A comparative investigation of using microbial- and CO 2 -induced carbonate minerals in sustainable soil Improvement

The structure and shape of carbonate crystals formed by microbial activity and carbon dioxide reaction were investigated in this work. The mineral carbonates treated sandy soil was subsequently determined using unconned compression tests (UCS). Sporoscarcina pasteurii bacteria were used to produce an aqueous solution of free carbonate ions (CO 32- ) under laboratory circumstances called microbial-induced carbonate precipitation (MICP). In CO 2 - induced carbonate precipitation, carbon dioxide was added to a sodium hydroxide solution to form free carbonate ions (CO 32- ) (CICP). Different carbonate mineral compositions were then provided by adding Fe 2+ , Mg 2+ , and Ca 2+ ions to carbonate ions (CO 32- ). In the MICP and CICP procedures, the results of scanning electron microscopy (SEM) and X-ray powder diffraction (XRD) revealed a distinct morphology of carbonate minerals. Vaterite (CaCO 3 ), siderite (FeCO 3 ), magnesium carbonate hydrate, or Nesquehonite (MgCO 3 (H 2 O) 3 ) and dolomite (CaMg(CO 3 ) 2 ) were produced in MICP. Calcite (CaCO 3 ), siderite (FeCO 3 ), magnesium carbonate hydrate or Nesquehonite (MgCO 3 (H 2 O) 3 ), and high-Mg calcite (CaMg(CO 3 ) 2 ) were produced in CICP. The results of UCS showed that siderite and high-Mg calcite /dolomite had more eciency in soil strength. The lowest value of strength was related to magnesium carbonate hydrate treated soils. Mineral-treated soils in CICP showed a slightly higher UCS strength than MICP, which could be attributable to greater crystal particle sizes and particle interlocking.


Introduction
Sustainable precipitation of carbonate minerals may provide green means of mitigating some geotechnical challenges associated with soils. Microbial-induced carbonate precipitation (MICP) and CO 2 -induced carbonate precipitation (CICP) can induce interparticle cementation and mineral precipitation in soil pore space to address geotechnical problems such as soil erosion and slope instability (Keykha et al. 2017, Keykha et al. 2018a, Romiani et al. 2021).
Biomineralization is the chemical modi cation of an environment that comes about within the precipitation of minerals by microbial activity which may be extracellular precipitation. The carbonate precipitation as a result of extracellular mineralization is a well-known phenomenon in all classes of living organisms. (Lowenstam and Weiner 1989). While precipitation of carbonate as an inorganic chemical reaction is di cult in natural environments, bacteria can cause mineral precipitation in microenvironments by (1)  The precipitation of carbonate minerals by CO 2 transformation is considered a promising option for carbon capture and storage (CCS) since (i) CO 2 can be accumulated permanently and (ii) industrial disposals (i.e. cement and lime kiln dust, steel, and stainless-steel slags, etc.) can be recycled into carbonate materials by controlling polymorphs and properties of the mineral carbonates (Chang et al. 2017). CO 2 may be stored and converted to carbonate minerals for a long time such as CaCO 3 and MgCO 3 because they are stable forms of carbon (Smit et al. 2014). In addition, mineral carbonation also has unique advantages because allows the utilization of industrial wastes (with hazardous effects), and given this providing an appropriate method for disposal or recycling is a signi cant environmental issue.
According to a recent study, CICP improves soil characteristics by precipitating carbonate minerals between soil particles in an environmentally friendly method. (Romiani et al. 2021). It entails capturing carbon dioxide (carbon sink) and producing carbonate minerals as a result (e.g. siderite, magnesite, and calcite). The carbonate minerals are environmentally safe and have the potential to physically and chemically bind soil particles. The developed CICP and their application in soil improvement is a promising result for new efforts because carbon dioxide is a contributing factor to the major global warming projected in future decades.
This research aims to evaluate the structure and morphology of carbonate crystals that form as a result of microbial activity and CO 2 reaction, as well as to compare the effects of minerals on soil treatment.

Microbial induced-ammonium free carbonate production
The Sporsarcina pasteurii (PTCC 1645) a urease-producing bacterium was used in this study. S. pasteurii was harvested using NH4 -YE, a medium made up of 20.0 g yeast extract in 100 ml distilled water, 10.0 g (NH 4 ) 2 SO 4 in 100 ml distilled water, and 0.13 mol/lit tris buffer (pH = 9.0). After that, the materials were autoclaved separately (121°C for 15 minutes), cooled, and then blended in an Erlenmeyer ask. The bacteria medium was cultured at a pH of 9.0, which is ideal for the growth of S. pasteurii. The bacteria were thereafter moved to the medium broth in an incubator (30°C) and shaken at 200 rpm under aerobic conditions for 2-3 days. The bacterium was extracted from an overnight culture by centrifugation (Sartoriun AG, sigma 3-18 K, Germany) at 8000 g for 10 min with an optical density of 1.6 at 600 nm (Labomed UVD 2950). During cultivation, cell concentration was obtained about 10 8 cells ml −1 by measuring the OD600 using a spectrophotometer (Thermo Electron Corporation, Madison, WI).
To prepare a free carbonate solution, 100 ml urea (1M) was added to 100 ml of the culture bacteria. The bacteria were exposed to the urea and produced carbonate ions (CO 3

CO 2 induced-carbonate production
To produce free carbonate ions (CO 3 2− ) in this study, a 2.5 kg industrial cylinder of CO 2 gas with relative density and solubility in water of 1.5 and 2000 mg/l was used respectively. Sodium hydroxide (NaOH) with molar masses of 39.99 g/mol and a density of 2.13 g/cm3 (as powder) was also used.
Free carbonate ions (CO 3 2− ) were produced through laboratory-scale carbon capture technology (Inoue et al. 2010). To prepare free carbonate ions (CO 3 2− ), the CO 2 gas was injected into a solution of sodium hydroxide (2 molars) at a 5 ml/min rate for a period of 72 h. In this process, when CO 2 reacts with water, in the presence of sodium hydroxide (NaOH), it forms Na 2 CO 3 according to equation 2. The concentration of carbonate ions was eventually measured by Thomas Combination Carbonate Ion-Selective Electrode (ISE) at the end of the process.

Carbonate minerals production
A schematic diagram of the carbonate mineral precipitation techniques is shown in Figure 1. Ferrous sulfate (FeSO 4 ), magnesium sulfate (MgSO 4 ), and calcium chloride (CaCl 2 ) were used to extract the ion components of Fe 2+ , Mg 2+ , and Ca 2+ , respectively. The chemical characteristics of the materials utilized are listed in Table 1.

Scanning electron microscope (SEM)
The size, shape, and crystalline formation of the mineral crystals were examined using scanning electron microscopy (SEM) on the various mineral products from MICP and CICP. The specimens were gold-coated to make them conductive. For these trials, the accelerating voltages were chosen between 10 and 15 kV.

Green soil improvement using carbonate minerals
In this experiment, clean silica Firoozkuh sand was employed. The sand grain size distribution curve is shown in Figure 2. The sand had a speci c gravity of 2.65, a uniformity coe cient of 2.75, and maximum and minimum void ratios of 0.91 and 0.57, respectively.
The soil specimens were prepared by mixing condensed carbonate minerals (i.e., siderite, magnesite, calcite, and calcite/dolomite). To obtain condensed carbonate minerals, the solutions were centrifuged at 8000 g for 10 min and were eventually ltered by a lter paper (Keykha et al. 2021). The carbonate minerals (with a water content of about 100%) were used as soil stabilizing agents in the uncon ned compression tests.

Uncon ned compressive strength test
The uncon ned compression test was used for obtaining the approximate strength of the soil specimens.
To perform the uncon ned compression tests, the pre-weighed amount of sand was mixed with condensed carbonate minerals and placed in the two-piece split mold by slightly tamping to obtain a homogenous specimen of 50mm diameter and 100 mm length by dry density equal to about 19.2 kN/m 3 . Before the uncon ned compressive strength tests, the specimens were allowed to cure at room temperature for 7 days. Table 2 shows a summary of the testing program and speci cations of specimens. The uncon ned compression tests were repeated three times for each type of mineral and production method. The accuracy analysis was performed for parallel tests. The uncon ned compressive strength (UCS) was taken as the peak stress with the corresponding axial strain at failure in the stress-strain curve (ASTM D2166/D2166M-16, 2016).
The carbonate mineral percentage of the samples was determined by using an acid washing technique (Keykha et al. 2018a). The soil samples were soaked into an acid solution (HCl 5 M). The dissolved carbonates and acid wash solution were ltered through a lter paper several times, allowing the dissolved salts to be rinsed from the soil while the soil was retained. Before and after the acid washing technique, the oven-dried mass of the soil samples across the specimen was measured. The mass of carbonate minerals was calculated as the difference between the two observed masses. Table 2 shows the results that were achieved.

Induced carbonate production (CO 3 −2 )
In this study, the free carbonate ions (CO 3 2− ) produced by microbial-induced carbonate (MICP) and CO 2induced carbonate (CICP) were compared in Figure 3. As it can be seen from the gure, a value of 42.1 and 34.8 g/lit were obtained for CICP and MICP methods, respectively.
The effect of Zeolite treatment on ammonium concentration for MICP to generate an ammonium-free carbonate solution is shown in Figure 4. The ammonium content of the solution was initially around 1.39g/lit, as seen in the gure. After 15 percent Zeolite treatment, the ammonium concentration dropped to an average of 3.4 105 g/lit. The concentration meets the requirement speci ed by some drinking-water standards (e.g. lower than 5 ×10 −4 g/lit, Australian Drinking Water Guidelines and Guidelines for Drinkingwater Quality Management for New Zealand, 2016).

Carbonate mineral production
Carbonate minerals precipitated by MICP and CICP techniques are shown in Figure 5. The results demonstrate that the carbonate minerals precipitated by CICP had higher values than those precipitated by MICP. The most carbonate minerals content were  Figure 6 illustrates the crystal morphology of CaCO 3 which was produced by MICP and CICP methods. As it can be seen, the CaCO 3 crystals have been grown spherical with diameters of 2-10 µm in MICP (Fig. 6a), while crystals were in rhombohedral shape and smooth with diameters of 2-6 µm in CICP (Fig. 6b). Figure  7 shows EDS and XRD analysis of CaCO 3 deposition. The EDS layered image presents the distribution of C, O, and Ca elements in the test area. The result of XRD justi ed vaterite minerals (100%) (Fig. 7a) and calcite (100%) (Fig. 7b) in MICP and CICP, respectively. In CICP, deposition and morphology of CaCO 3 are dependent on additives such as CO 2 and Ca 2+ which are precipitated in three anhydrous polymorphic modi cations (calcite, aragonite, and vaterite) and mostly is calcite (Kralj et al. 2004). MICP is largely dependent on metabolic processes by ureolytic bacteria that in uence crystal growth ( Figure8 shows the crystal morphology of FeCO 3 which was produced by MICP and CICP methods. There was a similar crystal shape (spherical) for both of them. The particle size of FeCO 3 in MICP was with diameters of 0.5-3 µm while the particle size was bigger (2-6 µm) in CICP. Figure 9 shows EDS and XRD analysis of FeCO 3 samples. The EDS layered image presents the distribution of C, O, Mg, and Fe elements in the test area. The result of XRD justi ed siderite minerals (100%) for MICP and CICP. Less information is available concerning the mineralization or morphogenesis of iron carbonates (siderite). This is probably because natural siderite is often associated with other coexisting elements (e.g., Mn, Mg, and Ca), and contains a small amount of hematite (Fe 2 O 3 ) due to partial oxidation in natural air (French 1971;Isambert et al. 2003). The biological aspects of iron biomineralization have been investigated by studying iron-bearing biominerals, owing to their signi cance in identifying microbe-sediment-water interactions, as well as mineral and biogenic origins (Frankel and Bazylinski 2003;Dong et al. 2009). Figure 10 shows the crystal morphology of calcium magnesium carbonate which is produced by a mixture of magnesium and calcium ions with free carbonate ions (CO 3 2− ). As it can be seen from the gure, there was a botryoidal crystal shape of calcium magnesium carbonate in the MICP method (Fig.  10a) while a ower-like shape was observed in the CICP method (Fig. 10b). There was the main difference between the shape and size of crystals in both methods. The calcium magnesium carbonate was precipitated with diameters of 5-16 µm and 2-5 µm in MICP and CICP, respectively. Minerals deposited in a larger size in MICP than in CICP, as can be seen plainly. The EDS layered image presents the distribution of C, O, Mg, and Ca elements in the test area. XRD analysis of samples showed Ca-Mg (CO 3 ) deposition while was as dolomite (40%) and calcite (60%) in MICP (Fig. 11a) and high-Mg calcite (61%) and magnesium carbonate hydrate (Nesquehonite) (39%) in CICP (Fig. 11b), respectively. Bacteria can adsorb Mg 2+ on their membranes during induce dolomite formation, while calcite precipitation is induced by preferential adsorption of Ca 2+ . González-Muñoz et al. (2010) found that Mg probably plays a key role in the development of the morphologies of the precipitates since these morphologies had never been observed in the absence of Mg. CaCO 3 mineral that has precipitated from seawater changes from calcite to aragonite was experimentally determined as a function of temperature and additives. Results indicate precipitation of high-Mg calcite and dolomite is largely dependent on Mg: Ca ratio over a relatively small temperature range in an aqueous solution (Morse et al. 1997). In conclusion, the morphology, type, and size of calcium magnesium carbonate minerals are in uenced by the chemical and biochemical environment in CICP and MICP. Figure 12 illustrates the crystal morphology of MgCO 3 which was produced by MICP and CICP methods.

Carbonate mineral characterization
As it can be seen, the MgCO 3 crystals have been grown in the radial needle ( Fig. 12 (a, c)) and radial blades shape (Fig. 12 (b, d)) by MICP and CICP, respectively. There was the main difference between the sizes of crystals in both methods. The thickness of radial needle crystals was diameters of 1-10 µm in MICP, while it was precipitated in a greater thickness of about 5-20 µm as radial blades in CICP. It was clearly shown that the potential of CICP to precipitate MgCO 3 is more than MICP. Figure 13  The mean size distribution of minerals was measured by SEM photos. Figure 14 shows the average size of carbonate minerals produced by MICP and CICP methods. As can be seen from the results, the average size of crystals was larger in CICP. Although there was only a small difference in calcium magnesium carbonate precipitation in MICP compared to CICP. The largest crystals in CICP, however, were for magnesium carbonate hydrate (Nesquehonite). In this work, CICP demonstrated a great capacity for precipitating carbonate minerals with larger sizes. The minerals that develop during biomineralization are frequently characterized by low crystallinity, structural well-ordering, and a limited size range. (Frankel and Bazylinski 2003). Under the same environmental conditions, biologically induced mineralization is equal to inorganic mineralization, and the minerals are expected to contain crystal-chemical characteristics that are indistinguishable from minerals formed by inorganic chemical reactions. However, cell walls of bacterial surfaces including bio lms, dormant spores, and slim sheaths can act as important sites for the adsorption of ions for growth and nucleation minerals (Bäuerlein 2003).
A summary of the characterization of carbonate minerals produced by MICP and CICP techniques is shown in Table 3. In both procedures, all carbonate minerals were precipitated in various sizes, shapes, and types. 3.5 Uncon ned compressive strength Figure 15 shows the results of the UCS tests carried out on carbonate minerals-treated soil. As can be seen from the gure, the treated specimens with condensed Ca-Mg(CO 3 ) and FeCO 3 minerals, produced by the MICP method gained about 58.5 kPa and 63.0 kPa UCS strength respectively (Fig. 15) marginally stronger than MICP, likely due to larger crystal particle sizes or particle interlocking. Figure 1 Schematic diagram of carbonate mineral production Page 18/27    The average size of precipitated carbonate minerals by MICP and CICP methods