4.1. Dolomite Exposure to Pb2+
The initial solution configured for this experiment contained a lead ion concentration of 1 mg/L, which can be converted to 4.86 × 10
−6 mol/L. According to the K
sp value of lead hydroxide, when the solution reaches equilibrium (i.e., the precipitation limit) the OH
− concentration is 2.94 × 10
−15 = 5.42 × 10
−8 mol/L, at which point the H
+ concentration in the solution can be calculated from kw to be 1.85 × 10
−7, which means that the solution pH is 6.73. In other words, when the solution pH is greater than 6.73, the solution is supersaturated concerning lead hydroxide and vice versa. The initial lead solution in this experiment was configured by dissolving lead chloride in hot water, and its initial pH was determined to be 6.33, which is less than 6.73, so the initial solution was not saturated for lead hydroxide. The initial solution pH is weakly acidic and the solid phase dolomite in this experiment should more readily react with the acidic solution. The reaction equation is as follows:
Disregarding how the carbonic acid equilibrium of the solution causes the dissolved CO32− to change, 1 mol of H+ in the solution is always able to dissolve 1 mol of CO32− from the dolomite into the solution. At this point, the concentration of carbonate in the solution at equilibrium can be calculated from the Ksp of lead carbonate and the initial lead concentration in the solution to be 3.02 × 10−8 mol/L. There is no carbonate present in the initial solution, so the carbonate in the solution comes from the reaction between H+ and dolomite, the volume of the solution remains unchanged during the reaction, and the volume of the solution is assumed to be 1L for the convenience of calculation. Then, according to Equation (1), the consumption of 3.02 × 10−8 mol of H+ in solution reacts with dolomite. Only then can the CO32− reach 3.02 × 10−8 mol. The initial solution pH is 6.33 and, after consuming 3.02 × 10−8 mol H+, the H+ concentration becomes 4.38 × 10−7 and the pH becomes 6.35. 6.35 < 6.73, at which point the solution is saturated with lead carbonate but not with lead hydroxide. As the dolomite dissolves, the solution is reduced in H+ and increased in CO32−, which means that the solution is oversaturated with lead carbonate and still not saturated with lead hydroxide during the reaction from pH = 6.35 to pH = 6.73. Therefore, this paper argues that the initial lead solution first reaches lead carbonate supersaturation as the dolomite dissolves, a state that facilitates the generation of lead carbonate on the dolomite surface. The solution is then supersaturated with lead hydroxide only as the dolomite dissolves. If a lead carbonate precipitate is generated first, reducing the concentration of lead ions in the solution, the solution will likely remain unsaturated for lead hydroxide. Therefore, the lead on the surface of the dolomite should primarily become lead carbonate rather than lead hydroxide.
Regarding the possible presence of sulfate in the solution, we believe that the only source of sulfate present in the solution is the dissolution of the barite component of the solid phase impurities. Although the Ksp of barite is much greater than that of dolomite but under weakly acidic conditions, H+ promotes the dissolution reaction of dolomite, which rapidly and releases large quantities of carbonate into the solution and consumes H+. The Ksp of PbSO4 is 1.82 × 10−8, while that of lead carbonate is 1.46 × 10−13, which shows that the Ksp of lead carbonate is much smaller than that of barium sulfate; therefore, in this work we assume that the solution is first supersaturated with lead carbonate rather than lead sulfate under the premise that the source of carbonate in the solution is sufficient. However, we cannot exclude the possible generation of PbSO4, which is reflected in the results of the XPS experiments, but we still believe that the dolomite surface production is mainly PbCO3.
The solution in the reaction system was measured using inductively coupled plasma emission spectrometry (ICP-OES) after 1 day. The Pb concentration in the solution was found to be below the confidence range of the instrument. This experimental result indicates that, after 1 day of dolomite reaction in either 1 mg/L PbCl
2 solution or 50 mg/L PbCl
2 solutions, the Pb in the aqueous solution was converted to the solid phase and the remaining Pb in the solution was less than 50 μg/L. XPS results of dolomite powder in 1mg/L PbCl
2 solution and 50 mg/L PbCl
2 solution were compared and revealed characteristic peaks at 143 eV and 138 eV for both samples, which were close to the Pb4f
5/2 and Pb4f
7/2 standard peak positions, respectively, identifying the appearance of a Pb-containing solid phase on the dolomite surface after exposure to the Pb-containing solution for 15 days. Using the XPS chemical shift peak positions of the previous standard Pb-containing compounds as a reference, we carried out a splitting of Pb4f
7/2. The splitting results indicate that the Pb-containing solid phase on the dolomite surface mainly consists of PbCO
3 with a little PbSO
4; see
Figure 3. A small amount of PbSO
4 is presumed to be produced from the reaction of the impurity BaSO
4 with Pb in the dolomite. Comparing the two sets of data, we can see that the intensity of the Pb4f peak in the dolomite surface is also higher for the more concentrated Pb solution, and the lower PbCO
3 equilibrium constant (log KPbCO
3 = −13.13, T = 298 K) means that the solution is relatively supersaturated for the solid phase; therefore, we believe that the enhanced Pb4f peak signal is the result of more PbCO
3 precipitation. In experiments on the adsorption of Pb on calcite {10
4} cleavage surfaces, calcite cleavage surfaces immersed in high concentrations of Pb solutions generated more PbCO
3 precipitates, thereby increasing the intensity of the corresponding Pb peaks in XPS [
30]. When SEM observations were carried out in conjunction with EDS surface composition surface analysis and compared with the regional electron energy spectroscopy data of dolomite exposed to deionized water, the dolomite surface showed precipitation of Pb components after exposure to PbCl
2 solution for 15 d, which corroborated the results obtained by XPS. XRD showed the same variation in surface composition to some extent. The total diffraction peaks were still characterized by a solid phase consisting of dolomite with associated barite and, in the Pb-bearing group, the characteristic peaks in the PbCO
3 fraction and the characteristic peaks in the BaSO
4 fraction appeared; see
Figure 4. The Pb-bearing dolomite powder diffraction data showed characteristic peaks in PbCO
3 that were not present in the non-Pb group, but the positions of the peaks in the X-ray diffraction spectra and the standard PbCO
3 peaks rarely corresponded to each other, which is different from the experimental results, where PbCO
3 replaced CaCO
3 in a highly concentrated Pb
2+ solution [
29]. Notably, a characteristic CaCO
3 diffraction peak was found at 27.9850° from Pb-bearing dolomite, and CaCO
3 reprecipitation may reflect a coupled dissolution-precipitation process due to the uneven dissolution of the dolomite [
31].
The SEM observations of the dolomite surface in the previous section show that the surface is a powder fractured dolomite surface rather than a conventional pure crystal cut flat surface, but the same changes in surface morphology are observed: in the non-Pb solution the dolomite surface morphology is generally composed of rhombohedral-shaped pits and macroscopic steps, whereas after 15 days in the Pb containing solution, the dolomite surface is characterized by a large reduction in the number of steps and pits and curved steps on a flatter surface were observed (
Figure 7c,e). As seen in
Figure 7a,b, the morphology formed by mechanical breakage is characterized by relatively small curvature and similarly shaped steps on each deconstruction surface within a region, as if similarly shaped steps were stacked on top of each other. We attribute this similar morphology with the above characteristics to mechanical damage. In contrast with experiments that were conducted and observed on the artificially cut deconstruction surface of calcite, the surface of the dolomite raw rock powder has more pits or steps formed by mechanical fragmentation, and the surface consists of multiple incomplete deconstruction surfaces. Such a complex surface makes it impossible to accurately distinguish between mechanically broken and dissolved forms. Regardless of the presence or absence of Pb solution in the system, as the dissolution process proceeds, the surface morphologies formed by mechanical forces will develop, lengthen, and annihilate each other, eventually resulting in increasing flat areas. Independent and relatively large curvature steps with relatively long lengths occur in flat areas, as well as pits with some curvature around the perimeter and connecting multiple pits (
Figure 7e). This surface morphology is similar to that observed on artificially cut surfaces. Athanasios Godelitsas observed calcite solution surfaces exposed to 1000 mg/L aqueous lead solution for one month. In situ experimental results showed the transformation of straight steps into curved steps on the deconstruction surface, the transformation of small adjacent diamond-shaped solution pits into large irregular solution pits, and the appearance of mound-like precipitation at the steps [
32]. On the one hand, the specific morphologies on the surface of dolomite exposed to Pb solutions are similar to those observed by Athanasios Godelitsas; on the other hand, the surface of dolomite (calcite type) with intact solution surfaces after dissolution in pure water consists of regular rhombohedral-shaped pits with straight and folded steps, which are different from the morphological features observed on the surface of dolomite exposed to Pb solutions. Comparing images of the initial protolith surface with those of dolomite exposed to Pb solution, this paper concludes that the appearance of curved steps and irregular solution pits with a certain curvature on the surface of dolomite exposed to Pb solution is not entirely attributable to dissolution following mechanical fragmentation. The surface morphology of any origin is influenced by the dissolution process and these morphologies are altered by the presence of Pb in solution. Thus, similar to artificially cut surfaces, the presence of Pb on incomplete dissociation surfaces after mechanical fragmentation during the dissolution process can also lead to the appearance of curved dissolution steps and irregular dissolution pits as specific surface morphologies.
4.2. The “Sweet Spot” of Adsorption on Dolomite Surfaces
CHAI Rukuan used the characteristics of suspended bonds and surface energy theory to establish a molecular dynamics model and quantitatively calculate the suspended bond density and surface free energy for different forms of calcite surface in the model. The results show that the suspended bond density and surface free energy of surface vacancies and surface projections are greater than those of ideal planes. The surface vacancies and protrusions have more active sites, which correspond to stronger reactivity; therefore, water molecules are preferentially adsorbed at the surface vacancies and protrusions [
33]. Pb
2+ ionic radius is larger than Ca
2+, so when Pb
2+ is adsorbed onto CaMg(CO
3)
2 crystals by ion exchange, it is necessary to consider not only the dehydration of the metal ion but also the rejection of Pb
2+ by the crystal lattice [
12]. According to
Figure 10, the number of first nearest neighbours is three at the kink (
Figure 10 (3) and two at the step (
Figure 10 (2)). The high number of vacant sites with high first nearest neighbours has more unbonded surfaces, meaning that the high levels of free energy promote the dehydration of metal ions in solution and the formation of ionic bonds at this position; however, Pb
2+ is repelled by the lattice because its ionic radius is large compared to Ca
2+. Combining these two reasons, it can be qualitatively assumed that the kink and step may be the ‘sweet spot’ for Pb adsorption on the dolomite surface. Based on the results of the EDS assay of the dolomite surface area exposed to 50 mg/L Pb
2+ solution, the apparent concentration of Pb in the selected dolomite surface step areas was higher than that in the flatter areas. This result is consistent with the above assumptions.
The surface position elements represent the macroscopic morphological features on the surface, as shown in
Figure 11, while the dissociation energy of these sites varies and the dissolved surface exfoliation process can be decomposed into surface morphological processes, such as dislocation into cavities, kink progression into steps and the mutual annihilation of step waves, in addition to the overall surface exfoliation process along the normal direction. Therefore, the active sites in the surface morphology (point errors, nuclei of dissolution pits, positive and negative kinks, etc.) control the development of the surface morphology and can play a role in controlling the surface dissolution exfoliation process [
34].
SEM observations did not reveal the appearance of significant precipitation at the steps. In contrast, in experiments using Atomic Force Microscopy (AFM) correlation, an ellipsoidal hill-shaped precipitation morphology appeared on the calcite surface at the dissolution steps in a 100 μM Pb solution [
30]. AFM observed the growth in heterogeneous nucleated PbCO
3 at the contact interface and a ‘pyramidal’ structure on the surface when calcite was dissolved in a 5 mM Pb solution [
35]. The SEM images of this experiment do not show the morphology of PbCO
3 precipitation on the surface. However, this work still suggests that a new solid phase of PbCO
3 or Pb-bearing carbonate solid solution is present on the dolomite surface. One reason for this is that experimental XPS and XRD data on surface composition corroborate the hypothesis that a new Pb-bearing solid phase was generated. The AFM images from the TENG H also did not show significant morphological features of a new solid-phase precipitation formation at the dissolution step [
36]. This inconsistency may be due to inconsistencies in the magnification and resolution of the instruments used for observation.
4.3. Form of Pb Presence on Dolomite Surfaces
The common crystal structure of PbCO
3 is not the same octahedral ligand structure as calcium carbonate or calcium magnesium carbonate, implying that PbCO
3 with a non-octahedral ligand structure has different crystal properties to dolomite. Neil C. Sturchio showed that the substitution of Pb for Ca can form the new solid-phase PbCO
3, implying that Pb enters the Ca site in the calcite octahedral coordination structure, and that Pb migrates from the outermost layer to the interior of the solid solution, which is the near-surface layer of dolomite [
37]. Infrared spectroscopy and Raman spectroscopy are different in principle, but both can reflect the elemental composition and crystal structure of a solid phase through the molecular structure and vibrational patterns of molecular groups within the solid phase.
Figure 5 shows the IR spectra of dolomite exposed to PbCl
2 solution and dolomite pretreated with PbCl
2 solution before being dissolved again in deionized water, with similar peak positions. Referring to the results of previous studies, the distribution of the peak positions of these two peaks is consistent with that of typical dolomite IR spectra [
38], and no obvious spectral peaks appear at the PbCO
3 IR spectral peaks, indicating that the solid phase is still internally dominated by dolomite and the crystal structure is still the octahedral ligand structure of dolomite.
Figure 6 similarly shows the Raman spectra of dolomite exposed to PbCl
2 solution and dolomite pretreated with PbCl
2 solution before being dissolved again in deionized water, adding the Raman spectrum of the initial dolomite as a control. All three show spectral peaks at 291 cm
−1 and 331 cm
−1 due to the alternating distribution of Ca
2+ and Mg
2+ in dolomite, causing two vibrational modes of bending outside the carbon and oxygen planes [
39]. The three peaks are also similar in other peak positions, indicating that the solid phase is still dominated by dolomite within the solid phase and the crystal structure is still the octahedral ligand structure of dolomite [
40]. However, for the dolomite exposed to PbCl
2 solution, the Raman spectra of both peaks at 168 cm
−1, 292 cm
−1, and 718 cm
−1 are shifted in the short-wave direction compared to the initial dolomite. The peak at 168 cm
−1 corresponds to the crystal internal lattice vibration, that at 292 cm
−1 corresponds to the CO
32− group out-of-plane bending vibration mode and that at 718 cm
−1 corresponds to the group in-plane bending vibration mode. The Raman shifts produced by the above three vibrational modes are proportional to the bond energy of the C-O bond. For carbonatite minerals, the larger the ionic radius of the metal ion, the smaller the bond energy of the C-O bond, and the corresponding Raman shift is shifted in the short-wave direction [
41]. Comparing the positions of the three Raman shift spectral peaks, it can be concluded that the surface crystal structure of the dolomite that was in contact with the solution is still the octahedral ligand structure of dolomite, but the Pb ions with larger ionic radii than Ca and Mg occupy the position of Ca or Mg in the original octahedral ligand structure; see
Figure 12.
No peaks of sufficient intensity appeared at the peak positions of the Raman and IR spectra corresponding to the non-octahedral ligand PbCO3. This suggests that the new solid phase generated on the surface of dolomite surface to the PbCl2 solution was mainly the octahedral ligand PbCO3. For polycrystalline powder X-ray diffraction, different crystal structures correspond to different diffraction peak spectra, and the XRD test results show that the positions of the diffraction peaks of the dolomite exposed to PbCl2 solution are mostly different from the dolomite (PbCO3) displayed on the PDF card, with only a few characteristic peaks in the same position. PbCO3 with an octahedral ligand structure is similar to PbCO3 with a non-octahedral ligand structure; however, due to the difference in their crystal structures, they exhibit different diffraction peaks. Therefore, as in the previous paper, we can deduce that the main product on the dolomite surface after exposure to PbCl2 solution is PbCO3 with an octahedral ligand structure, and the proportion of non-octahedral ligand structure of dolomite is so low that its characteristics are difficult to detect by various solid-phase characterization methods.
4.4. Effect of Pb2+ on the Dissolution Process of Dolomite
Atmospheric CO
2 is actively involved in the dissolution of carbonate rocks, and the partial pressure of CO
2 controls the dissolution process and solubility to a certain extent [
11]. CO
2 is involved in the stoichiometric equilibrium of the water–rock–gas reaction, which provides a carbon source to the system and is closely related to pH. In this experiment, pH, total alkalinity, and conductivity were used as indicators of the degree of carbonate dissolution. CO
2 is constantly replenished by the external environment in the open experimental system, whereas, in the closed experimental system, CO
2 is consumed as dissolution proceeds [
42]. In the open experimental system, the pH decreases with time and the total inorganic carbon values of the high concentration PbCl
2 solution are similar to those of the low-concentration PbCl
2 solution during the whole experimental cycle of 15 days. The Pb source for the open experimental system was not only derived from the added PbCl
2 solution but also from the Pb-zinc ore containing Pb. The Pb-Zn ore dissolved with the dolomite, probably because the dissolution of the sulfur fraction changed the pH of the solution, causing the pH to decrease as dissolution proceeded. Under acidic conditions with a high H+ content in the solution, dolomite dissolution is controlled by pH. In the closed experimental system, the pH increased as dissolution proceeded. In the group charged with CO
2 to reach CO
2 saturation, the HCO
3− and CO
32− in solution were supplied by both CO
2 and solid-phase dolomite as the reaction proceeded to near equilibrium. The total alkalinity and conductivity of the solution were much higher than in the non-CO
2 charged group, while the pH was lower than in the non-CO
2 charged group, indicating that CO
2 is involved in and contributes to the dissolution of dolomite. A comparison of the water chemistry data for the same gas groups with and without Pb elements shows that the total alkalinity of the dolomite dissolution process in the Pb-containing solution is smaller than the total alkalinity of the dolomite dissolution process in the Pb-free solution in the non-CO
2 charged system. The water chemistry trends in the CO
2-sufficient system are still similar to that of the uncharged system, and the difference in total alkalinity and conductivity is more pronounced in the group with or without Pb. John W. Morse systematically summarized the study of carbonate dissolution kinetics, summarizing previous studies linking solutions to carbonate surface dissolution morphology and suggesting that dissolution inhibitors have a profound effect on dissolution kinetics in the near-equilibrium phase of carbonates [
43]. As the whole system was in a closed state, not many factors affected the dolomite dissolution. Following the change in the pH of the whole dissolution process, 1mg/L of PbCl
2 solution did not affect the pH of the system, so the Pb in the solution did not affect the system dissolution process by changing the pH of the solution. The conductivity of the solution was positively correlated with dolomite dissolution, so the difference in solution conductivity also demonstrated an inhibitory effect of Pb on the near-equilibrium phase of dolomite dissolution.
Based on the results of this experiment’s solid and liquid phases, the previous hypothesis is confirmed. Pb adsorbs or precipitates at the active sites on the dolomite surface. The ionic bonding of the newly formed PbCO
3 or Ca, Mg, and Pb carbonate solid solutions possesses a sufficiently strong bonding energy to hinder the dissolution process at these sites (e.g., kink extension and deepening of the dissolution step extension), ultimately leading to the inhibition of dolomite dissolution in the near-equilibrium phase. The morphological features observed in this experiment are similar to the structure followed by Hongmei Tang [
44]. These generated PbCO
3 occupied active sites, causing the steps to deepen and expand at an uneven rate, allowing for linear steps to transform themselves into curved steps or adjacent linear steps to combine to create curved steps. Similarly, irregular dissolution pits formed on the surface; see
Figure 7e.
More importantly, the initial dolomite surface in this experiment was mechanically fragmented and the surface morphology existed on multiple incomplete deconfined surfaces, as opposed to the more complete and flat surfaces cut out of the extended deconfined surfaces chosen in previous experiments. It is not possible to quantitatively compare the mass, effective surface area, and active sites on the surface of the solid-phase material used in each of the previous experiments. However, a powder with the same amount of material must possess many more active sites than the whole crystal, and this difference may have several orders of magnitude. Therefore, we can qualitatively assume that there are more active sites present on the surface of the powder than on artificially cut planes. By comparing
Figure 7b with previous AFM or SEM images, it can be seen that the mechanically crushed dolomite surface selected for this experiment is far more complex than the artificially cut out calcite crystal surface, implying that the surface-active site density in this experiment is much higher than that of the artificially cut surface. Jin Ma investigated the variation in effective surface area (ESA) during the dissolution of a dolomite–CO
2–water system. When only the active sites on the dolomite surface were involved in the dissolution reaction, results from image-based stochastic Monte Carlo simulations showed that ESA nearly linearly increased [
45]. Unlike previous experiments, this experiment used dolomite powder as the sample, which can be approximated as a complete infiltration of the solid phase surface after immersion in the solution. At this point, only the active sites on the dolomite surface controlled the effective surface area of dolomite. According to Jin Ma’s Monte Carlo simulations of dolomite, the ESA will increase linearly as the reaction proceeds, meaning that powdered dolomite with more initial active sites will have a higher ESA than artificially cut dolomite surfaces during the entire dissolution process. The present experiments, characterized by a low Pb
2+ concentration (2.6 μM) with a high density of active sites on the surface, still exhibited the inhibition of dolomite dissolution by Pb
2+. We suggest that the reason for this is that Pb
2+ binds preferentially to the active sites on the surface. These active sites control the dissolution of adjacent areas and, by occupying these active sites, Pb
2+ controls the dissolution of adjacent areas, so that a small amount of Pb
2+ controls the dissolution of a ‘large’ area. The solid phase characterizations result also indicates that the main product on the dolomite surface is the octahedral ligand PbCO
3. Pb
2+ in solution preferentially adsorbs to the active sites on the wider surface rather than aggregating at a few sites and then producing non-octahedral ligands of PbCO
3 in heterogeneous nucleation. Therefore, the inference where Pb preferentially binds to the surface-active site may explain why no precipitation morphology was observed in SEM. In summary, similar to experiments with artificially cut crystal faces at relatively high molar concentrations of Pb solutions, the 2.6 μM (1 mg/L) Pb solution still acts as a solubility inhibitor for complex dolomite surfaces that possess incomplete cleavage surfaces. The preferential binding of Pb2+ causes this phenomenon at the active site and leads to the formation of octahedral ligands of PbCO
3 by occupying Ca
2+ or Mg
2+ positions. These new solids rely on their strong ionic bonds to hinder the development of regional dissolution morphology and ultimately inhibit the dissolution of dolomite in the near-equilibrium phase.
Calcite carbonates also dissolve slowly in pure water, with H
+ in solution crossing the liquid-phase diffusion layer and combining with carbonate on the surface of the solid phase, with the resulting HCO
3− breaking away from the lattice, and Ca
2+ also eventually breaking away from the lattice. Therefore, the involvement of atmospheric CO
2 in the chemical weathering of carbonates is essentially a process whereby the dissolution of atmospheric CO
2 leads to an increase in the concentration of H
+ in water, which in turn promotes the dissolution of the solid phase. The process noted by the red arrow in
Figure 13 is widely regarded as a net carbon sink for karst, where atmospheric CO
2 can be very slowly dissolved in water in the absence of carbonate dissolution. Atmospheric CO
2 dissolution and calcite carbonate dissolution are interrelated by H
+ and, therefore, contribute to each other’s positive reactions. Therefore, previous studies on the disturbance term accounting for atmospheric CO
2 carbon sinks have focused on H
+ that is “not generated by dissolved atmospheric CO
2”. Such H
+ is generally referred to as “exogenous acid”; however, exogenous acid may consist of two components. One component is the H
+ substituted into the system by “exogenous water”, which means that exogenous water with a low pH sinks into the system; the other is the reaction of non-carbonate minerals such as S and N in the solid phase with water or the presence of oxygen addition, which can produce H
2SO
4 or HNO
3, thus increasing the H
+ concentration in the solution. However, according to the inferences drawn in this paper, when exogenous water carries large amounts of heavy metal ions (Pb, etc.) into the system, these metal ions preferentially bind to the active sites on the surface of calcite group carbonates, forming MCO
3 with higher bonding energy. H
+ must still bond to CO
32− on the lattice before it can leave the lattice and enter the solution, and higher bond energies mean harder bond breakage; therefore, these MCO
3 act as inhibitors of calcite group carbonates. As mentioned earlier, atmospheric CO
2 solutions are facilitated in association with the chemical weathering of the calcite group, so when the calcite carbonate dissolution process is inhibited, it can be assumed that the amount of atmospheric CO
2 dissolved into the water is correspondingly reduced. This means that the heavy metal ions M
2+ in the exogenous water reduces the total karst carbon sink by inhibiting the calcite group dissolution process. In other words, the process by which heavy metal ions from outside the groundwater system enter the system and adsorb on the calcite group carbonate surface will diminish the karst carbon sink potential of the whole system.