Contents lists available at ScienceDirect Journal of CO2 Utilization journal homepage: www.elsevier.com/locate/jcou CO2 sequestration by pH-swing mineral carbonation based on HCl/NH4OH system using iron-rich lizardite 1T Gretta Larisa Aurora Arce Ferrufinoa,c,d,⁎, Sayuri Okamotoa, Jose Carlos Dos Santosa, João Andrade de Carvalho Jr.c, I. Avilac, Carlos Manuel Romero Lunab,d, Turibio Gomes Soares Netoa a Combustion and Propulsion Associated Laboratory, Brazilian Space Research Institute (LCP/INPE), Brazil b Production Engineering, Campus of Itapeva, São Paulo State University (UNESP), Brazil c Combustion and Carbon Capture Laboratory, Energy Department, Campus of Guaratinguetá, São Paulo State University (LC3/DEN/UNESP), Brazil d Advanced Materials and Nanotechnology Research Group, Faculty of Chemical and Metallurgical Engineering, Jose Faustino Sanchez Carrion National University (UNJFSC), Huacho, Lima, Peru A R T I C L E I N F O Keywords: Mining waste Lizardite 1T pH-swing mineral carbonation HCl/NH4OH system Carbonates CO2 sequestration A B S T R A C T In pH-swing mineral carbonation, several acid/base systems has been investigated. Currently the main acid/base systems employed are HCl/NaOH and NH4HSO4/NH4OH. However, the use of a HCl/NH4OH system was not yet elucidated. This study proposes to evaluate the feasibility of a pH-swing mineral carbonation based on HCl/ NH4OH system at atmospheric pressure and moderate temperatures using mining waste from asbestos pro- duction from Goiás State, Brazil (S-GO) for two conditions (i.e. stoichiometric conditions (T2E) and acid excess (T2)). Results indicated that the Fe3+ content in S-GO acted as a catalyst, due to FeCl3 hydrolysis in aqueous solutions. Thus, high Mg and Fe extraction efficiency (95 ± 2%), were achieved in the leaching stage for both conditions. The S1 solid residue was mainly SiO2 with 90 ± 1% purity content. In the purification stage 91.7 ± 1.9% of Fet were removed, however, a loss of Mg of 13.6 ± 2.3% was also detected. On the carbonation stage, high purity hydromagnesite was formed in T2E; this stage had a 85% efficiency, thus, 36.7% of CO2 was fixed. On T2, excess H2O and CO2 promoted dypingite formation and reduced hydromagnesite formation. After carbonation, the formation of crystals was observed in the NH4Cl aqueous solution at 25 °C, indicating NH4Cl supersaturation. The results of mass balance indicate that 4 ton of mineral waste will be employed for each ton of captured CO2, as well as 2.6 ton of HCl, and 4.5 ton of NH4OH. However, 1.7 ton of SiO2, 0.55 ton of iron oxides, and 2.7 ton of hydromagnesite could be produced. 1. Introduction Sterile is a residue from a mining step related to asbestos production, normally disposed of in landfills. However, since it has low chrysotile contents, it is considered a hazardous waste and its disposal represents a challenge for environmental engineering and public health [1–5]. According to Valuoma et al. [5], Fedoročková et al. [6], and Gadikota et al. [1], carbon capture and storage by mineralization (CCSM) using asbestos-containing materials is a technology that would have several environmental benefits, since it can permanently alter the asbestos fiber structure in these materials, as well as sequester carbon dioxide (CO2) in a safe and permanent manner, producing materials with high ag- gregated value in the market. This would avoid disposal issues normally associated to these mining wastes [7]. Among all CCSM methodologies, the pH-swing mineral carbonation is recently receiving more attention, since it presents a favorable ki- netics. Besides this, it would allow obtainment of useful products and subproducts, which could be commercialized, reducing elevated costs associated to such processes [28–13]. Although many challenges have been overcome by employing the pH-swing mineral carbonation, there are other issues that still remain unsolved. The major hurdle is related to the high energy consumption, making it difficult to insert this technology at industrial scale [13]. Great part of the energy consumption is linked to acid and base recovery, mainly used for pH control. High leaching efficiency is as- sociated to pH under 1 (for example, in the extraction of reactive ele- ments such as Mg, Ca, and Fe). However, carbonation efficiency is in- hibited at pH under 7. Thus, to maximize carbonates production, pH https://doi.org/10.1016/j.jcou.2018.01.001 Received 7 May 2017; Received in revised form 27 November 2017; Accepted 2 January 2018 ⁎ Corresponding author at: National Institute for Space Research (INPE) Combustion and Propulsion Associated Laboratory (LCP) Rodovia Presidente Dutra, km 40, Cachoeira Paulista, SP, CEP 12630-000, Brazil. E-mail addresses: gretta@lcp.inpe.br, grettagaf@yahoo.es (G.L.A. Arce Ferrufino). Journal of CO₂ Utilization 24 (2018) 164–173 Available online 05 January 2018 2212-9820/ © 2018 Elsevier Ltd. All rights reserved. T http://www.sciencedirect.com/science/journal/22129820 https://www.elsevier.com/locate/jcou https://doi.org/10.1016/j.jcou.2018.01.001 https://doi.org/10.1016/j.jcou.2018.01.001 mailto:gretta@lcp.inpe.br mailto:grettagaf@yahoo.es https://doi.org/10.1016/j.jcou.2018.01.001 http://crossmark.crossref.org/dialog/?doi=10.1016/j.jcou.2018.01.001&domain=pdf must be increased to levels above 9 [14–16]. Therefore, use of several acids and bases was thoroughly investigated in order to guarantee carbonates formation. Currently the main acid/base systems employed are HCl/NaOH and NH4HSO4/NH4OH [9,12,17–19]. The advantage of the HCl/NaOH system is the recognized high ef- ficiency of leaching stage when used hydrochloric acid (HCl), with a short reaction time of around 2 h [29,20] and a maximum of 30min on carbonation stage to form carbonates when sodium hydroxide (NaOH) is used [16]. However, the main disadvantage related to this system is a high energy consumption to recover these chemical reagents when electrolytic process is used (3MWh/tCO2). Recently, Yuen et al. [13] indicated that energy consumption from this system could be reduced by 1.2MWh/tCO2 if bipolar membrane electrodialysis (BMED) is em- ployed to recover HCl and NaOH as solutions. However, as mentioned by Bu et al. [21], their concentrations would be limited to 1mol L−1. The main advantage in the case of the NH4HSO4/NH4OH system is the possibility to broadly employ recovered chemical reagents. Moreover, previous CO2 capture stages would not be necessary [22]. However, this system presents two disadvantages which limit its use: (i) a long reaction time of leaching of around 3 h; and (ii) high energy consumption to recover NH4HSO4 and NH4OH from an aqueous solu- tion of ammonium sulfate (NH4)2SO4. Since low concentrations of NH4HSO4 solution is employed for leaching stage, large amounts of water are required, and hence, there is a considerable increase of en- ergy consumption on following stages associated to (NH4)2SO4 solution concentration processes [18]. Analysis of cost sensitivity indicate that if reaction time is reduced to 1 h, there is a possibility to reduce the capital cost by 50%, since the number and/or size of equipment can be reduced by half [13,22]. Thus, use of HCl would present advantages when compared to NH4HSO4. On the other hand, several carbonates present different morpholo- gies and their formation depends on temperature, pressure, and pH of the reaction medium [10,14,16,23–28]. CO2 solubility in water is in- hibited with an increase of temperature; however, formation of stable carbonates is favored when carbonation temperature is between 70 °C and 100 °C. Moreover, when solution pH is kept at values above 9, carbonate formation is also favored. Thus, for pH-swing mineral car- bonation, strong bases (NaOH) as well as weak bases (NH4OH) were employed. According to studies performed by Teir et al. [16]. Zhang et al. [28], both types of bases do not present advantages or dis- advantages between them, with regards to keeping the solution pH. However, strong bases are more chemically stable than weak bases. Thus weak bases normally have a smaller dissociation degree which would allow an easy and fast recovery in stage following carbonation, when compared to a strong base. Although several acid/base systems have been broadly studied by the aforementioned authors, the feasibility of a pH-swing mineral car- bonation process based on a strong acid and a weak base (HCl/NH4OH) was not yet elucidated. According to Zhang et al. [28], solutions pre- pared with 1.41mol L−1 magnesium chloride (MgCl2) might precipitate carbonates by using CO2/NH3.H2O in elevated pressures and moderate temperatures. A maximum carbonation efficiency of 73.6% was re- ported for a temperature of 100 °C and pressure of 6MPa. However, it must be highlighted that carbonation efficiency was not yet in- vestigated in atmospheric pressures. This study proposes to evaluate the feasibility of a pH-swing mineral carbonation process based on HCl/NH4OH system at atmospheric pressure and moderate temperatures. In order to accomplish this, an iron-rich mining waste from asbestos production process was used. The study investigates the silica production, selective precipitation of iron oxides, and carbonates production, as well as, the CO2 capture effi- ciency. 2. Materials and methods 2.1. Materials A mining waste was employed in this study, which came from the Minaçu Mine, located in Cana Brava, Goiás State, Brazil. The mining waste was named S-GO. It was crushed and grinded in order to obtain 150 μm particle size. This particle size range reports an acceptable energy requirement of 11 kW h by ton of processed mineral, in the previous stages of crushing and grinding [16,22,29,30]. Thus, the particle size classification in this study was done by using a number of sieves ASTM 120–80. 2.2. Characterization methods Samples of each material (S-GO, S1, S2, S3, L1, L2, L3) were prepared and underwent the following analyses: thermogravimetric analysis (TGA); X-ray diffraction (XRD); wavelength dispersion sequential fluorescence (XRF); and inductively coupled plasma optical emission spectroscopy (ICP-OES). Thermogravimetric analyses (TGA) were carried out to assess the materials thermal behavior. These TGA were carried out in a TA Instrument SDT TGA-DSC Q600 simultaneous system. Sample mass was approximately 30 ± 2mg and a 90 μL alumina crucible was used for all tests, with a dynamic nitrogen atmosphere as purge gas, at a flow rate of 100mL/min and a heating rate of 10 °C/min. The temperature range for S-GO and S1 were 30 °C−1000 °C; for S4 and L3 were be- tween 30 °C−600 °C; and 30 °C−300 °C, respectively. Mineralogical composition of the materials was determined using a PANalytical X’pert3 Powder model X-ray diffraction analyzer (XRD). This device uses Cu Kα radiation in a range of 6–90° 2θ. Diffractograms obtained were processed using the HighScore Plus software. A mass of 1 ± 0.5 g was used for this analysis. Sequential fluorescence with wavelength dispersion (XRF) was used for the quantitative analyses of materials chemical composition (SiO2, MgO, CaO, Fe2O3t, K2O, Na2O, Al2O3, TiO2, MnO, Cr2O3). It was carried out using a PANalytical Axios MAX-Advanced model device, with 4.0 kW operating power and 60 kV agitation. This device was employed to carry out a quantitative elemental chemical analysis of boron (B) and uranium (U). For such analysis, a 1 g mass sample was used for each material. ICP-OES analyses were conducted for determining the mate- rials elemental composition(Si, Mg, Ca, Fe, K, Na, Al, Ti, Mn, Cr) with an Arcos Spectro model, inductively coupled with a plasma optical emission spectroscope. 2.3. Process description A scheme of the process and its main reactions are presented on Fig. 1 and Table 1, respectively. The proposed process has four stages: 1) leaching, 2) purification, 3) carbonation, and 4) recovery of chemical reagents. Leaching stage extracts reactive elements (Mg, Fe) from the mining wastes (S-GO) employing HCl solutions, in which the main involved reaction is described as R1 and presented on Table 1. This stage gen- erates a leached solution (L1) with high chloride concentrations from those reactive elements, as well as a solid residue (S1) with elevated silica content [29,20]. The leached L1 solution is conducted to the purification stage, in which a NH4OH solution is employed for pH adjustment [27,30]. pH is increased during purification in order to precipitate Fe3+ and Fe2+in the form of hydroxides, as shown on reactions R2 and R3 (Table 1). Formed hydroxides in this stage (S2 and S3) are separated from the resulting purified solution (L2). Once Fet is separated, the purified L2 solution presents high MgCl2 concentration. The purified L2 solution is forwarded to the carbonation stage, in which it is mixed with an additional NH4OH solution and CO2, G.L.A. Arce Ferrufino et al. Journal of CO₂ Utilization 24 (2018) 164–173 165 producing carbonates (S4) according to reaction R4 presented on Table 1. Produced carbonate (S4) is precipitated and separated from NH4Cl (L3) aqueous solution. This solution is then sent to the last stage (re- covery of chemical reagents), in which NH4Cl can be separated into NH3 and HCl according Zhang’s methodology [28]. 2.3.1. Leaching experiments Mass/volume ratio for the experimental tests in this work were 74 g/L and, 37 g/L for stoichiometric conditions (T2E) and non-stoi- chiometric (T2 – acid excess), respectively. Initially, 100mL of a 2mol L−1 HCl solution was inserted into the vessel reactor of 500mL and heated until 100 °C. Once the acid solution was heated, S-GO powder was added to the vessel reactor, which was equipped with a Graham condenser in order to avoid HCl loss by evaporation, as well as a thermocouple for temperature control. The reactor was continuously agitated with a magnetic stirrer set at 600–700 rpm during 2 h. After each leaching process experimental test (T), two products were obtained, one solid residue (S1) and one leachate solution (L1). The solid residue S1 was separated from the leachate solution L1 through vacuum filtration. Following each experiment, residues (S1) were dried during 2 h at approximately 105 °C. XRD, XRF, and TGA analyses were then carried out. Leachate solutions (L1) were analyzed through ICP-OES in order to obtain Mg, Fe, and Si concentrations. Mg, Fe, and Si extraction in the leachate solutions (L1) were calculated based on their content within S-GO, as showed in the Equation below (1): = × × × − − X V C M i% 100i L SOL i S GO S GO (1) In which: Xi L represents the leaching efficiency, −i%S GO is the initial element content (“i”: Mg, Fe, and Si) in the S-GO sample, MS-GO is the initial S-GO mass used in the experiments, and VSOL is the leachate solution volume after 2 h of reaction. Ci is the concentration of elements in the leachate solution. 2.3.2. Impurities separation Leachate solutions (L1) were sent to the purification stage in order to separate the dissolved total iron (Fet). Separation was done in two steps, by controlled precipitation [9]. Thus, iron II (Fe2+) and iron III (Fe3+) were precipitated as hydroxides, when increased the solution pH. For this, a 30% w/w NH4OH solution was employed. pH mea- surement was recorded through a METTLER TOLEDO pHmeter, Seven Easy pHmeter S20, with a ± 0.01 at 25 °C precision. In the first purification step, NH4OH addition was finalized when solution pH (pHP1) reached a value of 5. Formation of a precipitate (S2) was observed in this step, thus all Fe3+ was removed of aqueous so- lution as Fe(OH)3 [9]. This precipitate was separated from the aqueous solution (or semi-purified – L’2) by vacuum filtration. After that, the semi-purified L’2 solution was sent to the second purification step in order to remove Fe2+. Thus, an additional amount of NH4OH in solu- tion was added until L’2 solution reached a pH value of 9 (pHP2). Once the pH was adjusted, formation of another precipitate (S3) was ob- served, which was separated from the purified solution (L2) by vacuum filtration. S2 and S3 solid precipitates were dried within an oven set at 105 °C during 12 h. They were then sent to elemental analysis to determine MgO, Fe2O3t, and SiO2 through XRF. Purified L2 solutions were sent for elemental analysis through ICP-OES in order to determine Mg, Fe, and Si concentrations. The calculation for purification efficiency was done by using Eq. (2) below: = − × − X m m m [ ] 100,i P i L i P i S GO (2) In which: Xi P, is the efficiency of the purification stage; i, is the Mg and Fe ions; mi L, is the concentration of each i in the leachate solution; mi P, is the concentration of each i in the purified L2 solution. 2.3.3. Formation of carbonates An alkaline pH is one determining factor to precipitate carbonates [14,16,27,28]. Thus, in the L2 purified solutions, an NH4OH solution was added at Mg:NH4OH mass proportion of 1:2.9. This addition was NH4OH NH3 HCl HCl L1 L2 L3Leaching (R1) Purification (R2,R3) Carbonation (R4) NH3.H2O CO2 S1 S2, S3 Recovery of chemical reagents Fig. 1. Process scheme with several stages to pH-swing mineral carbonation based on HCl- NH4OH system. Table 1 Chemical reaction and thermodynamic properties from different process stages. R.Q. Reaction Equations T ΔH ΔG °C kJ kJ R1 (Mg,Fe)3(Si,Fe)2O5(OH)4+ 6HCl= 3(Mg,Fe)Cl2+ 2FeCl3+ 2SiO2+ 5H2O 100 −321.3 −243 R2 FeCl2+ 2NH4OH=Fe(OH)2+2NH4Cl 25 −120.6 −97 R3 FeCl3+ 3NH4OH=Fe(OH)3+3NH4Cl 25 −101.2 −180 R4 5MgCl2+ 10NH4OH+4CO2=Mg5(OH)2(CO3)4*4H2O+10NH4Cl 90 −42.9 −98 G.L.A. Arce Ferrufino et al. Journal of CO₂ Utilization 24 (2018) 164–173 166 done at room temperature (25 °C) and mixed at 400 rpm, guaranteeing a pH equivalent to 11. Once NH4OH was added to the L2 purified solutions, the mixtures were conducted to the carbonation stage, in which the process tem- perature was set at 90 °C. However, before carbonation, the mixture was pre-heated to 60 °C, followed by CO2 injection at a flow rate of 20mL/min, until reaching 90 °C. CO2 injection was kept for 60min in order to guarantee total carbonates precipitation. The reactor was equipped with a Graham condenser to avoid losses by NH4OH evaporation. Besides this, a temperature and pH sensor was coupled to the reactor in order to have control of such variables during reaction. During 60min of carbonation, a precipitate (S4) was formed, which was separated from the liquid phase by vacuum filtration (car- bonated solution – L3). The S4 precipitates were dried in one oven at 105 °C during 12 h. Later on, the precipitates were characterized by XRF, XRD and TGA, respectively. On the other hand, the L3 solutions were analyzed by ICP- OES to determine Mg concentration. Carbonation and CO2 capture ef- ficiency calculations were obtained using Eqs. (3) and (4), respectively. = − × − X m m m ( ) 100Mg C Mg I Mg C Mg S GO (3) = − ×X m m m ( ) 100CO C CO In CO i C CO In2 2 2 2 (4) In which: XMg C , is the carbonation efficiency; mMg I , is the Mg mass in the L2 purified solution; XMg C , is the Mg mass in the L3 solution after the carbonation stage; −mMg S GO, is the Mg mass in the S-GO. On the other hand, XCO C 2 , is the CO2 capture efficiency, mCO In 2 is the CO2 injected mass within the reactor; mCO i C 2 , the fixed CO2 mass in the carbonate. The fixed CO2 mass within the carbonate was quantified by TGA. 3. Results and discussion 3.1. Materials characterization As it can be observed on Fig. 2, the S-GO mining waste presents a heterogeneous composition, with lizardite 1T as main mineral and secondary phases as brucite, magnesite, clinochrysotile and magnetite. According Arce et al. [2], this mineralogical characterization exhibits presence of clinochrysotile traces in the S-GO structure, indicating it as a hazardous waste. Rietveld quantification from different mineral phases resulted in 92% serpentines, 6.3% brucite, 1.09% magnesite, and 1.96% magnetite. Table 2 presents the S-GO chemical composition determined by XRF, which showed 43.33% MgO, 40.64% SiO2, and 12.61% Fe3O4. Other metals were also detected by ICP-OES, such as K, Na, Al, Ti, Mn, and Cr in less than 2.4% of the solid. The BET area corresponding to S-GO was determined to be 6.7 cm3/g. Fig. 3 shows TG/DTG curves for S-GO, which exhibits a total loss of mass around 15.5% between 30 °C and 1000 °C. Three decomposition events were observed which corresponded to mineral phases in the S- GO. DTG peaks at 382 °C correspond to brucite and magnesite dehy- droxylation, and DTG peaks at 600 °C and 659 °C correspond to lizardite 1T and clinochrysotile dehydroxylation, respectively [2]. As stated in Arce et al. [231]., the S-GO is derived from hydration of dunites (or olivine - (Mg,Fe)2SiO4), consequently there were brucite production along with serpentines. The phase diagram of the MgO- SiO2-H2O-FeO system can represent the serpentinization processes [12]. Besides that, depending on the P (atm) and T (°C), a minerals mixture (i.e. serpentines [serp - Mg3Si2O5(OH)4]+ brucites [Brc – Mg (OH)2]+magnetites [Mag – Fe3O4]) can be produced. Evans [32] suggested that the iron (Fe) – derived of the olivine ((Mg,Fe)2SiO4), that it is not embedded into of the serpentine structure by the cationic re- placement mechanism, could be precipitated as Fe3O4. In addition, in high SiO2 activities, Fe3O4 is not stable and the growth of Fe-rich ser- pentines is favored [32]. Moreover, Hostetler et al. [33] indicated that Fe inside brucite structure is related to exchange potential (μΔ (Fe2+Mg-1)). The μΔ (Fe2+Mg-1) exerts a control on the molar fraction of Fe2+ turning it a ferromagnesium minerals [32]. So, the brucite contained within S-GO had a replacement of Fe2+ by Mg2+ from iron rich olivine. There are structural differences between the serpentines; unlike the curve structures from chrysotile and antigorite, the lizardite have flat structures [34]. In the lizardite, the distance between flat layers is produced by cations replacement mechanism [35]. When the replace- ment of Si4+ by Al3+ in the tetrahedral sheets is realized, the forces of attraction between layers will become stronger [34]. Lizardite 1T has a flat structure that is more organized when compared with other poly- types, due to the affinity of lizardite 1T to replace Al3+ in the Si4+ 10 20 30 40 50 60 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 5500 L L LL L L S-GO L B L M m C B C C M Fig. 2. S-GO X-ray diffractogram. Table 2 Chemical composition of S-GO mining waste. XRF ICP Oxides Concentration (%) Elements Concentration (%) SiO2 40.64 Si 8.40 MgO 43.33 Mg 23.01 CaO 0.10 Ca 279.5a Fe2O3t 12.61 Fe 4.49 K2O n.d. K < 0.01a Na2O n.d. Na < 0.02a Al2O3 1.17 Al 0.35 TiO2 n.d. Ti 140.3a MnO 0.20 Mn n.d. Cr2O3 1.02 Cr n.d. n.d.= not detected. a mg/kg. G.L.A. Arce Ferrufino et al. Journal of CO₂ Utilization 24 (2018) 164–173 167 tetrahedral sheets. Therefore this affinity makes it less porous, and more stable [34,36]. Arce et al [2] suggested that lizardite 1T with Al2O3 content between 0.23–1.00% has less thermal and chemical stability. On the other hand, according to Lacinska et al. [37]. and Chamley [38], the lizardite stability can be affected by the Fe3+ re- placement into Mg2+ octahedral and/or Si4+ tetrahedral sheets, how- ever, the Fe3+ replacements into Mg2+ octahedral sheets could produce less stability [38]. According to Wicks and Wittacher [34], Fe2+ inside of the lizardite structure is unlikely since the atomic radius of Fe2+ would not allow the flat layer formation. So, crystalline structure of lizardite 1T presents higher iron content, mainly as Fe3+ and a low Al3+ content. 3.2. Leaching experiments L1 leachate solutions of both experimental tests, stoichiometric conditions (T2E) and acid excess (T2) were observed to be yellow; and solid residues (S1) were slightly gray. Total mass of Mg, Fe and Si contained in L1 leachate solutions, as well as, the efficiencies of Mg and Fe extraction from leaching stage are presented on Table 3, for both T2E and T2 tests. It can be observed that for both conditions, a high ex- traction efficiency were achieved using this type of mining waste, with an average values of 94 ± 2.0% for Mg and 95 ± 2.5% for Fe, re- spectively. Although L1 leachate solutions presented a higher amount of Mg than Fe, the amount of Fe extracted in T2E and T2 tests is significant when compared to other studies reported in the literature [20,39]. The S-GO behavior in the leaching stage can be attributed to the low content of Al3+, once this reduces the thermal and chemical stability from this serpentinite [2,40,41]. It must be mentioned that the pH of the L1 leached solutions (pHL) were kept very acidic, even after 2 h of reaction, as 0.45 for T2E test and 0.02 for T2 test (Table 4). For pH-swing mineral carbonation, the pH is inversely proportional to leaching efficiency. As reported by McCutcheon et al. [15]; pHs higher than 1 can reduce the Mg extraction from serpentines. On the other hand, direct mineral carbonation indicates that high Fe3+ content in Mg-bearing silicates which are also present in mining wastes, inhibits the formation of carbonates due to medium acidifica- tion [14,25,42]. In this study, Fe3+ present within S-GO is leached as ferric chloride (FeCl3). The ferric chloride (FeCl3) has promising catalytic abilities, then, it has been used as an acid catalyst in many processes [43,44]. The aqueous FeCl3 solutions have the property to acidify the reactional medium, due to hydrolysis processes, since contact between FeCl3 and H2O favors H+ release. Therefore, FeCl3 had a catalytic effect in the leaching stage, promoting dissolution of other reactive metals, such as Mg. It should be highlighted that probably part of Fe2+ has been oxi- dized to Fe3+, once the reaction was not realized in an inert atmosphere (N2 not added into the reactor). In addition, although not investigated here in detail, the kinetics studies considering the structural modifica- tion (i.e the crystal order, the surface area and the porosity) of the iron- 200 400 600 800 1000 84 86 88 90 92 94 96 98 TG DTG 382 600 659 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 Fig. 3. TG/DTG curves from S-GO. Table 3 The amounts of metals in the solutions and the efficiency obtained from several stages of the process. Stage Test Solutions Metal Mass Efficiency Name V (mL) Mg (mg) Fe (mg) Si (mg) Mg (%) Fe (%) S-GO T2E S-GO 100 1734a 357a 1000a — — T2 S-GO 100 891a 181a 507a Leaching T2E L1 90 1672 331 17 96.4 92.7 T2 L1 88 826 177 13 92.7 97.7 Purification T2E L2 97 1475 10 12 11.4 89.9 T2 L2 110 685 8 14 15.8 93.4 Carbonation T2E L3 115 5 10 12 84.8b 81c T2 L3 119 4 8 13 76.4b 42c a Mass of Mg, Fe, and Si, within S-GO mining waste. b carbonation efficiency. c CO2 capture efficiency. G.L.A. Arce Ferrufino et al. Journal of CO₂ Utilization 24 (2018) 164–173 168 rich mining wastes should be carried out in future studies, since this will allow knowing the impact of FeCl3 on the linkages of octahedral and tetrahedral sheets. In Figs. 4a, b, 5a and b , a large structural S-GO modification can be observed after leaching on diffractograms and TG/DTG curves of the S1 solid residues for both experimental tests (T2E and T2). Fig. 4a and b show curved patterns for S1 diffractograms obtained on experimental tests T2E and T2, respectively. S-GO crystalline structure was modified to amorphous SiO2, however, there was also the presence of crystalline silica mainly in the T2E stoichiometric test. It must be emphasized that although high Mg and Fe extraction were achieved in this stage, S1 solid residues diffractograms still in- dicate the presence of lizardite 1T, clinochrysotile, and brucite. This is corroborated by TG/DTG curves of S1 solid residues presented on Fig. 5a, b, e, which show DTG decomposition peaks referring to such minerals (382 °C for brucite, 600 °C for lizardite 1T, and 659 °C for clinochrysotile). Nevertheless, although that a high leaching efficiency has been obtained, the total S-GO dissolution was not reached due to the Al2O3 content into the rock, which is above 1.00%. On the Table 5, it is observed that the chemical composition of S1 solid residues obtained in both T2E and T2 tests, exhibits a content of MgO, Fe2O3t, and SiO2, approximately of 3.6–2.8%, 2.3–2.0% and 90.4–91.5%, respectively. Such S1 residues are not yet appropriate to be commercialized, since silica purity must be higher than 99% [6]. However, SiO2 content could be increased by using purification pro- cesses. 3.3. Impurities separation Once the leaching stage was concluded, L1 solutions were sent for pH adjustment until approximately 5 and 9, in order to perform se- lective precipitation of Fe3+ and Fe2+. Table 4 shows total volumes of NH4OH added to adjust pH (P1 and P2) in the purification stage. It can be observed that NH4OH added on T2 experimental test is three times higher than T2E experimental test. Large base consumption for pH adjustment during the T2 experimental test is due to excess acid used during the leaching stage, which was twice the stoichiometric value. It must be mentioned that the acid excess only increased Mg and Fe ex- traction by 5%. Thus, by using materials with S-GO characteristics, a stoichiometric condition would reduce the consumption of chemical reagents. Table 3 shows Mg, Fe and Si amounts within the L2 purified solu- tion, as well as precipitation efficiencies from the purification stage. It can be seen that an average of Fe precipitation efficiency of approxi- mately 91.7 ± 1.9% was obtained for both T2E and T2 experimental tests, moreover, it was also observed an average of 13.6 ± 2.3% Mg precipitated in this stage. On Table 5, in the purification stage can be observed a higher production (mp) of S2 precipitate than S3 precipitate, indicating a high Fe3+ content, this is congruent with lizardite 1T inherent characteristic [34]. S3 precipitate in small amount refers to ferrous brucite ([Mg,Fe (OH)2]) and/or clinochrysotile [2,45]. The concentration of Fe2O3t in S2 is of approximately 91.2% and 89.5% for experimental tests T2E and T2, respectively, and in the S3 precipitate is of approximately 90.7% and 88.3% for tests T2E and T2, respectively (Table 5). Table 4 The volume of NH4OH added and pH adjustment for each stage from the pH-swing mi- neral carbonation. test Leaching Purification Carbonation P1 P2 initial Final T2E NH4OH added (mL) 0.00 4.6 2.4 18.1 0.00 pH 0.45 5.04 9.00 10.85 8.85 T2 NH4OH added (mL) 0.00 8.7 12.8 8.9 0.00 pH 0.02 5.05 8.99 11.97 7.03 10 20 30 40 50 60 10 20 30 40 50 60 B L L S B S a) L B L S B S b) c) HH H H H H D DD d) H H H H H H D Fig. 4. Diffractograms for solid residues. a) S1 solid residue of T2E, b) S1 solid residue of T2, c) S4 solid residue of T2E and d) S4 solid residue of T2. G.L.A. Arce Ferrufino et al. Journal of CO₂ Utilization 24 (2018) 164–173 169 200 400 600 800 1000 84 86 88 90 92 94 96 98 b) Temperature (°C) 200 400 600 800 1000 a) 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 100 200 300 400 500 0 20 40 60 80 100 c) 100 200 300 400 500 600 T emperature (°C ) 0 5 10 15 20 d) 50 100 150 200 250 300 0 20 40 60 80 100 e) W /W o (% ) 50 100 150 200 250 300 f) 0 2 4 6 8 10 12 14 16 18 Fig. 5. TG curves (line) and DTG curves (dot lines) for residues obtained during several stages of the pH-swing mineral carbonation process. a) S1solid residue of T2E b) S1 solid residue of T2,c) S4 solid residues of T2E, d) S4 solid residues of T2, e)L3 solutions of T2E and f) L3 solution of T2. G.L.A. Arce Ferrufino et al. Journal of CO₂ Utilization 24 (2018) 164–173 170 3.4. Carbonates precipitation As soon as the L2 solutions were purified, they were sent to carbo- nation stage. In order to keep the alkaline pH of the L2 solution, NH4OH was added to both T2E and T2 experimental tests. Table 4 shows that the T2E test employed a larger amount of NH4OH, due to a higher amount of Mg contained in the T2E stoichiometric test when compared to T2 tests with excess acid. The pH value after base addition was 11.41 ± 0.56 for both experimental tests. After pH adjustment, the mixture (L2 solution and NH4OH) was preheated until 60 °C and then CO2 was injected. When this procedure was performed, immediate carbonates formation and precipitation were observed, as well as a decrease of solution pH, by two and/or three units during the first 30min of carbonation. A stable pH was obtained between 7 and 9, during a total reaction time equivalent to 60min. So, Table 3 shows a carbonation efficiency of 84.8 and 76.4% for both T2E and T2 tests, respectively. In pH-swing mineral carbonation, nesquehonite (MgCO3.3H2O) and hydromagnesite (Mg5(CO3)4(OH)2.4H2O) are frequently sintered as car- bonates [17,37,46]. According to Wilson et al. [46], to capture of CO2, the best sintered carbonates must have a molar ratio of MgO to CO2 equal to 1:1 (i.e. magnesite (MgCO3), nesquehonite (MgCO3.3H2O) among others), however, their environmental stability is low. On the other hand, al- though the carbonates with molar ratio equal to 5:4 (i.e. hydromagnesite (Mg5(CO3)4(OH)2.4H2O), dypingite ((Mg5(CO3)4(OH)2.5H2O)) do not capture as much CO2 as carbonates with molar ratio 1:1, their environ- mental stability is better [46]. Therefore, carbonates precipitation depend on both the environmental stability and the CO2 capture efficiency. In this study can be seen that the carbonates obtained in T2E and T2 experimental tests (S4) have a MgO content of 43.7% and 40.5%, re- spectively (Table 5). Figs. 4c, d, 5c and d show X-ray diffractograms and TG/DTG curves for carbonates obtained after the carbonation stage (S4). Narrow peaks were observed on S4 precipitate diffractograms (Fig. 4c and d) which reflect a high crystallinity of carbonates. Miner- alogical analysis of S4 precipitates obtained from T2E test (Fig. 4c) indicated that the main phase is hydromagnesite, however, there are traces of dypingite. According to Wilson et al. [46]., in direct mineral carbonation using mine waste, the hydrated carbonates with MgO:CO2 molar ratio of 5:4 are formed from hydrated carbonates with MgO:CO2 molar ratio of 1:1. Hence, the nesquehonite (MgCO3.3H2O) decomposition can produce dypingite (Mg5(CO3)4(OH)2.5H2O) and hydromagnesite (Mg5(CO3)4(OH)2.4H2O) [46]. Additionally Glasser et al. [8] indicated that the increase of dypingite and hydromagnesite from nesquehonite depends on the quantities of H2O and CO2 in the reaction. Studies related to pH-swing mineral carbonation indicated that dypingite can be produced when initial temperature of carbonation is about 60 °C [27]. In this study was produces dypingite, due to carbonates precipita- tion started at 60 °C [27]. On the other hand, the S4 precipitate dif- fractogram obtained from T2 test (Fig. 5d) has a higher dypingite content, probably due to excess H2O and CO2 in the reaction medium [8,47]. When TG/DTG curves from Fig. 5c and d are observed, it can be noticed that the carbonate from T2E test has a higher purity than the carbonate from T2 test, thus confirming the XRD patterns (Fig. 4c and d). From the carbonates mass loss (S4) were observed on TG curves, it can be noticed that the carbonate obtained in T2E test presents ap- proximately 36.7% of CO2 fixed as CO3 2−. However, the carbonate obtained in T2 presents 44.1% of CO2 fixed. Thus, the T2E condition was able to form carbonates with a CO2 capture efficiency of 81%, which is twice more captured CO2 when compared to T2 test condition (Table 3). Since the best CO2 capture efficiency was obtained for the T2E test, a mass balance was done for the whole process. Results are presented on Table 6. The calculation basis employed for this study was the capture of 1 ton of CO2. Calculations indicate that 4 ton of S-GO mineral waste will be employed for each ton of captured CO2, as well as 2.6 ton of HCl, and 4.5 ton of NH4OH. However, 1.7 ton of SiO2, 0.55 ton of iron oxide, and 2.7 ton of hydromagnesite could be produced. It must be mentioned that this process could be optimized to reduce con- sumption of chemical reagents. On the other hand, recovery of chemical reagent (HCl and NH4OH) could be realized by Zhang’s methodology [28]. However, it must be mentioned that, in this experimental study, after the carbonation pro- cess, the L3 solution was cooled to room temperature (25 °C). In this temperature was observed the formation of crystals in the L3 aqueous solution, indicating NH4Cl supersaturation. These L3 solutions were characterized by TGA. Fig. 5e and f show L3 solutions thermal de- composition for experimental tests T2E and T2, respectively. Total de- composition of the L3 aqueous solution was observed up to 250 °C. The first thermal decomposition event from the DTG curve refers to water evaporation. In this region, a mass loss is observed around 80% and 86%, for T2E and T2 tests, respectively. After water elimination, the DTG curve presents three peaks due to salt decomposition, indicating different events, which can be related to decomposition of NH4Cl in NH3 and HCl. Although this study demonstrated the feasibility to produce silica, iron oxide and carbonates with high purity in atmospheric pressure and moderate temperature using pH-swing mineral carbonation based on HCl/NH4OH system, there is evidently a need for further investigations about the recovery of chemical reagent (i.e. HCl and NH4OH) aiming a low energy consumption. This type of HCl/NH4OH system could make the pH-swing mineral carbonation process more efficient, since it can employ higher HCl concentrations. An increase of HCl concentration intensifies the ex- traction of reactive elements, generating a silica with high degree of purity [26]. Increasing acid concentration can also reduce water con- sumption, which is a critical point for the pH-swing mineral carbona- tion process, due to environmental limitations [13]. Table 5 Mass of metals and efficiencies obtained in several stages of the pH-swing mineral car- bonation process using HCl/NH4OH. Stage Test Solid Products Metal Concentration (%) Name mp (g) MgO Fe2O3t SiO2 S-GO T2E S-GO 7.4 43.33 12.61 40.64 T2 S-GO 3.7 43.33 12.61 40.64 Leaching T2E S1 2.8 3.6 2.3 90.4 T2 S1 1.5 2.8 2.0 91.5 Purification T2E S2 0.73 2.4 91.2 1.3 S3 0.18 0.6 90.7 0.5 T2 S2 0.51 4.4 89.5 1.6 S3 0.01 1.2 88.3 0.3 Carbonation T2E S4 4.53 43.7 0.013 0.5 T2 S4 2.1 40.5 0.007 0.1 Table 6 NH4OH added into aqueous solution for precipitation of the products. CO2SEQ ton S-GO ton HCl ton NH4OH ton SiO2 ton Fe ton Hydro-Mg ton CO2CAP % 1 4.0 2.6 4.5 1.7 0.55 2.7 81 G.L.A. Arce Ferrufino et al. Journal of CO₂ Utilization 24 (2018) 164–173 171 4. Conclusion This study evaluated the feasibility to produce silica, iron oxide and carbonates with high purity using one HCl/NH4OH system for pH-swing mineral carbonation, at atmospheric pressure and moderate tempera- tures (90 °C). Besides this, the possibility to recover chemical reagents was explored through NH4Cl thermal decomposition, which is formed after the carbonation stage. Results indicated that the S-GO mining waste was appropriate to be employed in pH-swing mineral carbonation, even though it presented lizardite 1T as the main phase. The Fe3+ content in S-GO acted as a catalyst, aiding the leaching of other reactive metals, since it generated a strong acidic medium during the reaction process. This is due to FeCl3 hydrolysis in aqueous solutions, causing an intrinsic HCl regeneration. High Mg and Fe extraction efficiency was achieved in the first process stage (95 ± 2%), even in stoichiometric conditions. The solid residue obtained after the leaching stage (S1) was mainly silica (SiO2) with 90% purity content. Such silica purity can be improved if HCl concentrations higher to 2mol L−1 were employed. Use of stoichiometric conditions (T2E) during the purification stage reduces the volume of NH4OH solution employed, by approximately three folds, when compared to acid excess conditions (T2). Efficiency of the purification stage was high, with iron removal of around 91.7 ± 1.9%. However, a loss of Mg was detected during this step, of approximately 13.6 ± 2.3%. Hydromagnesite was formed during the carbonation stage, which had a 80.6 ± 3.5% efficiency. 81% of the total CO2 injected was fixed by hydromagnesite inside the reactor, through crystallographic capture. This study showed a good yield of silica, iron oxides, and carbonates production. Besides this, the recovery study for chemical reagents in- dicated that NH4Cl crystallization temperature would favor its re- generation into NH3 and HCl, which can be conducted at temperatures below 250 °C. Acknowledgements The authors are thankful to FAPESP for post-doctorate project number 2013/21244-5 and as well as CNPq for project number 150868/2017-0, which supported this study. In addition, they ac- knowledge the company that provided study materials: Mineração Associada (SAMA S.A.). References [1] G. Gadikota, C. Natali, C. Boschi, A.-H.A. Park, Morphological changes during en- hanced carbonation of asbestos containing material and its comparison to magne- sium silicate minerals, J. Hazard. Mater. 264 (2014) 42–52, http://dx.doi.org/10. 1016/j.jhazmat.2013.09.068. [2] G.L.A.F. Arce, T.G. Soares Neto, I. Ávila, C.M.R. Luna, J.A. Carvalho Jr., Leaching optimization of mining wastes with lizardite and brucite contents for use in indirect mineral carbonation through the pH swing method, J. Clean. Prod. 141 (2017) 1324–1336, http://dx.doi.org/10.1016/j.jclepro.2016.09.204. [3] A. Bloise, T. Critelli, M. Catalano, C. Apollaro, D. Miriello, A. Croce, et al., Asbestos and other fibrous minerals contained in the serpentinites of the Gimigliano-Mount Reventino Unit (Calabria, S-Italy), Environ. Earth Sci. 71 (2014) 3773–3786, http://dx.doi.org/10.1007/s12665-013-3035-2. [4] J. Li, Q. Dong, K. Yu, L. Liu, Asbestos and asbestos waste management in the Asian- Pacific region: trends, challenges and solutions, J. Clean. Prod. 81 (2014), http:// dx.doi.org/10.1016/j.jclepro.2014.06.022. [5] A. Valouma, A. Verganelaki, P. Maravelaki-kalaitzaki, E. Gidarakos, Chrysotile as- bestos detoxification with a combined treatment of oxalic acid and silicates pro- ducing amorphous silica and biomaterial, J. Hazard. Mater. 305 (2016) 164–170. [6] A. Fedoročková, P. Raschman, G. Sučik, B. Plešingerová, Ľ. Popovič, J. Briančin, Processing of serpentinite tailings to pure amorphous silica, Ceramics 59 (2015) 275–282. [7] S.N. Nam, S. Jeong, H. Lim, Thermochemical destruction of asbestos-containing roofing slate and the feasibility of using recycled waste sulfuric acid, J. Hazard. Mater. 265 (2014) 151–157, http://dx.doi.org/10.1016/j.jhazmat.2013.11.004. [8] F.P. Glasser, G. Jauffret, J. Morrison, J.-L. Galvez-Martos, N. Patterson, M.S.- E. Imbabi, Sequestering CO2 by mineralization into useful nesquehonite-based products, Front. Energy Res. 4 (2016) 1–7, http://dx.doi.org/10.3389/fenrg.2016. 00003. [9] A. Hemmati, J. Shayegan, J. Bu, T.Y. Yeo, P. Sharratt, Process optimization for mineral carbonation in aqueous phase, Int. J. Min. Process. 130 (2014) 20–27, http://dx.doi.org/10.1016/j.minpro.2014.05.007. [10] A. Sanna, M. Uibu, G. Caramanna, R. Kuusik, M.M. Maroto-Valer, A review of mineral carbonation technologies to sequester CO2, Chem. Soc. Rev. 43 (2014) 8049–8080, http://dx.doi.org/10.1039/C4CS00035H. [11] J. Sipilä, S. Teir, R. Zevenhoven, Carbon Dioxide Sequestration by Mineral Carbonation - Literature Review Update 2005–2007, Espoo, Finland, (2008) (ac- cessed September 16, 2013), http://users.abo.fi/rzevenho/ MineralCarbonationLiteratureReview05-07.pdf. [12] X. Wang, M.M. Maroto-Valer, Dissolution of serpentine using recyclable ammonium salts for CO2 mineral carbonation, Fuel 90 (2011) 1229–1237, http://dx.doi.org/ 10.1016/j.fuel.2010.10.040. [13] Y.T. Yuen, P.N. Sharratt, B. Jie, Carbon dioxide mineralization process design and evaluation: concepts, case studies, and considerations, Environ. Sci. Pollut. Res. 23 (2016) 22309–22330, http://dx.doi.org/10.1007/s11356-016-6512-9. [14] P. Córdoba, Q. Liu, S. Garcia, M. Maroto-Valer, Understanding the importance of iron speciation in oil-field brine pH for CO2 mineral sequestration, J. CO2 Util. 16 (2016) 78–85, http://dx.doi.org/10.1016/j.jcou.2016.06.004. [15] J. McCutcheon, G.M. Dipple, S.A. Wilson, G. Southam, Production of magnesium- rich solutions by acid leaching of chrysotile: a precursor to field-scale deployment of microbially enabled carbonate mineral precipitation, Chem. Geol. 413 (2015) 119–131, http://dx.doi.org/10.1016/j.chemgeo.2015.08.023. [16] S. Teir, R. Kuusik, C.-J. Fogelholm, R. Zevenhoven, Production of magnesium car- bonates from serpentinite for long-term storage of CO2, Int. J. Min. Process. 85 (2007) 1–15, http://dx.doi.org/10.1016/j.minpro.2007.08.007. [17] S. Teir, S. Eloneva, C. Fogelholm, R. Zevenhoven, Fixation of carbon dioxide by producing hydromagnesite from serpentinite, Appl. Energy 86 (2009) 214–218, http://dx.doi.org/10.1016/j.apenergy.2008.03.013. [18] A. Sanna, J. Gaubert, M.M. Maroto-Valer, Alternative regeneration of chemicals employed in mineral carbonation towards technology cost reduction, Chem. Eng. J. 306 (2016) 1049–1057, http://dx.doi.org/10.1016/j.cej.2016.08.039. [19] A. Azdarpour, M. Asadullah, E. Mohammadian, H. Hamidi, R. Junin, M.A. Karaei, A review on carbon dioxide mineral carbonation through pH-swing process, Chem. Eng. J. 279 (2015) 615–630, http://dx.doi.org/10.1016/j.cej.2015.05.064. [20] S. Teir, H. Revitzer, S. Eloneva, C.-J. Fogelholm, R. Zevenhoven, Dissolution of natural serpentinite in mineral and organic acids, Int. J. Min. Process. 83 (2007) 36–46, http://dx.doi.org/10.1016/j.minpro.2007.04.001. [21] J. Bu, P. Bai, P. Sharratt, Carbon Dioxide Capture With Regeneration of Salt: International Publication Number WO 2012/050530 A1, (2012). [22] A. Sanna, X. Wang, A. Lacinska, M. Styles, T. Paulson, M.M. Maroto-Valer, Enhancing Mg extraction from lizardite-rich serpentine for CO2 mineral seques- tration, Min. Eng. 49 (2013) 135–144, http://dx.doi.org/10.1016/j.mineng.2013. 05.018. [23] G. Gadikota, K.J. Fricker, S.-H. Jang, A.-H.A. Park, Carbonation of silicate minerals and industrial wastes and their potential use as sustainable construction materials, Carbon Dioxide Util Closing Carbon Cycle, 1st ed., (2015), pp. 115–137, http://dx. doi.org/10.1016/B978-0-444-62746-9.00008-6 American C, Washington DC. [24] G. Gadikota, E.J. Swanson, H. Zhao, A.-H.A. Park, Experimental design and data analysis for accurate estimation of reaction kinetics and conversion for carbon mineralization, Ind Eng. Chem. Res. 53 (2014) 6664–6676, http://dx.doi.org/10. 1021/ie500393h. [25] G. Gadikota, J. Matter, P. Kelemen, A.-H.A. Park, Chemical and morphological changes during olivine carbonation for CO2 storage in the presence of NaCl and NaHCO3, Phys. Chem. Chem. Phys. 16 (2014) 4679–4693, http://dx.doi.org/10. 1039/c3cp54903h. [26] G. Gadikota, A.A. Park, Chapter 8 – accelerated carbonation of Ca- and Mg-bearing minerals and industrial wastes using CO2, in: P. Styring, Q. Elsje Alessandra, K. Armstrong (Eds.), Carbon Dioxide Util Closing Carbon Cycle, 1st ed., Elsevier, Amsterdam, 2015, pp. 115–130. [27] A. Hemmati, J. Shayegan, P. Sharratt, T.Y. Yeo, J. Bu, Solid products character- ization in a multi-step mineralization process, Chem. Eng. J. 252 (2014) 210–219, http://dx.doi.org/10.1016/j.cej.2014.04.112. [28] J. Zhang, R. Zhang, H. Geerlings, J. Bi, Mg-silicate carbonation based on an HCl- and NH 3-recyclable process: effect of carbonation temperature, Chem. Eng. Technol. 35 (2012) 525–531, http://dx.doi.org/10.1002/ceat.201100425. [29] A. Sanna, M. Dri, M.M. Maroto-Valer, Carbon dioxide capture and storage by pH swing aqueous mineralisation using a mixture of ammonium salts and antigorite source, Fuel. 114 (2013) 153–161, http://dx.doi.org/10.1016/j.fuel.2012.08.014. [30] X. Wang, M.M. Maroto-Valer, Integration of CO2 capture and mineral carbonation by using recyclable ammonium salts, ChemSusChem 4 (2011) 1291–1300, http:// dx.doi.org/10.1002/cssc.201000441. [31] G.L.A.F. Arce, T.G.S. Neto, I. Ávila, C.M.R. Luna, J.C. dos Santos, J.A. Carvalho, Influence of physicochemical properties of Brazilian serpentinites on the leaching process for indirect CO2 mineral carbonation, Hydrometallurgy 169 (2017) 142–151, http://dx.doi.org/10.1016/j.hydromet.2017.01.003. [32] B.W. Evans, Control of the products of serpentinization by the Fe2+Mg-1 exchange potential of olivine and orthopyroxene, J. Petrol. 49 (2008) 1873–1887, http://dx. doi.org/10.1093/petrology/egn050. [33] P.B. Hostetler, R.G. Coleman, F.A. Mumpton, Brucite in alpine Serpentinites, Am. Mineral. 51 (1966) 74–98, http://dx.doi.org/10.1163/_q3_SIM_00374. [34] F.J. Wicks, E.J.W. Whittaker, A reappraisal of the structures of the serpentine mi- nerals, Can. Mineral. 13 (1975) 227–243. [35] R.B. Frost, J.S. Beard, On silica activity and serpentinization, J. Petrol. 48 (2007) 1351–1368, http://dx.doi.org/10.1093/petrology/egm021. [36] C. Viti, M. Mellini, Contrasting chemical compositions in associated lizardite and G.L.A. Arce Ferrufino et al. Journal of CO₂ Utilization 24 (2018) 164–173 172 http://dx.doi.org/10.1016/j.jhazmat.2013.09.068 http://dx.doi.org/10.1016/j.jhazmat.2013.09.068 http://dx.doi.org/10.1016/j.jclepro.2016.09.204 http://dx.doi.org/10.1007/s12665-013-3035-2 http://dx.doi.org/10.1016/j.jclepro.2014.06.022 http://dx.doi.org/10.1016/j.jclepro.2014.06.022 http://refhub.elsevier.com/S2212-9820(17)30272-X/sbref0025 http://refhub.elsevier.com/S2212-9820(17)30272-X/sbref0025 http://refhub.elsevier.com/S2212-9820(17)30272-X/sbref0025 http://refhub.elsevier.com/S2212-9820(17)30272-X/sbref0030 http://refhub.elsevier.com/S2212-9820(17)30272-X/sbref0030 http://refhub.elsevier.com/S2212-9820(17)30272-X/sbref0030 http://dx.doi.org/10.1016/j.jhazmat.2013.11.004 http://dx.doi.org/10.3389/fenrg.2016.00003 http://dx.doi.org/10.3389/fenrg.2016.00003 http://dx.doi.org/10.1016/j.minpro.2014.05.007 http://dx.doi.org/10.1039/C4CS00035H http://users.abo.fi/rzevenho/MineralCarbonationLiteratureReview05-07.pdf http://users.abo.fi/rzevenho/MineralCarbonationLiteratureReview05-07.pdf http://dx.doi.org/10.1016/j.fuel.2010.10.040 http://dx.doi.org/10.1016/j.fuel.2010.10.040 http://dx.doi.org/10.1007/s11356-016-6512-9 http://dx.doi.org/10.1016/j.jcou.2016.06.004 http://dx.doi.org/10.1016/j.chemgeo.2015.08.023 http://dx.doi.org/10.1016/j.minpro.2007.08.007 http://dx.doi.org/10.1016/j.apenergy.2008.03.013 http://dx.doi.org/10.1016/j.cej.2016.08.039 http://dx.doi.org/10.1016/j.cej.2015.05.064 http://dx.doi.org/10.1016/j.minpro.2007.04.001 http://refhub.elsevier.com/S2212-9820(17)30272-X/sbref0105 http://refhub.elsevier.com/S2212-9820(17)30272-X/sbref0105 http://dx.doi.org/10.1016/j.mineng.2013.05.018 http://dx.doi.org/10.1016/j.mineng.2013.05.018 http://dx.doi.org/10.1016/B978-0-444-62746-9.00008-6 http://dx.doi.org/10.1016/B978-0-444-62746-9.00008-6 http://dx.doi.org/10.1021/ie500393h http://dx.doi.org/10.1021/ie500393h http://dx.doi.org/10.1039/c3cp54903h http://dx.doi.org/10.1039/c3cp54903h http://refhub.elsevier.com/S2212-9820(17)30272-X/sbref0130 http://refhub.elsevier.com/S2212-9820(17)30272-X/sbref0130 http://refhub.elsevier.com/S2212-9820(17)30272-X/sbref0130 http://refhub.elsevier.com/S2212-9820(17)30272-X/sbref0130 http://dx.doi.org/10.1016/j.cej.2014.04.112 http://dx.doi.org/10.1002/ceat.201100425 http://dx.doi.org/10.1016/j.fuel.2012.08.014 http://dx.doi.org/10.1002/cssc.201000441 http://dx.doi.org/10.1002/cssc.201000441 http://dx.doi.org/10.1016/j.hydromet.2017.01.003 http://dx.doi.org/10.1093/petrology/egn050 http://dx.doi.org/10.1093/petrology/egn050 http://dx.doi.org/10.1163/_q3_SIM_00374 http://refhub.elsevier.com/S2212-9820(17)30272-X/sbref0170 http://refhub.elsevier.com/S2212-9820(17)30272-X/sbref0170 http://dx.doi.org/10.1093/petrology/egm021 chrysotile in veins from Elba, Italy, Eur. J. Mineral. 9 (1997) 585–596 http:// eurjmin.geoscienceworld.org/content/9/3/585.short. [37] A. Sanna, A. Lacinska, M. Styles, M.M. Maroto-Valer, Silicate rock dissolution by ammonium bisulphate for pH swing mineral CO2 sequestration, Fuel Process. Technol. 120 (2014) 128–135, http://dx.doi.org/10.1016/j.fuproc.2013.12.012. [38] H. Chamley, Clay Sedimentology, Springer, New York, 1989, http://dx.doi.org/10. 1007/s13398-014-0173-7.2. [39] A. Sanna, X. Wang, A. Lacinska, M. Styles, T. Paulson, M.M. Maroto-Valer, Enhancing Mg extraction from lizardite-rich serpentine for CO2 mineral seques- tration, Min. Eng. 49 (2013) 135–144, http://dx.doi.org/10.1016/j.mineng.2013. 05.018. [40] A.M. Lacinska, M.T. Styles, K. Bateman, D. Wagner, M.R. Hall, C. Gowing, et al., Acid-dissolution of antigorite, chrysotile and lizardite for ex situ carbon capture and storage by mineralisation, Chem. Geol. 437 (2016) 153–169, http://dx.doi.org/10. 1016/j.chemgeo.2016.05.015. [41] C. Viti, Serpentine minerals discrimination by thermal analysis, Am. Mineral. 95 (2010) 631–638, http://dx.doi.org/10.2138/am.2010.3366. [42] G.P. Assima, F. Larachi, J. Molson, G. Beaudoin, Comparative study of five Québec ultramafic mining residues for use in direct ambient carbon dioxide mineral sequestration, Chem. Eng. J. 245 (2014) 56–64, http://dx.doi.org/10.1016/j.cej. 2014.02.010. [43] I. Muthuvel, S. Dineshkumar, K. Thirumurthy, S. Rajasri, G. Thirunarayanan, A New Solid Acid Catalyst FeCl 3 / Bentonite for Aldol Condensation Under Solvent-Free Condition 55 (2016), pp. 252–260. [44] Z. Wang, X. Sun, J. Wu, FeCl3: An efficient catalyst for reactions of electron-rich arenes with imines or aziridines, Tetrahedron 64 (2008) 5013–5018, http://dx.doi. org/10.1016/j.tet.2008.03.081. [45] H.C. Oskierski, B.Z. Dlugogorski, G. Jacobsen, Sequestration of atmospheric CO2 in a weathering-derived, serpentinite-hosted magnesite deposit: 14C tracing of carbon sources and age constraints for a refined genetic model, Geochim. Cosmochim. Acta 122 (2013) 226–246, http://dx.doi.org/10.1016/j.gca.2013.08.029. [46] S.A. Wilson, G.M. Dipple, I.M. Power, J.M. Thom, R.G. Anderson, M. Raudsepp, et al., Carbon dioxide fixation within mine wastes of ultramafic-hosted ore deposits: examples from the Clinton Creek and Cassiar Chrysotile deposits, Can. Econ. Geol. 104 (2009) 95–112, http://dx.doi.org/10.2113/gsecongeo.104.1.95. [47] I.M. Power, S.A. Wilson, G.M. Dipple, Serpentinite carbonation for CO2 seques- tration, Elements 9 (2013) 115–121, http://dx.doi.org/10.2113/gselements.9.2. 115. G.L.A. Arce Ferrufino et al. Journal of CO₂ Utilization 24 (2018) 164–173 173 http://eurjmin.geoscienceworld.org/content/9/3/585.short http://eurjmin.geoscienceworld.org/content/9/3/585.short http://dx.doi.org/10.1016/j.fuproc.2013.12.012 http://dx.doi.org/10.1007/s13398-014-0173-7.2 http://dx.doi.org/10.1007/s13398-014-0173-7.2 http://dx.doi.org/10.1016/j.mineng.2013.05.018 http://dx.doi.org/10.1016/j.mineng.2013.05.018 http://dx.doi.org/10.1016/j.chemgeo.2016.05.015 http://dx.doi.org/10.1016/j.chemgeo.2016.05.015 http://dx.doi.org/10.2138/am.2010.3366 http://dx.doi.org/10.1016/j.cej.2014.02.010 http://dx.doi.org/10.1016/j.cej.2014.02.010 http://refhub.elsevier.com/S2212-9820(17)30272-X/sbref0215 http://refhub.elsevier.com/S2212-9820(17)30272-X/sbref0215 http://refhub.elsevier.com/S2212-9820(17)30272-X/sbref0215 http://dx.doi.org/10.1016/j.tet.2008.03.081 http://dx.doi.org/10.1016/j.tet.2008.03.081 http://dx.doi.org/10.1016/j.gca.2013.08.029 http://dx.doi.org/10.2113/gsecongeo.104.1.95 http://dx.doi.org/10.2113/gselements.9.2.115 http://dx.doi.org/10.2113/gselements.9.2.115 CO2 sequestration by pH-swing mineral carbonation based on HCl/NH4OH system using iron-rich lizardite 1T Introduction Materials and methods Materials Characterization methods Process description Leaching experiments Impurities separation Formation of carbonates Results and discussion Materials characterization Leaching experiments Impurities separation Carbonates precipitation Conclusion Acknowledgements References