An atomic surface of an aluminum alloy induced by novel green chemical mechanical polishing using hybrid rare earth abrasives and mechanisms unraveled

Abstract

An aluminum (Al) alloy is a soft, plastic-like metal that is prone to embedding abrasives, scratches, corrosion pits, and deformation. Achieving an atomically smooth surface on an Al alloy presents a significant challenge. This study introduces a novel green chemical mechanical polishing (CMP) technique using hydrogen peroxide, tyrosine, sodium carbonate, and hybrid abrasives composed of silica, yttria, and ceria. The method produces a surface roughness (S_a) of 0.187 nm over a 50 × 50 μm² scanning area, with a material removal rate of 17.23 μm h⁻¹. Transmission electron microscopy (TEM) analysis shows a damaged layer thickness of 3.6 nm. To the best of our knowledge, this work reports the lowest S_a and damaged layer thickness for Al alloys to date. Molecular dynamics simulations are used to elucidate the mechanism of dynamic material removal during nanoscratching. As the cutting depth increases from 2.5 to 3 nm, the thickness of the damaged layer varies from 3.5 to 4.3 nm, aligning well with the TEM findings. The maximum von Mises stress recorded is 7.98 GPa, with the appearance of dislocation loops, vacancy defects, stacking faults, and an amorphous phase. The zeta potential measurements are −22.78, −9.46, −34.03, and −41.55 mV for silica in water, silica and yttria in water, silica at a pH of 10, and silica and yttria at a pH of 10, respectively. These correspond to polydispersity indices of 0.281, 0.412, 0.231, and 0.185. These data indicate that the hybrid abrasives of silica and yttria in a sodium carbonate solution exhibit the best stability and dispersion among the tested solutions. Characterization techniques, including X-ray photoelectron spectroscopy, Fourier transform infrared spectroscopy, and Raman spectroscopy, demonstrate the role of hydrogen peroxide in oxidizing the Al alloy surface to form aluminum hydroxide, alumina, and silica. Proposed complexing formulas suggest that tyrosine forms complexes with Al³⁺ ions, while ceria reacts with silica. The removal of oxides and complexes is facilitated by hybrid abrasives. The proposed novel green CMP method offers a new pathway for achieving atomic-level surface smoothness on Al alloys, potentially enhancing their application in high-performance devices.

---

1. Introduction

Aluminum (Al) alloys, particularly AlSi10Mg fabricated via additive manufacturing, are increasingly employed in aerospace optical systems due to their superior reflectivity, lightweight nature, and excellent formability.^1,2 However, the intricate microstructural characteristics of these alloys, including cellular substructures and the heterogeneous distribution of Si phases, pose significant challenges in achieving ultra-smooth surfaces required for high-precision optical applications.^3,4 Surface imperfections such as pits, scratches, and subsurface damage often arise from conventional mechanical processing due to the alloys’ inherent softness and chemical reactivity.^5,6 These defects not only compromise the optical performance but also limit the applicability of Al-based components in visible light imaging systems.⁷ Although advanced techniques such as single-point diamond turning,^8,9 ion beam polishing,¹⁰ and elastic emission machining¹¹ have been employed to improve the surface quality, they often suffer from drawbacks including surface waviness, low material removal rates, and high operational costs.¹² In this context, chemical mechanical polishing (CMP) has emerged as a promising alternative for Al alloy surface finishing, offering the potential for atomic-level smoothness through the synergistic effect of mechanical abrasion and chemical passivation.¹³ Despite its advantages, the CMP mechanism of additively manufactured Al alloys, particularly those with silicon-rich microstructures, remains inadequately understood. This highlights the need for in-depth investigations into the chemo-mechanical interplay at both the interface and subsurface levels.

Chemical components are crucial for the effectiveness of CMP, particularly for reactive metals like aluminum.^14,15 Previous studies have explored various oxidants, including potassium permanganate,¹⁶ ferric nitrate,¹⁷ and hydrogen peroxide (H₂O₂).¹⁸ These oxidants aim to create passivated oxide layers that allow for controlled material removal. However, some oxidants can be unsafe or lack selectivity towards aluminum, resulting in uneven polishing and surface defects. In recent studies, complexing agents like amino acids have been introduced to enhance the complexation and dissolution of aluminum oxide.^19,20 Nevertheless, most of these agents, such as glycine²¹ and serine,²² have limited solubility or weak interactions with oxidized aluminum, which restricts their effectiveness in alkaline environments. In this work, a novel chemical system is presented that combines hydrogen peroxide (H₂O₂) and tyrosine.²³ In this system, H₂O₂ serves as a mild oxidant that forms a soft oxide layer, while tyrosine acts as an effective complexing agent for aluminum ions. The synergistic interaction between these two agents ensures a controlled balance between oxidation and dissolution, leading to improved surface smoothness and reduced damage.

The selection and dispersion stability of abrasive particles are crucial for the effectiveness of the CMP technique. Most commonly used abrasives in CMP include silica (SiO₂),²⁴ alumina (Al₂O₃),²⁵ and ceria (CeO₂)²⁶, each chosen for their tunable hardness and surface reactivity. SiO₂ is preferred for its relatively gentle mechanical action, while CeO₂ is known for its redox-assisted reactivity with silicon (Si)-containing phases.²⁷ However, conventional abrasive systems often face challenges such as poor dispersion stability and particle agglomeration, which can lead to uneven surfaces and random scratches.²⁸ While several dispersants have been suggested, few provide long-term colloidal stability without compromising the slurry's reactivity.^29,30 Furthermore, many studies neglect the importance of interface compatibility between abrasives and multiphase alloys, which is essential for effectively polishing composites like AlSi10Mg. To address these limitations, yttria (Y₂O₃) is proposed as a co-abrasive. Under alkaline conditions, Y₂O₃ enhances the dispersion of SiO₂ through electrostatic and steric stabilization.³¹ The inclusion of CeO₂ also provides targeted chemical reactivity with Si-rich phases. This combination aims to achieve uniform material removal while minimizing polishing defects.

Subsurface damage (SSD) is another critical factor influencing the performance of Al–Si alloy components, particularly in precision optical applications.^32,33 During CMP and other material removal processes, localized stress concentrations can lead to plastic deformation, dislocation generation, and in severe cases, amorphization beneath the polished surface.^34,35 Previous studies have investigated SSD through both experimental techniques, such as transmission electron microscopy (TEM), and computational methods, including molecular dynamics (MD) simulations.^36,37 These investigations have provided valuable insights into dislocation nucleation, fault formation, and the transition from crystalline to amorphous states under mechanical loading.³⁸ However, much of the previous research has concentrated on monolithic or Si-dominant systems,³⁹ with limited attention given to the complex, multiphase nature of Al–Si alloys and their heterogeneous mechanical responses. Moreover, the coupling between abrasive-induced stress fields and phase-specific damage mechanisms remains poorly understood. A more comprehensive understanding of SSD evolution at different removal depths and across phase interfaces is essential for optimizing slurry design and minimizing degradation in optical performance.

This work developed a novel CMP slurry system tailored for the surface finishing of the AlSi10Mg alloy. This innovative slurry comprises H₂O₂, tyrosine, sodium carbonate (Na₂CO₃), SiO₂, Y₂O₃, CeO₂, and deionized water. It is chemically benign and compositionally optimized to promote stable dispersion, controlled chemical reactivity, and mechanical compatibility with multiphase Al–Si microstructures. In this study, a series of advanced testing and surface characterization techniques was employed to elucidate the material removal mechanisms and subsurface responses of AlSi10Mg. This integrated approach provides a comprehensive framework for understanding and optimizing the chemo-mechanical interactions during CMP, ultimately aiming to achieve defect-free, ultra-smooth surfaces.

2. Experimental details

2.1. Materials and methods

Spherical abrasives of SiO₂, Y₂O₃, and CeO₂, having an average diameter of approximately 20 nm, were purchased from Zhongke New Materials Co., Ltd. Specifically, H₂O₂ (≥30% content), tyrosine (≥99% purity), and Na₂CO₃ (≥99.8% purity) were sourced from Shanghai National Pharmaceutical Reagent Co., Ltd. All experimental solutions were prepared using deionized water to ensure consistent purity and reproducibility. AlSi10Mg alloy specimens with dimensions of 10 mm × 10 mm × 3 mm were used in this work. The detailed elemental composition of the AlSi10Mg alloy is provided in Table 1.

Table 1 Main elements and contents of the AlSi10Mg alloy

Element	Si	Mg	Fe	Mn	Al
Content (wt%)	9–11	0.2–0.45	≤0.55	≤0.45	Bal.

---

2.2. CMP experiments

Both grinding and polishing experiments were conducted using a precision lapping and polishing machine (UNIPOL-1200S, Shenyang Kejing Automation Equipment Co., Ltd). To ensure the process stability, three AlSi10Mg specimens were uniformly mounted onto the carrier plate. The slurry used for lapping consisted of 6 wt% SiO₂ (500 nm) dispersed in deionized water. The grinding process aims to remove larger scratches and protrusions, thereby enhancing the efficiency of the subsequent polishing process. The lapped surface serves as the initial surface for the subsequent polishing process. Before polishing, the CMP abrasives were ultrasonically dispersed for 40 minutes using an ultrasonic cleaner. Chemical reagents were then added to the CMP abrasives to create the CMP slurry. To ensure the stability and uniformity of the CMP slurry, a magnetic stirrer was used for continuous stirring throughout the polishing process. The polishing was conducted at a pressure of 30 kPa and a rotational speed of 75 rpm for 15 minutes, with a slurry supply rate of 12 mL min⁻¹. After polishing, the sample surfaces were promptly cleaned with deionized water and ethanol.

As illustrated in Table 2, eight experiments with different parameters were designed to investigate the influence of the components of the CMP slurry on polishing performance. Among them, slurries S1–S4 explored how additives enhanced the dispersion of SiO₂ abrasives, thereby reducing scratches and pits on the polished surface. Slurries S5–S8 investigated the chemical reaction mechanisms of the additives, aiming to achieve an excellent S_a and material removal rate (MRR).

Table 2 Composition and concentration of the CMP slurry

Group	SiO2 mass fraction (wt%)	Y2O3 mass fraction (wt%)	CeO2 mass fraction (wt%)	Na2CO3 pH	H2O2 mass fraction (wt%)	Tyrosine mass fraction (wt%)
S1	4.0	0	0	0	0	0
S2	4.0	1.0	0	0	0	0
S3	4.0	0	0	10.0	0	0
S4	4.0	1.0	0	10.0	0	0
S5	4.0	1.0	0	10.0	1.25	0
S6	4.0	1.0	0	10.0	0	1.5
S7	4.0	1.0	0	10.0	1.25	1.5
S8	4.0	1.0	1.0	10.0	1.25	1.5

---

MRR is an important metric for assessing the polishing performance of the CMP slurry. The mass of the specimens was measured before and after polishing using a precision electronic balance (ME204E, Mettler-Toledo), with each measurement being repeated three times for accuracy. The formula used to calculate the MRR is as follows:⁴⁰

where, Δm represents the mass difference of the sample before and after polishing, ρ (2.7 g cm⁻³) is the sample density, S denotes the polishing area, and t is the polishing time.

2.3. Simulation model and method

In this study, an atomic-scale simulation model of the Al–Si alloy system was developed using the ATOMSK. Al atoms were arranged in a face-centered cubic (FCC) lattice configuration, while Si atoms followed a diamond cubic crystal structure, consistent with their respective crystallographic characteristics.⁴¹ For the aluminum domain, a lattice constant of a = 4.05 Å was adopted, with the model dimensions set to 35a × 39a × 20a. Similarly, the Si region was modeled using a lattice constant of b = 5.43 Å and dimensions of 10b × 29b × 15b. The constructed Al and Si models were merged along the X-axis to create a composite Al–Si–Al configuration. This configuration represents a simplified model of the segregated microstructure, where Si phases are embedded within an Al matrix, resembling the typical microstructural features of AlSi10Mg alloys, known for Si segregation within the Al-rich matrix. The atoms in the model were categorized into three distinct regions from bottom to top, namely the boundary layer, the thermostat layer, and the Newtonian layer. The dimensional mismatch between Al and Si regions in the Y and Z directions was less than 0.6%, which is sufficiently small to apply periodic boundary conditions without causing significant lattice distortion.

Molecular dynamics (MD) simulations of the constructed model were conducted using the Large-scale Atomic/Molecular Massively Parallel Simulator (LAMMPS).⁴² The initial structure was first relaxed using the conjugate gradient method to reduce the potential energy, thereby eliminating structural artifacts and relieving internal stress. To ensure adequate system equilibration, a relaxation process was first performed using the canonical ensemble (NVT) at 300 K for 50 ps. This was followed by a second relaxation phase under the isothermal-isobaric ensemble (NPT), also maintained at 300 K and 0.1 bar, for an additional 50 ps. Upon completion of these two sequential steps, the system achieved equilibrium, thereby allowing for the commencement of subsequent cutting simulations.

The particle was modeled using a repulsive force potential, as described by eqn (2).⁴³ This approach has been extensively applied in MD simulations involving nanoindentation and abrasive particle cutting.

where K is the specified force constant, which is set to 10.0 eV Å⁻³, r represents the distance of an atom from the virtual abrasive grain center, and R denotes the abrasive particle radius.

The interatomic interactions in the simulation model were described using the Modified Embedded Atom Method (MEAM) potential. The general form of the MEAM potential is given by eqn (3).⁴⁴

where F represents the embedding energy dependent on the local electron density ρ, ϕ denotes the pair interaction, and r_ij is the distance between atom i and atom j.

2.4. Characterization techniques

The morphology and particle size of the abrasive were characterized using field emission scanning electron microscopy (FESEM; NOVA Nano SEM 450, FEI, USA). The value of S_a was assessed with a non-contact optical profiler (SuperView W1 series, CHOTEST, China) and atomic force microscopy (AFM; Nanowizard4XP, Bruker, Germany). Optical microscopy (MX 40, Olympus, Japan) was employed to compare surface topography before and after polishing. To evaluate the thickness of the subsurface damage layer on AlSi10Mg, high-resolution transmission electron microscopy (HRTEM; JEM-F200, JEOL, Japan) was utilized, with specimens prepared using a dual-beam focused ion beam system (FIB; Helios G4 UX, FEI, USA).

The properties of the slurry were evaluated using a nanoparticle size and zeta potential analyzer (Zetasizer Pro, Malvern, UK) to determine the zeta potential and polydispersity index (PI). The electrochemical behavior was examined through Tafel polarization and electrochemical impedance spectroscopy (EIS), conducted on an electrochemical workstation (CHI760E, Shanghai Chenhua Instrument Co., Ltd). The measurements utilized a three-electrode configuration, consisting of a saturated calomel electrode as the reference electrode, a platinum counter electrode, and the sample acting as the working electrode.

The surface elemental composition was examined using X-ray photoelectron spectroscopy (XPS; K-Alpha, Thermo Scientific, USA), with the carbon (C) 1s peak at 284.8 eV serving as the reference for charge correction. Changes in functional groups were analyzed through Fourier-transform infrared spectroscopy (FTIR; Nicolet iS20, Thermo Scientific, USA), while structural alterations were characterized by Raman spectroscopy (DXR, Thermo Fisher, USA).

3. Results and discussion

3.1. CMP experimental results

To identify the optimal CMP slurry formulation for simultaneously minimizing S_a and maximizing the MRR, four single-factor experiments were performed, each varying by only one parameter. Fig. 1 demonstrates the influence of these factors on the S_a and MRR. Specifically, Fig. 1(a) reveals a direct proportionality between the MRR and the SiO₂ concentration, with 4 wt% SiO₂ exhibiting the highest polishing efficiency. Lower concentrations of the abrasive struggled to effectively remove surface defects, resulting in diminished polishing efficiency. Conversely, higher concentrations led to a significant abrasive accumulation, exacerbating surface defects and increasing S_a. As illustrated in Fig. 1(b), an appropriate increase in pH enhanced both the MRR and S_a. However, at a pH of 11, the corrosion of the material surface intensified, and the effects of chemical corrosion outweighed mechanical removal, leading to a poorer S_a. The presence of H₂O₂ induced the formation of an oxide film on the surface that protected the substrate, as shown in Fig. 1(c). With a low mass fraction of H₂O₂, this oxide film was relatively thin and delivered limited protection. In contrast, a higher mass fraction of H₂O₂ resulted in a thicker oxide film, which hindered the progress of chemical reactions. The impact of the mass fraction of tyrosine on the MRR and S_a is depicted in Fig. 1(d), revealing 1.5 wt% as an optimal concentration. Lower concentrations of tyrosine formed complexes with Al ions to facilitate the reaction, while higher concentrations adhered to the material surface, inhibiting the reaction. In summary, by carefully controlling the mass fraction of SiO₂ (4 wt%), slurry pH (pH = 10), H₂O₂ concentration (1.25 wt%), and tyrosine concentration (1.5 wt%), this combination of parameters effectively balances S_a reduction and MRR enhancement, leading to significant improvements in both surface quality and polishing efficiency.

Fig. 1 The impact of different components on the S_a and MRR. (a) SiO₂ concentration. (b) Slurry pH. (c) H₂O₂ concentration. (d) Tyrosine concentration.

To enhance the dispersion and stability of nano-SiO₂ abrasives, SEM characterization was performed on SiO₂ slurries containing various additives. The zeta potential and PI values were also measured and analyzed.⁴⁵ The measurement and characterization results for slurries S1–S4 are presented in Fig. 2. The SEM images of the abrasives in slurries S1–S4 are presented in Fig. 2(a–d), highlighting variations in particle size and aggregation degree, which directly impact the performance of the CMP process. It is evident from Fig. 2(a) that pure nano-silica abrasives exhibited significant aggregation and uneven dispersion, leading to larger agglomerates that caused non-uniform surface quality during polishing. However, the addition of Y₂O₃ abrasives exacerbated this situation, as shown in Fig. 2(b). In contrast, Fig. 2(c) illustrates that the dispersion of SiO₂ abrasives is enhanced under alkaline conditions. Interestingly, in Fig. 2(d), the presence of Y₂O₃ abrasives further boosted the dispersion of SiO₂ abrasives under these alkaline conditions. Overall, both the alkaline environment and the synergistic effect of Y₂O₃ contributed to increased electrostatic repulsion, significantly improving dispersion stability. This enhancement is essential for achieving high-quality polishing performance.

Fig. 2 SEM images and property characterization of the abrasives in slurries S1–S4. (a) S1 slurry. (b) S2 slurry. (c) S3 slurry. (d) S4 slurry. (e) PI value. (f) Zeta potential.

Fig. 2(e) illustrates that the PI value of slurry S4 was the lowest, indicating that the dispersion of the SiO₂ abrasives, sodium carbonate, and the Y₂O₃ abrasive slurry was relatively effective. Additionally, the absolute zeta potential of slurry S1 increased from 22.78 mV to 41.55 mV upon the addition of sodium carbonate and Y₂O₃ abrasives. This implies that these additives enhanced the dispersion and stability of the SiO₂ abrasives, as observed from Fig. 2(f). A comprehensive analysis of both the PI value and zeta potential measurements revealed that combining Y₂O₃ abrasives and sodium carbonate solution with SiO₂ abrasives enhanced the dispersion and stability of pure SiO₂ abrasives. This improvement can be attributed to the transportation of a negative charge on the surface of Y₂O₃ abrasives under alkaline conditions, which collectively increased the electrostatic repulsion between the similarly charged SiO₂ abrasives. These findings underscore the importance of optimizing both the chemical composition and dispersion properties of CMP slurries for the effective polishing of AlSi10Mg alloys.

The polishing efficiency of various slurry formulations, as shown in Table 2, was systematically investigated using AlSi10Mg substrates through controlled experimental trials. The resulting surface morphology and S_a outcomes are presented in Fig. 3. The S_a value was precisely measured over an area of 50 × 50 μm². Notably, slurry S8 achieved the optimal S_a and effectively improved the polishing efficiency, with S_a reaching an impressive 0.187 nm and the MRR achieving an exceptional 17.23 μm h⁻¹.

Fig. 3 Surface morphology and S_a of AlSi10Mg after polishing with S1–S8 slurries. (a–h) Surface morphology after polishing with S1–S8 slurries. Optical images (i) before and (j) after polishing with slurry S8. AFM morphology after polishing with slurry S8 is shown in (k) 2D and (l) 3D patterns.

After polishing with slurries S1–S4, the value of S_a continuously decreased, and the number of scratches on the surface significantly reduced. This indicated that improving the dispersion of the abrasives enhanced polishing performance. Both Fig. 3(e) and (f) demonstrate that the addition of H₂O₂ and tyrosine improved S_a and accelerated the removal of surface protrusions. This phenomenon may be attributed to the reaction between H₂O₂, tyrosine, and the material surface, forming a soft layer. Fig. 3(g) provides strong evidence of the significant impact of the synergistic effect of H₂O₂ and tyrosine on S_a. While this finding is important, the underlying mechanism of this synergistic effect still requires further investigation and analysis. With the inclusion of CeO₂ abrasives, S_a reached its lowest value, as displayed in Fig. 3(h). This result was closely related to the favorable chemical reaction between CeO₂ and Si. The optical images clearly revealed numerous scratches on the surface before polishing, while the surface became extremely smooth after polishing, as shown in Fig. 3(i) and (j). The AFM analysis in Fig. 3(k) and (l) demonstrated the exceptional surface quality achieved after polishing, with a measured S_a of 0.201 nm and peak-to-valley height variation of approximately 0.12 nm. Extensive experimental results confirmed that the optimized CMP slurry successfully balanced the competing demands of minimal S_a and enhanced MRR, which is a critical requirement for precision surface finishing.

To evaluate the surface and subsurface features of AlSi10Mg before and after polishing, TEM samples were prepared using a focused ion beam (FIB) equipment. The cross-sectional view of the unpolished surface obtained via FIB preparation is shown in Fig. 4(a). Similarly, Fig. 4(b) revealed a damaged surface, with the thickness of the damage layer measuring approximately 315 nm. A comparison between the damaged and undamaged regions pinpointed a structural transformation from the original single-crystalline phase to a polycrystalline structure, as illustrated in Fig. 4(c) and (d). Meanwhile, a precipitate phase of Si was also identified in the undamaged area, which was confirmed through single-crystal electron diffraction patterns, as evident from Fig. 4(e) and (f). Further observations of the damaged region in Fig. 4(g) showed a high density of crystal defects in the outermost layer of the damage zone. High-resolution imaging of this damaged region, illustrated in Fig. 4(h), exposed a high density of edge dislocations within individual crystalline grains. These observations demonstrate that coarse grinding induces significant grain fragmentation and stress concentration at grain boundaries.⁴⁶ Notably, the residual stress exhibited a gradual attenuation with increasing depth from the surface. Post-polishing characterization showed remarkable improvement, with the damage layer thickness being reduced to 3.6 nm. The high-resolution images also displayed well-defined lattice fringes with a spacing of 0.268 nm, as depicted in Fig. 4(i). This confirms that there is minimal subsurface impact after polishing and that the crystal integrity is maintained, emphasizing the superior performance of slurry S8 in the CMP process.

Fig. 4 TEM characterization analysis of the AlSi10Mg surface before and after polishing. (a) Original FIB sample. (b) A cross-sectional view of a low magnification TEM image. Selected area electron diffraction patterns of (c) the damaged area and (d) the undamaged area are indicated by a red square and a light blue square in (b) respectively. (e) Precipitate phase of Si is marked by the orange square in (b). (f) Selected area electron diffraction pattern of Si is marked by the yellow square in (e). (g) Magnified HRTEM image of the crystal edge dislocations is marked by the light green square in (b). (h) The magnified HRTEM image of the edge dislocations is indicated by the green square in (g). (i) HRTEM image after CMP.

3.2. Subsurface damage mechanism analysis

The microstructural characteristics of the AlSi10Mg alloy are illustrated in Fig. 5, providing insights into its solidification features and elemental distribution. As shown in Fig. 5(a), the SEM image of the transverse cross-section revealed a distinct cellular structure, typical of alloys processed via laser powder bed fusion (LPBF). This morphology stemmed from rapid solidification and thermal gradients, forming α-Al-rich cellular cores surrounded by a continuous eutectic Si network at the boundaries.⁴⁷ These microstructural features significantly affected the mechanical properties and were also expected to influence the material's response during surface processing.

Fig. 5 Microstructural characteristics of the AlSi10Mg sample. (a) SEM image illustrating the cellular structure in the transverse cross-section. (b) STEM image of the longitudinal cross-section. (c and d) STEM-EDS elemental maps corresponding to (b).

A high-resolution scanning transmission electron microscopy (STEM) image of the longitudinal section is depicted in Fig. 5(b), highlighting the directional cellular growth along the build path. The corresponding STEM-EDS elemental mappings in Fig. 5(c) and (d) further confirmed the segregation behavior during solidification: Al dominated the cell interiors, while Si was enriched along the boundaries. The localized enrichment of Si at cell boundaries introduced chemical and mechanical heterogeneity across the surface, which was a critical factor during the CMP process.

From a CMP perspective, the heterogeneous microstructure of the AlSi10Mg alloy implied that its surface comprised regions with varying hardness and chemical reactivity.⁴⁸ The relatively soft α-Al matrix was more prone to mechanical abrasion and chemical oxidation, whereas the Si-rich boundaries, being solid and chemically inert, resisted removal and potentially acted as polishing endpoints or defect initiation sites. Therefore, understanding the distribution and morphology of these phases is essential for optimizing the slurry chemistry and process parameters. These observations highlight the necessity of tailoring CMP strategies for Al–Si alloys, particularly those fabricated by additive manufacturing, as microstructural anisotropy can further complicate surface planarization.

To elucidate the formation mechanism of subsurface damage layers during the CMP process of the AlSi10Mg alloy, abrasive particle cutting simulations were performed on the Al–Si alloy model. The SEM and STEM observations guided the establishment of this model. The Al–Si alloy model illustrated in Fig. 6 was constructed based on the characteristic microstructure of the AlSi10Mg alloy. The atomic-scale simulation results were post-processed and rendered using the Open Visualization Tool (OVITO) for defect analysis and visualization.⁴⁹ A virtual abrasive particle with a radius of 35 Å was employed in the simulation. In the cutting simulation, the virtual abrasive particle initially penetrated the material along the negative z-axis at a velocity of 25 m s⁻¹ until it reached the designated depth. Subsequently, it was transported laterally along the negative x-axis at a velocity of 100 m s⁻¹. To examine the effect of the indentation depth on the cutting behavior and material response, simulations were conducted at six distinct depths: 0.5 nm, 1.0 nm, 1.5 nm, 2.0 nm, 2.5 nm, and 3.0 nm.

Fig. 6 Parameters and region division of the simulation model.

Fig. 7 displays the tangential and normal forces exerted on the abrasive particle as it traverses across the AlSi10Mg surface at varying indentation depths. The x-axis denotes the lateral displacement of the abrasive, corresponding to its trajectory across different material regions, while the y-axis represents the measured forces during the cutting process. The force profiles demonstrated distinct mechanical responses across these regions. In Al domains (0–8 nm and 13–21 nm), both tangential and normal forces exhibited relatively smooth and moderate increases with greater penetration depth, indicating ductile cutting behavior. In contrast, a sharp spike in both forces was observed in the Si region (8–13 nm), particularly at deeper indentation depths. This abrupt rise reflected the high resistance of a solid and more brittle Si phase, suggesting localized stress concentration and potential initiation of brittle fracture.⁵⁰ Such segmented force evolution underscores the challenges of planarizing dual-phase systems like AlSi10Mg, where the embedded Si phase limits the ability to achieve low S_a and efficient material removal. The increased cutting resistance in the Si zone confirmed that Si was significantly more difficult to remove through purely mechanical means, which can lead to subsurface damage or inefficient planarization if not appropriately addressed. These findings emphasize the need for a selective chemical approach to assist in the removal of the Si phase during CMP.

Fig. 7 Stress curve of the abrasive grain in the cutting process. (a) Tangential force curve. (b) Normal force curve.

Fig. 8 illustrates the surface topography (a–f) and corresponding von Mises stress distributions (g–l) of the AlSi10Mg alloy under varying abrasive penetration and cutting depths. As the penetration and cutting depth increased, a clear progression in material deformation and damage morphology was observed.⁵¹ At shallow depths (0.5–1.0 nm), the surface displayed only minor ploughing marks with limited material pile-up. The contact zone remained relatively localized, suggesting primarily elastic to incipient plastic deformation. The associated von Mises stress distribution was concentrated directly beneath the abrasive tip, exhibiting a hemispherical stress field confined within the Al matrix.

Fig. 8 Surface morphologies at depth of cut of (a) 0.5, (b) 1, (c) 1.5, (d) 2, (e) 2.5, (f) 3 nm and their corresponding stress distributions (g)–(l).

As the cutting depth increased from 1.5 to 2.0 nm, the surface features became more pronounced, exhibiting broader material uplift and asymmetric groove profiles. This indicated enhanced plastic flow and directional shearing. In such cases, the von Mises stress fields expanded both laterally and vertically, forming stress “tails” that extended toward the Al–Si interface. Notably, the stress intensity was higher near the Si region, implying that interface mismatch began to influence the local mechanical response. At the maximum cutting depth between 2.5 and 3.0 nm, severe material distortion was observed. The grooves became deeper, with sharp material extrusion and increased surface roughness. The corresponding von Mises maps showed extensive stress accumulation along both the cutting direction and beneath the abrasive tip. The accumulation of significant stress directly affected the nature and extent of subsurface damage in the material, thereby influencing its performance during service.

The effect of stress accumulation on SSD was further analyzed using the lattice structure diagram of the model. The lattice structure analysis presented in Fig. 9 visually revealed SSD, including dislocations, dislocation loops, vacancies, atomic clusters, and stacking fault defects. As the cutting depth increased, the nature and extent of subsurface damage became increasingly complex. At lower cutting depths (0.5–1.0 nm), the primary damage modes observed in the Al phase were stacking faults and incipient dislocation loops. These features emerged due to the relatively soft and ductile nature of Al, which allowed for plastic accommodation through dislocation nucleation and glide.⁵² The presence of stacking faults—particularly noticeable near the cutting path—indicated that partial dislocations were activated in response to shear stresses generated by the abrasive particle.

Fig. 9 Distribution and types of subsurface damage under different cutting depths. (a) 0.5 nm. (b) 1 nm. (c) 1.5 nm. (d) 2 nm. (e) 2.5 nm. (f) 3 nm.

As the cutting depth increased to 1.5–2.0 nm, dislocation loops and stacking faults became more prominent and began to penetrate deeper into the material. Notably, closed dislocation rings were observed around the cutting trajectory in both the Al and Si regions, suggesting the presence of localized stress concentration and zones of plastic relaxation. This transition marked the onset of cross-interface interaction, where the stress field began to significantly affect the more brittle Si phase.

At the highest cutting depths of 2.5–3.0 nm, extensive lattice distortion was observed in both phases. In the Al region, dislocation tangles and dense stacking fault networks developed, indicating significant plastic deformation. Meanwhile, in the Si region, the damage mode transitioned from isolated dislocations to partial amorphization near the cutting zone. The loss of long-range order, as evident by irregular atomic arrangements and the absence of clear lattice planes, suggested that a mechanically induced amorphous layer was formed.⁵³ This phenomenon is characteristic of severe shear deformation in brittle materials subjected to high contact stress.

In summary, the formation of stacking faults, dislocation rings, and amorphous regions is strongly dependent on the cutting depth. These defects initiate sequentially with increasing depth, reflecting a transition from elastic/plastic deformation to fracture-dominated removal, particularly in the Si phase. The evolution of these subsurface features offers critical insight into the material removal mechanism in Al–Si alloys and reinforces the need for chemical-assisted cutting strategies when targeting the embedded Si phase.

3.3. Mechanism characterization of the CMP slurry

The chemical corrosion of the polishing slurry has a direct impact on the S_a and MRR.⁵⁴ To investigate these parameters post-CMP, the corrosion behavior of the AlSi10Mg alloy in different slurries under static conditions was studied using electrochemical measurements. The Tafel curves after immersion in slurries S4 to S7 are presented in Fig. 10(a), while the corresponding corrosion potentials and current densities are depicted in Fig. 10(b). The highest corrosion current was observed in the S4 slurry, indicating that the AlSi10Mg alloy experienced significant corrosion under alkaline conditions. However, the addition of either H₂O₂ or tyrosine substantially reduced the corrosion current, suggesting that both agents reacted with AlSi10Mg to form a protective layer, which inhibited further corrosion. Notably, the corrosion current and potential observed after immersion in the S7 slurry suggest that the synergistic effect of H₂O₂ and tyrosine in inhibiting corrosion was more pronounced than their individual effects.

Fig. 10 Electrochemical tests of AlSi10Mg specimens in different slurries. (a) Tafel plots. (b) Corrosion current densities and corrosion potentials. (c) Nyquist plots. (d) Bode plots.

The EIS results presented in Fig. 10(c) and (d) further corroborate the protective capabilities of different CMP slurries. In the Nyquist plot, the smallest semicircle observed for the S4 slurry indicated the lowest charge transfer resistance (R_ct), suggesting severe corrosion under alkaline conditions.⁵⁵ The addition of H₂O₂ and tyrosine in the S5 and S6 slurries increased R_ct, reflecting the formation of passivation or complexation layers. Remarkably, the S7 slurry exhibited the largest R_ct and impedance modulus in the Bode plot, along with the highest phase angle, confirming that the synergistic effect of H₂O₂ and tyrosine led to the formation of a compact and protective interfacial layer that effectively inhibited the corrosion of AlSi10Mg.

The XPS full spectra of AlSi10Mg analyzed before and after immersion in the CMP slurry are shown in Fig. 11(a). After immersion in slurry S5, a significant increase in oxygen (O) intensity was observed, confirming the strong oxidative effect of H₂O₂. Following immersion in slurry S6, a noticeable reduction in the Al signal was observed, indicating that the slurry constituents chemically interacted with the Al surface. Furthermore, the appearance of a nitrogen (N) peak was attributed to the presence of tyrosine.⁵⁶ A significant change in the content of Si and SiO₂ before and after polishing suggests that the CMP process effectively removed the surface oxide layer on Si, revealing a smoother underlying Si surface, as presented in Fig. 11(b) and (c).

Fig. 11 XPS spectral analysis of AlSi10Mg samples in the developed CMP slurry. (a) XPS full spectrum analysis of samples before and after immersion. (b) Si 2p spectra before polishing, (c) Si 2p spectra after polishing. (d–f) Al 2p spectra before and after immersion. (g–i) O 1s spectra before and after immersion.

The Al 2p spectra of the samples before and after immersion in the CMP slurry are shown in Fig. 11(d)–(f). The peaks located at 74.38, 74.58, and 74.58 eV indicated the presence of Al³⁺, whereas those at 71.88, 72.28, and 72.18 eV were associated with metallic Al.⁵⁷ The decreased intensity of Al indicates that the Al surface was either oxidized or complexed. Both H₂O₂ and tyrosine exerted beneficial regulatory effects on Al. Similarly, as shown in Fig. 11(g)–(i), the peaks located at 531.28, 531.78, and 531.68 eV were associated with Al₂O₃, and those at 531.98, 532.08, and 532.18 eV were attributed to Al(OH)₃. Before immersion, a dense oxide layer composed primarily of Al₂O₃ dominated the surface of AlSi10Mg. However, the introduction of H₂O₂ reversed this trend, resulting in a predominance of Al(OH)₃.⁴¹ The peak at 535.08 eV corresponded with the complexation between Al and tyrosine. Notably, the enhanced intensity of the Al–tyrosine complex and the reduced intensity of Al₂O₃ suggested that tyrosine dissolved the dense oxide layer before forming a chelate with Al ions. This reaction occurred continuously, enabling the rapid removal of surface asperities, which contributed to improved S_a and MRR values.

The surface chemical states of the AlSi10Mg alloy were analyzed both before and after immersion in the CMP slurry using XPS, FTIR, and Raman spectroscopy, as illustrated in Fig. 12. The survey spectra in Fig. 12(a) showed enhanced signals related to Ce following immersion, indicating the adsorption or deposition of Ce-containing species on the alloy's surface. This supports the role of CeO₂ abrasive in the surface modification process. The Ce 3d spectra in Fig. 12(b) exhibited characteristic peaks for both Ce³⁺ and Ce⁴⁺ after immersion, suggesting the presence of redox-active Ce species.⁵⁸ These redox interactions facilitated the chemical softening and removal of native silicon oxide layers, thereby accelerating the polishing of Si-rich regions and improving planarization. The chemical reaction equation between the CMP slurry and AlSi10Mg can be expressed as follows:⁵⁹

2Al + 3H₂O₂ → Al₂O₃ + 3H₂O (4)

Al₂O₃ + 3H₂O → 2Al(OH)₃ (5)

Al(OH)₃ + OH⁻ → [Al(OH)₄]⁻ (6)

Si + H₂O₂ → SiO₂ + H₂↑ (7)

Ce⁴⁺ + Si → Ce³⁺ + SiO₂ (8)

CeO₂ + SiO₂ + H₂O₂ → Ce³⁺ + SiO₂·H₂O (9)

Fig. 12 Chemical reaction mechanism analysis of AlSi10Mg during CMP. (a) XPS full spectra and (b) Ce 3d spectra before and after immersion in slurry S8. (c) FTIR spectra before and after polishing and immersion with the S8 slurry. (d) Raman spectra before and after immersion.

Fig. 12(c) displays the FTIR spectra of AlSi10Mg surfaces treated with S8 slurries. The broad absorption features around 3200 cm⁻¹ and 1450 cm⁻¹ were respectively associated with the O–H stretching and bending vibrational modes. Moreover, the absorption bands around 2930 cm⁻¹, 1580 cm⁻¹, and 1320 cm⁻¹ corresponded to the stretching vibrations of N–H, C=O, and C=C bonds, respectively.^60,61 These bands implied the existence of functional groups originating from tyrosine on the polished surface, indicating successful complexation and adsorption. Additionally, novel peaks observed at ∼560 cm⁻¹ were linked to Ce–O vibrations, further confirming CeO₂ deposition.⁶² These complementary spectroscopic results collectively demonstrated that the CeO₂ abrasive in the S8 slurry actively participated in surface reactions, particularly with Si components, while tyrosine and H₂O₂ contributed to aluminum surface modification. Likewise, Fig. 12(d) presents the Raman spectra of the surface before and after immersion. Subsequent immersion in the S8 slurry led to a significant change in the peak intensity and position, including the emergence of new peaks associated with Al–O and Ce–O bonds.⁶³ Overall, these spectroscopic results provide strong evidence for the synergistic roles of H₂O₂, tyrosine, and CeO₂ in forming a chemically active, protective, and removable surface layer that enhances CMP performance.

The complexation process between Al ions and varying numbers of tyrosine molecules is illustrated in Fig. 13. This interaction is pivotal for understanding the mechanism of corrosion inhibition, as the coordination of Al³⁺ with tyrosine molecules contributed to the formation of a protective layer on the AlSi10Mg surface. With the increasing number of tyrosine molecules involved, the stability of the resulting complexes was enhanced, effectively reducing the risk of further oxidation or corrosion of the Al substrate. This coordination was primarily expedited through the carboxyl (–COOH) and phenolic hydroxyl groups (–OH) of tyrosine, which exhibited strong affinity for Al ions. In some instances, the amino group (–NH₂) of tyrosine also participated in coordination, leading to a tridentate binding mode. Consequently, stable chelate rings were formed, effectively isolating the Al³⁺ ions from the corrosive environment.

Fig. 13 The complexation process between aluminum ions and different numbers of tyrosine molecules. (a) Ball-and-stick model of a tyrosine molecule. (b) Ball-and-stick model of ionized tyrosine. (c) Complexation reaction.

The electron-donating capacity of the tyrosine functional groups not only contributed to complex stability but also altered the local electronic structure of the aluminum surface, further enhancing passivation. Such a multilayered complexation and adsorption model provided a plausible explanation for the synergistic inhibition effect observed in the presence of both tyrosine and H₂O₂, as validated by the electrochemical and XPS analyses. This proposed mechanism underscores the significance of molecular design in CMP slurry formulations, where tailored ligands like tyrosine can serve dual functions: corrosion inhibition and surface modification. The multidentate complexation of tyrosine not only enhances aluminum ion immobilization but also favors the formation of a dense organic–inorganic hybrid film, which explains the improved surface smoothness (low S_a) and MRR after CMP using the S7 slurry.

The CMP mechanism of the AlSi10Mg alloy is illustrated in Fig. 14, based on findings from electrochemical testing, XPS, FTIR, and Raman spectroscopy. The CMP process for this Al–Si alloy, utilizing a composite slurry, operated through a synergistic removal mechanism governed by both chemical and mechanical interactions. Initially, H₂O₂ oxidized the surface of the material, creating a softened layer. Both Na₂CO₃ and Y₂O₃ significantly enhanced the dispersion of SiO₂ particles, which helped minimize abrasive agglomeration. This reduction in agglomeration decreased localized defects and enabled consistent material removal across the Al and Si regions.

Fig. 14 CMP mechanism for the developed optimal CMP slurry.

The chemical contributions of tyrosine and cerium oxide further modulated the CMP behavior. Tyrosine formed stable coordination complexes with surface Al atoms, weakening the native oxide layer and promoting efficient removal of the Al matrix under mild mechanical loading. This chelation-assisted polishing mechanism was manifested in the smoother Al regions shown in the magnified left inset of Fig. 12. In contrast, CeO₂ enhanced the oxidative dissolution of the more brittle Si phase through redox interactions with Si–O bonds. The right inset revealed local amorphization and nanoscale material removal in Si-rich areas, indicating the active chemical participation of CeO₂. Together, these processes created a balanced CMP mechanism wherein chelation facilitated Al removal, oxidation accelerated surface degradation of Si, and uniformly dispersed abrasives provided the mechanical force necessary for material removal. This integrated approach ensures high-quality surface finishing while maintaining selectivity between the softer Al matrix and the harder Si inclusions.

4. Conclusions

In this study, a novel CMP slurry was developed and evaluated for the ultra-precision finishing of the additively manufactured AlSi10Mg alloy. The proposed slurry exhibits excellent polishing performance and effective subsurface damage control. The following conclusions can be drawn:

(1) The optimized slurry yielded a minimum S_a of 0.187 nm within a 50 × 50 μm² area and a high MRR of 17.23 μm h⁻¹, meeting the demanding surface requirements for aerospace optical components.

(2) Under alkaline conditions, Y₂O₃ enhances the stability and dispersion of SiO₂ abrasives, as evidenced by a significant increase in zeta potential and a reduction in polydispersity index. This enhancement mitigated abrasive agglomeration and reduced surface defects during polishing.

(3) CeO₂ exhibited redox-assisted reactivity with Si phases, promoting selective material removal at phase interfaces and contributing to overall surface uniformity and polishing efficiency.

(4) The combination of hydrogen peroxide and tyrosine ensured a balanced oxidation–dissolution mechanism. Electrochemical tests confirmed their synergistic corrosion inhibition, while XPS, FTIR, and Raman analyses revealed the formation of surface oxide and complexation layers.

(5) TEM characterization and MD simulations revealed the evolution of subsurface damage under different cutting depths. Stacking faults, dislocation loops, and amorphization zones were observed, particularly near phase interfaces, offering valuable insight into phase-specific deformation responses during polishing.

Overall, this work presents a sustainable and high-efficiency CMP strategy tailored for Al–Si alloys with complex microstructures. The findings provide a robust framework for future slurry design and polishing optimization in high-precision optical and aerospace applications.

Author contributions

Zeyun Wang: investigation, formal analysis, data curation, and visualization. Zhenyu Zhang: funding acquisition, project administration, methodology, supervision, and conceptualization. Pengfei Hu: formal analysis and investigation. Ganggang Liu: formal analysis and investigation. Jianjun Hu: formal analysis, data curation, and visualization. Jianan Xu: formal analysis and data curation. Huaxiang Cai: formal analysis and data curation. Zehong Pang: formal analysis and investigation. Peng Ding: formal analysis and data curation.

Conflicts of interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data availability

The authors are unable or have chosen not to specify which data have been used.

Acknowledgements

The authors acknowledge the financial support from the Key Research and Development Program of Yunnan Province (202402AB080009), the Fundamental Research Funds for the Central Universities (DUT24YG209), the Science and Technology Special Fund of Hainan Province (ZDYF2024GXJS026), and the Changjiang Scholars Program of the Chinese Ministry of Education.

References

1. Q. Gu, Z. Zhang, H. Zhou, J. Yu, D. Wang, J. Feng, C. Shi, J. Yang and J. Qi, A novel approach of jet polishing for interior surface of small-grooved components using three developed setups, Int. J. Extreme Manuf., 2024, 6, 025101,  DOI:10.1088/2631-7990/ad1bba.
2. Z. Wu, J. Shen, X. Wu, Y. Peng, S. Lin, X. Huang, J. Liu and L. Huang, Research on polishing aluminum alloy optical elements with a new solid flexible bonnet tool, J. Manuf. Process., 2023, 103, 168–180,  DOI:10.1016/j.jmapro.2023.08.037.
3. J. Feng, Z. Zhang, S. Yu, X. Chen, D. Wang, Q. Gu, C. Zhou, T. Zhang and B. Liu, Novel multiphase jet polishing for complicated structured components produced by laser powder bed fusion, Addit. Manuf., 2023, 72, 103634,  DOI:10.1016/j.addma.2023.103634.
4. Q. Yan, B. Song and Y. Shi, Comparative study of performance comparison of AlSi10Mg alloy prepared by selective laser melting and casting, J. Mater. Sci. Technol., 2020, 41, 199–208,  DOI:10.1016/j.jmst.2019.08.049.
5. Z. Zhu, Z. Hu, H. L. Seet, T. Liu, W. Liao, U. Ramamurty and S. M. Ling Nai, Recent progress on the additive manufacturing of aluminum alloys and aluminum matrix composites: Microstructure, properties, and applications, Int. J. Mach. Tools Manuf., 2023, 190, 104047,  DOI:10.1016/j.ijmachtools.2023.104047.
6. C. Wu, J. Wen, J. Zhang, B. Song and Y. Shi, Additive manufacturing of heat-resistant aluminum alloys: a review, Int. J. Extreme Manuf., 2024, 6, 062013,  DOI:10.1088/2631-7990/ad7f2e.
7. J. Schröder, T. Fritsch, B. Ferrari, M. L. Altmann, G. Bruno and A. Toenjes, Laser powder bed fusion: Defect type influences critical porosity re-growth during reheating after hot isostatic pressing, J. Mater. Process. Technol., 2025, 340, 118839,  DOI:10.1016/j.jmatprotec.2025.118839.
8. Z. Xue, M. Lai, F. Xu and F. Fang, Influence factors and prediction model of surface roughness in single-point diamond turning of polycrystalline soft metal, J. Mater. Process. Technol., 2024, 324, 118256,  DOI:10.1016/j.jmatprotec.2023.118256.
9. Z. Li, F. Jiang, Z. Jiang, Z. Tian, T. Qiu, T. Zhang, Q. Wen, X. Lu, J. Lu and H. Huang, Energy beam-based direct and assisted polishing techniques for diamond: A review, Int. J. Extreme Manuf., 2023, 6, 012004,  DOI:10.1088/2631-7990/acfd67.
10. Z. Fan, Z. Li, R. Wang, N. Yu, M. Ren, X. Zhang and L. Zhu, A novel and efficient multi-jet plasma polishing process for optical fabrication, J. Mater. Process. Technol., 2025, 337, 118735,  DOI:10.1016/j.jmatprotec.2025.118735.
11. W. Ma, J. Li, X. Zhang, M. Ren and X. Hou, Manufacture of ultra-smooth surface with low damage by elastic emission machining, J. Mater. Process. Technol., 2025, 338, 118780,  DOI:10.1016/j.jmatprotec.2025.118780.
12. S. Gong, N. Takata, M. Kobashi and S. E. Shin, Wear properties of aluminum alloys fabricated by laser powder bed fusion, Tribol. Int., 2023, 187, 108769,  DOI:10.1016/j.triboint.2023.108769.
13. F. Zhao, Z. Zhang, H. Zhou, L. Chen, K. Hai, L. Wu, J. Yu, Z. Zhang and C. Fan, Unprecedented developed composite polishing system to achieve atomic surface integrating rough and fine polishing using a novel hyper-conjugated pad through controlling the temperature of a proposed green slurry, Adv. Compos. Hybrid Mater., 2025, 8, 234,  DOI:10.1007/s42114-025-01306-0.
14. Z. Luo, Z. Zhang, F. Zhao, C. Fan, J. Feng, H. Zhou, F. Meng, X. Zhuang and J. Wang, Advanced polishing methods for atomic-scale surfaces: A review, Mater. Today Sustain., 2024, 27, 100841,  DOI:10.1016/j.mtsust.2024.100841.
15. K. Yang, N. Huang, H. Di and P. Zhou, Modeling of surface microtopography evolution in chemical mechanical polishing considering chemical-mechanical synergy, Tribol. Int., 2025, 201, 110206,  DOI:10.1016/j.triboint.2024.110206.
16. L. Liao, Z. Zhang, F. Meng, D. Liu, B. Wu, Y. Li and W. Xie, A novel slurry for chemical mechanical polishing of single crystal diamond, Appl. Surf. Sci., 2021, 564, 150431,  DOI:10.1016/j.apsusc.2021.150431.
17. S. Mandal, E. L. H. Thomas, L. Gines, D. Morgan, J. Green, E. B. Brousseau and O. A. Williams, Redox agent enhanced chemical mechanical polishing of thin film diamond, Carbon, 2018, 130, 25–30,  DOI:10.1016/j.carbon.2017.12.077.
18. H. Yan, Y. Peng, Y. Huang, M. Shen, X. Wei, W. Zou, Q. Tong, N. Zhou, J. Xu, Y. Zhang, Y.-X. Ye and G. Ouyang, Enhancing Photosynthesis Efficiency of Hydrogen Peroxide by Modulating Side Chains to Facilitate Water Oxidation at Low–Energy Barrier Sites, Adv. Mater., 2024, 36, 2311535,  DOI:10.1002/adma.202311535.
19. X. Han, S. Zhang, R. Liu, F. Wang, B. Tan, X. Zhao, J. Zhao and Y. Shi, The effect of amino acid addition in CeO₂-based slurry on SiO₂/Si3N4 CMP: Removal rate selectivity, morphology, and mechanism research, J. Mol. Liq., 2024, 412, 125855,  DOI:10.1016/j.molliq.2024.125855.
20. T. T. H. Tran, J. Gichovi, J. Commane, E. J. Podlaha and J. Seo, Examining amino acids as environmentally friendly corrosion inhibitors for Cu and Co chemical mechanical planarization, J. Environ. Chem. Eng., 2024, 12, 113669,  DOI:10.1016/j.jece.2024.113669.
21. H. Sun, Z. Zhang, Z. Xue, H. Zhou, Z. Geng, C. Cheng, L. Chen and Y. Tian, Atomic surface of titanium alloy using novel chemical mechanical polishing with rare earth composite abrasives, J. Manuf. Process., 2025, 134, 79–89,  DOI:10.1016/j.jmapro.2024.12.036.
22. Z. Wang, Z. Zhang, H. Zhou, D. Han, C. Shi, L. Chen, J. Yao, S. Yu and J. Xu, Close atomic surface on aluminum alloy using green chemical mechanical polishing with synergistic effect between yttria and silica abrasives, Appl. Surf. Sci., 2025, 681, 161586,  DOI:10.1016/j.apsusc.2024.161586.
23. F. Li, L. Wei, Y. Liu, H. Deng, J. Cui, J. Wang and T. Xiao, Characterization of dissolved organic matter in rivers impacted by acid mine drainage: Components and complexation with metals, Sci. Total Environ., 2024, 926, 171960,  DOI:10.1016/j.scitotenv.2024.171960.
24. Y. Dong, H. Lei and W. Liu, Effect of mixed-shaped silica sol abrasives on surface roughness and material removal rate of zirconia ceramic cover, Ceram. Int., 2020, 46, 23828–23833,  DOI:10.1016/j.ceramint.2020.06.159.
25. S. Dai, J. Fu, H. Lei and Y. Chen, Study on the interaction between SiO₂ and ZrO2 in the chemical mechanical polishing of zirconia ceramic with colloidal silica, Ceram. Int., 2021, 47, 21642–21649,  DOI:10.1016/j.ceramint.2021.04.177.
26. K. Khivantsev, H. Pham, M. H. Engelhard, H. A. Aleksandrov, L. Kovarik, M. Bowden, X. S. Li, J. Tian, I. Z. Koleva, I. Song, W. Hu, X. Wei, Y. Sun, P. Tran, T. R. Graham, D. Jiang, D. P. Dean, C. J. Breckner, J. T. Miller, G. N. Vayssilov, J. Szanyi, A. Datye and Y. Wang, Transforming ceria into 2D clusters enhances catalytic activity, Nature, 2025, 640, 947–953,  DOI:10.1038/s41586-025-08684-x.
27. W. Zhang, P. Zhang, Y. Wu, X. Li, X. Wang, M. Wang, S. Jin, W. Bi, Y. Zhao, W. Zhou and D. Zhang, Preparation of CeO₂ Polishing Powder and Its Performance and Mechanism for Chemical Mechanical Polishing of Optical Glass, ACS Appl. Mater. Interfaces, 2025, 17, 20394–20410,  DOI:10.1021/acsami.4c21540.
28. N. Xu, Y. Lin, Y. Luo, Y. Huo, K. Gao, Z. Gao and Y. Pu, Effect of surfactants with different ionizing properties on dispersion stability and PCMP properties of CeO₂ nanoparticle polishing slurry, Ceram. Int., 2025, 51, 856–864,  DOI:10.1016/j.ceramint.2024.11.070.
29. S. Gao, B. Hu, X. Song, Z. Dong and R. Kang, Dispersal mechanism of different dispersants and its effect on performance of 4H-SiC polishing slurry, Ceram. Int., 2025, 51, 3943–3952,  DOI:10.1016/j.ceramint.2024.11.373.
30. A. Venkataronappa, J. Bankaitis and J. Seo, Enhancing dispersion stability and recyclability of ceria slurry with polyacrylic acid for improved glass polishing performance, J. Ind. Eng. Chem., 2024, 138, 623–631,  DOI:10.1016/j.jiec.2024.04.040.
31. Á.I Szabó, A. Csík, T. Fodor, K. Vad, M. Marsicki and Á.D Tóth, Tribological Investigation of the Surface Protective Layer-Forming Effect of a Nano-Sized Yttria–Silica Mixture as a Lubricating Oil Additive, Lubricants, 2025, 13, 28,  DOI:10.3390/lubricants13010028.
32. Y. Gu, L. Han, J. Lin, B. Sun, Y. Zhang, T. Gao, B. Fu and B. Yu, Surface modification mechanism of laser-assisted grinding process for high silicon aluminum alloy: A molecular dynamics study, Surf. Interfaces, 2025, 56, 105753,  DOI:10.1016/j.surfin.2025.105753.
33. S. Wang, Q. Zhang, Q. Zhao and M. Zhou, Investigate on surface/subsurface damage mechanisms and manufacturability of ultra-smooth surface in ultra-precision ductile grinding of sapphire optics, J. Mater. Res. Technol., 2025, 34, 1966–1989,  DOI:10.1016/j.jmrt.2024.12.213.
34. P.-C. Le, T.-T. Do, T.-H. Fang and C.-I. Lee, Polishing-induced material attrition in surface-texturing AlN using a nanoscale polishing tool: An atomic-scale understanding, Tribol. Int., 2024, 192, 109254,  DOI:10.1016/j.triboint.2024.109254.
35. S. Lu, J. Xia, J. Yu and Z. Wang, Investigating the partial plastic formation mechanism of typical scratches on silicon wafers induced by rogue particles during chemical mechanical polishing, Mater. Sci. Semicond. Process., 2024, 181, 108666,  DOI:10.1016/j.mssp.2024.108666.
36. J. Li, L. Dong, X. Zang, X. Zhang, W. Zhao and F. Wang, Study on micro-crack propagation behavior of single-crystal α-Ti under shear stress based on molecular dynamics, Mater. Today Commun., 2020, 25, 101622,  DOI:10.1016/j.mtcomm.2020.101622.
37. D. Cheng, J. Hong, D. Lee, S.-Y. Lee and H. Zheng, In Situ TEM Characterization of Battery Materials, Chem. Rev., 2025, 125, 1840–1896,  DOI:10.1021/acs.chemrev.4c00507.
38. B. Chen, Z. Wu, C. Liu, J. Zhang, X. Chen, J. Xiao and J. Xu, Molecular dynamics simulation for nanometric cutting of NiTi shape memory alloys at elevated temperatures, J. Manuf. Process., 2024, 124, 581–589,  DOI:10.1016/j.jmapro.2024.06.031.
39. Z. Song, X. Guo, W. Zhang, Z. Li, G. Fan and R. Kang, Material removal rate model of single crystal diamond near atomic scale chemical mechanical polishing based on SOA, J. Mater. Res. Technol., 2025, 36, 762–776,  DOI:10.1016/j.jmrt.2025.03.109.
40. C. Shi, Y. Fan, Z. Zhang, X. Deng, J. Yu, H. Zhou, F. Meng and J. Feng, Development of core–shell SiO₂@A-TiO2 abrasives and novel photocatalytic chemical machinal polishing for atomic surface of fused silica, Appl. Surf. Sci., 2024, 652, 159293,  DOI:10.1016/j.apsusc.2024.159293.
41. Y. Li, Z. Zhang, X. Deng, H. Zhou, Y. Gu, C. Shi, S. Liu, W. Wang and W. Zhang, Novel green chemical mechanical polishing for an aluminum alloy and mechanisms interpreted by molecular dynamics simulations and measurements, Surf. Interfaces, 2023, 42, 103493,  DOI:10.1016/j.surfin.2023.103493.
42. F. Zhao, Z. Zhang, X. Deng, J. Feng, H. Zhou, Z. Liu, F. Meng and C. Shi, Atomic surface achieved through a novel cross-scale model from macroscale to nanoscale, Nanoscale, 2024, 16, 2318–2336,  10.1039/D3NR05278H.
43. J. Li, L. Dong, X. Dong, W. Zhao, J. Liu, J. Xiong and C. Xu, Study on wear behavior of FeNiCrCoCu high entropy alloy coating on Cu substrate based on molecular dynamics, Appl. Surf. Sci., 2021, 570, 151236,  DOI:10.1016/j.apsusc.2021.151236.
44. B. Jelinek, S. Groh, M. F. Horstemeyer, J. Houze, S. G. Kim, G. J. Wagner, A. Moitra and M. I. Baskes, Modified embedded atom method potential for Al, Si, Mg, Cu, and Fe alloys, Phys. Rev. B:Condens. Matter Mater. Phys., 2012, 85, 245102,  DOI:10.1103/PhysRevB.85.245102.
45. X. Chen, Z. Zhang, L. Chen, X. Yang, B. Ju, W. Liu, X. Deng, M. Wang and Y. Xu, Atomic surface of fused silica with high material removal rate produced by novel chemical mechanical polishing using composite rare earth oxides, Colloids Surf., A, 2025, 712, 136420,  DOI:10.1016/j.colsurfa.2025.136420.
46. D. Liu, Z. Zhang, J. Feng, Z. Yu, F. Meng, C. Shi, G. Xu, S. Shi and W. Liu, Environment-friendly chemical mechanical polishing for copper with atomic surface confirmed by transmission electron microscopy, Colloids Surf., A, 2023, 656, 130500,  DOI:10.1016/j.colsurfa.2022.130500.
47. A. Hadadzadeh, B. S. Amirkhiz, S. Shakerin, J. Kelly, J. Li and M. Mohammadi, Microstructural investigation and mechanical behavior of a two-material component fabricated through selective laser melting of AlSi10Mg on an Al-Cu-Ni-Fe-Mg cast alloy substrate, Addit. Manuf., 2020, 31, 100937,  DOI:10.1016/j.addma.2019.100937.
48. L. Song, S. Yuan, L. Zhao, Y. Zhu, S. Liang, M. Huang, A. Simar and Z. Li, Deciphering phase stress partition and its correlation to mechanical anisotropy of laser powder bed fusion AlSi10Mg, Addit. Manuf., 2023, 71, 103594,  DOI:10.1016/j.addma.2023.103594.
49. H. Dai, F. Zhang and J. Chen, A study of ultraprecision mechanical polishing of single-crystal silicon with laser nano-structured diamond abrasive by molecular dynamics simulation, Int. J. Mech. Sci., 2019, 157–158, 254–266,  DOI:10.1016/j.ijmecsci.2019.04.027.
50. L. Zhao, L. Song, J. G. Santos Macías, Y. Zhu, M. Huang, A. Simar and Z. Li, Review on the correlation between microstructure and mechanical performance for laser powder bed fusion AlSi10Mg, Addit. Manuf., 2022, 56, 102914,  DOI:10.1016/j.addma.2022.102914.
51. A. K. Gain and L. Zhang, Tribological behavior of eutectic Al–12Si alloy manufactured by selective laser melting, Wear, 2023, 522, 204679,  DOI:10.1016/j.wear.2023.204679.
52. K. Zhu, L. Song, L. Zhao, Y. Zhu, S. Liang, M. Huang and Z. Li, Numerical study on the effect of crystallographic orientation on mechanical behavior and its anisotropy of laser powder bed fusion AlSi10Mg, J. Alloys Compd., 2024, 976, 173284,  DOI:10.1016/j.jallcom.2023.173284.
53. M. S. Bisht, V. Gaur and I. V. Singh, On mechanical properties of SLM Al–Si alloy: Role of heat treatment-induced evolution of silicon morphology, Mater. Sci. Eng., A, 2022, 858, 144157,  DOI:10.1016/j.msea.2022.144157.
54. H. Liu, Z. Zhang, X. Deng, J. Yu, C. Shi, H. Zhou, F. Meng and J. Feng, Atomic surface of cobalt-chromium-molybdenum alloy induced by novel green chemical mechanical polishing through controlling pH values and oxidation processes, Appl. Surf. Sci., 2024, 652, 159356,  DOI:10.1016/j.apsusc.2024.159356.
55. S. Fajardo, L. Miguélez, M. A. Arenas, J. De Damborenea, I. Llorente and S. Feliu, Corrosion resistance of pulsed laser modified AZ31 Mg alloy surfaces, J. Magnesium Alloys, 2022, 10, 756–768,  DOI:10.1016/j.jma.2021.09.020.
56. Z. Zhang, J. Cui, J. Zhang, D. Liu, Z. Yu and D. Guo, Environment friendly chemical mechanical polishing of copper, Appl. Surf. Sci., 2019, 467–468, 5–11,  DOI:10.1016/j.apsusc.2018.10.133.
57. X. Yu and G. Li, XPS study of cerium conversion coating on the anodized 2024 aluminum alloy, J. Alloys Compd., 2004, 364, 193–198,  DOI:10.1016/S0925-8388(03)00502-4.
58. X. Cui, Z. Zhang, S. Yu, X. Chen, C. Shi, H. Zhou, F. Meng, J. Yu and W. Wen, Unprecedented atomic surface of silicon induced by environmentally friendly chemical mechanical polishing, Nanoscale, 2023, 15, 9304–9314,  10.1039/D3NR01149F.
59. F. Li, Y. Bai, H. Hu, G. Qiao, L. Li, F. Zhang and X. Zhang, pH-driven interfacial bond dynamics enable high-efficiency low-damage polishing of fused silica with CeO₂ based slurries, J. Mater. Process. Technol., 2025, 341, 118896,  DOI:10.1016/j.jmatprotec.2025.118896.
60. J. Lu, R. Chen, H. Liang and Q. Yan, The influence of concentration of hydroxyl radical on the chemical mechanical polishing of SiC wafer based on the Fenton reaction, Precis. Eng., 2018, 52, 221–226,  DOI:10.1016/j.precisioneng.2017.12.011.
61. S. Zhang, F. Wang, B. Tan, X. Wang, H. Du, X. Han and R. Liu, Diethylenetriamine and serine as synergists used in FA/OII based alkaline solution to realize enhanced cleaning efficiency of the organic corrosion inhibitor, Colloids Surf., A, 2023, 674, 131952,  DOI:10.1016/j.colsurfa.2023.131952.
62. G. Xu, Z. Zhang, F. Meng, L. Liu, D. Liu, C. Shi, X. Cui, J. Wang and W. Wen, Atomic-scale surface of fused silica induced by chemical mechanical polishing with controlled size spherical ceria abrasives, J. Manuf. Process., 2023, 85, 783–792,  DOI:10.1016/j.jmapro.2022.12.008.
63. F. Li, Y. Bai, H. Hu, L. Li, F. Zhang, X. Luo and X. Zhang, Atomic-scale insights into the material removal mechanism of cerium oxide polished fused silica based on ReaxFF-MD, J. Manuf. Process., 2024, 132, 339–352,  DOI:10.1016/j.jmapro.2024.10.068.

---