Raw materials

Chrysotile samples were taken from the Mangya asbestos mine in Qinghai Province, China. The samples were crushed and hand-picked under a binocular microscope to remove magnetite and other clearly identifiable impurities. The purified samples were examined using light microscopy (LM), X-ray diffraction (XRD) analysis and scanning electron microscopy (SEM) to ensure that the samples were >98% pure. The fibrous samples were white, filamentous, and pliable, with lengths ranging from a few millimeters to several centimeters. The lizardite samples were obtained from the Beiwagou Cuiyu jade mine, Xiuyan County, Liaoning Province, China. The samples were crushed, ground, passed through a 200-mesh sieve and rinsed ultrasonically three times with anhydrous ethanol to remove impurities.

Material characterization

Specific surface and porosity analyzer Brunauer - Emmett - Teller (BET)

The specific surface area of the samples was analyzed and tested using a McASAP 2460 instrument (Norcross, GA, USA). X-ray diffraction (XRD) was used to determine the mineralogical changes in the samples before and after the mineral–water interfacial reaction. XRD data were obtained using a D/max-IIIA diffractometer manufactured by Rigaku (Tokyo, Japan). A copper target with a sum-tube voltage of 40 kV and a sum-tube current of 40 mA accompanied in this study by a slit system DS1/2(°). The powdered samples were analyzed between 3 and 80°2θ, with a step size of 0.02°2θ, and the scan rate of 2 s per step. FTIR was used to determine the changes in crystal chemistry before and after the mineral–water interfacial reaction. The FTIR spectra were measured using a Frontier spectrometer from Perkin Elmer Instruments (Suzhou, China), using the KBr pressed pellet technique. For this, a 0.9 mg aliquot of a dried subsample was mixed with 80.0 mg of oven-dried, spectroscopic-grade KBr salt (refractive index 1.559, particle diameter 5–20 μm), and the mixture was then ground for 1 min. The mixture was pressed in a die with 10 t of pressure applied for 1 min to form a disc or pellet. The spectra were measured at room temperature, in a test range of 4000–400 cm^–1 and 64 scans. The cation content (Si, Mg, Fe) in solution was measured by an inductively coupled plasma emission spectrometer (ThermoiCAP 6500, Waltham, MA, USA). The ²⁹Si magic angle spinning nuclear magnetic resonance (MAS NMR) experiments were performed on an Agilent 600M spectrometer (Santa Clara, CA, USA) with a ²⁹Si single pulse magic angle spin resonance frequency of 199.13 MHz, a magic angle spin speed of 8 kHz, and a π/2 pulse width of 4 μs. The chemical shifts of ²⁹Si were calibrated with reference to the TMS solution. Scanning electron microscopy (SEM) was conducted to study the micro-morphology of the samples, using a ZEISS Supra 55 Sapphire (Jena, Thuringen, Germany) field emission scanning electron microscope. The samples were spread onto conductive carbon tape, and the surface was carbon-sprayed before being placed in the sample chamber, vacuumed and observed at an accelerating voltage of 30 kV. Transmission electron microscopy (TEM) was carried out on an American-FEI-Talos F200S (Hillsboro, OR, USA) microscope, accelerated at 200 kV. The chrysotile and lizardite samples were sonicated and dispersed in ethanol for 10 min, then the sample suspension drops were placed on a 200-mesh carbon-coated Cu grid for at least 10 min before being transferred to the microscope.

The average rate of dissolution of each ion in the mineral can be calculated from the following equation:

(1) $$ {r}_i(t)=\frac{C_{\mathrm{i}}\times {m}_1}{t\times {m}_0\times SSA\times {\eta}_i}, $$

where C _i is the concentration of solute i (Si, Mg, Fe; mmol kg^–1) in the sample recovered at time t; m ₁ is the mass of the sample solution before sampling (in kg); t is the sample reaction time (s); m ₀ is the initial mass of the mineral (g); SSA (m² g^–1) is the specific surface area of the mineral; and η_i is the stoichiometric coefficient of element i in the mineral.

Experimental procedure

At room temperature, 1 g of chrysotile and 1 g of lizardite were each placed in separate 150 mL conical flasks (there were nine flasks for chrystotile and nine for lizardite, for each of the nine reaction times). Then, 100 mL of a 1.0 mol L^–1 sulfuric acid solution was added to each flask, at a solid-to-liquid ratio of 1:100. The reaction times were 1, 2, 4, 12, 24, 72, 120, 168, or 216 h. In each case, the conical flask was placed in a shaker at a constant temperature, and the reaction was allowed to proceed fully. The reacted suspension was filtered to obtain the solid product and washed three times with distilled water until the pH of the filtrate was neutral. The solid product was dried in an oven at 60°C and set aside in a desiccator. The filtrate was collected in a 50 mL volumetric flask and left for subsequent testing. The sample labels were KS-T, where K is the mineral sample number, i.e., chrysotile is XS and lizardite is LS, S is the sulfuric acid medium, and T is the reaction time. To explore the commonalities and differences in the dissolution processes of chrysotile and lizardite, XRD, FTIR, inductively coupled plasma mass spectrometry (ICP-MS), and other analytical techniques were used. The mineral–water interface reaction rates of chrysotile and lizardite in sulfuric acid solution were compared by calculating the dissolution concentrations and dissolution rates of Si, Mg, and Fe in chrysotile and lizardite.

Process of dissolution of Si, Mg, and Fe from chrysotile and lizardite in sulfuric acid solutions

Mineral–water interfacial reactions are related to the formation and transfer of charges which occur when substances on the mineral surface break bonds with the lattice, forming ions in the solution (Knauss and Wolery, Reference Knauss and Wolery1988). Table 2 shows the ion dissolution content in the water interface reaction process of chrysotile and lizardite, and the ion dissolution concentration and dissolution rate are obtained by calculation. In sulfuric acid, the concentrations of Mg and Fe, from the chrysotile and lizardite sample solution, consistently exhibited an increasing trend as the reaction time increased (Fig. 2). When chrysotile was reacted in sulfuric acid for 216 h, the dissolved concentration of Mg was 43.16 mmol L^–1, and the dissolved concentration of Fe was 0.78 mmol L^–1. Additionally, when lizardite was reacted for 216 h, the dissolved concentration of Mg was 71.63 mmol L^–1, accompanied by a dissolved concentration of 2.74 mmol L^–1 for Fe. Interestingly, the dissolved concentration of Si peaked at 72 and 120 h for chrysotile and lizardite, respectively. Dissolved Si then decreases due to the partial adsorption of Si by the pores of the amorphous SiO₂ and the attachment of Si to the surface of the crystal structure for renucleation. Chrysotile and lizardite first exhibited the dissolution of Mg from the octahedral sheets in the mineral–water interfacial reaction. Si in the tetrahedral sheets then dissolved (Thom et al., Reference Thom, Dipple, Power and Harrison2013), followed by Fe. For lizardite the concentration of dissolved Fe exceeded that of Si at the late stage of the mineral–water interfacial reaction, and the dissolution concentration of each ion was greaterf than that of the chrysotile for the same reaction times. Regarding the ion dissolution concentrations, Mg leached easily from the mineral crystals in acidic media, and the dissolution concentration was greatest. The greater bond energy of Si–O compared with Mg–O led to easier bond breakage of the latter, and the complexation of Mg with the sulfuric acid solution, which promoted further dissolution of the octahedral sheets (Rozalen and Huertas, Reference Rozalen and Huertas2013).

Table 2. Elemental analysis of the liquid phase at the water interface of chrysotile and lizardite (mg)

XSS and LSS are defined in the experimental procedure section.

Figure 2. Variation in the dissolved concentration of ions in serpentine in sulfuric acid after various reaction times: (a) chrysotile; (b) lizardite.

In mineralogy, the macroscopic ion dissolution rate is usually related to the normalized reaction surface area. Based on the mineral, different models for normalizing the reaction surface area have been proposed: (1) the reaction surface area is usually assumed to be either the Brunauer–Emmett–Teller (BET) surface area of the mineral or proportional to the BET surface area (Ganor et al., Reference Ganor, Mogollón and Lasaga1999), and (2) the reaction surface area is confined to certain crystalline surfaces or is mainly controlled by surface defects. This work uses the BET surface area as the mineral–water interfacial reaction surface area. The measured BET specific surface area was 13.8 m² g^–1 for chrysotile and 17.8 m² g^–1 for lizardite. Chrysotile and lizardite react at the mineral–water interface in sulfuric acid, and the average dissolution rate of each ion continues to decrease as the reaction time increases, accompanied by an average dissolution rate: Mg>Fe>Si (Fig. 3). During the first 72 h of the reaction, the ions dissolved rapidly, and the mineral–water interface reaction took place vigorously. Over 72–216 h, the ion dissolution rate leveled off. In the presence of sulfuric acid, the dissolution concentrations of Mg, Si, and Fe and the average dissolution rates of lizardite were larger than those of chrysotile.

Figure 3. Variation in the rate of dissolution of ions from serpentine in sulfuric acid after different reaction times: (a) chrysotile; (b) lizardite.

Structural changes of chrysotile and lizardite during reaction in sulfuric acid solutions

The XRD results for chrysotile demonstrate three diffraction peaks at d ₀₀₂ = 7.37 Å, d ₀₀₄ = 3.67 Å, and d ₂₀₂ = 2.45 Å (Fig. 4a). Lizardite is the main phase in the original sample, showing diffraction peaks at d ₀₀₁ = 7.29 Å, d ₁₁₀ = 4.76 Å, and d ₀₀₂ = 3.64 Å (Fig. 4b). A small amount of chlorite was detected, with diffraction peaks at d ₀₀₁ = 14.1 Å and d ₀₀₄ = 3.67 Å. On the one hand, the intensities of the characteristic peaks of both chrysotile and lizardite decrease with increasing mineral–water interfacial reaction time in sulfuric acid. This means that the octahedral sheets of both minerals are destroyed and Mg is gradually dissolved. On the other hand, the diffraction peak intensity of lizardite decreases more rapidly than that of chrysotile, suggesting that lizardite undergoes structural changes earlier during the reaction.

Figure 4. XRD results of serpentine products after various reaction times in sulfuric acid: (a) chrysotile; (b) lizardite.

The FTIR results demonstrate that the intensity of the characteristic peaks of chrysotile (1071, 1031, and 955 cm^–1) and lizardite (1079 and 956 cm^–1) decreases with reaction time (Fig. 5). This indicates a disruption of the mineral structure and a gradual change to an amorphous SiO₂ structure (Farmer, Reference Farmer1974), leading to amorphization, which is related to the emergence of a new peak of lizardite at 830 cm^–1. The shift of the band at 1079 cm⁻¹ to 1091 cm⁻¹ (Fig. 5b) is associated with the vibrations of n(Si-O-Si) in amorphous silica (Rozalen and Huertas, Reference Rozalen and Huertas2013). The peak at 1031 cm^–1 is attributed to the Si-O-Si vibration within the crystal perpendicular to the direction of the structure layer. The gradual disappearance of the peak strength proves that the environment of Si has changed substantially after the mineral–water interface reaction (Pentrák et al., Reference Pentrák, Madejová and Komadel2010). Moreover, the increase of the band intensity at 3432 cm⁻¹ (chrysotile) and 3448 cm⁻¹ (lizardite) confirms that the siliceous by-product is hydrated (Rozalen and Huertas, Reference Rozalen and Huertas2013; Lacinska et al., Reference Lacinska, Styles, Bateman, Wagner, Hall, Gowing and Brown2016; Beglaryan et al., Reference Beglaryan, Isahakyan, Zulumyan, Melikyan and Terzyan2023). The intensity of the external Mg–OH bond (3690 cm⁻¹) stretching vibration decreases more rapidly than that of the internal Mg–OH bond (3644 and 3646 cm⁻¹) following the mineral–water interfacial reaction between the chrysotile and lizardite and the sulfuric acid solution, suggesting that the exposed external Mg is more susceptible to dissolution in acidic media (Liu et al., Reference Liu, Peng, Liu, Zhang and Wang2022). Moreover, the internal Mg-OH in chrysotile was still present at 120 h, while that of lizardite disappeared at 72 h. This indicates that the mineral–water interfacial reaction rate of lizardite is faster than that of chrysotile under the same conditions, which corresponds to the XRD data. The gradual disappearance of the characteristic peaks at 605 cm⁻¹ for fibrous chrysotile and at 610 cm⁻¹ for lizardite also confirms the gradual dissolution of Mg from the structure (Rozalen and Huertas, Reference Rozalen and Huertas2013).

Figure 5. FTIR results of serpentine products after various reaction times in sulfuric acid: (a) chrysotile; (b) lizardite.

The ²⁹Si MAS NMR results show that the main peaks of the Si spectra of chrysotile and lizardite are located at –95.88 and –93.99 ppm, respectively, attributed to the serpentine Q³ type signal (Fig. 6) (Balucan, Reference Balucan2012). After 4 h of reaction time in the sulfuric acid medium for both minerals, the intensity of the main peak of chrysotile decreased (Fig. 6a), and no new peaks appeared. Lizardite showed a decrease in the intensity of the main peak (Fig. 6b), and a new peak at –82 ppm, which belongs to the chemical shift of ²⁹Si Q² (Magi et al., Reference Magi, Lippmaa, Samoson, Engelhardt and Grimmer1984). This implies that lizardite, compared with chrysotile, undergoes a chemical shift from the Q³ structure to the Q² structures before the chemical change under sulfuric acid. This is consistent with the experimental results from ICP and XRD. Lizardite is more aggressively attacked by H⁺, preferentially undergoing structural changes. Lizardite undergoes a chemical shift from the Q³ to the Q² structure, probably due to the dissolution of Si leading to the breakage of the bridging oxygen bonds linking SiO₂ in the tetrahedra. Thus, lizardite is more aggressively attacked by H⁺, preferentially undergoing structural changes.

Figure 6. ²⁹Si MAS NMR results of serpentine products after 4 h of reaction in sulfuric acid: (a) chrysotile; (b) lizardite.

Micromorphological changes of chrysotile and lizardite during reaction in sulfuric acid solutions

SEM results show that chrysotile (Fig. 7a) has an obvious tubular structure with a smooth surface. Lizardite has a complete lamellar structure with some small mineral particles attached to the surface (Fig. 7b). When chrysotile reacts with sulfuric acid at the mineral–water interface for 24 h, the surface smoothens and the tubular structure gradually disappears and becomes adherent to itself (Fig. 7c). After 24 h of the reaction of lizardite with sulfuric acid, the mineral is corroded by H⁺, destroying part of the surface edge lamellar structure (Fig. 7d). Thus, the process of structural destruction of lizardite in acidic media proceeds from the surface to the interior. However, the destruction occurs preferentially at locations where the structure is weak or has excess surface energy and does not occur uniformly on all exposed surfaces. After 120 h of reaction between the two minerals in the sulfuric acid solution (He et al., Reference He, Cao and Duan2019), the edges, ends, and outer surfaces of the chrysotile were corroded (Fig. 7e). The corrosion can be attributed to the fracture of the external octahedral sheet M- (O, OH), resulting in bending and fracture. However, the fibrous tubular morphological features of chrysotile can still be seen. In contrast, lizardite was only partially laminated after 120 h of reaction, with most of the lamellar structure destroyed into amorphous SiO₂ aggregates (Fig. 7f).

Figure 7. SEM images of serpentine products after different reaction times in sulfuric acid: (a) chrysotile; (b) lizardite; (c) chrysotile reacted for 24 h; (d) lizardite reacted for 24 h; (e) chrysotile reacted for 120 h; and (f) lizardite reacted for 120 h.

Chrysotile has a single inner diameter of 7 nm and is structurally intact (Fig. 8a). The mineral–water interfacial reaction between chrysotile and sulfuric acid leads to the corrosion and dissolution of the edges of the fibrous tubes and the breakage and fragmentation of the long fibrous tubes, resulting in the formation of short tubes (Fig. 8c). Additionally, the inner wall of the fiber tube is also damaged by H⁺ attack, leading to an increase in the inner diameter from 7 to 12 nm. Lizardite can clearly be seen as several lattice streaks in different directions, indicating a good crystallinity of lizardite (Fig. 8b). The lattice streaks become faint, indicating a decrease in lizardite crystallinity following a mineral–water interfacial reaction with sulfuric acid (Fig. 8d).

Figure 8. TEM images of serpentine products after 4 h of reaction in sulfuric acid: (a) chrysotile; (b) lizardite; (c) chrysotile reacted for 4 h; and (d) lizardite reacted for 4 h.

In summary, the mechanism of water interface reaction of chrysotile and lizardite in sulfuric acid medium is analyzed. The mineral–water interfacial reaction of lizardite in acidic media is stronger and faster than that of chrysotile. This can be attributed to the different structure of lizardite compared with chrysotile. Chrysotile has a tubular structure where the hydroxyl group in the outer octahedral sheet reacts with H⁺ and the subsequent reaction is blocked by the inner silica-oxygen tetrahedral sheet blocking the subsequent reaction. The greater bond energy of Si–O compared with Mg- (O, OH) led to the dissolution rate of the silica-oxygen tetrahedral sheet being slower than that of the magnesium-oxygen octahedral sheet in acidic media. It is more stable, resulting in a weaker and less rapid reaction (Fig. 9a). Lizardite has a flat layer structure with small particles. After H⁺ from the acidic medium enters the interlayer domain, it can react with the (OH)^– group on each octahedral sheet, and the Mg in the mineral structure can dissolve quickly. The dissolution of Mg is not affected by the mineral structure. This results in the difference of dissolution between chrysotile and lizardite. The Si–O bond in the tetrahedral sheet is stronger than the Mg–(O, OH) bonds in the octahedral sheet. Therefore, Mg is dissolved preferentially (Fig. 9b).

Figure 9. Schematic diagram of the serpentine structure attacked by H⁺: (a) chrysotile; (b) lizardite.