Carbonation of Natural Wollastonite at Non-Ambient Conditions Relevant for CCS—the Possible Use as Cementitious Material in Wellbores

Abstract

The reaction of wollastonite with CO₂ accompanied by SO₂ and NO₂ in the presence of a chloride-rich brine (230 g/L NaCl, 15 g/L CaCl₂, 5 g/L MgCl₂) at temperatures relevant to injection conditions (333 K) in carbon capture and storage (CCS) were investigated within the joint BMWi (Federal Ministry for Economic Affairs and Energy) research project CLUSTER. The reaction which describes the formation of wollastonite during metamorphism is reversed and shows a strong temperature dependence. Wollastonite reacts in the presence of CO₂ ($\overline{C}$) in aqueous conditions to form calcium carbonate and amorphous silicon oxide. At 333 K and 2 MPa the carbonation reaction of wollastonite ($\overline{C}CS$) is fast (<24 h). To determine the conversion rate of the reaction quantitatively different methods were used and compared: Powder X-ray diffraction (PXRD) with the Rietveld method and differential scanning calorimetry with thermogravimetry, coupled with a mass spectrometer (DSC-TG/MS) for quantitative phase analysis and for determination of the carbonation. The carbonation (CO₂ accompanied by SO₂ and NO₂) of natural wollastonite at 333 K in presence of chloride-rich brine was rather fast (almost complete after 24 h reaction time).

1. Introduction

The Intergovernmental Panel on Climate Change (IPCC) climate report, 2018 [1] concluded that reduction of CO₂ emissions was necessary, because the current CO₂ concentrations are still considerably rising and currently amount to 408 ppm [2]. Therefore, it is a mandatory challenge to reduce CO₂ emissions. One possible option being currently investigated and discussed [3] is carbon capture and storage (CCS). Using CCS, CO₂ emissions from different sources (energy production, steel, and cement industries) could be reduced by finding and exploring suitable deep geologic formations (e.g., sandstones) for CO₂ storage. The formations considered for injection could be on- or offshore which could be in part already exploited former gas fields [3]. During and after injection, the casing and sealing of boreholes at storage sites must feature high reliability, because CO₂ has an impact on the stability of conventional hydrated cements [4,5], i.e., it leads to degradation due to carbonation.

Therefore, the right choice of suitable sealing systems is challenging. The material has to resist the harsh conditions at the point of injection considering pressure, temperature, acidic environment, and the composition of aggressive fluids (chloride-rich brines, CO₂ accompanied by NO₂ and SO₂). The high amount of CO₂ and fluids will impose massive impact on cement properties. The commonly used ordinary Portland cements (OPC) are often limited due to their instability under these conditions for sealing wellbores.

The current study is based on the investigation of carbonation as a hardening method to obtain carbonates as reaction products by using wollastonite. Although this reaction has been investigated [6,7,8], its use for CCS is still not widely evaluated.

Wollastonite has a Ca–Si ratio of 1:1 and, therefore, during industrial production, a better CO₂ balance than conventional cements with Ca–Si ratio 2:1 and 3:1 [6,9].

However, resources are limited. To overcome this bottleneck of wollastonite supply, it can be produced by adjusting a conventional cement rotary kiln production process. Additionally, CO₂ will be redeemed during this process and improve the CO₂ balance.

The present study considers the back reaction by carbonation of wollastonite forming CaCO₃ and SiO₂ as a base [10] within the joint BMWi (German Federal Ministry for Economic Affairs and Energy) research ([6,7,8,11,12] CLUSTER). The conditions, especially non-ambient ones, for the reaction of CaSiO₃ (CS) with CO₂ () were investigated.

The carbonation reaction of wollastonite is described as extremely slow in the absence of water at low temperatures [13]. However, at 333 K and ambient pressure, the $\overline{C}CS$ (CO₂ + CaSiO₃) reaction is accelerated (several hours up to one day) in the presence of H₂O [6,7,13]. Microstructure development of the carbonation was investigated by Villani et al. [14] and Sahu and DeCristofaro [15]. Both studies showed a carbonation starting from the surface to the center of the wollastonite grains, analogous to the model of the shrinking core [16]. In the presence of CO₂ in aqueous conditions, CaSiO₃ is consumed and calcium carbonate is formed. Adding an internal standard of known weight to the mixture, it is possible to determine the amorphous silicon oxide formation through X-ray diffraction [6]:

$${\mathrm{CaSiO}}_3+{\mathrm{CO}}_2\stackrel{{\mathrm{H}}_2\mathrm{O}}{\to }{\mathrm{CaCO}}_3+{\mathrm{SiO}}_2\left(\mathrm{amorphous}\right)$$

(1)

The chosen temperature (333 K) and brine composition (230 g/L NaCl, 15 g/L CaCl₂, 5 g/L MgCl₂) defined within this joint project [11] simulated conditions in deep geological formations envisaged for CCS. According to Gartner and Hirao [13] and Longo et al. [17], the presence of water is crucial for the carbonation reaction. Our experiments were performed using chloride-rich brine as the carbonation media, because aqueous and chloride-rich solutions are expected to be present at the point of injection in wellbores.

The use of a cementitious material, based on the carbonation of wollastonite (CS), seems very promising [18,19,20], because CO₂ has an impact on the stability of conventional hydrated OPCs and wellbore cements [4,5,6].

In previous studies, the carbonation of calcium silicates in general [21,22] and wollastonite in particular [6,7,8,13,17,23] in the presence of pure water and pure CO₂ was investigated.

In this study, carbonation of wollastonite in a chloride-rich brine (230 g/L NaCl, 15 g/L CaCl₂, 5 g/L MgCl₂) as aqueous solution with CO₂ accompanied by 70 ppm SO₂ and 70 ppm NO₂ at 333 K was investigated. The conditions were defined within the project CLUSTER as a realistic scenario during injection of CO₂ in a geologic formation in CCS.

2. Materials and Methods

A series of non-ambient experiments were performed and different analytic techniques were applied, to determine the reaction of wollastonite with carbon dioxide (CO₂) in the presence of a chloride-rich brine. For the characterization of educts and products, powder X-ray diffraction (PXRD) and differential scanning calorimetry with thermogravimetry (DSC-TG) coupled with a mass spectrometer (MS) and an infrared spectrometer (IR) were applied.

2.1. Materials

Steel vessels with Teflon liners from the company Büchi AG (capacity: 100 mL) were used. According to the experiments of Huijgen et al. [23] and Svensson et al. [6], the carbonation experiments were performed at 2 MPa. The temperature was chosen at 333 K. The composition of the used CO₂ (doped with ~70 ppm SO₂ and ~70 ppm NO₂) and synthetic brine were identical to the studies of the authors [6,7,8,12]. The natural wollastonite, i.e., CaSiO₃ was supplied by Alfa Aeser (calcium silicate, meta, reagent grade, <20 µm powder). As described by Huijgen et al. (2006), the chosen grain size (20 µm) of the wollastonite raw material benefits the carbonation reaction. The chemical composition of the used raw material was determined by X-ray fluorescence spectroscopy (XRF) [6]. The main elements in the used natural CaSiO₃ were SiO (50.55 wt.%) and CaO (45.95 wt.%), while the impurities (Fe₂O₃, Al₂O₃, MgO, MnO, TiO₂, K₂O, SrO) were <1.5%. The loss of ignition was determined at 1323 K and accounted for 2.09 wt.%.

The composition of the used raw material is given in

Table 1. Phase composition of the wollastonite raw material (CaSiO₃), qualitatively and quantitatively analyzed, using powder X-ray diffraction (PXRD) and the Rietveld method [24].

Phases	Content [wt.%]
Wollastonite	90.7 ± 0.8
Calcite	1.9 ± 0.2
Aragonite	1.0 ± 0.5
Amorphous	6.4 ± 0.5

and Figure 1. Besides wollastonite (90.7 wt.%), the calcium carbonates, calcite (1.9 wt.%) and aragonite (1.0 wt.%), could be determined as well. For the determination of the X-ray amorphous content rutile (TiO₂) as internal standard was added. The amorphous part amounts to 6.4 wt.%.

An ISCO-pump (Teledyne ISCO, Model 500D Syringe pump) was used for supplying the gases and adjusting the pressure. CO₂ (purity 4.5; 99.995%), SO₂ (purity 3.0; 99.9%), and NO₂ (purity 1.8; 98%) were supplied by Westfalen AG. The chloride-rich brine (H₂O with 230 g/L NaCl, 15 g/L CaCl₂, 5 g/L MgCl₂) was produced using NaCl (Carl Roth GmbH + Co. KG), CaCl₂ (AppliChem GmbH; Panreac Quimica SLU; ITW Companies), and MgCl₂ (Carl Roth GmbH + Co. KG).

2.2. Analytical Methods

Powder X-ray diffraction (PXRD) was used to analyze the educts. The parameters of the PXRD measurements were identical to the work of Svensson et al. [6] (PANalytical X’Pert³ Powder with PIXcel 1D detector; 45 kV, 40 mA; apertures: 0.04 rad Soller and fixed slits (0.125° and 0.25°)). Diffractograms range from 10° to 75° 2Θ with 49.725 s per step and step size 0.013° 2Θ [6]. For qualitative phase analysis, the ICDD (International Centre for Diffraction Data) database [25] implemented within the HighScore Plus suite [26] and, for quantitative phase analysis, the Rietveld method [24] with the software package Profex-BGMN [27,28] were applied. The amorphous content was determined by adding crystalline TiO₂ (Kronos 2900–TiO₂; rutile, supplied by KRONOS TITAN GmbH) of known weight (~10 wt.%).

The degree of carbonation of the samples was investigated depending on experimental time. The decrease of wollastonite accompanied by the increasing of calcium carbonate and amorphous contents had to be quantified. Different approaches were applied to identify and evaluate educts and products quantitatively. The weight fractions of wollastonite, calcite, aragonite, and X-ray amorphous material were determined with the Rietveld method [24]. To quantify the fraction of CaCO₃, heat flow coupled with thermogravimetric analysis (DSC-TG; STA 449 F3 Jupiter NETZSCH) were carried out. The gas release during the heating regime was additionally recorded with MS (QMS 403D Aeolos NETZSCH). The atom mass units were counted from 1 to 100 u. The thermoanalytical measurements were carried out in the temperature interval from 296 K (23 °C) to 1273 K (1000 °C). The heating rate was 10 K per minute. Approximately 15–25 mg of sample material was used per measurement [6].

The products were milled before the PXRD and DSC-TG/MS measurements were carried out. An internal standard (rutile) was added for PXRD measurements.

The amount of X-ray amorphous content of CaCO₃ in the product could be analyzed as well with DSC-TG/IR/MS.

The content of carbonate (crystalline and amorphous) was determined by the loss of CO₂ which is given in Equation (2):

$${\mathrm{CaCO}}_3\left[\mathrm{wt}.\%\right]=\frac{{\mathrm{mass}\mathrm{loss}\mathrm{CO}}_2\left[\mathrm{wt}.\%\right]}{\left(\frac{{\mathrm{mass}\mathrm{CO}}_2\left[\mathrm{g}/\mathrm{mol}\right]}{{\mathrm{mass}\mathrm{CaCO}}_3\left[\mathrm{g}/\mathrm{mol}\right]}\right)}$$

(2)

2.3. Experimental Setup

The reaction of wollastonite with CO₂ at conditions close to geological storage conditions (333 K, CO₂ accompanied by SO₂ and NO₂, and chloride-rich brine) was investigated. In the experiments 5 mL chloride-rich brine (H₂O with 230 g/L NaCl, 15 g/L CaCl₂, 5 g/L MgCl₂) was added to 1 g of sample material. The pH value of the brine was ~6. This mixture was put in steel autoclaves for further treatment, i.e., 25 mL CO₂ accompanied by SO₂ (2000 ppm, 0.1 MPa), 25 mL CO₂, accompanied by NO₂ (2000 ppm, 0.1 MPa), and 25 mL CO₂ (5.5 MPa) were filled with an ISCO-pump into the reaction vessel. End pressure was adjusted to 2 MPa inside the reactor. For heating, the autoclaves were placed in a drying oven. A series of experiments were carried out by varying time intervals. The pressure (2 MPa) and temperature (333 K) remained constant. As preliminary tests, wollastonite raw material was carbonated in brine for 24 h using pure CO₂ (purity 4.5; 99.995%) and wollastonite raw material was carbonated in water for 24 h using pure CO₂ (purity 4.5; 99.995%) (

Table 2. Overview of performed series of experiments at constant pressure and temperature and varied time.

Temperature [K]	Carbonation Medium	Gas Composition	Pressure [MPa]	Reaction Time [h]
333	water	CO2	2	24
333	brine	CO2	2	24
333	brine	CO2 + SO2 + NO2	2	0, 3, 6, 12, 18, 24

).

After carbonation, the samples were washed with pure water to remove the salts and dried at 303 K (30 °C).

3. Results

The carbonation of wollastonite in a chloride-rich brine (H₂O with 230 g/L NaCl, 15 g/L CaCl₂, 5 g/L MgCl₂) with pure CO₂ and CO₂ accompanied by ~70 ppm SO₂ and ~70 ppm NO₂ was investigated.

After 24 h reaction time at the selected conditions, the experiments were terminated [6]. As carbonation products, Svensson et al. [6] observed calcite, aragonite, and amorphous content. In addition, small contents of wollastonite were observed, due to the reaction process, which can be described by the shrinking core model [17].

3.1. PXRD Analysis

In Table S1, the phase composition of the carbonated samples (chloride-rich brine, 333 K, 2 MPa), determined by PXRD and Rietveld method [24], obtained in this study is presented.

Contrary to the results of the reference experiment (

Table 3. Results of the differential scanning calorimetry with thermogravimetry (DSC-TG) measurements.

Reaction Time [h]	Gas Composition	Weight Loss [%]	CO2 Content [wt.%]	Calculated Content of CaCO3 [wt.%]
0	---	2.1	1.7	3.9
3	CO2 + SO2 + NO2	7.6	7.0	16.1
6	CO2 + SO2 + NO2	13.4	12.2	28.0
12	CO2 + SO2 + NO2	20.2	18.6	42.7
18	CO2 + SO2 + NO2	22.4	21.4	49.0
24	CO2 + SO2 + NO2	25.9	24.4	56.0
24	CO2	25.7	24.3	55.8

), no additional newly formed aragonite was observed. This was due to the chloride-rich brine, which contained high amounts of sodium chloride (NaCl). Sodium chloride is known to inhibit the formation of aragonite [29,30].

3.2. DSC-TG/MS Measurements

DSC-TG/MS measurements of the wollastonite raw material (educt) and the carbonated wollastonite (CO₂ + SO₂ + NO₂, 333 K, 2 MPa) were carried out to determine the full content of water and carbonate (Table 3). This was necessary because X-ray amorphous carbonate could not be determined directly with PXRD, when additional phases (SiO₂) contributed to the total amorphous content [6].

DSC-TG/MS measurements (Figure 2a) were carried out by Svensson et al. [6] with the raw material. Figure 2b represents the results of DSC-TG/MS measurements of carbonated wollastonite (24 h CO₂, 333 K, 2 MPa). During the measurement of the raw material, the weight loss was 2.1%, from which 1.7 wt.% could be associated with CO₂. Hence the original content amounts to 3.9 wt.% carbonate (CaCO₃) in the sample (Figure 2a, Equation (2)). The Rietveld analysis (5 wt.%) and the DSC-TG (4 wt.%) measurements of the raw material for the CaCO₃ exhibited similar values [6].

DSC-TG/MS/IR analysis of the carbonated wollastonite (24 h CO₂, 333 K, 2 MPa), shown in Figure 2b, was carried out. During heating, a weight loss of 24.4% was recorded. From which 23.9 wt.% could be attributed to CO₂. According to Equation (2), the original content of carbonate was 54.8% (Figure 2b). The results of the Rietveld method and the DSC-TG measurements of the carbonated wollastonite for the content of carbonate showed again similar results (Rietveld: 50%; DSC-TG: 55%) [6].

Thermal (DSC-TG) and spectroscopic (MS) analyses on previously treated wollastonite with CO₂ and SO₂/NO₂ (~70 ppm each) at 333 K and 2 MPa is shown exemplarily in Figure 3. During heating, 25.9% of the weight was lost; 24.4% could be attributed to CO₂, i.e., to 56 wt.% carbonate in the sample (Figure 3, Table 3, Equation (2)). The MS curve (green) detected a maxima of the mass flow of H₂O (Figure 3) at approximately 400 K, which decreased distinctly at 500 K. From 500 to 700 K the release of H₂O increased again and dropped afterwards until 1050 K. The MS curve of CO₂ (Figure 3, magenta line) remained at a constant level until 700 K and increased slightly until 900 K, then it increased more distinctly and held a plateau until 950 K. Afterwards, a maximum at 1100 K was reached. At temperatures beyond 1100 K, no significant release of CO₂ could be observed. The TG curve (weight loss, Figure 3, black line) decreased with a small rate until 900 K. The loss of CO₂ could be observed clearly in the curve between 900 and 1080 K. Beyond 1100 K, the release of CO₂ stopped. The DSC signal (Figure 3, blue line) showed a maximum (endothermic) amplitude at 1080 K and correlated with the detected MS maximum of CO₂. At approximately 1150 K, a positive amplitude for the DSC signal could be observed and indicated the formation of Ca₂SiO₄ (i.e., C₂S, larnite), which was confirmed by PXRD of the TG sample. The formation of larnite was similar to the work published by Svensson et al. [6].

The quantitative PXRD and the DSC-TG analysis of the carbonated wollastonite (CO₂ + SO₂ + NO₂, 333 K, 2 MPa) for the content of carbonate were of similar magnitude (Rietveld: 58.4 ± 0.5 wt.%; DSC-TG: 56 ± 2.8 wt.%).

Thermal (DSC-TG) and spectroscopic (MS) analyses on previously treated wollastonite with CO₂ at 333 K and 2 MPa is shown exemplarily in Figure 4. During heating, 25.7% of the weight was lost; 24.3% could be attributed to CO₂, i.e., to 55.8 wt.% carbonate in the sample (Figure 4, Table 3, Equation (2)). The MS curve (green) detected a maxima of the mass flow of H₂O (Figure 4) at approximately 400 K, which decreased distinctly at 500 K. From 550 to 650 K the release of H₂O increased again and dropped afterwards until 1050 K. The MS curve of CO₂ (Figure 4, magenta line) remained at a constant level until 700 K and increased slightly until 900 K, then it increased more distinctly and held a plateau until 950 K. Afterwards, a maximum at 1080 K was reached. At temperatures beyond 1080 K, no significant release of CO₂ could be observed. The TG curve (weight loss, Figure 4, black line) decreased with a small rate until 900 K. The loss of CO₂ could be observed clearly in the curve between 900 and 1080 K. Beyond 1100 K, the release of CO₂ stopped. The DSC signal (Figure 4, blue line) showed a maximum (endothermic) amplitude at 1080 K and correlated with the detected MS maximum of CO₂. At approximately 1150 K, a positive amplitude for the DSC signal could be observed and indicated the formation of Ca₂SiO₄ (i.e., C₂S, larnite), which was confirmed by PXRD of the TG sample. The formation of larnite was similar to the work published by Svensson et al. [6].

The quantitative PXRD and the DSC-TG analysis of the carbonated wollastonite (CO₂ + SO₂ + NO₂, 333 K, 2 MPa) for the content of carbonate were of similar magnitude (Rietveld: 58.4 ± 0.5 wt.%; DSC-TG: 56 ± 2.8 wt.%).

4. Discussion

The overall carbonation reaction at 333 K, 2 MPa CO₂ pressure (accompanied by 70 ppm SO₂ and 70 ppm NO₂), and in brine were in good agreement with the results of Svensson et al. [6,7] (Figure 5, 333 K, 2 MPa pure CO₂ pressure, in water; Figure 6).

The re untreated wollastonite raw material at 333 K was more or less fully converted after 18 h (Figure 5a) [6]. The content of wollastonite (black squares) decreased exponentially, while the amounts of the products (calcite, aragonite, total amount of CaCO₃, and X-ray amorphous content) increased exponentially. The “full” (~95%) conversion of purified (HCl, 10%) wollastonite at 333 K was already achieved after 18 h (Figure 5b) [7]. The quantities of wollastonite, calcite, the total amount of CaCO₃, and the X-ray amorphous content developed in a similar manner compared to the untreated wollastonite raw material (Figure 5). However, in the products of the previously purified wollastonite raw material, no aragonite was observed.

The conversion of wollastonite under conditions expected at injection sites for CCS (CO₂ accompanied by 70 ppm SO₂ and 70 ppm NO₂, 333 K, chloride-rich brine) was practically fully completed after 24 h (Figure 6). The amount of calcite (magenta circles) and the total amount of CaCO₃ measured by thermogravimetry (cyan circles) and the amorphous content (blue triangles) increased exponentially, whereas the vice versa amount of wollastonite (black squares) decreased exponentially. The content of aragonite was constant during the series of experiments and no newly formed aragonite was observed. The content of halite was less than 1 wt.% which functioned as a monitoring of the successful washing procedure after the carbonation reaction.

Compared to the work of Svensson et al. [6,7,8,9] (Figure 5), higher amounts of calcite and lower amounts of X-ray amorphous content were observed. This was possibly due to the washing process, which was necessary to remove the salts (NaCl, CaCl₂, and MgCl₂) from the samples. No additional aragonite formation was observed, which was expected due to previous work [6,8,13]. According to the work of Kitano [29] and Kitano et al. [30], the chloride-rich brine, containing high amounts of sodium chloride (230 g/L), inhibited the aragonite formation.

5. Conclusions

Carbonation of wollastonite in chloride-rich brine using pure CO₂ and CO₂ accompanied by 70 ppm SO₂ and 70 ppm NO₂ was investigated. According to previous studies [6,8,13], the formation of aragonite was expected. However, no additional aragonite formation was observed, which was, according to the work of Kitano [29] and Kitano et al. [30], inhibited by the high amounts of sodium chloride in the brine.

The carbonation of wollastonite, depending on the reaction time (Table 3 and Figure 6), was in a good agreement with the work of Svensson et al. [6,7] (Figure 5). Higher amounts of calcite and lower amounts of X-ray amorphous content were observed. This was possibly due to the washing process, which was necessary to remove the salts (NaCl, CaCl₂, and MgCl₂) from the samples.

The performed experiments showed that the carbonation reaction of wollastonite under conditions relevant for CCS (333 K, chloride-rich brine, CO₂ accompanied by 70 ppm SO₂ and 70 ppm NO₂) was rather fast (24 h). The use of wollastonite as a CS-Cement in CCS seems possible. Even the aggressive fluid (H₂O with 230 g/L NaCl, 15 g/L CaCl₂, 5 g/L MgCl₂) had no negative impact on the conversion of wollastonite.

The long-term stability is still under investigation and has to be evaluated in depth. As assumed and reported by Svensson et al. [6], the most likely use of wollastonite as a cementitious material in CCS application would be in finishing operations for sealing the borehole after the injection of CO₂ in the reservoir is completed. Further concrete technology considerations will be the next step towards sustainable mixtures, which will be poured into the borehole.