CO₂ capture, geological storage, and mineralization using biobased biodegradable chelating agents and seawater

Abstract

Geological storage and mineralization of CO₂ in mafic/ultramafic reservoirs faces challenges including limited effective porosity, permeability, and rock reactivity; difficulties in using seawater for CO₂ capture; and uncontrolled carbonation. This study introduces a CO₂ capture, storage, and mineralization approach with the utilization of biobased biodegradable chelating agents and seawater. An acidic chelating agent solution is used to increase effective porosity and permeability through enhanced mineral dissolution. For instance, applying an acidic N,N-Bis(carboxymethyl)-L-glutamate solution to a porous basalt increased effective porosity by 16% and permeability by 26-fold in 120 hours. Subsequently, alkaline chelating agent–containing seawater improves CO₂ capture and storage by inhibiting mineralization, thus maintaining injectivity while providing ions for mineralization and further expanding storage space. Last, controlled mineralization is achieved by adjusting chelating agent biodegradation. Promising CO₂ storage and mineralization capacities two orders higher than current techniques, this approach reduces required reservoir volume while enhancing efficiency.

Biobased and biodegradable chelating agents are used in CO₂ capture, geological storage, and mineralization.

INTRODUCTION

To address climate change, the Paris Agreement advocates restraining the average global warming to 1.5° to 2°C relative to preindustrial levels (1). To meet this goal, greenhouse gas emissions must decline by 43% by 2030, compared to 2019 levels, which will require scaling up the removal and storage to ~10 billion tonnes (Gt) of CO₂ annually by 2050 and 20 Gt of CO₂ by 2100 (2–4). Among various options for CO₂ storage, geological storage emerges as a promising means of meeting the ambitious global climate target because of the large storage potential and abundance of storage reservoirs, which include sedimentary formations and unconventional reservoirs (mafic and ultramafic rock reservoirs) (5–7).

Injecting CO₂ into sedimentary formations for storage is a relatively mature technology. However, as most of the injected CO₂ is stored as a free phase, there is a potential risk of CO₂ leakage from the storage site back to the atmosphere from faults or fractures (8–9). Another option is mineral trapping of CO₂ in mafic and ultramafic rock reservoirs (10–13), in which the CO₂ dissolved in water is mineralized into stable carbonate minerals such as calcite (CaCO₃), dolomite (CaMgCO₃), or magnesite (MgCO₃) (14). Basaltic rocks are particularly attractive for CO₂ mineralization, because of their enrichment in divalent metal ions and global abundance (15). Basaltic rocks can sequester up to 60,000,000 Gt of CO₂ through carbonation if the resource is economically accessible and ultimately fully carbonated (6). The first injection of CO₂ into a basalt reservoir was carried out at the Carbfix pilot project in Iceland, in which 95% of the CO₂ injected was mineralized within 2 years (16).

However, there are concerns regarding the capacity and injectivity (i.e., ease of fluid flow) of basaltic rock reservoirs for CO₂ storage. First, the interconnected pore spaces (i.e., effective porosity) and permeability of basaltic rocks are not always sufficiently high. Callow et al. (17) noted that ~40 to 80% of the total porosity of basalt at the Carbfix site is isolated porosity and cannot be used for CO₂ storage. The permeability and, thus, the injectivity may be further reduced by precipitation of carbonates during CO₂ injection (18–20). Therefore, CO₂ mineralization should be minimized during the injection process. Another concern is the reactivity of the rocks. While the injection of CO₂-charged acidic water can promote the dissolution of basalt by providing more H⁺, this effect is small at the typical temperatures of basalt reservoirs targeted for CO₂ storage, which ranges from 20° to 50°C (21–23). Last, the use of large quantities of fresh water for CO₂ dissolution has been regarded as one of the main drawbacks of geological CO₂ storage. Although the use of seawater instead of freshwater has been proposed, its efficiency is even lower (reduced by about 25%) (24).

The effective utilization of biobased and biodegradable chelating agents provides a solution to these challenges. Chelating agents can form complexes with metals over a wide range of pH values. For this reason, they are widely used in industrial, domestic, and agricultural applications. They can enhance mineral dissolution, promoting the leaching of both metals and silicon (25–27), and they have been applied to improve permeability in oil-field stimulation (28) and are potentially promising in geothermal field stimulation (26, 29–32). Among the various types of chelating agents, some are biobased and readily biodegradable, such as N,N-Bis(carboxymethyl)-L-glutamic acid (GLDA), a derivative of glutamic acid, competitively priced relative to other chelating agents and attractive for use in natural environments (33). Biobased chelating agents are particularly attractive in CO₂ storage because their production involves absorption of large amounts of CO₂ from the atmosphere through photosynthesis.

In this study, we propose an approach for enhanced CO₂ capture, geological storage, and mineralization using biobased and biodegradable chelating agents. Figure 1 illustrates the operational processes, using GLDA as a representative chelating agent. This method consists of the following steps: First, as the effect of GLDA on mineral dissolution is more pronounced under acidic conditions (27), an acidic GLDA solution is injected to enhance rock dissolution. This accelerates the leaching of divalent metals from minerals for CO₂ mineralization and increases the pore volume and connectivity, as well as the permeability of the rock, thereby improving the capacity and injectivity for CO₂ storage. An improved CO₂ storage reservoir can store either CO₂ or CO₂-charged water. To maximize CO₂ capture, storage, and mineralization efficiencies, an alkaline GLDA seawater solution is suggested for the second step. The alkaline and pH-buffering characteristics of commercial GLDA-Na₄ solution can serve as an effective absorbent for CO₂ capture. The use of seawater as a solvent for GLDA not only overcomes the challenge of the substantial consumption of fresh water but also provides divalent metals and microbes for CO₂ mineralization and GLDA biodegradation, respectively. Furthermore, the injection of GLDA solution with CO₂ inhibits mineralization, because almost all the metal ions in the solution are bound to GLDA as chelates and, therefore, do not contribute to the generation of secondary minerals. The presence of GLDA also inhibits the crystallization of carbonate; even if a small amount of carbonate is formed, it is nano-sized (34). This allows the CO₂ storage process to proceed without substantially decreasing the effective porosity and permeability. The injection of the solution may actually increase the effective porosity to some degree through enhanced mineral dissolution. Last, with GLDA biodegradation in the reservoir, both metal ions (those leached from basalt or in the initial seawater) and CO₂ (captured from the atmosphere during GLDA production) are released, which triggers CO₂ mineralization. Therefore, the rate of CO₂ mineralization can be controlled by adjusting the rate of GLDA degradation, for instance, by modifying the number of microbes in the injection fluid or reservoir.

Fig. 1. Diagram of enhanced CO₂ capture, geological storage, and mineralization through utilization of a biobased and biodegradable chelating agent, GLDA solutions.

This approach addresses the challenges in existing technologies, including potentially poor effective porosity, permeability, and rock reactivity, as well as difficulties in the use of seawater for CO₂ capture and uncontrolled carbonation (referring to the difficulty of controlling the timing of carbonation). The purpose of this study is to investigate the feasibility of this approach and, through laboratory experiments, to elucidate the appropriate operating conditions.

RESULTS

Effectiveness of GLDA in enhancing basalt dissolution

To determine the suitable conditions to enhance basalt dissolution, a series of batch experiments was carried out using powders of both unaltered and altered basalt. Altered basalt is more abundant and more readily available than fresh basalt but has been considered unsuitable for CO₂ storage due to its lower reactivity (35), which is expected to be improved with GLDA. When minerals are present as powders, the dissolution rate is expected to be higher than in the actual reservoir, where minerals exist as larger particles; however, the implications for suitable operating conditions will remain unchanged. The GLDA-Na₄ concentrations varied from 0 to 20 wt %, initial pH values from 1 to 10, and temperatures from 20° to 60°C (typical of basalt reservoirs targeted for CO₂ storage). Temporal changes in elemental concentrations are illustrated in figs. S3 to S5. Figure 2 summarizes the concentrations of elements leached after 6-hour reactions.

Fig. 2. Increase in elemental concentrations following a 6-hour batch dissolution reaction in GLDA solutions.

(A) Unaltered basalt powders reacted in a 0 to 20 wt % GLDA-Na₄ solution at pH 4 and 35°C; (B) unaltered basalt powders reacted in a 5 wt %GLDA-Na₄ solution at pH values of 1 to 10 and 35°C; (C) unaltered basalt powders reacted in a 5 wt % GLDA-Na₄ solution at pH 2 and 20° to 60°C; (D) altered basalt powders reacted in a 0 to 20 wt % GLDA-Na₄ solution at pH 4 and 35°C; (E) altered basalt powders reacted in a 10 wt % GLDA-Na₄ solution at pH values of 1 to 10 and 35°C; and (F) altered basalt powders reacted in a 10 wt % GLDA-Na₄ solution at pH 2 and 20° to 60°C. A darker background color represents a stronger enhancement.

As illustrated in Fig. 2 (A and D), the presence of GLDA enhanced the leaching of the dominant elements (i.e., Mg, Al, Si, K, Ca, and Fe) from both unaltered and altered basalts by up to three orders of magnitude. This enhancement was generally more significant at higher GLDA concentrations, indicating a greater dissolution enhancement of most minerals in the rocks. Although excessive amounts of GLDA led to a slight weakening of this enhancement owing to adsorption of chelating agent molecules to the mineral surface and inhibition of the diffusion process (36), a wide range of GLDA concentrations was suitable for injection. From the unaltered basalt, Fe exhibited a high leaching concentration, indicating that the reactivity of Fe-bearing minerals, such as augite [(Ca,Mg,Fe)₂Si₂O₆] and olivine [(Mg,Fe)₂SiO₄], was notably elevated in the presence of GLDA. The dissolution of augite and olivine in GLDA solution can be represented, for instance, by

$$\begin{array}{c}{(\text{Ca},\text{Mg},\text{Fe})}_2{\text{Si}}_2{\mathrm{O}}_6+{\text{4H}}^{+}+{\text{2H}}_2\mathrm{O}+{\text{2GLDA}}^{4-}\hfill \\ \rightleftharpoons 2(\text{Ca},\text{Mg},\text{Fe})-{\text{GLDA}}^{2-}+{\text{2H}}_4{\text{SiO}}_4\hfill \end{array}$$ (1)

$$\begin{array}{c}{(\text{Mg},\text{Fe})}_2{\text{SiO}}_4+{\text{4H}}^{+}+{\text{2GLDA}}^{4-}\hfill \\ \rightleftharpoons 2(\text{Mg},\text{Fe})-{\text{GLDA}}^{2-}+{\mathrm{H}}_4{\text{SiO}}_4\hfill \end{array}$$ (2)

On the other hand, a substantial amount of Ca was leached from altered basalt, reaching a concentration up to 20 times higher than that from the unaltered basalt. This was attributed to the preferred and rapid dissolution of Ca-montmorillonite [Ca_0.3(Al,Mg)₂(Si₄O₁₀)(OH)₂·nH₂O] by GLDA, as evidenced by the markedly diminished characteristic peaks of Ca-montmorillonite in the x-ray diffraction (XRD) patterns of the reacted basalt (fig. S6). The preferred and enhanced dissolution of clay minerals with the use of chelating agents expands the range of potential storage reservoirs to include altered basalt, which is more abundant and more readily available than fresh basalt (35, 37).

Basalt dissolution was also influenced by the pH of the GLDA solution, particularly at pH values below 6 (Fig. 2, B and E). Lower pH values resulted in further enhanced mineral dissolution. For instance, Fe leaching concentrations increased by a factor of 50 when pH was lowered from 6 to 1. This was attributed to the combined effect of H⁺ attack and chelation on the mineral surface. The leaching of Fe and associated improvement in the chemical and hydraulic properties was particularly significant at lower pH values. The enhancement in Ca leaching from montmorillonite in altered basalt was notable at all pH values, suggesting that a wider range of pH values may be effective for altered basalt. Last, whereas GLDA was effective in promoting mineral dissolution across the entire temperature range of 20° to 60°C (Fig. 2, C and F), higher temperatures further promoted the dissolution of unaltered and altered basalts, particularly Fe-bearing minerals.

Pore properties and permeability improvement

After determining the suitable conditions for injection through batch experiments using powders, a chelating agent flooding experiment was conducted using basalt core samples to understand the behavior of the whole rock. The flooding experiment was conducted using the experimental apparatus described in fig. S2. Specifically, a 5 wt % GLDA-Na₄ solution with an initial pH value of 2, which is effective in enhancing basalt dissolution, was selected for injection into an unaltered basalt core sample at 35°C and a flow rate of 0.03 ml/min. The effectiveness of acidic GLDA solution injection in improving pore properties and permeability was evaluated.

Figure 3A presents the results of differential pressure monitoring between the inlet and outlet of the sample. Over a 120-hour injection, the difference in pressure decreased continuously and notably. The final differential pressure was calculated to represent a permeability of 1.1 × 10^–16 m², a 26-fold increase from the initial value of 4.1 × 10^–18 m² measured with water before the experiment. With the increase in permeability, substantial leaching was observed, particularly for the elements Fe, Mg, and Si (fig. S7). Notably, the elemental concentrations in the flooding experiments were generally higher than those in the powder experiments. This increase can be partly attributed to the higher rock-to-water ratio in the flooding experiments. The amount of Fe leached was the highest among all the elements, which is consistent with the results of the powder experiments (Fig. 2A), indicating the high reactivity of Fe-bearing minerals. The Mg and Ca concentrations also attained values of 565 and 71 mg/liter, respectively, indicating the potential for CO₂ mineralization after GLDA degradation. The permeability is expected to be further improved if the flooding experiment continues, driven by the continuous dissolution of minerals.

Fig. 3. Changes in the properties of a basalt core sample during flooding with a 5 wt % GLDA-Na₄ solution with an initial pH value of 2 at 35°C and an injection flow rate of 0.03 ml/min.

(A) Changes in differential pressure during the flooding experiment. (B) X-ray computed tomography (CT) images showing the three-dimensional distribution of isolated pores, and (C) x-ray CT slice images of the basalt sample before and after the flooding experiment. BR, before reaction; AR, after reaction; aug, augite; ol, olivine.

The pore properties of the basalt core samples were evaluated using x-ray computed tomography (CT) and Molcer Plus 3-D image visualization and processing software (White Rabbit Corp., Tokyo, Japan) (38). Noteworthy changes were observed before and after the flooding experiment. The total porosity increased from 15.6 to 16.1%, whereas the fraction of isolated pores decreased from 20.1 to 10.5%. The combination of increased total porosity and increased connected pore fraction resulted in an increase in effective porosity from 12.4 to 14.4%, a 16% increment within 120 hours. Large isolated pores (>0.1 mm³) were less abundant after the injection of the GLDA solution (as illustrated by the red areas in Fig. 3B), indicating enhanced connectivity between large pores. Comparison of x-ray CT slice images before and after reaction revealed selective dissolution of Fe-bearing augite and olivine, shown as relatively small light-colored particles in Fig. 3C, consistent with the fluid chemistry analysis results. Selective dissolution of small augite and olivine particles contributed to the opening of pathways between the pores. By injecting an acidic GLDA solution, the capacity and injectivity of the CO₂ storage reservoir improved, yielding suitable conditions for the subsequent injection of CO₂ or CO₂-charged seawater.

CO₂ storage with alkaline GLDA seawater solution

To maximize CO₂ capture, storage, and mineralization efficiencies, the use of a biobased and alkaline chelating agent seawater solution for CO₂ capture and storage is attractive. We experimented with a commercial GLDA-Na₄ solution that was alkaline (pH > 12) and exhibited buffering properties. When using this GLDA solution for CO₂ absorption until the pH decreases to 8, an equivalent molar amount of CO₂ is expected to be absorbed as HCO₃^– (e.g., 20 wt % GLDA absorbs 0.65 M CO₂).

To identify suitable conditions for injecting this CO₂-charged GLDA seawater solution for CO₂ storage, batch experiments were conducted using unaltered basalt powders and GLDA-Na₄ solutions at concentrations of 1 to 20 wt %. The pH of all solutions was adjusted to 8 by CO₂ dissolution. Figure S8 summarizes the elemental leaching concentrations after 6-hour reactions. The higher leaching concentrations of Fe and Si in the experiments with higher GLDA concentrations indicate greater enhancement of mineral dissolution. Effective enhancement of mineral dissolution during CO₂ injection is still possible. However, marked decreases in the leaching concentrations of Ca and Mg were observed when the GLDA-Na₄ concentration was increased from 2 to 5 wt %, owing to the formation of Ca and Mg carbonates. Fourier transform infrared (FTIR) spectroscopy (fig. S9) and transmission electron microscopy (TEM) measurements of the powders collected after the reactions (Fig. 4A) confirmed the presence of fine particles of Ca and Mg carbonates on the surface of the basalt covered by amorphous silica. Owing to enhanced basalt dissolution in the presence of GLDA, a large amount of Si was also leached, which generated amorphous silica covering the carbonates, thus inhibiting their crystallization into larger particles (39, 40). Figure 4B presents a graphical description of the possible carbonation process during CO₂-charged seawater injection. Given that these carbonate particles were on the nanometer scale, they are not expected to have a significant impact on pore connectivity.

Fig. 4. Formation of fine carbonate particles during CO₂-charged GLDA solution injection into basaltic rocks.

(A) Transmission electron microscopy (TEM) image and energy-dispersive x-ray (EDX) measurement of basalt powders collected after a 6-hour reaction with CO₂-charged GLDA seawater solution with a GLDA-Na₄ concentration of 5 wt % and pH 8 at 35°C. Here, P1 and P2 are the points for EDX measurement. (B) Graphical description of the carbonation process during CO₂-charged GLDA solution injection into the rock.

As the solution with higher GLDA concentration captured more CO₂ and resulted in a greater enhancement of basalt dissolution, a 20 wt % GLDA-Na₄ seawater solution was used for CO₂ capture until the pH decreased to 8 and was then used for a flooding experiment. The captured CO₂ concentration was estimated to be ~0.65 M, equal to that of 20 wt % GLDA-Na₄. During injection of the alkaline CO₂-charged GLDA seawater solution, there was an initial increase in the extracted Ca concentration followed by a decrease, as well as an initial decrease in the extracted Mg concentration, followed by an increase, implying the formation of Ca and Mg carbonates (fig. S10), consistent with the findings from the powder experiments. A slight early-time increase in the differential pressure was observed owing to the formation of secondary minerals, but then it remained stable with continued injection (fig. S11A). An increase in the total porosity from 18.4 to 18.5% and a decrease in the isolated pore fraction from 6.5 to 6.2% were observed after a 120-hour injection (fig. S11B). Both appearance and disappearance of isolated pores, particularly larger pores, were observed at different locations in the rock samples, suggesting simultaneous mineral dissolution and carbonate formation. As most metals complex with GLDA, carbonate generation during CO₂ injection was not high. The use of GLDA seawater solutions enabled the efficient capture and storage of large amounts of CO₂ under stable hydraulic conditions.

CO₂ mineralization with GLDA biodegradation and overall evaluation

GLDA was characterized by a high biodegradation rate. More than 60% biodegraded to H₂O, CO₂, and ammonium in less than 60 days (33, 41). The degradation of GLDA increased the CO₂ amount in the reservoir, e.g., if a solution containing 20 wt % GLDA-Na₄ (C₉H₉NNa₄O₈) was used for injection, then CO₂ concentration is expected to increase 10-fold, from 0.65 M to 0.65 + 0.65 × 9 (number of carbon atoms in GLDA) = 6.5 M. With biodegradation of the GLDA chelating agent in several months, via microorganisms in the initial seawater or reservoir environment, metal cations were liberated, subsequently initiating carbonation with dissolved CO₂ species. When the amount of CO₂ in the fluid was greater than the availability of divalent metal ions, the remaining CO₂ existed in NaHCO₃ or (NH₄)₂CO₃ form in the fluid.

A first-order assessment of the CO₂ storage capacity was made using the pore property parameters of basalt samples from the Carbfix site in Iceland provided by Callow et al. (17). Specifically, the total porosity ranged from 10.99 to 15.23%, of which only 18.7 to 57.5% were connected porosity, resulting in a mean effective porosity of 4.8%. Assuming an enhancement in pore properties comparable to that observed in the experiment with injection of the acidic GLDA solution, the effective porosity was anticipated to increase from 4.8 to 5.7%. When filling all effective porosity with a CO₂-charged 20 wt % GLDA-Na₄ seawater solution (CO₂ concentration estimated as 0.65 M), the reservoir exhibited a CO₂ storage capacity of 1.5 Gt/km³. With complete degradation of GLDA, the concentration of CO₂ in the fluid would attain a value of 6.5 M, resulting in a reservoir capable of storing 15 Gt of CO₂/km³, two orders of magnitude higher than the value reported by Callow et al. (17) based on CO₂ capture and storage using fresh water (a mean value of 0.18 Gt of CO₂/km³). Up to 20 Gt of CO₂ annually could potentially be sequestered in a 1.3 km³ reservoir volume, which is 1/25,000 of the active rift zone in Iceland (area of 33,000 km², thickness of 1 km). The stored CO₂ originates directly or indirectly from the atmosphere, because GLDA is a biobased chelating agent in which 55.6% of the carbon is green carbon from plants or microorganisms (42). With a GLDA seawater solution, higher concentrations of CO₂ can be stored under alkaline conditions, resulting in high concentrations of bicarbonate and carbonate ions, thereby facilitating easier carbonate precipitation. Moreover, as GLDA promotes metal leaching and seawater contains Ca and Mg, there should be a notable enhancement in the CO₂ mineralization process. Complete mineralization of the injected CO₂ (including that from the biodegradation of GLDA) is also possible through reactions with metals gradually leached from basalt.

DISCUSSION

This study demonstrates that the efficiency, capacity, and safety of CO₂ capture, geological storage, and mineralization can be greatly enhanced through the effective use of biobased and biodegradable chelating agents such as GLDA and seawater. The injection of an acidic chelating agent solution is effective in improving rock pore properties and permeability. This reduces the injection pressure required and prevents localized pressure buildup, minimizing the risk of fracturing the rock and potential seismic events, as well as ensuring a more controlled injection process, thereby improving injection safety and reducing operational costs. As chelating agents promote the dissolution of various minerals, including clay minerals, this approach can be effective not only in unaltered basalt reservoirs but also in altered rocks and sandstones. The use of seawater can minimize freshwater resource consumption. We further suggest an alkaline GLDA seawater solution for substantial CO₂ capture and injection, which can also improve the storage capacity and inhibit the formation of carbonates, thus ensuring that CO₂ is transported over longer distances. The rate of biodegradation of the chelating agent can be controlled by adjusting the number of microorganisms in the injected solution, thus allowing control of the CO₂ mineralization rate.

Biodegradation of chelating agents leads to an increase in the free phases of both metals and CO₂, triggering substantial CO₂ mineralization. In cases where the rock does not contain sufficient metals to mineralize all the CO₂, we recommend the use of chelating agent solutions to treat Ca- or Mg-bearing industrial by-products, such as steel slag and fly ash, to increase the metal content before CO₂ capture and injection. This would simultaneously address both CO₂ emissions and waste disposal problems. Regarding operational costs, it is possible that the use of chelating agents may lead to increased chemical expenses. However, chelating agents are widely used across various applications, and their cost is relatively low. Using chelating agents to enhance reservoir injectivity and storage capacity can greatly reduce the costs of CCS in other respects. For instance, applying this approach can reduce the substantial expenses associated with exploration and development (e.g., injection wells construction) of CCS sites. Although it is difficult to quantitatively demonstrate the cost benefits of this approach at present, the extensive use of chelating agents like GLDA in preliminary studies, particularly in improving the well/reservoir productivity and enhancing the hydrocarbon recovery from sandstone and carbonate reservoirs, provides indirect evidence of its economic viability (43). Further studies are required to understand CO₂ storage and mineralization using chelating agents at the field scale through simulations, quantitative evaluation of CO₂ mineralization, and quantitative evaluation of secondary mineral generation during biodegradation of chelating agents.

MATERIALS AND METHODS

Materials

Unaltered basalt samples with mineralogical properties similar to those of Carbfix were obtained from Daikon Island, Japan. The main mineral compositions were plagioclase (CaAl₂Si₂O₈), augite [(Ca,Mg,Fe)₂Si₂O₆], olivine [(Mg,Fe)₂SiO₄], and glassy materials with small quantities of opaque minerals, accounting for 47.3, 34.0, 6.0, 10.0, and 2.8%, respectively. The total porosity was measured to be 14 to 20%. The permeability was 4.0 × 10^–18 to 2.0 × 10^–16 m². Altered basalts were sampled from Niigata, Japan. Their main mineral compositions were plagioclase, augite, olivine, montmorillonite [(Na,Ca)_0.3(Al,Mg)₂(Si₄O₁₀)(OH)₂·nH₂O], and glassy materials. The bulk chemical compositions of the unaltered and altered basalt samples were determined by x-ray fluorescence spectrometry (Rigaku, ZSX Primus IV). The results are presented in table S1. Two core samples of unaltered basalt with diameters of 25 mm and lengths of 25 mm (as shown in fig. S1A) were prepared for the flooding experiments. Both unaltered and altered basalt were ground and sieved to obtain the <100-μm-sized fractions for the batch dissolution experiments.

Aqueous solutions of N,N-Bis(carboxymethyl)-L-glutamate (GLDA) at various concentrations were prepared from a 40 wt % aqueous solution of GLDA-Na₄ (C₉H₉NNa₄O₈) purchased from Tokyo Chemical Industry Co. Ltd. (Tokyo, Japan). The molar concentration of the initial GLDA-Na₄ solution was calculated to be ~1.3 M. The pH of these solutions was adjusted to the desired values of 1, 2, 4, 6, 8, and 10 through the addition of a nitric acid (HNO₃) solution. Artificial seawater purchased from Tomita Pharmaceutical Co. Ltd. (Japan) was used to prepare the alkaline CO₂-charged solution. The main composition of artificial seawater is listed in table S2.

Powder experiments

To understand the dissolution characteristics of basalt in the presence of GLDA and determine the suitable conditions for injection in each step, batch dissolution experiments were conducted using both unaltered and altered basalt powders with particle sizes of less than 100 μm. In a typical run, 2.0 g of the basalt powders was mixed with 100 ml of the prepared GLDA-Na₄ solution at the desired pH and concentration in a plastic sample bottle, which was then placed in a thermostatic shaker for continuous shaking. Dissolution experiments were conducted at a constant temperature for 6 hours at a shaking speed of 150 rpm. During the experiments, a small amount of the sample (~1 ml) was intermittently sampled using a syringe via filtering with a 0.45-μm syringe filter. Five 1-ml samples were extracted at a reaction time of 0 hours (before adding the basalt powder) and 1, 2, 4, and 6 hours during each experiment. Once the reaction was completed in 6 hours, both the solution and solid components in the bottle were collected separately using a 0.45-μm membrane filter.

First, the effectiveness of acidic GLDA solution in promoting basalt dissolution was evaluated. Acidic GLDA-Na₄ solutions with concentrations adjusted by Milli-Q water to 1 to 20 wt % and pH adjusted to 1 to 10 by HNO₃ were prepared for dissolution at 20° to 60°C. Subsequently, to identify the optimal conditions for CO₂-charged GLDA seawater injection, solutions containing GLDA-Na₄ at concentrations of 1, 2, 5, 10, and 20 wt % with artificial seawater as the solvent were prepared. These solutions were used for CO₂ capture by bubbling until the pH decreased from 12 to 8. An equivalent molar amount of CO₂ with the GLDA solution was estimated to be dissolved as HCO₃^–. Therefore, the corresponding theoretical CO₂ concentrations in the 1, 2, 5, 10, and 20 wt % GLDA seawater solutions were calculated to be ~0.0325, 0.065, 0.162, 0.325, and 0.65 M, respectively.

The collected solution samples were analyzed for pH and concentrations of Mg, Al, Si, K, Ca, and Fe (including both free cations and those combined with GLDA) using inductively coupled plasma optical emission spectrometry (ICP-OES; Agilent 5100). The measurements were conducted thrice and showed good reproducibility with an error margin of 3%. The mineral composition of the basalt powders before and after reaction was measured using XRD (Multiflex, Rigaku, Japan) with Cu Kα radiation (λ = 1.54 Å) operated at 40 kV and 20 mA and with a 2θ step size of 0.02° from 3° to 50°. FTIR (PerkinElmer, Japan) and TEM (JEOL JEM-2100F) analyses of the unaltered basalt powders collected before and after reacting with the CO₂-charged GLDA seawater solution were also performed.

Flooding experiments

Chelating agent flooding experiments were conducted using unaltered basalt core samples (fig. S1). The experimental system is shown in fig. S2. Before the experiment, the initial permeabilities of the samples were measured using pure water. A Viton-sleeved sample with end plugs attached to both the inlet and outlet faces was placed horizontally in a pressure vessel maintained at 35°C using a mantle heater. The sample was then subjected to a confining pressure of 6 MPa, which represents the confining stress in CO₂ storage reservoirs, by pumping water into the vessel at a constant pressure. The initial permeabilities (k) of the samples were determined using Darcy’s law (Eq. 3) by first injecting pure water into the sample at a constant flow rate of 0.03 ml/min using a pump

$$k=\frac{Q\mathrm{\mu }L}{\mathrm{\pi }{r}^{\mathit{2}}∆P}$$ (3)

where Q is the flow rate, μ is the dynamic viscosity of the injected fluid, ΔP is the differential pressure of the fluid between the inlet and outlet faces of the sample, and r and L are the radius and length of the sample, respectively. At 35°C, the dynamic viscosity of a 5 wt % GLDA-Na₄ solution at pH 2 was measured to be 9.72 × 10⁻⁴ Pa·s, while that of a 20 wt % GLDA-Na₄ solution at pH 8 was 2.39 × 10⁻³ Pa·s. These values are 1.35 and 3.32 times the viscosity of water at the same temperature and pH values. During this process, water flowed out of the sample through a backpressure regulator adjusted to 1 MPa. After determining the initial permeability, the sample was removed from the vessel and dried at 60°C in an oven for 1 day to remove all of the water inside the sample. Subsequently, the sample was re-subjected to the same vessel for injection of the GLDA solution.

To investigate the effectiveness of using an acid GLDA solution to improve the porosity properties and permeability of unaltered basalt, an acid GLDA solution with a concentration of 5 wt % and pH adjusted to 2, based on the results of the powder experiments, was injected at a constant temperature of 35°C for ~120 hours. The experiment was conducted at a confining pressure of 6 MPa, a back pressure of 1 MPa, and a constant injection flow rate of 0.03 ml/min. These parameters were selected to ensure that the differential pressure between the two ends of the sample was within a suitable range for observation. On the basis of the calculated effective pore space (~1.5 cm³) and the injection flow rate (0.03 ml/min), the reaction time between the fluid and the rock samples was determined to be ~50 min.

For the injection of the alkaline CO₂-charged GLDA seawater solution, a GLDA-Na₄ solution with a concentration adjusted by artificial seawater to 20 wt % was used for CO₂ capture until the pH decreased to 8 and was subsequently injected. The dissolved CO₂ concentration was ~0.65 M. The flooding experiment was conducted at 35°C, with a confining pressure of 6 MPa, a back pressure of 1 MPa, and a GLDA solution injection flow rate of 0.006 ml/min, with the reaction time between the fluid and the rock samples determined to be ~350 min. The total injection period was 120 hours.

Effluent samples were collected throughout the experiment. Their pH and elemental concentrations were analyzed using ICP-OES. X-ray CT was conducted on all samples before and after the chelating agent flooding experiments at a tube voltage of 120 kV, a tube current of 150 μA, and a voxel size of 12 μm by 12 μm by 12 μm. Subsequently, the Molcer Plus 3-D image visualization and processing software (White Rabbit Corp., Tokyo, Japan) (38) was used to visualize the distributions of large pores with sizes similar to or larger than the voxel size. X-ray CT slice images at the same locations at the basalt sample were obtained before and after the flooding experiments for comparison to investigate the mineral dissolution behavior.

Supplementary Materials

This PDF file includes:

Tables S1 and S2

Figs. S1 to S11