• About US +
• For Contributors +
• Journal Search +
Journal Search Engine
ISSN : 1225-309X(Print)
ISSN : 2288-7172(Online)
Journal of the mineralogical society of korea Vol.30 No.4 pp.205-217
DOI : https://doi.org/10.9727/jmsk.2017.30.4.205

# Mineral Carbonation of Serpentinite: Extraction, pH swing, and Carbonation

Seung-Woo LEE1, Hyein Won1, Byoung-Young Choi2, Soochun Chae1, Jun-Hwan Bang1, Kwon Gyu Park2*
1Center for carbon mineralization, Korea Institute of Geosciences and Mineral Resources
2Center for CO2 geological storage, Korea Institute of Geosciences and Mineral Resources
Corresponding author : +82-42-868-3250, kgpark@kigam.re.kr
20171216 20171218 20171228

## Abstract

Mineral carbonation by indirect method has been studied by serpentinite as cation source. Through the carbonation of CO2 and alkaline earth ions (calcium and magnesium) from serpentinite, the pure carbonates including MgCO3 and CaCO3 were synthesized. The extraction solvent used to extract magnesium (Mg) was ammonium sulfate ((NH4)2SO4), and the investigated experimental factors were the concentration of (NH4)2SO4, reaction temperature, and ratio of serpentinite to the extraction solvent. From this study, the Mg extraction efficiency of approximately 80 wt% was obtained under the conditions of 2 M (NH4)2SO4, 300°C, and a ratio of 5 g of serpentinite/75 mL of extraction solvent. The Mg extraction efficiency was proportional to the concentration and reaction temperature. NH3 produced from the Mg extraction of serpentinite was used as a pH swing agent for carbonation to increase the pH value. About 1.78 M of NH3 as the form of NH4+ was recovered after Mg extraction from serpentinite. And, the main step in Mg extraction process of serpentinite was estimated by geochemical modeling.

# 사문암(Serpentinite)을 이용한 광물탄산화: Mg 추출과 pH swing 및 탄산화

이 승우1, 원 혜인1, 최 병영2, 채 수천1, 방 준환1, 박 권규2*
1탄소광물화사업단 한국지질자원연구원
2지중저장연구단 한국지질자원연구원

## 초록

간접 탄산화(indirect method) 및 양이온 공급원으로 사문암(serpentinite)을 이용하여 광물탄산화 연구를 수행하였다. 이산화탄소와 사문암 내 알칼리 토금속(칼슘과 마그네슘)의 탄산화 반응을 통해 고 순도의 탄산칼슘과 탄산마그네슘을 합성할 수 있었다. 마그네슘 추출을 위해 황산암모늄을 사용하였고 Mg 추출률 향상을 위해 황산암모늄 농도, 반응온도 및 사문암과 추출 용매의 비(고액비) 등 여러 반응 변수를 검토하였다. 본 연구로부터 2 M 황산암모늄을 사용하여 300°C 반응온도에서 고액비(5 g/66 mL) 실험을 진행한 경우 약 80 wt% 이상의 Mg를 얻을 수 있었다. Mg 추출률은 추출 용매 농도 및 반 응온도와 비례하여 증가하였다. 사문암의 Mg 추출 과정에서 얻어진 암모니아(NH3)는 회수하여 탄산화 과정에서 필요한 pH 복원제(pH swing agent)로 활용하였다. 본 연구를 통해 약 1.78 M 암모니아를 회 수할 수 있었고 지구화학 모델링을 통해 사문암의 Mg 추출 과정의 핵심 단계를 해석하고자 하였다.

Ministry of Science, ICT and Future Planning

## Introduction

Serpentinites are rocks consisting mostly of serpentine group minerals. Minerals in this group are formed by serpentinization, a hydration and metamorphic transformation by Earth surface conditions and hydrothermal temperatures (Evans et al., 2013). Serpentine group minerals have the approximate formula, (R62+) Si4O10 (R: Mg2+, Fe2+, Mn2+, or Ni2+), and mainly consists of antigorite, lizardite, and chrysotile (Bailey, 1988).

During the last two decades the interest in how to keep the captured CO2 permanently away from the atmosphere has increased, especially at the point of CCS (Carbon Capture & Storage) and mineral carbonation (Carneiro et al., 2011; Zhang and Bachu, 2011; Wolterbeek et al., 2013). Mineral carbonation is a technology to reduce CO2 on the basis of alkaline metals in natural minerals or industrial wastes. It can proceed through direct method (carbonation) and indirect method (extraction and carbonation) (Azdarpour, 2015).

Research investigating CCS has been demonstrated in the USA, England, and Australia, countries which have a commercialized gas or oil field. On the other hand, research investigating mineral carbonation could be profitable for the countries that do not have enough sites for CO2 storage but produce Ca- or Mg-rich rock through mining. The advantages of CCS are the availability of a large amount of CO2 storage and a relatively lower processing cost compared to mineral carbonation. The merit of mineral carbonation is CO2 storage ability in a thermodynamically stable form, carbonate minerals, and the minimization of future costs associated with the monitoring of the risk of CO2 leakage. Of course, there are disadvantages of mineral carbonation. Especially, the weakest point of mineral carbonation (indirect method) is that the supply cost of extraction solvent is high at the viewpoint of economic feasibility (Azarpour et al., 2015; Lee et al., 2016). Thus, the development of technology that can reduce the processing cost is so important.

At the point of mineral carbonation, Mg-silicate rock including serpentinite has the potential to be used as materials for carbonation because of its higher Mg content and relatively lower reaction temperature compared to other minerals (Béarat et al., 2002; Lin et al., 2008; Wang and Maroto-Valer, 2011). However, natural minerals including serpentinite need to be mined and pulverized. Thus, the total cost of mineral carbonation using natural minerals or rocks would be increased due to material pretreatment. To create a cost-effective process for mineral carbonation, there are several prerequisites. First, the maximum extraction efficiency for alkaline earth metals is an essential. For example, the extraction of magnesium from Mg-silicate rocks during mineral carbonation can determine the amount of stored CO2 as carbonate minerals. The second requirement is the recovery and reuse of the extraction solvent for a cost-effective mineral carbonation process. Finally, there is the development of high value-added product. Researchers investigating mineral carbonation have to consider how to apply the carbonate materials such as MgCO3 and CaCO3 after CO2 fixation to relevant industrial fields such as paper, paint, building structures, fire proofing, fire extinguishing compositions, and so on. There are many studies on mineral carbonation using various methods worldwide. However, the most important studies are those that propose a cost-effective process for mineral carbonation. The process will unify the entire CO2 market related to mineral carbonation.

In the study, we have tried to propose a model of a cost-effective process using serpentinite and (NH4)2SO4. The optimum experimental factors including the concentration of ammonium sulfate, ratio of serpentinite to extraction solvent, and reaction temperature were identified for the maximum Mg extraction efficiency at 300℃ of reaction temperature, the recovery and the reuse of NH3 as a pH swing agent for carbonation.

## Materials and Method

### Materials

The serpentinite used in the experiment was obtained from Andong serpentinite mine (South Korea), and was pulverized. The serpentinite pulverized below 75 μm was used for the Mg extraction. The elemental analysis of serpentinite was carried out by inductively coupled plasma optical emission spectroscopy (ICP-OES, Optima 5300 DV, Perkin Elmer).

### Quantitative Analysis using X-ray diffraction pattern

Rietveld analysis was employed in the quantitative phase analysis of the serpentinite. All calculations were performed with Diffrac. TOPAS (ver. 4.2.0.2, Bruker AXS GmbH). The XRD pattern of the serpentinite was acquired by a high resolution X-ray diffractometer (D8 Discover, Bruker AXS GmbH). Diffraction patterns of sufficient resolution for Rietveld analysis were obtained under the following conditions: step interval, 0.05° between 5 and 120° (2 theta); and wavelength of 1.893 Å. The refinement was carried out until the agreement factor (good of fitness (GOF)) was below 1.5.

### Magnesium extraction

To increase the Mg extraction efficiency, several experimental factors were investigated including the concentration of the extraction solvent (NH4)2SO4, reaction temperature, and ratio of serpentinite (g) to extraction solvent (mL). (NH4)2SO4 was used as the extraction solvent to extract magnesium from serpentinite. The concentration range of (NH4)2SO4 was from 0.5 M to 2 M. And the range of reaction temperature was from 200℃ to 300℃. The interval of the reaction temperatures was set at 50℃. The ratios of serpentinite (g) to extraction solvent (ml) investigated were 5 g/100 mL, 5 g/75 mL, and 5 g/50 mL, and N2 gas was injected into the reactor until a pressure of 5 bar to check the sealing of the reactor. The degree of Mg extraction was calculated by equation (1).

$η M g ( % ) = C M g ( g / m o l ) × V ( l ) M s e r p e n t i n i t e ( g ) × X M g ( g ) × 100$
(1)

CMg is the Mg concentration in solution after the extraction, and it was analyzed by ICP-OES. Mserpentinite is the mass (g) of serpentinite used in the experiment, and XMg is the Mg concentration of serpentinite before extraction. V indicates the volume of solution including (NH4)2SO4 and distilled water.

To extract Mg from serpentinite, the autoclave (C-276) made by Hastelloy, containing a nickel-molybdenum-chromium alloy was used (III in Fig. 1C). The holding time at each reaction temperature is 10 seconds, and the power of the reactor was cut off after 10 seconds. After the reactor was stopped and cooled down, the produced gas including N2 and NH3 was introduced into 75 mL of distilled water for the recovery and reuse of NH3. The solid and liquid in the extraction solvent after the experiment were separated by filtration.

### Recovery of NH3 for reuse as a pH swing agent

The reuse of the extraction solvent is an important technique to obtain a cost-effective process for mineral carbonation. The (NH4)2SO4 used as the extraction solvent could be decomposed at approximately 300℃. As shown in reaction (2), the serpentinite reacted with ammonium sulfate reacted, and then 6 M NH3 can be produced from 3 M (NH4)2SO4 above 250℃.

(2)

After Mg extraction from serpentinite, the solid (serpentinite residue) and liquid (solution) were separated by filtration. The pH value of the extraction solvent (NH4)2SO4 before and after Mg extraction was analyzed by a pH meter (Orion 3 Star, Thermo Scientific). The concentration of ammonium ion existing in the extraction solvent after Mg extraction was analyzed by ion chromatography (883 Basic IC Plus, Metrohm). The NH4+ rich solution was used as a pH swing agent for the carbonation between CO2 and the Mg-rich solution.

### Characterization after Mg extraction and carbonation

After Mg extraction using (NH4)2SO4, the content of magnesium in the extraction solvent was analyzed by inductively coupled plasma optical emission spectroscopy (ICP-OES, Optima 5300 DV, Perkin Elmer). The morphologies of the synthesized carbonation minerals (MgCO3 and CaCO3) were analyzed by a field emission scanning electron microscope (FE-SEM, MERIN, Carl Zeiss), and the surface elements of the carbonate minerals were analyzed by an energy dispersive spectrometer (EDS, Xflash 6, Bruker).

### Geochemical reaction path model

To trace the geochemical reactions during the experiments, reaction path modeling was performed using PHREEQC v. 3 (Parkhurst and Appelo, 2017) for the case showing the highest extraction efficiency of Mg: 2M (NH4)2SO4, 5 g/75 mL. Thermodynamic database used for the water-rock interactions were THERMODDEM.DAT (Blanc et al., 2012) because it has the thermodynamic data for all minerals identified in this study. Instead, ammonia redox state was decoupled from nitrogen gas to inhibit the complete oxidation of ammonium into nitrogen gas. In this study, the dissolution/precipitation of all minerals was simulated considering the dissolution kinetics of minerals and the gradual increase of temperature up to 300℃. The dissolution kinetics was based on the transition state theory (Lasaga, 1998) as the following equation:(3)

$r n = ± k n A n | 1 − ( Q n K n ) θ | η$
(3)

where n is mineral index, r is the kinetic rate for dissolution (positive values) and precipitation (negative values), k is the rate constant (moles per unit mineral surface area and unit time), A is the specific reactive surface area per gram of mineral, Q is the reaction quotient, and K is the equilibrium constant. The parameters θ and η are commonly taken equal to unity. The kinetics of minerals is pH-dependent and is catalyzed by H+, H2O, and OH-. The overall reaction rate is the sum of the rates of the individual reaction mechanism and is calculated as follows (Palandri and Kharaka, 2004):(4)

$k = k 25 n u exp [ − E a n u R ( 1 T − 1 298.15 ) ] + k 25 H exp [ − E a H R ( 1 T − 1 298.15 ) ] α N n H + k 25 O H exp [ − E a O H R ( 1 T − 1 298.15 ) ] a O H n O H$
(4)

where nu, H, and OH represent the neutral, acid and base mechanisms, respectively. E is activation energy, K25 is rate constant at 25℃, R is the gas constant, T is the absolute temperature, a is activity of the species, and n is the power constant. Most kinetic parameters were taken from Palandri and Kharaka (2004) and the rate constants of serpentinite minerals were from Critelli et al. (2015) (Table 1).

## Results and Discussion

### Characterization of serpentinite

As shown in ICP-OES (Table 2), the serpentinite used in the experiment has a higher content of magnesium (MgO : 37.2 wt%) than calcium (CaO : 2.91 wt%). The serpentinite contains 34 wt% of SiO2 and approximately 11 wt% of iron including Fe2O3 and FeO. The first objective in the study was to extract Mg from the serpentinite as effectively as possible.

From the Rietveld refinement (Table 3), the crystal phases of the serpentinite showed that antigorite (41.49%) and lizardite (18.91%) occupied approximately 60% of total crystalline minerals, and approximately 2.77% of the total crystalline minerals was identified as calcite. The CaCO3 is inferred from the result of the weathering of serpentinite.

### Mg extraction

To obtain the highest Mg extraction efficiency from the serpentinite, the concentration of ext- raction solvent, (NH4)2SO4, ratio of serpentinite to extraction solvent, and reaction temperature were investigated (Table 4). As shown in Table 4, the higher the concentration of (NH4)2SO4 and reaction temperature, the greater the Mg extraction efficiency. However, the ratio of serpentinite to extraction solvent was not directly proportional to the Mg extraction efficiency. In the case of 2 M (NH4)2SO4 as the extraction solvent, the Mg extraction efficiency at the ratio 5 g/75 mL was higher than that at the ratio 5 g/ 100 mL (Fig. 2). The best Mg extraction efficiency in the study was 83.0 wt% on average, and the experiment was carried out three times for the identification of reproducibility. The prime experimental factor determining the Mg extraction efficiency was the reaction temperature, which is related to the decomposition of the extraction solvent, (NH4)2SO4. The melting point range of (NH4)2SO4 is 235-280℃. At that point, (NH4)2SO4 can be thermally decomposed and reacted with serpentinite. Thus, a reaction temperature above 300℃ was an important factor to extract higher amounts of Mg from serpentinite compared to temperatures of 200℃ and 250℃. From the measurement of the color of the slurry including serpentinite and (NH4)2SO4 (supplementary information, Fig. S1), a difference in the color was identified before and after Mg extraction. The color of the slurry before extraction was blue celeste. On the other hand, the color of the slurry after the experiment was red corresponding with the increase in the reaction temperature. This effect could be inferred to be due to the dissolution of iron (Fe2O3, 11.4%) in serpentinite (Table 2). As the reaction temperature was increased, the color of the slurry was changed to reddish.

The pH value of the Mg-rich extraction solvent after extraction was measured. The range of the pH values according to the reaction temperature was 10.69-11.04, and the average of pH was 10.88 (supplementary information, Fig. S2). The high pH value resulted from the production of NH3 due to the decomposition of (NH4)2SO4.

At the point of CO2 sequestration, the process related to mineral carbonation needs a high extraction efficiency of calcium and/or magnesium because the amount of earth metal ions determines the mass of CO2 to be stored. Thus, the development of a high extraction method for Mg and/or Ca would be one of prerequisites for the proposal of a cost-effective CO2 utilization process. In this study, we analyzed experimental factors including the concentration of the extraction solvent (NH4)2SO4, ratio of serpentinite to extraction solvent, and reaction temperature to obtain the highest Mg extraction efficiency from the serpentinite. We found the best experimental conditions to be 2 M (NH4)2SO4, a ratio of 5 g/ 75 mL, and 300℃. Using these experimental conditions, around an 83 wt% Mg extraction efficiency was obtained.

### Reuse of the extraction solvent as a pH swing agent

Through the dissolution of NH3 (gas phase) into distilled water during Mg extraction at 300℃ (reaction temperature), a NH4+ rich solution can be prepared (Dissolution part in Fig. 2). From the analysis of ion chromatography (IC) after Mg extraction, the quantity of NH4+ was measured to approximately 30,000 ppm (Table 5). Based on the reaction (equation 2) of serpentine group (Mg3Si2O5(OH)4) and 3(NH4)2SO4 (extraction solvent) equation, the measured quantity of NH4+, 30,000 ppm, was calculated from the amount of NH3 produced from the reaction of serpentinite and (NH4)2SO4.

The amount of NH3 was 1.78 mol, meaning 44% of ammonia was able to be produced through equation (2). Considering the solubility (33% w/w) of NH3 at 25℃ (Oxtoby et al., 2012), a large amount of NH3 would be dissolved in the (NH4)2SO4 solvent after Mg extraction. Previous researchers reported the potential of ammonium salt for mineral carbonation (Park and Fan, 2004; Mirjafari et al., 2007; Lee et al., 2010; Wang and Martoto-Valer, 2011; Sanna et al., 2013). The pH value of solution including dissolved NH3 and extraction solvent was approximately 11 because of the presence of NH4+ in solution.

To develop a cost-effective process for mineral carbonation, we propose a new process focused on the recovery and the reuse of the NH3 after Mg extraction at 300℃ as a pH swing agent (blue line in Fig. 3). The NH3 after Mg extraction was dissolved in distilled water and supplied to carbonation reactor to increase pH value. The important factor of the carbonation process is the pH value. The conversion of gaseous carbon dioxide (CO2) into a carbonate ion (CO32-) has to proceed through several steps, as shown in reaction (5-9) (Cartwright et al., 2012). First, gaseous carbon dioxide has to be dissolved.

$CO 2 ( g ) ↔ CO 2 ( aq )$
(5)

Then, aqueous CO2 reacts with water to form carbonic acid(6)

(6)

In the next step, carbonic acid dissociates to bicarbonate and carbonate ions.(7)(8)

$H 2 CO 3 ↔ H + + HCO 3 -$
(7)

$HCO 3 - ↔ H + + CO 3 2-$
(8)

Finally, in the presence of calcium cations, calcium carbonate forms and precipitates.(9)

$Ca 2 + + CO 3 2 − ↔ CaCO 3$
(9)

Under acidic conditions, carbonation cannot occur because bicarbonate ions (HCO3 -) could not be converted into carbonate ions (CO32-) but would be converted to carbonic acid (H2CO3) instead. Consequently, CaCO3 cannot be precipitated in the above acidic condition. The advantage of basic conditions for carbonation is that it is easy to control the conversion and purity of CaCO3.

### Carbonation

The effect of the pH swing agent, a NH4+-rich solution, on carbonation was identified. The first acid ionization constant of CO2, Ka1, is 4.3 × 10-7 at 25℃, and the concentration of hydrogen ion (H+) is 2.1 × 10-4 (Mirjafari et al., 2007). Based on 2 M (NH4)2SO4 (100 mL) and 5 g of serpentinite, the average Mg extraction is 6,600 ppm at a reaction temperature of 300℃. On the other hand, the average Ca extraction is approximately 1,000 ppm.

From the electron microscope analysis of the precipitates after carbonation, distinct shaped particles obtained from carbonation were identified (Fig. 4). The carbonation of the Mg-rich solution obtained from serpentinite and CO2 showed synthesized particles with a rhombohedral form (I in Fig. 4a) and sphere-shaped form (II in Fig. 4a). Additionally, cubic-like particles (III in Fig. 3c) were identified through the carbonation. From the EDS analysis (Table 6), we found that the particles with angled forms (I in Fig. 4a) contain relatively higher Mg contents compared to the other shaped particles, whereas the sphere-shaped particles (I in Fig. 4a) have relatively higher Ca contents. The existence of calcium in the sphere-shaped particles resulted from the extraction of Ca contained in serpentinite. On the other hand, the cubic-like particles (III in Fig. 4c) showed the typical atomic ratio of CaCO3 (Table 6).

There are three anhydrous crystalline polymorphs of CaCO3, which are ordered by their decreasing thermodynamic stabilities under atmospheric conditions: calcite, aragonite, and vaterite (Cartwright et al., 2012). Among them, calcite is the most thermodynamically stable and is the crystal structure (trigonal) typically found in nature. On the contrary, vaterite has a higher free energy than calcite, and calcite is usually formed from the transformation of vaterite or amorphous calcium carbonate (Rodriguez-Blanco et al., 2011). However, the kinetics and mechanisms still need to be studied. From previous studies (Song et al., 2014; Luo et al., 2015), the shape of calcite has been reported to be cubic and that of vaterite has been reported to be spherical. Thus, the synthesized sphere-shaped particles and cubic particles can be considered to be vaterite and calcite, respectively. Based on the atomic ratio from EDS (Table 6), the particles with a rhombohedral form (I in Fig. 4) were determined to be MgCO3. As the results were obtained from this study, the carbonation at atmospheric condition preferred CaCO3. The preference of CaCO3 for carbonation as compared with MgCO3 could be understood by solubility product constant. The solubility product constant (Ksp) of CaCO3 (3.36 × 10-9) at 25℃ is lower than that of MgCO3 (6.82 × 10-6) (Lide and Frederikse, 1998). The characteristics of carbonates including the solubility product constant have a significant effect on the rate of precipitation of CaCO3 and MgCO3, respectively.

The control of the morphology CaCO3 and MgCO3 is one of the important factors for the industrial application of synthesized carbonation minerals from mineral carbonation, and the transformation has significant effects on the morphological changes of carbonate minerals. Thus, future research should include a study to investigate the morphological changes via transformation as well as the optimization of the proposed process.

### Geochemical reaction path model

The simulation results are presented in Fig. 5. In the experiment, total pressure in reaction vessels approaches up to 100 bars at 300℃. The simulated pressure also increases up to near 100 bars at 300℃ (Fig. 5a). The modeling result also shows that total pressure is mostly composed of vapor pressure. The contribution of NH3 (g) to total pressure is small but slightly increases with the increase in temperature. The simulated Mg concentrations are presented with the measured values (Fig. 5b). As shown the figure, the modeling results coincide with the proximity to the experiment results. Similarly, the simulated NH4+ concentrations approach to the measured values (Fig. 5c). According to the modeling result, the concentration of NH4+ decreases whereas that of NH3 increases with the increase in temperature. This indicates that NH4+ decomposes to NH3, finally NH3 (g). To identify the contributions of the minerals to Mg extraction, we calculate the percentage of Mg leached from each mineral to the total leached Mg (Fig. 5d). This result shows that most Mg is released from the dissolution of antigorite. However, modeling result also shows that the contribution of chlorite increases with the increase in temperature above 150℃. Thus, our modeling results shows that NH4+ decomposes into NH3 (g) with the increase in temperature and most Mg is leached from antigorite and chlorite.

## Conclusion

To develop a cost-effective process for mineral carbonation, magnesium extraction from serpentinite has been studied by evaluating the experimental factors, including the concentration of (NH4)2SO4, reaction temperature, and ratio of serpentinite/extraction solvent. From this study, the following conclusions were made:

Maximum Mg extraction from serpentinite is approximately 80 wt%, and the best experimental conditions are 2 M (NH4)2SO4 as the concentration of the extraction solvent, 5 g/100 mL as the ratio, and 300℃ as the reaction temperature.

The amount of recovered NH4+ through the decomposition of (NH4)2SO4 is approximately 1.78 M, which is used as a pH swing agent for carbonation to induce the precipitation of carbonation materials, such as MgCO3 and CaCO3.

After carbonation, MgCO3 and CaCO3 (vaterite) from the extraction solvent of serpentinite are synthesized by reaction with CO2, and CaCO3 (calcite) from the extraction solvent of slag is formed by carbonation. From the results of carbonation, we can identify that the use of a pH swing agent is effective for the carbonation step.

There are several conditions that still need to be improved for the development of a cost-effective process for mineral carbonation. First, Mg-rich industrial wastes, such as ferro-Ni slag, have to be used as the raw material for mineral carbonation instead of natural minerals. Second, various ammonium solvents need to be tested to obtain the best Mg extraction efficiency. Finally, the high value-added conversion of synthesized MgCO3 and CaCO3 can be applied to various industries, which will aid the goal of developing a cost-effective process for mineral carbonation.

## Acknowledgements

This research was supported by the Basic Research Project of the Korea Institute of Geoscience and Mineral resources (KIGAM) funded by the Ministry of Science, ICT and Future Planning of Korea.

## Figure

Schematic illustration of the extraction system for magnesium from serpentinite (A: water bath, B: controller, C: reactor including stirrer, and D: gas (N2 or CO2). The dotted line indicates the main part of the extraction system. That part consists of a stirrer (i) cooled by a water bath, a pressure gauge (ii), a reactor (iii), and a furnace (iv). N2 gas was inserted into the reactor for extraction, and CO2 gas was inserted into the reactor for carbonation.

Photos before and after extraction: (a) slurry solution including serpentinite and (NH4)2SO4 before extraction, (b) slurry solution after extraction at 200℃, (c) slurry solution after extraction at 250℃, and (d) slurry solution after extraction at 300℃.

Extraction yield of magnesium (experimental conditions: 2 M ammonium sulfate, 200 RPM, 30 0℃). The ratio of serpentinite (g) to extraction solution (L) (NH4)2SO4 was fixed to 5/100 and 5/75, and the values (76.3 ± 1.53 and 83.0 ± 1.02) are the averages the second deviation of three experiments.

pH value of the solution after extraction. Depending on the reaction temperature, the pH value shows a range of 10.88 ± 0.18.

The pH swing process on serpentinite (Mg-silicates) and blast furnace slag (Ca-silicates). The concept of this process is the reuse of a NH4+-rich solution for carbonation. The NH4+-rich solution was obtained from the Mg extraction of serpentinite. To synthesize CaCO3 precipitate, a high pH value above 10 is needed, and the NH4+-rich solution was used as the pH swing agent for the carbonation processes of both serpentinite and BF slag after Mg and Ca extraction. The ⓥ means valve of CO2 gas.

FE-SEM images after carbonation of a Mg-rich solution (serpentinite) (a and b) and a Ca-rich solution (BF slag) (c and d). The scale bar indicates 10 μm (a and d) and 20 μm (b and c), respectively. The characters I, II, and III are the points analyzed by EDS.

The results of geochemical modeling; The change in (a) pressure, (b) Mg concentration, (c) N species (NH4+ and NH3) concentrations, and (d) contribution of each mineral to Mg concentration, respectively. The solid and dotted line indicate the modeling result. The solid circle indicates the measured data.

## Table

Kinetic parameters of minerals used in this model

a: from reference (Parkhurst and Appelo, 2017),
b: from reference (Critelli et al., 2015).

Elemental analysis of serpentinite before extraction (units: wt%). All elements except sulfur are presented as oxide forms

Crystalline composition of serpentinite obtained from Rietveld refinement

Extraction efficiency of magnesium from serpentinite. (NH4)2SO4 was used as the extraction solution for Mg extraction. Experimental parameters were as follows: concentration of the extraction solution, ratio of serpentinite to the extraction solution, and reaction temperature

Concentration of ammonium ions after Mg extraction from serpentinite. (NH4)2SO4 was used as the solution for extraction. The concentration (*) of ammonium ions was analyzed in the extraction solution after Mg extraction. The NH4+ was analyzed by ion chromatography (IC)

EDS analysis of particles after the carbonation of serpentinite and BF slag (unit: atomic ratio)

## Reference

1. AzdarpourA. AsadullahM. MohammadianE. HamidiH. JuninR. KaraeiM.A. (2015) A review on carbon dioxide mineral carbonation through pH-swing process. , Chem. Eng. J., Vol.279 ; pp.615-630
2. BaileyS.W. , BaileyS.W. (1988) Hydrous phyllosilicates (exclusive of mica)., MSA Reviews in Mineralogy, Vol.Vol. 19 ; pp.9-37
3. BA(c)aratH. MckelvyM.J. ChizmeshyaA.V. SharmaR. CarpenterR.W. (2002) Magnesium hydroxide dihydroxylation/ carbonation reaction processes: Implication for carbon dioxide mineral sequestration. , J. Am. Ceram. Soc., Vol.4 ; pp.742-748
4. BlancP. LassinA. PiantoneP. AzaroualM. JacquemetN. FabbriA. GaucherE.C. (2012) THERMODDEM: A geochemical database focused on low temperature water/rock interactions and waste materials. , Appl. Geochem., Vol.27 ; pp.2107-2116
5. CarneiroJ.F. BoavidaD. SilvaR. (2011) First assessment of sources and sinks for carbon capture and geological storage in Portugal. , Int. J. Greenh. Gas Control, Vol.5 ; pp.538-548
6. CartwrightJ.H. ChecaA.G. GaleJ.D. GebauerD. Sainz-DiazC.I. (2012) Calcium carbonate polymorphism and its role in biomineralization: how many amorphous calcium carbonates are there? , Angew. Chem. Int. Ed., Vol.51 ; pp.11960-11970
7. CritelliT. MariniL. SchottJ. MavromatisV. ApollaroC. RinderT. De RosaR. OelkersE.H. (2015) Dissolution rate of antigorite from a whole-rock experimental study of serpentinite dissolution from 2 < pH < 9 at 25A C: Implications for carbon mitigation via enhanced serpentinite weathering. , Appl. Geochem., Vol.61 ; pp.259-271
8. EvansB.W. HattoriK. BaronnetA. (2013) Serpentinite: What, Why, Where? , Elements, Vol.9 ; pp.99-106
9. LasagaA.C. (1998) Kinetic theory in the earth science., Princeton University Press,
10. LeeS. KimJ. ChaeS. BangJ.H. LeeS.W. (2016) CO2 sequestration technology through mineral carbonation: An extraction and carbonation of blast slag. , Journal of CO2 Utilization, Vol.16 ; pp.336-345
11. LeeS.W. ParkS.B. JeongS.W. LimK.S. LeeS.H. TrachtenbergM.C. (2010) On carbon dioxide storage based on biomineralization strategies. , Micron, Vol.41 ; pp.273-282
12. LideD.R. FrederikseH.P. (1998) CRC handbook of chemistry and physics., CRC Press,
13. LinP.C. HuangC.W. HsiaoC.T. TengH. (2008) Magnesium hydroxide extracted from a magnesium- rich mineral for CO2 sequestration in a gas-solid system. , Environ. Sci. Technol., Vol.42 ; pp.2748-2752
14. LuoJ. KongF. MaX. (2015) Role of aspartic acid on the synthesis of spherical vaterite by the Ca(OH)-CO reaction. , Cryst. Growth Des., Vol.16 ; pp.728-736
15. MirjafariP. AsghariK. MahinpeyN. (2007) Investigation the application of enzyme carbonic anhydrase for CO2 sequestration purposes. , Ind. Eng. Chem. Res., Vol.46 ; pp.921-926
16. OxtobyD.W. GillisH.P. CampionA. (2012) Principle modern chemistry., Brooks Cole,
17. PalandriJ.L. KharakaY.K. (2004) A compilation of rate parameter of water-mineral interaction kinetics for application to geochemical modeling., ; pp.1-64p
18. ParkA.H. FanL.S. (2004) CO2 mineral sequestration: physically activated dissolution of serpentinite and pH swing process. , Chem. Eng. Sci., Vol.59 ; pp.5241-5247
19. ParkhurstD.L. AppeloC.A. (2017) usgs.gov/tm/06/a43
20. Rodriguez-BlancoJ.D. ShawS. BenningL.G. (2011) The kinetics and mechanisms of amorphous calcium carbonate (ACC) crystallization to calcite, via vaterite. , Nanoscale, Vol.3 ; pp.265-271
21. SannaA. DriM. Maroto-ValerM.M. (2013) Carbon dioxide capture and storage with mineralization using recyclable ammonium salts. , Energy, Vol.51 ; pp.431-161
22. SongK. JangY.N. KimW. LeeM.G. ShinD. BangJ.W. JeonC.W. ChaeS.C. (2014) Factors affecting the precipitation of pure calcium carbonate during the direct aqueous carbonation of flue gas desulfurization gypsum. , Energy, Vol.65 ; pp.527-532
23. WangX. Maroto-ValerM.M. (2011) Dissolution of serpentinite using recyclable ammonium salts for CO2 mineral carbonation. , Fuel, Vol.90 ; pp.1229-1237
24. WangX. Maroto-ValerM.M. (2011) Integration of CO2 capture and mineral carbonation by using recyclable ammonium salts. , ChemSusChem, Vol.4 ; pp.1291-1300
25. WolterbeekT.K. PeachC.J. SpiersC.J. (2013) Reaction and transport in wellbore interfaces under CO2 storage conditions: Experiments simulating debonded cement-casing interfaces. , Int. J. Greenh. Gas Control, Vol.19 ; pp.519-529
26. ZhangM. BachuS. (2011) Review of integrity of existing wells in relation to CO2 geological storage: What do we know? , Int. J. Greenh. Gas Control, Vol.5 ; pp.826-840