In this study, ReaxFF MD simulations (2.5 ns) are employed to systematically investigate the interfacial reactivity of albite's different surfaces ((100), (010), (001)) in contact with CaSO4 solutions. The anisotropic dissolution mechanisms and sulfate-mediated surface transformations are unraveled through analysis of reaction species distribution, hydrogen bonding networks, atomic density profiles, charge states, and mean square displacements. These findings provide a valuable perspective for refining the understanding of fluid-solid reactions at the molecular scale. This study advances the fundamental understanding of albite-solution interactions while providing predictive insights for optimizing mineral stability in subsurface engineering applications.
The structural configurations and calculated surface energies for the (100), (010), and (001) crystallographic surfaces of albite are summarized in Table 1. The surface energies follow the order: (001) > (010) > (100). This disparity predominantly originates from distinct atomic configurations and coordination environments exposed during cleavage processes. The (100) surface exhibits the lowest surface energy, attributable to its continuous [AlSiO] framework network (as shown in Fig. 2a(iv)) formed by tetrahedrally coordinated Al/Si centers interconnected through bridging oxygen atoms. All oxygen species maintain their bridging configuration in this surface architecture, effectively preventing the formation of undercoordinated metal centers. Furthermore, interstitial Na⁺ remains encapsulated within oxygen-constructed cavities, achieving charge compensation through long-range electrostatic interactions while maintaining low defect density. In contrast, the moderately higher surface energy of the (010) surface correlates with cleavage-induced disruption of Al-O octahedral chains. This structural perturbation induces a coordination number increase (from four to five coordination, as shown in Fig. 2b(i)(ii)) in partial Al centers within tetrahedral frameworks, creating localized charge asymmetry that necessitates additional energy expenditure for electrostatic stabilization. The (001) surface demonstrates the highest surface energy due to its cleavage mechanism that directly disrupts the covalent Si-O-Al network, exposing high densities of terminal non-bridging oxygen species and unsaturated Si/Al centers. Moreover, Na⁺ cations within framework channels become fully exposed at this interface, where their unscreened positive charges generate substantial electrostatic repulsion with adjacent electronegative terminal oxygen atoms. The synergistic effects of these atomic-scale structural defects and charge imbalances necessitate complex electronic redistribution mechanisms for surface stabilization, thereby substantially elevating surface energy. The observed positive correlation between surface reactivity towards CaSO solutions and surface energy across different albite surfaces primarily derives from variations in exposed coordination states and charge distribution patterns. Detailed mechanistic discussions will be presented in subsequent sections.
Figure 1 illustrates the dynamic interfacial structural evolution of albite (100), (010), and (001) crystal surfaces interacting with CaSO solution at the initial reaction stage (0 ns) and after 2.5 ns. The corresponding top-view surface morphology and magnified details of characteristic structures at 2.5 ns are shown in Fig. 2. Comparative analysis reveals that all surfaces undergo significant surface reconstruction through interactions with the solution, while exhibiting distinct reaction pathways and kinetic characteristics. On the (100) surface, the exposed aluminum atoms (initially undercoordinated) and the oxygen atoms of the SiO tetrahedra serve as the primary reactive sites. On the one hand, water molecules directly coordinate with the Al atoms through coordination interaction. On the other hand, the water molecules dissociate into OH and H species, and the OH groups combine with Al atoms to achieve their hydroxylation, while the oxygen atoms on the surface silicon-oxygen tetrahedra act as proton acceptors for the dissociated H. This process drives the coordination evolution of Al from initial Al to Al/Al (The superscript numbers represent the coordination number). This is schematically illustrated by the (i) (ii) structural configuration in Fig. 2a. Notably, the [AlSiO] framework network (Fig. 2a(iv)) contains substantial porosity, enabling limited water penetration into subsurface regions. While SO migrate to bond with surface Al atoms, Ca adsorption is inhibited by Na shielding effects.
The (010) surface features reactive tri-coordinated Si atoms as active centers. Dissociated OH preferentially binds with Si to form dense silanol groups (Fig. 2b(iii)), surpassing other surfaces in hydroxyl density. The Si-O-Al bridging oxygen acts as a proton acceptor, forming Si-OH-Al configurations. Interestingly, the absence of SO adsorption is observed on this surface, which is attributed to the absence of directly exposed Al atoms at the interface, leading to a lack of binding sites for SO. The adsorption of Ca is associated with the hydroxylation of surface Si atoms, where the hydroxylated surface electrostatically attracts Ca. In contrast, the (001) surface demonstrates the highest reactivity with pronounced structural disordering compared to other surfaces. Enhanced adsorption behaviors of SO with Ca were observed on the (001) surface. Notably, Na leaching behavior was uniquely identified on this facet, a phenomenon absent in both (100) and (010) surfaces. This is because, on one hand, the high surface energy promotes the hydroxylation of surface Si and Al, weakening the bonding of Na at the central sites, and on the other hand, the typical ion exchange mechanism facilitates Na release. Additionally, the exposed Al atoms demonstrate a metastable dissolution tendency. Surface Al atoms undergo a coordination transformation from three-coordinate to Al/Al/Al configurations (as shown in Fig. 2c(i)-(iii)), accompanied by partial hydration dissolution, evidenced by significant deviations of the Al coordination geometry from the surface reference plane. Nevertheless, the dissolved Al species maintain strong bonding with surface terminal oxygen through inner-sphere complexation, indicating that complete dissolution of Al remains kinetically constrained within the limited simulation timeframe (2.5 ns). This phenomenon suggests that the dissolution process of feldspar minerals likely follows a stepwise mechanism of local activation-progressive detachment, where initial surface modifications precede gradual structural disintegration. This concept is consistent with the stepwise, site-specific dissolution mechanisms proposed based on quantum chemical calculations and is supported by experimental observations of interfacial dissolution-reprecipitation processes in silicates and aluminosilicates at the nanoscale.
Hydrogen bonding, as a prototypical example of intermolecular weak interactions in mineral-solution interfacial reactions, demonstrates density and strength parameters that directly characterize the interaction characteristics between water molecules and mineral surfaces. Figures 3 and 4 present the calculated 1D and 2D hydrogen bond density and intensity profiles for albite surfaces interacting with CaSO solution at 2.5 ns. To define a hydrogen bond, the distance cutoff is set as 3.5 Å, and the angle is set as 30.0°. The (100) system shown in Fig. 3a demonstrates much higher peak values at the interface, and the number density is about 0.062/Å compared with 0.054/Å for the (010) system and 0.043/Å for the (110) system. This discrepancy can be attributed to the heterogeneous behavior of different surfaces in response to the solution (as illustrated in snapshots in Fig. 4). The (100) surface atomic arrangement features high-density bridging oxygens (Si-O-Al, Si-O-Si) and active sites consisting of terminal non-bridging oxygens in silicon-oxygen tetrahedra, along with a well-ordered structure and relatively flat reactive surfaces that facilitate strong hydrogen bond network formation with water molecules. In contrast, the (010) surface aligns parallel to silicon-oxygen tetrahedral chains, adopting a dominant [SiO] configuration with silicon atoms as the primary exposed species. The sparse exposure of active oxygen sites reduces hydrogen bond strength and density compared to the (100) surface. The (001) surface exposes two terminal non-bridging oxygens and Si-O-Al bridging oxygens from tetrahedra, but its irregular reactive surface topology results in short-range, low-density hydrogen bond networks at the interface.
Furthermore, two-dimensional color contour maps reveal spatial distribution details of hydrogen bond density across XY and YZ surfaces for each crystal system (left and middle columns of Fig. 4), illustrating how hydrogen bond density propagates along albite's edge geometry with different crystallographic orientations. The (100) surface shows significant interfacial hydrogen bond density variations with distinct localized high-density regions, demonstrating pronounced interfacial aggregation characteristics. Comparatively, the (010) surface exhibits more uniform density distribution while retaining observable local hydrogen bond clustering at the interface, suggesting that the (010) surface possesses stronger hydrogen bond-forming capabilities in certain localized regions. In contrast, the (001) surface displays relatively homogeneous hydrogen bond distribution from interface to bulk solution without significant localized aggregation features observed in the (010) and (100) systems.
The atomic density distribution maps provide spatial characterization of various atom types within minerals and at their interfaces with solutions. By analyzing changes in atomic density profiles during dissolution, these maps visually reveal the evolution of mineral surface structures and reaction dynamics. Figure 5 presents the atomic density distribution characteristics of albite (100), (010), and (001) surfaces after interaction with CaSO solution. The density profiles reveal significant atomic density fluctuations in the mineral/solution interfacial region of the (100) surface, where the steep interfacial feature indicates distinct phase separation between mineral and solution phases, associated with the lower surface energy of this crystallographic surface. Notably, both H and Ow atoms exhibit higher density peaks at the (100) surface compared to the (010) and (001) surfaces, with the (001) surface showing the lowest peak values for both H and Ow atomic densities. These observations align with the hydrogen bond network analysis presented in Section "H-bond density and strength analysis". Higher surface hydrogen bond density facilitates tighter water molecule adsorption, forming a stable hydration layer that not only prevents undesirable surface reactions but also reduces dissolution and corrosion potentials. A distinct Os density peak emerges in the mineral region of the (100) and (001) surfaces, indicating sulfate ion adsorption at the mineral/solution interface. Characteristic Ow and H atom density peaks observed in the regions of all three surfaces respond to the gradient diffusion behavior of water molecules along mineral surfaces. Particularly, the (100) and (010) surfaces maintain well-ordered density distributions of O, Al, and Si atoms post-reaction, demonstrating remarkable structural stability during interfacial processes. In contrast, the (001) surface displays disordered atomic density fluctuations at the interface, attributable to its higher surface energy that enhances susceptibility to solution-induced structural reorganization. Furthermore, a distinct sodium density peak emerged in the solution phase of the (001) system, which indicates the occurrence of Na dissolution behavior on the (001) surface.
To further analyze the characteristic coordination of surface atoms with water and ions during reactions between different albite surfaces and CaSO₄ solution. The radial distribution function (RDF) is defined as the ratio of the average number of atoms per unit volume at a distance r from a reference atom to the overall average atomic number density of the system. The calculation formula is as follows:
where represents the average atomic number density of the system (where N is the total number of atoms and V is the volume); dN(r) denotes the average number of other atoms within a spherical shell at a distance r+dr from the reference atom; is the volume of the spherical shell.
Figure 6a presents the RDF of Si-Ow in albite (100), (010), and (001) surface systems. Distinct characteristic peaks emerge at 1.65 Å for (010) and (001) surfaces, with this bond length closely matching the typical coordination distance of Si-Ow during silica tetrahedron hydrolysis, indicating the formation of stable coordination structures between surface silicon atoms and water molecules. The (010) surface exhibits a prominent peak, attributable to under-coordinated Si atoms (tri-coordinated state) with strong electronegativity on this surface, which readily coordinate with water molecules to form Si-OwH surface hydroxyl groups. In contrast, the weaker peak intensity observed for the (001) surface suggests reduced interaction strength at this coordination distance. No discernible peaks are detected for the (100) surface within this bond length range, directly correlated with the coordination saturation of four-coordinated Si atoms on this surface.
The Al-Ow RDF analysis in Fig. 6b reveals first coordination peaks at 1.83 Å for (100) and (001) surfaces, consistent with results reported by Jabraoui et al.. The peak intensity follows the order: (001) > (100), while the (010) surface shows no significant peaks at short distances (<3 Å). This discrepancy originates from structural characteristics: Al atoms on the (010) surface are predominantly embedded within the silica tetrahedral framework, forming stable Al-O tetrahedral configurations with low surface Al/O ratios that effectively suppress Al-Ow interactions. The layered structure of the (001) surface exposes higher proportions of under-coordinated Al atoms, which form stable Al-Ow coordination through strong interactions with oxygen atoms in water molecules. Notably, the lower surface energy of the (100) surface provides limited active sites through defects or localized Al exposure, explaining its reduced coordination peak intensity compared to (001). The Na-Ow hydration characteristics (Fig. 6c) demonstrate significant crystallographic orientation dependence, with primary peaks at 2.41 Å showing intensity sequence (100) > (010) > (001), consistent with experimental observations. This trend correlates with structural differences in mineral-solution interfacial hydrogen bond networks, where the unique Na⁺ spatial distribution pattern on the (100) surface likely enhances hydration.
For calcium coordination behavior (Fig. 6d), the main Ca-Ow RDF peak at 2.5 Å aligns closely with experimental and MD simulation results (average 2.47 Å) reported by Jalilehvand et al., validating computational model reliability. Detailed analysis (Fig. 6e) shows both (001) and (010) surfaces exhibit Ca-O coordination peaks at 2.41 Å, but with broader peak width for (001). This difference arises from distinct oxygen species exposure: terminal oxygen atoms with higher negative charge density on the (001) surface readily form coordination bonds with Ca, while bridge oxygen and Al-coordinated oxygen on the (010) surface exhibit lower coordination activity. The denser atomic arrangement on the (010) surface restricts Ca adsorption sites, reflected in a narrower peak width. The (100) surface shows no detectable peaks due to Na electrostatic shielding effects.
In terms of SO adsorption behavior (Fig. 6f), the surface of (001) shows a prominent Al-Os coordination peak at 1.89 Å, which is very close to the X-ray test result of 1.879 Å, indicating stable surface complex formation. In contrast, the Al-Os coordination bond length on the (100) surface measures 2.01 Å, attributed to the electrostatic shielding effect induced by Na on the (100) surface, which weakens the adsorption interaction between SO and Al. Notably, no Al-Os peaks appear on the (010) surface, where dense silica tetrahedral networks effectively isolate Al from the solution phase.
The change in the charge state of atoms or ions reflects the electron transfer process of reactions occurring at the mineral-solution interface. Monitoring the charge state helps determine the hydration or hydroxylation degree of surface atoms, identify the formation of hydrated products in reactions, and understand the overall reaction pathways involved in mineral dissolution. Figure 7a-c present the atomic distribution along the z-axis and the surface charge distribution of albite (100), (010), and (001) surface systems at 2.5 ns. To elucidate the charge transfer mechanism during interfacial reactions, a comparative analysis of the charge values of representative metal cations under distinct structures on each surface was conducted. The z-axis charge distribution reveals non-identical surface adsorption groups and charge patterns across different crystallographic surfaces. This study statistically analyzed the trend of charge changes in Si and Al atoms within the solid region (24-27 Å) near the solution reaction interface on different surfaces.
In the interface region of the (100) surface, the average positive charge of Si atoms increased by 1.2% (Δq = +0.017e) compared to the overall system, while Al atoms increased by 7.7% (Δq = +0.08e). Conversely, the (010) surface showed a reverse trend with Si atom charges decreasing by 7.97% (Δq = -0.1e) and Al atoms decreasing by 5.77% (Δq = -0.075e). The (001) surface exhibited differential changes, with Si atom positive charges decreasing by 5.71% (Δq = -0.071e) and Al atoms increasing by 3.38% (Δq = +0.04e). These charge distribution differences are closely related to the post-reaction interface structure. The (010) surface's charge reduction originates from hydroxylation of exposed SiO groups (SiO → SiO-OH, Fig. 2b(iii)), where Si accepts hydroxyl electrons. Meanwhile, the protonation of SiO groups on the (001) surface (SiO → SiOH) weakens the electron attraction of Si (Fig. 2c(vi)), reducing its positive charge. This is further confirmed by comparing the charge values of Si/Si structures in Fig. 7b, c. Notably, the decrease in the charge of Al atoms on the (010) surface is also related to this electron transfer mechanism. However, the 100 surface, due to its low surface reactivity and dense hydrogen bond network that inhibits protonation reactions, has only a minimal portion of Si-O tetrahedra with protonated O, resulting in minimal overall charge fluctuations at the interface. Additionally, the number of neighboring oxygen atoms (O atoms) to Si atoms is positively correlated with their positive charge (comparison of Si/Si in Fig. 7a), a pattern consistent with the results reported in the literature.
Unlike the charge changes dominated by the coordination environment of Si atoms, the charge state of Al atoms is primarily regulated by coordinating molecules. It can be observed that the charge of Al atoms increases with the number of neighboring molecules (i.e., coordination with Ow, Os, O), as illustrated in the right panel of Fig. 7. In Fig. 7a, b, Al → Al, and in Fig. 7c, Al → Al, all exhibit an increase in positive charge for centrally coordinated Al atoms. When coordinated with the Ow of water molecules or the Os atoms of surface sulfate groups, Al atoms lose electrons, leading to an increase in positive charge (Δq > 0), while the coordinating Ow/Os gain additional electrons (Δq < 0). This charge transfer mechanism indicates that the coordination of interfacial water molecules with surface Al sites plays a crucial role in charge redistribution.
In the study of reaction kinetics at mineral-solution interfaces, the mean square displacement (MSD) is a crucial dynamic parameter that characterizes atomic migration behavior, effectively revealing the diffusion characteristics and interaction mechanisms of ions during interfacial reaction. The expression for the calculation of MSD is as follows:
where N is the number of particles and r (t) is the position of the ith particle at time t. r (0) is the initial moment position. Figure 8a, b presents the evolution of MSD and its vertical component (Z-MSD) for surface Na at different crystal surfaces of the albite-CaSO interface system. Computational data reveal that the (001) surface exhibits significantly higher MSD values compared to (100) and (010) surfaces, with its kinetic curve demonstrating steep initial ascent and intense fluctuations throughout the temporal domain, indicating superior diffusion activity of Na on this surface. This dynamic characteristic likely originates from relatively weak lattice confinement effects on the (001) surface, enabling enhanced susceptibility of surface sodium ions to thermodynamic perturbations from the solution environment. In contrast, the (100) surface maintains relatively stable MSD evolution, suggesting better kinetic stability in its diffusion process, while the persistently low MSD values of the (010) surface reflect strong constraints imposed by three-dimensional chemical bonding networks. Z-MSD analysis (Fig. 8b) reveals significant coupling between vertical diffusion behavior and overall dynamics for the (001) surface. The MSD-Z curve displays two rapid ascending phases (0-71 ps and 157-221 ps), corresponding to stepwise dissolution processes of Na. Subsequent stages enter a slow-growth regime, though persistent fluctuations may correlate with dynamic reconstruction of layered structures or localized defect formation mechanisms. Comparative analysis shows the (100) surface achieves kinetic equilibrium after initial rapid response, while the near-zero MSD-Z values of (010) surface directly relate to steric hindrance effects from its interlayer close-packed structure. Combined with density distribution analysis from Fig. 5 (showing absence of sodium characteristic peaks in solution layer), this confirms sodium dissolution has not occurred on (100) and (010) surfaces. Comprehensive analysis demonstrates significant anisotropic diffusion characteristics among albite crystal surfaces: The (001) surface exhibits superior diffusion capacity due to layered structure and weak vertical confinement, particularly showing enhanced interfacial reactivity through high Z-direction activity; The (100) surface maintains limited 3D diffusion capacity with suppressed vertical motion; The (010) surface effectively restricts ionic migration through strong chemical bonding interactions.
Figure 8c, d reveals the anisotropic characteristics and dynamic evolution mechanisms of water molecule diffusion at the interface between different albite surface systems and calcium sulfate solution. Figure 8c demonstrates that the Ow-MSD curves of three surface systems initially exhibit rapid ascending trends, corresponding to the dynamic reconstruction processes of surface ion exchange and the initial interfacial hydration layer. During subsequent evolution stages, the MSD curves of the (100) and (010) surface systems gradually stabilize, indicating the achievement of dynamic equilibrium in their interfacial reactions. Notably, the Ow-MSD values of the (001) surface system present two distinct growth inflection points: rapid increase during 0-71 ps, slow growth in 71-485 ps, followed by another rapid increase in 485-575 ps before reaching system stabilization. This phenomenon suggests that after the initial rapid surface-solution interaction, subsequent dynamic reconstruction of metastable structures or localized concentration gradient effects induced by continuous leaching of surface cations (e.g., Na) may provide additional kinetic driving forces for water molecule diffusion.
Further analysis of z-axis directional MSD reveals more pronounced differences in the (001) system compared to (100) and (010) systems. The mean squared displacement of interfacial water molecules along the Z-direction and the simulation time required to reach equilibrium in the (001) system significantly exceed those of other surfaces, further confirming its superior interfacial reactivity and reaction extent. Additionally, the (010) surface system maintains the minimum mean squared displacement of interfacial water molecules throughout the simulation, demonstrating exceptional interfacial stability compared to other surfaces.
This section investigates the Na leaching behavior and underlying mechanisms on the (001) surface. Figure 9a-d illustrates snapshots of the dissolution characteristics of Na at different surface sites of the (001) surface. The migration trajectories of Na⁺ from various dissolution sites are depicted by black trace lines in Fig. 10. Based on kinetic trajectory analysis, two dominant mechanisms are identified: (1) leaching induced by SO (Fig. 9a-c), and (2) ion-exchange mechanisms (Fig. 9d).
Figure 9a-c respectively displays the Na leaching behavior dominated by SO at three different sites. During initial hydration, the three Na remain in equilibrium positions constrained by lattice energy barriers, exhibiting vibrational motion within confined regions (as shown by trajectories in Fig. 10a-c). When SO diffuses from solution and approaches the vibrating Na ions, they facilitate the detachment of Na from the lattice surface through electrostatic interactions and coordination effects. Notably, while the Na in Fig. 9a, b exhibit rapid dissolution in early reaction stages, the Na⁺ in Fig. 9c undergoes approximately 350 ps of localized motion before achieving final desorption through cooperative interaction with adjacent SO. This reveals that the Na leaching process is not only constrained by the local environment but also exhibits dynamic randomness over the reaction time. Remarkably, SO exhibits catalyst-like behavior through reversible adsorption-desorption cycles that continuously promote Na separation. In contrast, Fig. 9d demonstrates an ion-exchange behavior between Na and H at the (001) surface. This exchange process, widely reported in the literature, can occur either within a single SiO tetrahedron or between oxygen atoms of adjacent tetrahedra. As shown in Fig. 9d, a proton from the aqueous solution bonded with an oxygen atom in the SiO tetrahedron at 2.7 ps, forming a silanol group (Si-OH). Concurrently, the Na initiated dissociation from the reaction zone and diffused into the solution phase, achieving full desorption by 27 ps. The trajectory visualization in Fig. 10d corroborates this dynamic process. Charge compensation by protons from the aqueous solution was observed in both mechanisms to neutralize residual negative charges after Na dissolution. Additionally, surface reconstruction of the atomic structure near the leaching Na is observed, where the bridging oxygen bonds connecting Si-O-Al break and reform into Si-O-Si bonds. This surface chemical bond reorganization may result from the dissolution effect induced by the hydroxylation of surface Al atoms. Notably, following the initial Na leaching, the shielding effect of the Si-O-Al framework on bulk Na⁺ and the inherently slow hydrolysis of Si-O-Al bonds (relative to ion leaching/water diffusion) may lead to stabilization of the hydrolysis front, potentially halting further dissolution. Compared to Na dissolution in pure aqueous systems, the leaching dynamics in sulfate solutions appear predominantly governed by SO interactions.
Figure 11 presents a comparative schematic diagram illustrating hydrolysis reaction pathways and characteristic structures on different albite surfaces. The surface hydrolysis pathways exhibit significant structural dependence, primarily governed by the dual regulation of surface atomic coordination states and topological configurations. For the (100) surface (Fig. 11a), hydroxyl groups (OH) generated through interfacial water dissociation coordinate with exposed Al atoms to form Al-OH configurations, while released protons (H) are captured by bridge oxygen atoms in Al-O-Si frameworks to create Si-O(H)-Al structural states (Al-O-Si + HO → Al(OH)-O(H)-Si). The (100) surface demonstrates relatively stable Al-O-Si bridging bond networks due to its lower surface energy, resulting in less extensive hydrolysis compared to other surfaces. On the (010) surface (Fig. 11b), partially unsaturated coordination states emerge from fractured silica-oxygen tetrahedral units. During hydrolysis, dissociated water molecules generate hydroxyl groups that directly coordinate with Si atoms to form surface silanol groups ([SiO] + H₂O → [SiO(OH)] + H). Released protons migrate through surface diffusion to adjacent Al-O-Si bridge oxygen sites, forming protonated bridge oxygen structures (Al-O(H)-Si). Notably, no sulfate ion participation was observed in surface reactions here, consistent with previous discussions. The (001) surface exhibits distinct site-selective hydrolysis mechanisms (Fig. 11c). Upon hydrolysis initiation, hydroxyl groups coordinate with surface Al atoms to form Al-OH complexes ([AlO] + HO → [AlO(OH)] + H), while most protons are captured by terminal oxygen atoms in [SiO] units to generate Si-OH terminations (Fig. 2c(vi)). This site selectivity reflects the screening effect of the topological structures of different albite surfaces on proton migration pathways.
In this study, ReaxFF MD simulations were performed to investigate the anisotropic interfacial reaction mechanisms of albite with sulfate-rich solutions across its (100), (010), and (001) crystallographic surfaces. The simulation results reveal that fundamental differences in the stability and early-stage interfacial kinetic behavior of different surfaces of the albite mineral are governed by atomic coordination states, surface topological configurations, and sulfate interactions. Specifically, the (001) surface exposes a significant number of non-bridging oxygen atoms and undercoordinated Al atoms, promoting sulfate coordination adsorption and Na leaching. Dynamic analysis reveals that Na desorption is partially driven by sulfate-mediated coordination reactions, while another portion occurs through ion exchange mechanisms, accompanied by the rupture and reconstruction of surface Si-O-Al bonds. In contrast, the (100) surface maintains structural stability through a dense hydrogen bond network and Na shielding effects, which suppress Ca adsorption and hydrolysis reactions. Meanwhile, the (010) surface restricts ion migration through the dense formation of silanol groups, with its surface Al atoms, embedded within the silica tetrahedral framework, showing almost no interaction with sulfate ions. Atomic-scale charge analysis highlighted electron transfer mechanisms. Al atoms lost electrons through coordination with water or sulfate, increasing their positive charge, while Si charge variations depended on neighboring oxygen atom density. MSD analysis further revealed anisotropic diffusion behaviors; the (001) surface displayed enhanced vertical mobility due to its layered structure and weak lattice constraints, whereas the (010) surface strongly suppressed ion migration due to the confinement of the dense structural network.
In summary, this study provides a novel perspective for investigating the subtle dissolution mechanisms at mineral-solution interfaces and pioneers initial pathways toward establishing a comprehensive theoretical framework for phenomena related to the different surface stabilities in mineral phases. These atomic-scale insights illuminate how crystallographic orientation dictates initial surface reconstruction dynamics and intrinsic stability. Regarding engineering implications, while the high initial reactivity of the (001) surface could accelerate early-stage mineral carbonation in CO sequestration, its propensity to be rapidly consumed or passivated during prolonged dissolution necessitates careful evaluation of long-term sustainability, potentially requiring strategies to harness the stability of less reactive faces like (100). Conversely, the demonstrated ion-shielding capability of the (100) surface aligns with the long-term geochemical stability demands of nuclear waste barrier materials. Future research will further integrate experimental validation and extend to multicomponent solution environments and longer simulation timescales to comprehensively evaluate mineral evolution in complex geological and engineered scenarios.