression is obtained: [Fe(II)] = [Fe(II)]0−CX0 YXS ⋅(e μ max⋅t −1)(4) In Eq. 4, C X0 can be referred to the final biomass concentration of the inoculum C Xf ino : CX0 =CXf ino⋅Vino Vcul (5) where V ino and V cul are the volumes of the inoculum and the culture, respectively. Moreover, C Xf ino can be calculated as: CXf ino =YXS⋅[Fe(II)]0 ino (6) where [Fe(II)] 0 ino is the initial ferrous iron concentration of the inoculum. In Eq. 6, it is assumed that the initial cell concentration of the inoculum must be considerably lower than C Xf ino and, in consequence, was neglected. In addition, it is considered that the final Fe(II) concentration in the inoculum is 0 g/L. Substituting Eq. 6 in Eq. 5 and separating variables, the ratio C X0 /Y XS (a) can be obtained from known values as follows: a=CX0 YXS = [Fe(II)]0 ino⋅Vino Vcul (7) Therefore Eq. 4 can be written as: [Fe(II)] = [Fe(II)]0−a⋅(e μ max⋅t −1)(8) From the ORP vs time curves it can be obtained the time when [Fe (II)] ≈0 as the time when an ORP value of 605 mV is reached (t ORP=605 ). At this time Eq. 8 reads: [Fe(II)]0=a⋅(e μ max⋅tORP =605 −1)(9) and from Eq. 9, the specific growth rate can be calculated as follows: μ max =ln([Fe(II)]0 a+1) tORP =605 (10) 2.3. Continuous operation A FPBB for continuous Fe(II) biooxidation was operated for more than 60 days without interruptions in three campaigns planned with different objectives: - Campaign 1 (553.5 h): to test the influence of total iron concentration at total conversion of Fe(II). Total iron concentration was increased (variable Fe(III) concentration and Fe(II) concentration close to 9 g/L) step by step from 9 to close to 60 g/L. pH had to be decreased from 2 to 1 to avoid precipitation of Fe. Air flow rate was 250 mL/min and liquid flow rate was about 125 mL/h. Air was injected by a ceramic diffuser. - Campaign 2 (577.5 h): at total iron concentration of 40 g/L and pH 1.2, liquid and air flow rates were simultaneously and progressively increased for preventing Fe(II) and oxygen limitations, respectively. - Campaign 3 (504 h): to test the influence of pH at total iron concentration of 40 g/L. In this campaign, pH was progressively decreased from 1.25 to 0.4, adjusting pH in feed with H 2 SO 4 . Air flow rate was 1500 mL/min and hydraulic retention time for the liquid was 3 h. A 2 mm open pipe was used as air diffuser. The FPBB consisted of an 8.4 cm diameter column containing a packed bed of 30 cm in height (floor 1 in Fig. 3). The bed consisted of siliceous stone particles ranging from 4 to 6 mm in diameter. Bed porosity was 0.42. The bed rested on a plastic grid of 5 mm mesh size. A ceramic diffuser and a 2 mm internal diameter pipe were used for injecting air. These diffusers were placed in floor 0 (Fig. 3) without support particles in order to promote turbulence by air injection. The geometrical characterization of the FPBB used in this experimental work is presented in Table 1. The start-up of the FPBB required 10 days, following the protocol described by Mazuelos et al. (2001). FPBB was operated at room temperature. FPBB monitoring consisted of the control of flow dynamics: liquid flow rate, air flow rate, and density, and the measurement of chemical parameters: pH, Fe and Fe(II) concentrations. Conversion of Fe(II) in the FPBB was calculated by Eq. 11: Conversion =[Fe(II)]inlet − [Fe(II)]outlet [Fe(II)]inlet (11) Biooxidation rate was calculated with the Eq. 12: Fig. 3. Schematic of a flooded packed-bed bioreactor for continuous Fe(II) biooxidation. Table 1 Dimensions of the FPBB operated. Dimension FPBB Diameter (cm) 8.4 Floor 0 height (cm) 5 Floor 0 vol (mL) 277 Floor 1 height (cm) 28 Floor 1 vol (mL) 652 Overflow volume (mL) 0 Total volume (mL) 930 A. Mazuelos et al. Minerals Engineering 204 (2023) 108408 5 Biooxidation rate =([Fe(II)]inlet − [Fe(II)]outlet)⋅Liquid flow rate Total volume of the FPBB (12) Oxygen solubility values were estimated by the model described in Mazuelos et al. (2017). And gas hold-up ( ε ) was calculated as follows: ε =Volume of air Total volume of the FPBB (13) To determine the volume of air for each flow rate the following procedure was carried out: •The bioreactor was slowly filled with liquid medium (9 K medium) avoiding bubble accumulation in the packed-bed. •Air stream was fed at the set flow. •The liquid displaced by the air, which overflowed by the outlet pipe, was gathered and weighted. 2.4. Analytics Total Fe concentration was determined by atomic absorption spectrophotometry at 248.33 nm in air-acetylene flame. Fe(II) concentration was determined by automatic potentiometric titration with K 2 Cr 2 O 7 0.05 N. Fe(III) concentration was determined by the sulfosalicylic acid method (Karamanev et al., 2002). 3. Results and discussions 3.1. Batch tests A set of preliminary batch biooxidation tests was carried out in order to study the influence of total iron concentration and pH on biooxidation kinetics. 3.1.1. Influence of Fe(III) concentration Tolerance and adaptation of the inoculum to different Fe concentrations were studied. In tolerance (T) tests, the inoculum was exposed for the first time to different Fe(III) concentrations (Table 2). Subcultures from the previous tolerance tests were carried out in the same conditions to study the adaptation (A) of the microorganisms to Fe(III) (Table 2). All the tests were conducted at pH 1.25 and an initial Fe(II) concentration of 9.09 g/L. Table 2 shows the initial Fe(III) concentration, the Fe(II) depletion time (t [Fe(II)]→0 ) and the specific growth rate for both tolerance and adaptation tests. Fig. 4 shows ORP vs time curves for the tolerance cultures. Slight differences for Fe(II) depletion times (t [Fe(II)]→0 ) and maximum Table 2 Tolerance (T) and adaptation (A) preliminary tests to high Fe(III) concentrations. Fe(II) initial concentration 9.09 g/L, pH 1.25, 180 rpm, 31 ◦C. Culture [Fe(III)] 0 (g/L) t [Fe(II)]→0 (h) μ max (h −1 ) 0Fe(III)-T 0.82 56 0.048 10Fe(III)-T 9.91 53 0.051 20Fe(III)-T 19.0 57 0.047 30Fe(III)-T 28.1 61 0.044 0Fe(III)-A 0.90 48 0.052 10Fe(III)-A 10.81 44 0.056 20Fe(III)-A 20.74 47 0.053 30Fe(III)-A 30.65 46 0.054 Fig. 4. ORP vs time curves for the tolerance tests to Fe(III). ( ) 0Fe(III)-T; ( ) 10Fe(III)-T; ( ) 20Fe(III)-T;( ) 30Fe(III)-T. Fe(II) initial concentration 9.09 g/L, pH 1.25, 180 rpm, 31 ◦C. A. Mazuelos et al. Minerals Engineering 204 (2023) 108408 6 specific growth rates were observed for the tolerance cultures (Table 2). These differences were even diminished in the adaptation study (Table 2). The small influence of the Fe(III) concentration on the biooxidation kinetic allows to affirm that the culture used as inoculum is easily adaptable to iron for concentrations up to 40 g/L. 3.1.2. Influence of pH An analogous set of experiments was carried out for testing the influence of pH. Table 3 shows the initial pH, the Fe(II) depletion time (t [Fe (II)]→0 ) and the maximum specific growth rate ( μ max ) for the tolerance (T) and adaptation (A) tests. All tests were conducted with initial Fe(II) and Fe(III) concentrations of 9.09 and 28.1 g/L, respectively. ORP vs. Table 3 Tolerance (T) and adaptation (A) preliminary tests to low pH. Initial Fe(II) and Fe(III) concentrations 9.09 and 28.1 g/L, respectively, 180 rpm, 31 ◦C. Culture pH t [Fe(II)]→0 (h) μ max (h −1 ) 1.25pH-T 1.25 61 0.044 1.00pH-T 1.00 60 0.045 0.75pH-T 0.75 73 0.037 0.50pH-T 0.50 129 0.021 1.25pH-A 1.25 46 0.054 1.00pH-A 1.00 45 0.055 0.75pH-A 0.75 50 0.049 0.50pH-A 0.50 73 0.034 Fig. 5. ORP vs time curves for the tolerance tests to pH. ( ) 1.25pH-T; ( ) 1.00pH-T; ( ) 0.75pH-T;( ) 0.50pH-T. Initial Fe(II) and Fe(III) concentrations 9.09 and 28.1 g/L, respectively, 180 rpm, 31 ◦C. Fig. 6. Operating conditions to study the influence of Fe concentration. Total Fe concentration (×) and Fe(II) concentration in the inlet stream (▴) and outlet stream (△). Liquid flow rate (◊) values are also presented. Air flow rate was 250 mL/min. Aeration by a ceramic diffuser. A. Mazuelos et al. Minerals Engineering 204 (2023) 108408 7 time curves for the tolerance cultures are shown in Fig. 5. Similar biooxidation rates were obtained for the tolerance tests with initial pH values of 1.25 and 1.00 (Table 3). Nevertheless, for the tolerance cultures with initial pH values below 1, the biooxidation rate is greatly diminished (Table 3). Similar results were obtained in the adaptation tests (Table 3). Biooxidation kinetics for the cultures with initial pH values of 1.25 and 1.00 are practically identical and the biooxidation rates were lower for the adapted cultures at pH below 1.00. However, the adapted cultures showed a kinetic improvement regarding the tolerance cultures (Table 3). The Fe(II) depletion times (t [Fe(II)]→0 ) and maximum specific growth rates ( μ max ) for the adapted cultures at pH 1.25 and 1.00 are the similar that those obtained for the adapted culture to Fe(III) concentration of 30.65 g/L (Tables 2 and 3), showing that, after an adaptation step, the highest biooxidation rate is reached. 3.2. Continuous operation Continuous biooxidation tests were carried out in three campaigns to study the influence of Fe concentration, liquid and air flow rates and pH. 3.2.1. Influence of Fe concentration Figs. 6 and 7 show the history of operating conditions for testing the Fig. 7. Ph values in the inlet stream (▴) and outlet stream (△) of the fpbb during the time of operation for the study of the influence of fe concentration. air flow rate was 250 mL/min. Aeration by a ceramic diffuser. Fig. 8. Study of the influence of Fe concentration: biooxidation rate (■) and conversion (◆) calculated for the FPBB during the time of operation. Air flow rate was 250 mL/min. Aeration by a ceramic diffuser. A. Mazuelos et al. Minerals Engineering 204 (2023) 108408 8 influence of iron concentration. In the inlet liquid stream Fe(III) concentration was progressively increased while Fe(II) concentration was kept at values close to 9 g/L (Fig. 6). Iron can precipitate in the boundary layers that form close to the solids (support particles, air diffusers, walls, and pipes). Biofilm growth in these locations. As the biofilm is a sink of protons, pH close to the biofilm can be expected to be different from the pH in the bulk liquid. To prevent iron precipitation, pH was decreased from 1.5 to 1.1 (Fig. 7). From the results shown in Fig. 6, Fe(II) conversion and biooxidation rate were calculated by equations 11 and 12, respectively (Fig. 8). Fe(II) conversion was 100 % at practically all times except for the first 72 h, which were attributed to the period of growth and maturation of the biofilm after the start-up process (Fig. 8). Fe(III) concentration and pH did not affect the performance of the reactor in the range tested. These results are in accordance with those obtained in the batch tests (Tables 2 and 3). In this reactor, biooxidation rates were close to 1.5 g/L⋅h, which is lower than the reported one in the literature (Mazuelos et al., 2000). This can be explained by substrate (Fe(II)) limitation. It is worth noting the stability of the bioreactor for more than 500 h of continuous operation without interruptions, reaching steady state for each Fe(III) concentration studied. 3.2.2. Influence of liquid and air flow rates Under non Fe(II) limiting conditions, the productivity of the bioreactor should be very sensitive to changes in the aeration conditions, especially those regarding to the flow rate at which air is injected. An increase of this variable usually has the positive effect of increasing the gas–liquid interfacial area through the increase of gas hold-up (Mazuelos et al., 2002). The air flow rate not only modifies gas hold-up but also introduces turbulence and mixing in the liquid phase. Fig. 9 shows gas hold-up values calculated by Equation 13 as a function of the air flow rate for the FPBB operated. The FPBB was continuously operated for 25 days without Fig. 9. Gas hold-up (•) values calculated for the FPBB at different air flow rates. Fig. 10. Operating conditions in the study of the influence of air and liquid flows rates. Air (•) and liquid (×) flow rates of the FPBB during the time of operation. Total Fe concentration 40 g/L, Fe(II) concentration 10 g/L and pH 1.10. At 750 h (vertical line), the ceramic diffuser was changed by a 2 mm open pipe. A. Mazuelos et al. Minerals Engineering 204 (2023) 108408 9 interruptions in order to study the influence of air and liquid flow rates. To prevent substrate and oxygen limitations, the liquid and air flow rates were increased progressively from 134 to 886 mL/h and from 250 to 1500 mL/min, respectively (Fig. 10). A ceramic diffuser was used for the first 750 h. The air passage channels in the ceramic diffuser were progressively closing due to the deposition of ferric precipitates and biofilm, requiring more air pressure to maintain the desired flow rate. As a result, during the FPBB operation it was necessary to replace the ceramic diffuser with a 2 mm open pipe. The inlet liquors had the following set point values: total iron concentration 40 g/L, Fe(II) concentration 10 g/L and pH 1.10. Fig. 11 shows biooxidation rate and conversion as a function of time. Increasing the air flow rate from 250 to 750 mL/min allowed the liquid flow rate to be increased from 134 to 291 mL/h maintaining the complete conversion of Fe(II) and increasing the biooxidation rate up to 3.27 g/L⋅h. Figs. 10 and 11 include the data obtained for similar operational conditions at 457.5 h and 481.5 h in the campaign 1 for the study of the influence of Fe concentration.. The change of the ceramic diffuser to a 2 mm open pipe (at 750 h) brought a decrease in conversion and biooxidation rate (Figs. 10 and 11). In this low height reactor, the influence of the diffuser plays an important role in the generation of gas–liquid interfacial area (Mazuelos et al., 2002). However, the progressive increase in air flow rate from 750 to 1500 mL/min is associated with an increase in biooxidation rate and conversion that corrects the aforementioned decrease in the FPBB performance due to the diffuser change. After 890 h, the conversion decreased drastically as a consequence of the increase in liquid flow rate up to 886 mL/h, maintaining the Fig. 11. Study of the influence of air and liquid flow rates: biooxidation rate (■) and conversion (◆) calculated for the FPBB during the time of operation. Total Fe concentration 40 g/L, Fe(II) concentration 10 g/L and pH 1.10. Fig. 12. Operating conditions in the study of the influence of acidity. pH values in the inlet stream (▴) and outlet stream (Δ) of the FPBB during the time of operation. Total Fe concentration 40 g/L, Fe(II) concentration 10 g/L, liquid flow rate 290 mL/h and air flow rate 1500 mL/min. 2 mm open pipe as air diffuser. A. Mazuelos et al.