Novel chemical mechanical polishing assisted by photocatalysis and shear-thickening for a free surface blade of a Ti alloy using ceria nano-abrasives

Abstract

Titanium (Ti) alloys have low thermal conductivity, suffer from tool wear and deformation of workpieces and are difficult-to-machine metals. This contributes to surface roughness, S_a > 240 nm of Ti alloys after mechanical polishing with a low material removal rate (MRR). With the addition of assisting energy fields, the MRR is usually lower than 7 μm h⁻¹. Nevertheless, there is a high demand to achieve S_a < 50 nm on a free surface blade to save energy and reduce the resistance of fluids. To address this challenge, novel photocatalytic shear-thickening chemical mechanical polishing (PSTCMP) was developed using a custom-made polisher. The new PSTCMP slurry contained ceria, corn starch, sodium bicarbonate and deionized water. After PSTCMP, the S_a and thickness of the damaged layer of a free surface blade of a Ti alloy decreased from 501.71 to 38.46 nm and from 634.79 to 7.83 nm, respectively, representing reductions of 92% and 99%. The MRR is 12.52 μm h⁻¹. To the best of our knowledge, both the S_a and MRR are the best published to date for a Ti alloy blade with a free surface. PSTCMP mechanisms were interpreted using first-principles molecular dynamics, X-ray photoelectron spectroscopy and Fourier transform infrared spectroscopy. Hydroxyl radicals were generated under ultraviolet irradiation on ceria with a size of 4.2 nm, oxidizing the surface of the Ti alloy and forming Ti–OH and Ti–O groups. A Ce–O–Ti interface bridge was produced between Ti–OH and Ce–OH, induced by the hydrolysis of ceria. Our findings provide a new way to fabricate nanometer-scale surface roughness on a free surface blade of a Ti alloy with a high MRR.

---

1. Introduction

Titanium (Ti) alloys exhibit high strength, excellent toughness, and exceptional fatigue resistance, maintaining stable mechanical properties even under extreme service conditions such as high temperature, high pressure, and severe corrosion.^1–3 Due to these characteristics, Ti alloys have been widely applied in aero-engine blades.^4,5 For aero-engine blades, superior surface quality is paramount for ensuring optimal engine performance and operational lifespan.^6,7 For example, in compressor blades, higher surface roughness increases viscous friction losses, fuel consumption, and noise.^8,9 Moreover, the damage layer induced by machining changes the microstructure and mechanical properties, decreasing fatigue life and resulting in crack initiation.¹⁰

As a key step in the manufacturing processes of a blade, polishing can effectively remove defects and improve surface roughness.^11–13 Current commonly used polishing methods for blades primarily include manual polishing, computerized numerical control polishing, robotic polishing, abrasive flow polishing, and immersion polishing. Zhang et al.¹⁴ proposed a blade polishing method based on a multi-axis machine tool, and the S_a of the Ti alloy blade after polishing was 0.4 μm. In robot polishing, Luo et al.¹⁵ made reasonable progress in planning for a robot abrasive belt polishing system, which reduced the S_a of the polished blade from 3.21 μm to 0.24 μm. Wang et al.¹⁶ used abrasive flow technology to Polish an integral impeller, and the results showed that the S_a was decreased from 0.76 μm to 0.30 μm after polishing. For an integral impeller, Liao et al.¹⁷ developed immersed chemical mechanical polishing, and the results reveal that the S_a was decreased from 1.6 μm to 0.5 μm after polishing. Although the above methods can effectively improve the surface quality of blades and impellers, it is difficult for these methods to achieve an S_a of less than 50 nm.

With the increasing use of non-Newtonian fluid materials, surface polishing technologies based on the shear-thickening effect have shown advantages in the polishing of alloys, ceramics, quartz, and other materials.^18–21 This technique utilizes a non-Newtonian fluid to form a flexible fixed-abrasive tool for polishing, ensuring good conformity to surface profiles. Shear thickening polishing (STP) has great potential in further improving the surface quality of blades. Compared with conventional polishing methods, a workpiece after STP has excellent surface quality, but the polishing effect is poor when the initial surface roughness is high. To improve the material removal ability, many researchers have applied reactive energy fields or introduced chemical reagents to STP.^22–26 Tian et al. combined the magnetorheological effect with the shear-thickening effect for the polishing of cylindrical Ti alloy surfaces, which improved the MRR while ensuring the surface quality.²⁷ For a Ti alloy cylinder, Wang et al. developed electrochemically assisted STP. After polishing, the S_a decreased from 100.10 nm to 10.20 nm.²⁸ In addition, they incorporated KHPO₄ into the STP slurry and optimized the slurry pH, reducing the S_a from 124.00 nm to 8.60 nm and increasing the MRR to 6.42 μm h⁻¹.²⁹ However, such methods still have certain limitations in improving the material removal ability and expanding the machinable range. Meanwhile, the above research is also limited to common cylindrical workpieces and attention to the subsurface damage layer of the workpiece after polishing is lacking.

In recent years, photocatalysis, as an advanced oxidation technology, has been applied to the polishing field to enhance polishing performance.^30–32 For example, Liu et al.³³ introduced ultraviolet (UV) light into a diamond polishing device. By controlling the graphitization rate of diamond, the MRR reached greater than 7.0 μm h⁻¹. Ge et al. developed a photocatalysis-assisted pulse jet polishing method. In polishing experiments on cobalt alloys, this method increased the MRR by approximately 20% and reduced the S_a by about 33.9%.³⁴ During polishing, hydroxyl radicals (˙OH) generated during photocatalytic reactions oxidize the workpiece surface to form a readily removable oxide layer, thereby enhancing the polishing efficiency. Compared with oxidation by chemical reagents such as hydrogen peroxide and sodium hypochlorite, photocatalytic oxidation is milder and can mitigate excessive corrosion of the workpiece surface,^35,36 thereby improving surface quality. Nevertheless, traditional photocatalysts have low utilization rates of light, and photogenerated electrons and holes are easy to reunite, which leads to limited photocatalytic reaction.³⁷ At the same time, the reaction mechanism between ˙OH and workpieces has not been studied.

Aimed at the polishing of Ti alloy blades, this study introduces UV light to the STP process and develops a photocatalytic shear-thickening chemical mechanical polishing (PSTCMP) method. Three kinds of nano CeO₂ with different particle sizes were prepared as abrasives and photocatalysts. The PSTCMP slurry consists solely of the abrasive, sodium bicarbonate, and corn starch, thereby avoiding the use of toxic or hazardous reagents. Additionally, this study investigates the effect of abrasive addition on rheological properties. Polishing experiments were conducted to evaluate the effects of abrasive type and concentration on polishing performance, and the damage layer was characterized by transmission electron microscopy. The chemical reaction mechanism in PSTCMP was revealed by first-principles molecular dynamics simulation and surface phase analysis. The results show that PSTCMP achieves high quality polishing of Ti alloy blades. The key mechanism involves surface oxidation induced by ˙OH, and the chelation reaction between CeO₂ abrasives and the oxide layer. Due to their higher chemical activity and photocatalytic properties, smaller-sized CeO₂ abrasives exhibit more pronounced advantages in PSTCMP.

2. Experimental and simulation methods

2.1. PSTCMP method

In the PSTCMP method (Fig. 1a), UV light is introduced as an auxiliary energy field to the STP, where the spindle of the equipment and the workpiece rotate in the same direction. Due to the relative motion between the workpiece and the slurry, the region of the workpiece surface in contact with the slurry undergoes shear thickening. In this region, the sharp increase in viscosity promotes the aggregation of solid particles into compact “particle clusters”, which encapsulate the CeO₂ abrasives (Fig. 1b). Within the slurry, CeO₂ functions in a dual role. It acts as an abrasive for material removal, and it also serves as a photocatalyst. As shown in Fig. 1c, the CeO₂ reacts with the slurry to generate ˙OH under UV light. The specific reactions are as follows:^30,38,39

CeO₂ + hν → e⁻ + h⁺ (1)

h⁺ + H₂O → ˙OH + H⁻ (2)

h⁺ + OH⁻ → ˙OH (3)

e⁻ + O₂ → ˙O₂⁻ (4)

˙O₂⁻ + H₂O → H₂O₂ + ˙OH (5)

Fig. 1 PSTCMP method: (a) schematic of the polishing apparatus, (b) schematic of the polishing mechanism, and (c) photocatalytic mechanism.

The ˙OH generated from the photocatalytic reactions chemically reacts with the workpiece surface to form an oxide layer. Polishing of the workpiece is achieved by the continuous mechanical removal of this oxide layer through the action of the CeO₂ abrasive. Meanwhile, studies have shown that CeO₂ abrasives can not only remove material through mechanical action but also eliminate it via reaction complexation due to their high surface chemical activity.^40–42

2.2. Preparation of the CeO₂ abrasive

The chemical reagents used for the preparation of the CeO₂ abrasive and the slurry included 30% aqueous ammonia (NH₃·H₂O) and anhydrous ethanol (C₂H₅OH), both purchased from China National Pharmaceutical Group Chemical Reagent Co., Ltd. Cerium nitrate hexahydrate (Ce(NO₃)₃·6H₂O) and sodium bicarbonate (NaHCO₃) were purchased from Aladdin Biochemical Technology Co., Ltd. All reagents were of analytical grade purity.

The CeO₂ abrasive was prepared via a chemical precipitation method. First, 300 g of Ce(NO₃)₃·6H₂O was added to a mixed solution of 500 mL of DI H₂O and 100 mL of ethanol, followed by ultrasonic stirring at room temperature for 15 min. After the complete dissolution of Ce(NO₃)₃·6H₂O, the mixed solution was placed on a magnetic stirrer with the stirring speed set to 600 rpm and the temperature maintained at 91 °C. Subsequently, 30 wt% NH₃·H₂O was added, and the reaction was allowed to proceed for 1 h. The resulting suspension was then centrifuged, washed, and dried, followed by calcination at 500 °C for 1 h to obtain the CeO₂ abrasive. Three different particle sizes of CeO₂, designated as 1-CeO₂, 2-CeO₂, and 3-CeO₂, were obtained by varying the amount of NH₃·H₂O added.

2.3. Experimental details

Fig. 2a–c present the PSTCMP experimental apparatus, which mainly includes a UV light, a polishing barrel, two drive motors, and a control panel. The slurry was composed of corn starch, abrasives, and NaHCO₃. Corn starch served as the thickening phase, with a mass ratio of corn starch to DI H₂O of 1.3 : 1. NaHCO₃ was used as a pH regulator. The detailed parameters of the PSTCMP experiment are presented in Table 1. After PSTCMP, the surface of the blades was rinsed with DI H₂O, and subsequently dried with compressed air. Fig. 2d shows the locations where the S_a was measured on the blade. The S_a was measured at nine points each on the pressure side and the suction side of the blade. To measure the MRR during polishing, scratches with a depth of approximately 30 µm were created on the blade surface (Fig. 2e). As shown in Fig. 2f, the MRR was evaluated by calculating the difference in scratch depth before and after polishing.

Fig. 2 PSTCMP apparatus and polishing performance test method: (a)–(c) photographs of the PSTCMP apparatus, (d) S_a measuring point of the blade, (e) 3D surface profile of scratches, and (f) schematic diagram of MRR calculation.

Table 1 Detailed parameters of the PSTCMP experiment

Experimental condition	Parameters
Spindle speed (rpm)	50
Workpiece speed (rpm)	10
Radius of polishing barrel (mm)	250
Abrasive type	1. SiC, SiO2, 1-CeO2, 2-CeO2, 3-CeO2
Abrasive concentration (wt%)	5, 10, 15, 20, 25
Polishing slurry pH	8
Polishing time (min)	100
Ultraviolet light power (W)	100

---

2.4. Simulation details

First-principles molecular dynamics simulations were performed using the CP2K software packages to investigate the chemical interactions between the slurry and the Ti alloy.⁴³ The visualization of the simulation process was carried out using the VESTA and OVITO software packages.^44,45 Since CeO₂ can react with the slurry under UV light to generate ˙OH, the primary components of the slurry in PSTCMP are CeO₂, H₂O, and ˙OH. Corn starch and sodium bicarbonate are the thickening phase and pH adjusting phase, respectively, which do not participate in chemical reaction, so they are ignored. Based on the above analysis, a molecular dynamics model comprising CeO₂, H₂O, ˙OH, and a Ti alloy substrate was constructed using the Materials Studio software. As shown in Fig. 3, the model dimensions are 15 Å × 15 Å × 30 Å. The Ti alloy substrate adopts a hexagonal close-packed structure, while CeO₂ has a fluorite structure. The model contains a total of 172 atoms, including 90 Ti atoms, 6 Ce atoms, 28 O atoms, and 48 H atoms.

Fig. 3 Molecular dynamics model of the chemical interaction system of the slurry and the Ti alloy surface: (a) perspective view, and (b) top view.

Structural optimization and molecular dynamics calculations were performed using the PBE functional with the Grimme D3 dispersion correction, in conjunction with the DZVP-MOLOPT-SR-GTH basis set and the Goedecker–Teter–Hutter pseudopotentials.^46–48 The plane-wave energy cutoff was set to 500 Ry. Molecular dynamics simulations were performed in the NVT ensemble. The initial simulation temperature was set to 300 K, and the integration time step was 0.5 fs. The simulations were divided into two main stages: (i) fixing CeO₂ in position and relaxing the system for 50 ps to allow sufficient interaction between the slurry and the Ti alloy surface; and (ii) moving the CeO₂ along the negative z-axis at a velocity of 0.05 Å ps⁻¹ for 200 ps, followed by holding for 50 ps. Finally, the CeO₂ was moved again along the negative z-axis at the same velocity for 200 ps.

2.5. Characterization

Transmission electron microscopy (Titan Themis G3, Thermo Fisher Scientific, USA), X-ray diffraction (SmartLab 9 kW, Rigaku Corporation, Japan), and physical adsorption analysis (Autosorb iQ, Anton Paar, USA) were used to characterize the surface morphology, phase composition, and specific surface area of the abrasive, respectively. Thermogravimetric analysis of the abrasive was conducted using a thermogravimetric analyzer (TGA/DSC3+, Mettler Toledo, Switzerland) with a heating temperature of 600 °C and a heating rate of 10 °C min⁻¹. X-ray photoelectron spectroscopy (Axis Supra+, Shimadzu, Japan) and Fourier transform infrared spectroscopy (Vertex 70, Bruker, USA) were used to analyze the chemical changes on the surfaces of the blade and the abrasive. UV-visible absorption spectra of the abrasive were measured using a UV-visible photometer (U4150, Hitachi Limited, Japan). Photoluminescence spectra of the abrasive were measured using a fluorescence spectrometer (FLS1000, Edinburgh, UK) with an excitation wavelength of 365 nm. The type and number of free radicals were determined using electron paramagnetic resonance (EMXplus-6/1, Bruker, Germany) using 5,5-dimethyl-1-pyrroline N-oxide as a trapping agent. The rheological properties of the slurry were tested using a stress-controlled rheometer (DHR-2, TA Instruments, USA). The particle size distribution of the abrasive during polishing was measured by a nano-particle size analyzer (Zetasizer Nano ZS90, Malvern, Britain). An optical microscope (MX 40, Olympus, Japan) and a scanning electron microscope (JSM-7900F, JEOL, Japan) were used to characterize the surface morphology of the blade. A focused ion beam microscope (Helios G4 UX, Thermo Fisher Scientific, USA) was employed to prepare TEM cross-sectional samples of the blade. The S_a was measured using a white light interferometer (WLI SuperView W1, Chotest, China).

3. Results and discussion

3.1. Morphology and characteristics of the abrasive

Fig. 4a–c show the three CeO₂ abrasives with different particle sizes prepared by the chemical precipitation method. Overall, these CeO₂ abrasives exhibit irregular polygonal morphologies. Compared with regular spherical abrasives, irregular polygonal abrasives are more favorable for improving polishing efficiency. Particle size statistics were obtained by randomly selecting 100 abrasives (Fig. 4d), and the results indicated that the average particle sizes of the three abrasives were 34.1 ± 9.2 nm, 17.1 ± 4.6 nm, and 4.2 ± 0.7 nm. Among them, the 3-CeO₂ exhibited the narrowest size distribution, reflecting higher particle size uniformity. It should be noted that CeO₂ exists in the form of clusters rather than as isolated particles. Low-magnification TEM images (Fig. S1a–c) revealed that the cluster sizes of CeO₂ with different particle sizes ranged from 100 to 200 nm. Compared with the isolated distribution of CeO₂, this cluster distribution has better mechanical removal ability. Fig. 4e presents the XRD spectra of the three CeO₂ abrasives, showing distinct diffraction peaks at 28.55°, 33.08°, 47.48°, 56.34°, 59.09°, 69.40°, 76.70°, and 79.07°, which correspond to the (111), (200), (220), (311), (222), (400), (331), and (420) crystal planes of CeO₂, respectively. In addition, to compare the polishing performance of different abrasives, commercial SiC (200 nm) and SiO₂ (200 nm) abrasives were used as controls (Fig. S1d and e).

Fig. 4 Morphology and phase composition of the three CeO₂ abrasives: (a) TEM image of the 1-CeO₂ abrasive, (b) TEM image of the 2-CeO₂ abrasive, (c) TEM image of the 3-CeO₂ abrasive, (d) particle size distributions of the three CeO₂ abrasives, and (e) XRD spectra of the three CeO₂ abrasives.

For the 1-CeO₂, 2-CeO₂, and 3-CeO₂ abrasives, the chemical activities are different because of their different particle sizes. Fig. 5a–c present the Ce 3d fine spectra of the 1-CeO₂, 2-CeO₂, and 3-CeO₂ abrasives. Based on the peak positions, the Ce 3d fine spectra can be deconvoluted into characteristic peaks corresponding to Ce³⁺ and Ce⁴⁺. Specifically, v₀, v′, u₀, and u′ correspond to the characteristic peaks of Ce³⁺, while v, v″, v‴, u, u″, and u‴ correspond to those of Ce⁴⁺. The Ce³⁺ content in CeO₂ can be calculated using eqn (6):⁴⁹

where W is the percentage of Ce³⁺ and S is the peak area of the characteristic peak. As shown in Table 2, the proportions of Ce³⁺ in the 1-CeO₂, 2-CeO₂, and 3-CeO₂ abrasives are 27%, 32%, and 39%, respectively. Among them, the 3-CeO₂ abrasive exhibits the highest Ce³⁺ content, which is mainly attributed to its smaller particle size. For CeO₂ clusters of the same dimensions, the smaller particle size of 3-CeO₂ results in a larger grain boundary area. These grain boundaries weaken the covalency of the Ce–O bond and reduce the binding strength of the lattice to surface oxygen, thereby promoting the formation of oxygen vacancies and driving the conversion of Ce⁴⁺ to Ce³⁺.⁵⁰

Fig. 5 Chemical activity of the three CeO₂ abrasives: (a) Ce 3d fine spectrum of the 1-CeO₂ abrasive, (b) Ce 3d fine spectrum of the 2-CeO₂ abrasive, (c) Ce 3d fine spectrum of the 3-CeO₂ abrasive, (d) thermogravimetric curves of the three CeO₂ abrasives, (e) specific surface areas of the three CeO₂ abrasives, (f) –OH surface densities of the three CeO₂ abrasives, (g) ultraviolet-visible spectra of the three CeO₂ abrasives, (h) photoluminescence spectra of the three CeO₂ abrasives, and (i) electron paramagnetic resonance spectra of the three CeO₂ abrasives.

Table 2 Peak areas of individual peaks of Ce 3d for the CeO₂ abrasives

Peak	Ce^3+ (%)				Ce^4+ (%)
	v 0	v′	u 0	u′	v	v″	v‴	u	u″	u‴
1-CeO2	2.61	14.06	1.75	8.23	19.14	8.50	16.45	9.99	7.65	11.62
2-CeO2	7.10	13.95	3.18	7.71	15.07	7.95	17.08	11.77	4.29	11.90
3-CeO2	5.15	18.28	5.04	10.46	13.71	5.73	14.30	10.01	6.39	10.93

---

Fig. 5d shows the thermogravimetric curves of the three CeO₂ abrasives. With increasing temperature, the three CeO₂ abrasives exhibit a decrease in weight. In the range of 40–120 °C, the weight loss of CeO₂ is attributed to the removal of surface-adsorbed H₂O, while in the range of 120–600 °C, it results from the dehydration and condensation of surface –OH groups. Furthermore, based on the specific surface areas of the three CeO₂ abrasives (Fig. 5e) and eqn (7), the surface density of –OH groups on CeO₂ are calculated as follows:⁵¹

where W_t1 is the weight at 120 °C, W_t2 is the weight at 600 °C, N_A is the Avogadro constant, SSA is the specific surface area, and is the relative molecular mass of H₂O. As shown in Fig. 5f, the –OH surface densities of 1-CeO₂, 2-CeO₂, and 3-CeO₂ are 0.37 OH nm⁻², 0.41 OH nm⁻², and 0.76 OH nm⁻², respectively. Compared with the 1-CeO₂ and 2-CeO₂, the –OH surface density of the 3-CeO₂ increases by 105% and 76%, respectively. The higher Ce³⁺ fraction on the surface of 3-CeO₂ is an important reason for its significantly larger –OH surface density.

In PSTCMP, the photocatalytic performance of CeO₂ also plays an important role in the final polishing performance. Fig. 5g shows the ultraviolet-visible (UV-vis) spectra of the three CeO₂ abrasives. Overall, the three CeO₂ abrasives exhibit strong light absorption capability. Compared with the 1-CeO₂ and 2-CeO₂, the 3-CeO₂ shows a slightly broader absorption range, with an absorption edge at approximately 460 nm. The photoluminescence (PL) spectrum can reflect the separation efficiency of electrons and holes in photocatalysts, and it is generally accepted that weaker peak intensity indicates a lower recombination efficiency of electrons and holes.⁵² As shown in Fig. 5h, the 3-CeO₂ exhibits a markedly lower peak intensity, indicating that the recombination of photogenerated electron–hole pairs is suppressed, which is beneficial for promoting the generation of ˙OH.

Fig. 5i shows the electron paramagnetic resonance (EPR) spectra of the three CeO₂ abrasives under UV light, which reflect the types and quantities of reactive radicals generated in the polishing slurry. The three CeO₂ abrasives exhibit four peaks in their EPR spectra, with peak intensity ratios from left to right of 1 : 2 : 2 : 1, which is a typical characteristic of ˙OH. Meanwhile, the EPR results are consistent with the UV-vis and PL spectra. A comparison of the peak intensities shows that 3-CeO₂ exhibits the highest intensity, while 1-CeO₂ and 2-CeO₂ show little difference. This indicates that under UV light, the slurry containing 3-CeO₂ generates the largest amount of ˙OH.

In PSTCMP, the rheological properties of the slurry are critical factors influencing the polishing performance. Fig. 6a presents the rheological curves of polishing slurries containing different abrasives. For the slurry containing the SiC, the peak shear rate is nearly identical to that of the base slurry, whereas the peak viscosity decreases from 140 Pa s to 42 Pa s. This reduction is mainly attributed to the lubricating effect of the abrasives between the solid particles, which leads to a lower peak viscosity.⁵³ Unlike SiC, the addition of CeO₂ and SiO₂ increases the peak shear rate of the slurry. Meanwhile, the peak viscosity of the slurry containing CeO₂ and SiO₂ is also lower than that of the slurry containing SiC. Overall, within a certain shear rate range (20–300 s⁻¹), all curves exhibit a shear-thickening effect. Although the addition of CeO₂ slightly alters the peak shear rate and reduces the peak viscosity, the slurry still retains its shear-thickening characteristics.

Fig. 6 Rheological properties of the polishing slurry: (a) rheological curves of slurries with different types of abrasive, and (b) rheological curves of slurries with different abrasive concentrations.

Since the rheological curves of the slurries containing the three CeO₂ abrasives are similar, 3-CeO₂ is selected as a representative to investigate the effect of CeO₂ concentration on slurry rheology. As shown in Fig. 6b, when the CeO₂ concentration is in the range of 5–20 wt%, the effect on slurry rheology is minor. When the concentration increases to 25 wt%, a large amount of CeO₂ disperses in the slurry, which significantly enhances its lubricating effect and consequently reduces the peak viscosity.

3.2. Effects of abrasive type and concentration on polishing performance

Polishing experiments with different abrasives are conducted to investigate the effects of the abrasive type on the S_a and MRR. (The abrasive concentration is 15 wt%.) As shown in Fig. 7a, the initial S_a of the blade is 534.32 nm. For SiC and SiO₂ (Fig. 7b and c), the polished surfaces exhibited relatively poor quality, with S_a values of 154.84 nm and 176.21 nm, respectively. In contrast, the three CeO₂ abrasives demonstrate superior polishing quality. After polishing with 1-CeO₂, 2-CeO₂, and 3-CeO₂, the S_a values of the blades are 106.12 nm, 74.22 nm, and 48.91 nm, respectively (Fig. 7d–f).

Fig. 7 The effect of different factors on polishing performance: (a) 3D surface profile before polishing, (b) 3D surface profile after polishing with SiC, (c) 3D surface profile after polishing with SiO₂, (d) 3D surface profile after polishing with 1-CeO₂, (e) 3D surface profile after polishing with 2-CeO₂, (f) 3D surface profile after polishing with 3-CeO₂, (g) effect of abrasive type on S_a and MRR, (h) effect of abrasive concentration on S_a and MRR, and (i) variation of S_a with polishing time.

The S_a and MRR with different abrasives are summarized in Fig. 7g, and the S_a was calculated as the average of eighteen measuring points, with the sampling positions described in section 2.3. The results for the S_a are consistent with the above analysis. For SiC and SiO₂, since they cannot undergo photocatalytic reactions, material removal mainly relies on the mechanical action of the abrasive. Consequently, the MRR is relatively low. In particular, SiO₂, owing to its intrinsically lower mechanical strength, cannot effectively remove surface asperities, resulting in the lowest MRR (1.79 μm h⁻¹). Compared with SiC and SiO₂, the MRR for CeO₂ is obviously improved. The MRRs of 1-CeO₂, 2-CeO₂, and 3-CeO₂ are 7.86 μm h⁻¹, 8.02 μm h⁻¹, and 11.21 μm h⁻¹, respectively. Among them, the 3-CeO₂ not only yields the lowest S_a but also the highest MRR, indicating the best polishing performance. The differences in polishing behavior among the three CeO₂ abrasives may be related to variations in chemical activity induced by their different particle sizes.

Furthermore, the effect of abrasive concentration on polishing performance was studied by using the 3-CeO₂ abrasive. As shown in Fig. 7h, when the abrasive concentration increases from 5 wt% to 20 wt%, the S_a after polishing decreases from 74.12 nm to 39.67 nm, while the MRR increases from 4.26 μm h⁻¹ to 12.52 μm h⁻¹, indicating a significant improvement in polishing performance. However, when the abrasive concentration further increases to 25 wt%, the S_a increases to 42.41 nm and the MRR decreases to 10.87 μm h⁻¹. In the slurry, increasing the abrasive concentration provides more active sites for photocatalytic reactions, leading to a higher concentration of ˙OH and thereby accelerating the reaction rate, which facilitates the formation of the oxidation layer. Nevertheless, excessive abrasive concentration reduces the peak viscosity of the slurry, weakening the confinement of the abrasive by solid phase particles. As a result, the mechanical removal effect of the abrasives diminishes, making it difficult to effectively remove the oxidation layer on the blade surface, which ultimately leads to a decrease in the MRR and an increase in the S_a.

Fig. 7i shows the evolution of the S_a during PSTCMP with different abrasive types. Overall, the S_a gradually decreases with polishing time, but clear differences are observed among the abrasives. For SiC and SiO₂, the polishing performance is relatively weak, and even after 60 min of polishing, the S_a remains above 300 nm. In contrast, the three CeO₂ abrasives exhibit superior polishing performance. Specifically, when using the 3-CeO₂, the S_a decreased to below 150 nm after 60 min and further dropped to below 40 nm at 100 min.

The experimental results shown in Fig. 7 indicate that adding 20 wt% of 3-CeO₂ abrasive to the slurry achieves the optimal overall polishing performance. Fig. 8 presents a comparison of the surface morphology before and after polishing under these conditions. The blade surface exhibits a distinct matte appearance before polishing (Fig. 8a–c). Optical micrographs and SEM images reveal obvious scratches, pits, and other defects. After polishing with a slurry containing 20 wt% 3-CeO₂, the surface quality of the blade is significantly improved. As shown in Fig. 8d–f, the blade surface presents a mirror-like finish, with no apparent defects observed. The 3D surface profile data further confirm this result (Fig. 8g and h), showing that the S_a decreases from 501.71 nm to 38.46 nm after polishing. Moreover, the blade surface exhibited good uniformity. The difference in the S_a values between the 18 measurement points on the pressure side and the suction side was less than 5 nm (Fig. 8i).

Fig. 8 Surface characteristics of the Ti alloy blade before and after polishing: (a) photograph of the blade before polishing, (b) optical micrograph of the blade before polishing, (c) SEM image of the blade before polishing, (d) photograph of the blade after polishing, (e) optical micrograph of the blade after polishing, (f) SEM image of the blade after polishing, (g) 3D surface profile of the blade before polishing, (h) 3D surface profile of the blade after polishing, and (i) S_a at different positions of the blade.

To further investigate the subsurface damage of the blade before and after PSTCMP, cross-sectional specimens were prepared using a FIB. As shown in Fig. 9a, the subsurface damage layer before PSTCMP is approximately 634.79 nm. A magnified view of the damaged region (Fig. 9b) reveals the presence of an amorphous layer of about 18.24 nm at the outermost part of the damage layer, which is primarily attributed to the combined mechanical and thermal effects.⁵⁴ Within the damage layer region, the inverse fast Fourier transform images reveal numerous dislocations along with a few stacking faults (Fig. 9c). Moreover, the selected area electron diffraction patterns in Fig. 9a and b indicate that the damaged region before polishing exhibits polycrystalline characteristics, whereas the undamaged matrix retains a single-crystal structure. In addition, Moiré fringes are observed in the HRTEM images of the damage layer region (Fig. 9d). These fringes arise from the overlap or intersection of two grains, further confirming the typical characteristics of a polycrystalline structure.⁵⁵ As shown in Fig. 9e, dislocation pile-ups are also observed within the damage layer. With the progression of strain, such dislocation pile-ups can induce the formation of sub-grain boundaries inside the grains, eventually leading to grain fragmentation. A higher magnification of the dislocation accumulation region reveals the formation of two grains at the pile-up sites, with an amorphous phase present at their interface (Fig. 9f). This phenomenon is attributed to insufficient grain growth and deformation at the boundaries, resulting in structural disorder.

Fig. 9 Analysis of the subsurface damage layer before and after PSTCMP: (a) TEM image of the subsurface region before polishing, (b) enlarged TEM image of the rectangle in (a), (c) inverse fast Fourier transform image of the rectangle in (b), (d) HRTEM image of the rectangle in (b), (e) enlarged TEM image of the rectangle in (a), (f) HRTEM image of the rectangle in (e), (g) TEM image of the subsurface region after PSTCMP, (h) enlarged TEM image of the rectangle in (g), and (i) inverse fast Fourier transform image of the rectangle in (h).

Fig. 9g and h show the subsurface damage after PSTCMP with slurry containing 20 wt% 3-CeO₂. Compared with before polishing, only an amorphous layer of approximately 7.83 nm is observed after polishing. Moreover, the inverse fast Fourier transform analysis reveals no apparent dislocations (Fig. 9i). This decrease can be attributed to the combined effect of the shear thickening effect and photocatalytic oxidation in the PSTCMP process. On the one hand, as the polishing progresses, the surface of the workpiece gradually becomes smoother, and the shear force exerted by the workpiece on the fluid decreases. According to shear-thickening theory,^56–58 this weakens the shear-thickening effect, which in turn reduces the stress exerted by the fluid on the abrasive, significantly preventing the occurrence and evolution of dislocations and stacking faults. On the other hand, under UV light, the ˙OH generated by the photocatalytic reaction oxidizes the surface of the workpiece, forming a reaction layer. This reaction layer acts as a buffer, effectively reducing the impact of the abrasives on the subsurface damage layer. To summarize, the above results show that there are many dislocations and stacking faults in the subsurface region before polishing. PSTCMP can effectively reduce the thickness of the damaged layer and remove the original crystal defects on the subsurface.

In the PSTCMP, the stability of the polishing slurry is also an important factor affecting polishing performance. Based on the polishing experimental results, adding 20 wt% 3-CeO₂ results in the best polishing performance. Therefore, this was selected as the research subject to study the stability of the polishing slurry. As shown in Fig. 10a, the morphology of the polishing slurry was observed after standing for 0 min, 25 min, 50 min, 75 min, and 100 min. Compared to the polishing slurry at 0 min, the other four time points showed only a very thin layer of water on the surface. This indicates that the prepared slurry has good stability, with no significant precipitation occurring within 100 min.

Fig. 10 Stability of the polishing slurry: (a) morphology of the polishing slurry after standing for 0 min, 25 min, 50 min, 75 min and 100 min, and (b) particle size analysis in the polishing process.

Furthermore, polishing slurries at different time points during the polishing process were collected and the particle size was analyzed. As shown in Fig. 10b, the overall curve exhibits two peaks at 138 nm and 5.48 μm, corresponding to CeO₂ abrasive and corn starch particles, respectively. According to the analysis in section 3.1, CeO₂ exists in the form of clusters, and the particle size analysis results correspond to those in Fig. S1c. As the polishing time increases from 0 to 100 min, the curve shows little change, indicating that the CeO₂ abrasives and corn starch particles maintain good stability during the 100 min polishing period.

3.3. Mechanism analysis of PSTCMP

The chemical reaction mechanism in PSTCMP was revealed by first-principles molecular dynamics simulation. As shown in Fig. 11a, in the initial stage (0–5 ps), H₂O molecules in the slurry gradually migrate toward the CeO₂ abrasive surface, and the first Ce–H₂O structure appears at 5 ps. With the simulation time extended to 35 ps, the number of binding sites between CeO₂ and H₂O molecules increases significantly, leading to the formation of four stable Ce–H₂O structures (Fig. 11b and c). Notably, at 35 ps, the ˙OH is first observed to bond with Ti atoms on the Ti alloy surface, forming a Ti–OH structure, and at 50 ps, a Ti–O structure appeared on the Ti alloy surface (Fig. 11d). Its formation primarily occurs through the transfer of H atoms among ˙OH groups, where the H atom in a Ti–OH structure is first transferred to an H₂O molecule, and subsequently the H atom migrates from the H₂O molecule to another ˙OH group. The transfer equation of the H atom is shown in eqn (8):⁵⁹

˙OH + H₂O + ˙OH → O + 2H₂O (8)

Fig. 11 Snapshots, total system energy, and radial distribution function variations during the free reaction stage: (a) 5 ps, (b) 20 ps, (c) 35 ps, (d) 50 ps, (e) total system energy, and (f) radial distribution function curves at 0 ps and 50 ps.

In addition, at 50 ps (Fig. 11d), a significant structural evolution occurs on the CeO₂ surface, where the previously formed Ce–H₂O structures gradually transform into a Ce–OH structure. According to previous studies,^51,60,61 Ce³⁺ ions on the CeO₂ surface serve as active sites in aqueous environments, facilitating the dissociation of H₂O molecules to form Ce–OH structures, as shown in eqn (9):

Ce³⁺ + H₂O → Ce–OH + H⁺ (9)

The Ce–OH structure can form a bridge bond with –OH or –O sites on the workpiece surface, which promotes material removal. The variation of the total system energy is shown in Fig. 11e. During the simulation period of 30–40 ps, the formation of Ce–OH and Ti–OH structures leads to a significant decrease in energy, indicating that the process is exothermic.

To further investigate the effect of Ti–OH and Ti–O formation on the Ti alloy substrate, radial distribution function (RDF) analysis was performed (Fig. 11f). The results reveal that the characteristic peaks of the initial Ti–Ti bond lengths are located at 2.91 Å and 4.15 Å. With the formation of Ti–OH and Ti–O structures, the Ti–Ti bond length distribution becomes significantly broadened, indicating that some crystal lattices of the Ti alloy are distorted. This structural distortion weakens the bonding strength between certain surface Ti–Ti atoms, thereby facilitating the selective removal of materials during the subsequent mechanical polishing process.⁶²

Based on Fig. 11d, a velocity of 0.05 Å ps⁻¹ was applied to the CeO₂ in the negative z-axis direction to further simulate the interaction between the Ti alloy and the slurry. As shown in Fig. 12a and b, when the CeO₂ approaches the Ti alloy surface, a Ce–OH structure on the CeO₂ surface reacts with a Ti–OH structure on the Ti alloy surface to form a Ce–O–Ti structure (Ce–O₁–Ti). By keeping the CeO₂ fixed and extending the simulation for 50 ps, it is observed in Fig. 12c that O atoms in CeO₂ directly combine with Ti atoms to form a Ce–O–Ti structure (Ce–O₂–Ti). Subsequently, the CeO₂ was returned to its initial position at the same velocity. As shown in Fig. 12d, at 370 ps, the O₂–Ti bond within the Ce–O₂–Ti structure broke, and the Ti atom was not removed. In contrast, as the CeO₂ moved along the positive z-axis, the Ce–O₁–Ti structure remained, and two Ti atoms bonded to the O₁ atom were removed (Fig. 12e). The removal of Ti atoms is beneficial to further weaken the mechanical properties of the oxide layer and improve the MRR.

Fig. 12 Snapshots and system energy variations during the approach and withdrawal stages of CeO₂: (a) 60 ps, (b) 250 ps, (c) 300 ps, (d) 370 ps, (e) 500 ps, and (f) total system energy.

Fig. 12f shows the evolution of the system energy during the motion of CeO₂ along the z-axis. The curve exhibits two stages of energy increase, corresponding to (i) the approach of CeO₂ toward the Ti-alloy surface with formation of Ce–O–Ti linkages, and (ii) the withdrawal of CeO₂ from the surface, accompanied by the removal of Ti atoms. In addition, both processes are endothermic reactions, and the required energy is mainly provided by the force generated by the shear thickening effect and the relative kinetic energy of CeO₂.

It should be noted that the molecular dynamics model used in this study primarily focuses on the interactions between molecules or atoms. However, the polishing process encompasses not only microscopic chemical and physical phenomena but also fluid dynamics and the interactions between the abrasive and the workpiece. Therefore, while the molecular dynamics model in this study is highly valuable for investigating atomic-scale interactions, it may fall short in capturing the collective effects involved in the actual polishing process.

Although molecular dynamics simulation focuses on the interaction between atoms and molecules, it provides valuable insights for understanding the polishing mechanism and the optimization of abrasive type and size. Specifically, due to the inherent properties of the Ti alloy, traditional STP methods typically result in a low material removal rate and poor surface quality. This study introduces UV light into STP and develops a new polishing slurry for polishing Ti alloy blades. However, the polishing mechanisms with this slurry are difficult to analyze directly through macroscopic tests, and enough theoretical guidance for optimizing factors such as abrasive type and size cannot be provided.

In this study, molecular dynamics simulations reveal the dynamic processes of reaction layer formation and chelation in PSTCMP. According to the analysis in section 3.3, promoting these processes is crucial for PSTCMP and directly contributes to material removal. Moreover, the photocatalytic and chelation reaction are closely related to the abrasive type and size. Therefore, while ensuring sufficient mechanical removal capability of the abrasive, the type and size of the abrasive should be optimized toward smaller CeO₂ particles with high photocatalytic performance and chemical activity.

Under UV light, the Ti alloy blade was immersed in a polishing slurry containing 3-CeO₂ for 5 min to investigate the surface phase change. Before and after immersion, the XPS full spectrum exhibited characteristic peaks of Ti, O, C, and Al (Fig. 13a). Because the Ti alloy readily reacts with oxygen in air to form an oxide layer, the full spectrum before immersion also shows the characteristic peak of O. By comparing the peak intensities of O 1s, it is evident that the O 1s peak after immersion is significantly stronger than that before immersion. Furthermore, the FTIR results show that the characteristic peaks at 737 cm⁻¹, 936 cm⁻¹, and 1128 cm⁻¹ represent the structures of Ti–O, Al–O, and V–O, respectively.^63–65 Compared with before immersion, the absorption peak of the Ti–O bond corresponding to Ti oxides is markedly enhanced after immersion (Fig. 13b). This indicates that an oxidation reaction occurs between the Ti alloy blade surface and the slurry after immersion. As shown in Fig. 13c and d, the O 1s fine spectra before and after immersion were compared. Before immersion, the O 1s spectrum mainly exhibited characteristic peaks of Ti–O and Al–O bonds, whereas after immersion, a characteristic peak corresponding to the Ti–OH structure was also observed.^66–68Fig. 13e and f show the Ti 2p fine spectra. Before immersion, the Ti alloy blade surface is mainly composed of Ti⁰ (metallic Ti) and Ti⁴⁺ (TiO₂).^69,70 After immersion, in addition to Ti⁰ and Ti⁴⁺, characteristic peaks of Ti²⁺ and Ti³⁺ are also detected. These results indicate that after immersion, a chemical reaction occurs between the Ti alloy blade surface and the slurry, and the products consist of Ti–OH structures and Ti oxides with multiple valence states.

Fig. 13 Chemical reaction mechanism analysis: (a) XPS full spectra of the Ti alloy blade before and after immersion, (b) FTIR spectra of the Ti alloy blade before and after immersion, (c) O 1s fine spectrum of the Ti alloy blade before immersion, (d) O 1s fine spectrum of the Ti alloy blade after immersion, (e) Ti 2p fine spectrum of the Ti blade alloy surface before immersion, (f) Ti 2p fine spectrum of the Ti alloy blade after immersion, (g) XPS full spectra of the 3-CeO₂ abrasive before and after polishing, (h) O 1s fine spectrum of the 3-CeO₂ abrasive before polishing, and (i) O 1s fine spectrum of the 3-CeO₂ abrasive after polishing.

Fig. 13g–i show the XPS full spectra and O 1s fine spectra of the 3-CeO₂ abrasive. Compared with before polishing, the survey spectrum of the 3-CeO₂ abrasive after polishing exhibits a Ti 2p characteristic peak in the range of 450–470 eV. The O 1s fine spectrum indicates that, in addition to the characteristic peak of the Ce–O structure in CeO₂, a characteristic peak corresponding to the Ce–O–Ti structure also appears after polishing.⁷¹ It is notable that multiple Ti oxide (a Ti atom bonded with one or more O atoms), Ti–OH, and Ce–O–Ti structures are likewise observed in the simulations shown in Fig. 11 and 12. The results of XPS and FTIR experimentally confirm the conclusions drawn from the simulations.

Based on the above analysis, the PSTCMP mechanism shown in Fig. 14 can be summarized in three stages: surface oxidation, chelation reaction, and mechanical removal. Under UV light, ˙OH generated during photocatalytic reactions oxidizes the Ti alloy blade surface, forming a readily removable oxide layer (Fig. 14a). Meanwhile, as shown in Fig. 14b, CeO₂ undergoes hydrolysis, forming Ce–OH structures with high activity. The Ce–OH groups on the CeO₂ surface chelate with the Ti alloy blade surface oxidized by ˙OH. As CeO₂ slides across the Ti alloy blade surface, Ti atoms are removed from the oxide layer. This removal further weakens the mechanical strength of the oxide layer and thereby enhances the polishing efficiency. Finally, under the shear-thickening effect, CeO₂ efficiently removes the oxide layer, achieving polishing of the Ti alloy blade.

Fig. 14 Schematic diagram of the surface oxidation–reactive complexation–mechanical removal mechanism of PSTCMP.

Out of the three CeO₂ abrasives, the higher Ce³⁺ content on the surface of 3-CeO₂ furnishes more active sites for the formation of Ce–OH structures. The greater number of Ce–OH structures (–OH nm⁻²) facilitates the complexation and removal of Ti atoms from the oxide layer during the PSTCMP. Furthermore, compared to 1-CeO₂ and 2-CeO₂, the superior light absorption capacity and electron–hole separation efficiency of 3-CeO₂ enable it to generate more ˙OH radicals, weakening the surface of the Ti alloy blade. These two aspects are the primary reasons for the superior polishing performance of 3-CeO₂ compared to 1-CeO₂ and 2-CeO₂.

4. Conclusion

In conclusion, this study integrates photocatalytic oxidation with the shear-thickening effect to develop a PSTCMP method and apply it to the polishing of Ti alloy blades. The PSTCMP slurry contains only CeO₂, sodium bicarbonate, corn starch, and water, thereby avoiding toxic reagents. After polishing for 100 min with a slurry containing 20 wt% 3-CeO₂, the S_a decreases from 501.71 nm to 38.46 nm, the MRR reaches 12.52 μm h⁻¹, and the subsurface damage layer is reduced from 634.79 nm to 7.83 nm. The content of CeO₂ has an important influence on the polishing performance. When the content of CeO₂ exceeds 20 wt%, the polishing performance becomes worse due to the obvious decrease of the slurry viscosity.

According to the results of first-principles molecular dynamics, XPS, and FTIR, the removal mechanism of PSTCMP can be summarized in three stages: surface oxidation, reactive chelation, and mechanical removal. First, under ultraviolet light, ˙OH generated by CeO₂ photocatalysis oxidizes the Ti alloy blade surface and forms an oxide layer containing Ti–O and Ti–OH structures. Subsequently, the Ce–OH generated by the hydrolysis of CeO₂ reacts with Ti–OH to generate the Ce–O–Ti structure, and the Ti atoms are removed as CeO₂ slips on the surface of the Ti alloy blade, further weakening the mechanical strength of the oxide layer. Finally, under the shear thickening effect, the viscosity of the slurry increases rapidly, and CeO₂ removes the oxide layer efficiently, thus realizing the polishing of the Ti alloy blade. At the same time, XPS, PL, and EPR results show that CeO₂ with a smaller particle size has a higher material removal efficiency and better surface quality in PSTCMP because it has more Ce³⁺ sites and stronger photocatalytic activity.

Author contributions

Shuai Zhang: investigation, formal analysis, data curation, visualization, methodology. Zhenyu Zhang: funding acquisition, project administration, conceptualization, supervision. Zhibin Yu: data curation, conceptualization. Junde Guo: investigation, formal analysis. Zhenghong Liu: formal analysis, software, resources. Feng Tian: visualization, resources, data curation. Yujie Chen: formal analysis, data curation, visualization. Xingqiao Deng: formal analysis, data curation, resources. Xiaofei Yang: formal analysis, resources, visualization.

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 data that support the findings of this study are available from the corresponding author upon reasonable request.

Acknowledgements

The authors acknowledge the financial support from the Frontier Technology Research and Development Project of Jiangsu Province (BF2025039), Integrated Outcomes of Qian Science and Technology (2025ZD110), Science and Technology Program Project of Shanghai Municipality (25JC3200200), Open Fund of State Key Laboratory of Baiyunobo Rare Earth Resource Researches and Comprehensive Utilization (2025B2725-QZ-10), and the Fundamental Research Funds for the Central Universities (DUT25YG268).

References

1. K. Wang, L. Wang, K. Zheng, Z. He, D. J. Politis and G. Liu, et al., High-efficiency forming processes for complex thin-walled titanium alloys components: state-of-the-art and perspectives, Int. J. Extreme Manuf., 2020, 2, 032001.
2. T. Song, Z. Chen, X. Cui, S. Lu, H. Chen and H. Wang, et al., Strong and ductile titanium-oxygen-iron alloys by additive manufacturing, Nature, 2023, 618, 63–68.
3. Z. Liu, Y. Wang, K. Song, K. Zhou, R. Jiang and Z. Liao, et al., Achieving finished surface and anti-fatigue performance improvements of titanium alloy through precision abrasive belt grinding-polishing process, Tribol. Int., 2026, 215, 111467.
4. C. Li, G. Bin, J. Li and Z. Liu, Study on the erosive wear of the gas-solid flow of compressor blade in an aero-turboshaft engine based on the Finnie model, Tribol. Int., 2021, 163, 107197.
5. Q. Li, Y. Yang, Y. Yang, P. Li, J. Liu and M. Wu, Enhancing the wear performance of WC-6Co tool by pulsed magnetic field in Ti-6Al-4V machining, J. Manuf. Process., 2022, 80, 898–908.
6. Q. Miao, W. Ding, J. Xu, L. Cao, H. Wang and Z. Yin, et al., Creep feed grinding induced gradient microstructures in the superficial layer of turbine blade root of single crystal nickel-based superalloy, Int. J. Extreme Manuf., 2021, 3, 045102.
7. Q. Shao, B. Lyu, J. Yuan, X. Wang, M. Ke and P. Zhao, Shear thickening polishing of the concave surface of high-temperature nickel-based alloy turbine blade, J. Mater. Res. Technol., 2021, 11, 72–84.
8. P. Gilge, A. Kellersmann, J. Friedrichs and J. Seume, Surface roughness of real operationally used compressor blade and blisk, Proc. Inst. Mech. Eng., Part G, 2019, 233, 5321–5330.
9. C. Liu, Y. Cao, S. Ding, W. Zhang, Y. Cai and A. Lin, et al., Effects of blade surface roughness on compressor performance and tonal noise emission in a marine diesel engine turbocharger, Proc. Inst. Mech. Eng., Part D, 2020, 234, 3476–3490.
10. M. Omari, H. Mousa, F. Al-Oqla and M. Aljarrah, Enhancing the surface hardness and roughness of engine blades using the shot peening process, Int. J. Miner., Metall. Mater., 2019, 26, 999–1004.
11. B. Zhao, Y. Wang, J. Peng, X. Wang, W. Ding and X. Lei, et al., Overcoming challenges: advancements in cutting techniques for high strength-toughness alloys in aero-engines, Int. J. Extreme Manuf., 2024, 6, 062012.
12. D. Zhu, S. Luo, L. Yang, W. Chen, S. Yan and H. Ding, On energetic assessment of cutting mechanisms in robot-assisted belt grinding of titanium alloys, Tribol. Int., 2015, 90, 55–59.
13. B. Zhang, S. Wu, D. Wang, S. Yang, F. Jiang and C. Li, A review of surface quality control technology for robotic abrasive belt grinding of aero-engine blades, Measurement, 2023, 220, 113381.
14. Y. Zhang, X. Wang, Z. Chen, Z. Zhu and H. Ye, A synchronous polishing method for aero-engine blade based on multi-spindle machine tool, Int. J. Adv. Manuf. Technol., 2022, 123, 1669–1678.
15. L. Luo, H. Zhao, H. Wang, X. Li and H. Ding, Pose optimization and tool path planning for robotic grinding of complicated curved surface, J. Mech. Eng., 2022, 58, 284–294.
16. H. Wang, Y. Fu, H. Gao and X. Wang, Study on prediction of surface roughness for abrasive flow machining, J. Mech. Eng., 2022, 58, 188–197.
17. L. Liao, Z. Zhang, F. Meng, D. Liu, J. Liu and Y. Li, et al., Novel rotary chemical mechanical polishing on an integral impeller, J. Manuf. Process., 2021, 66, 198–210.
18. H. Chen, H. Wan, B. Hong, W. Hang, T. Zhu and P. Zhang, et al., A novel liquid film shearing polishing technique for silicon carbide and its processing damage mechanisms, Appl. Surf. Sci., 2025, 688, 162317.
19. H. Chen, B. Hong, M. Wu, L. Wang, B. Lyu and Y. Yang, et al., A novel method of liquid film shearing polishing for achieving high-quality and low-damage polycrystalline tungsten, J. Manuf. Process., 2025, 149, 165–178.
20. M. Li, B. Lyu, J. Yuan, C. Dong and W. Dai, Shear-thickening polishing method, Int. J. Mach. Tools Manuf., 2015, 94, 88–99.
21. X. He, L. Yang, K. Zhang, R. Li and Y. Peng, Research on the shear thickening polishing CaF₂ with textured hollow polishing tool, J. Manuf. Process., 2024, 119, 193–203.
22. X. Li, Z. Hu, Y. Tu, S. Wu, H. Zou and X. Chen, et al., Simulation and experimental study of ultrasonic-assisted shear thickening polishing of cemented carbide, J. Manuf. Process., 2023, 102, 106–118.
23. W. Yao, Z. Chen, B. Lyu, L. Jiang and L. Zhang, Corona discharge assisted shear thickening polishing: non-Newtonian elastohydrodynamics and electrostatics, Int. J. Mech. Sci., 2025, 289, 110048.
24. W. Fang, J. Gu, Z. Tang, Z. Song, W. Zhao and B. Lyu, Effect of chemical action on shear thickening polishing of YAG, Ceram. Int., 2024, 50, 28775–28785.
25. J. Wang, Z. Tang, S. Goel, Y. Zhou, Y. Dai and J. Wang, et al., Mechanism of material removal in tungsten carbide-cobalt alloy during chemistry enhanced shear thickening polishing, J. Mater. Res. Technol., 2023, 25, 6865–6879.
26. M. Shen, L. Wu, M. Wei, H. Chen, J. Yuan and B. Lyu, et al., High-efficiency free-damage electrochemical shear-thickening polishing of single-crystal silicon carbide, J. Manuf. Process., 2024, 132, 532–543.
27. Y. Tian, Z. Ma, S. Ahmad, C. Qian, X. Ma and X. Yuan, et al., Theoretical and experimental investigation of material removal rate in magnetorheological shear thickening polishing of Ti-6Al-4V alloy, J. Manuf. Sci. Eng., 2024, 146, 031002.
28. J. Wang, Y. Zhou, Z. Qiao, S. Goel, J. Wang and X. Wang, et al., Surface polishing and modification of Ti-6Al-4V alloy by shear thickening polishing, Surf. Coat. Technol., 2023, 468, 129771.
29. J. Wang, B. Lyu, L. Jiang, Q. Shao, C. Deng and Y. Zhou, et al., Chemistry enhanced shear thickening polishing of Ti-6Al-4V, Precis. Eng., 2021, 72, 59–68.
30. Z. Yu, Z. Zhang, Z. Zeng, C. Fan, Y. Gu and C. Shi, et al., Atomic surface of diamond induced by novel green photocatalytic chemical mechanical polishing with high material removal rate, Int. J. Extreme Manuf., 2025, 7, 025102.
31. Q. Zhang, J. Pan, Z. Zhuo, M. Xiang and Q. Yan, Abrasion behavior of TiO₂ catalyzing H₂O₂ to synergistically remove single crystal 6H-SiC under ultraviolet irradiation, Surf. Interfaces, 2023, 38, 102781.
32. T. Wang, Y. Chen, A. Chen and Y. Chen, Development of carbon sphere/ceria (CS/CeO₂) heterostructured particles and their applications to functional abrasives toward photochemical mechanical polishing, Appl. Surf. Sci., 2022, 593, 153449.
33. N. Liu, H. Wang, J. Zhang, X. Chen, L. Lei and H. Jiang, et al., Highly efficient medium vacuum UV-assisted polishing of diamond via C-C bond breakage stimulated graphitization, Tribol. Int., 2025, 209, 110702.
34. J. Ge, Z. Zhang, Y. Lin, X. Chen, Y. Li and X. Li, et al., Research on photochemically-assisted pulsed jet polishing method for cobalt-based alloys, Tribol. Int., 2025, 211, 110845.
35. W. Liu, Q. Xiong, J. Lu, X. Wang and Q. Yan, Tribological behavior of single crystal diamond based on UV photocatalytic reaction, Tribol. Int., 2022, 175, 107806.
36. Y. Ishikawa, Y. Matsumoto, Y. Nishida, S. Taniguchi and J. Watanabe, Surface treatment of silicon carbide using TiO₂(IV) photocatalyst, J. Am. Chem. Soc., 2003, 125, 6558–6562.
37. M. Wang, Z. Mu, T. Wang, Y. Chen and A. Chen, Double-layered core-shell heterostructures of mSiO₂@CdS@CeO₂ abrasive systems toward photochemical mechanical polishing (PCMP) applications, Appl. Surf. Sci., 2023, 614, 156274.
38. W. Wang, B. Zhang, Y. Shi, J. Zhou, R. Wang and N. Zeng, Improved chemical mechanical polishing performance in 4H-SiC substrate by combining novel mixed abrasive slurry and photocatalytic effect, Appl. Surf. Sci., 2022, 575, 151676.
39. X. Zhu, Y. Gui, H. Fu, J. Ding, Z. Mo and X. Jiang, et al., Photocatalytic assisted chemical mechanical polishing for silicon carbide using developed ceria coated diamond core-shell abrasives, Tribol. Int., 2024, 197, 109827.
40. J. Li, W. Ma and X. Hou, Synergistic effect of physics and chemistry on material removal of fused quartz in nanoparticle jet polishing, Ceram. Int., 2024, 50, 46600–46610.
41. N. Xu, Y. Luo, J. Ma, Y. Lin, X. Zhu and Y. Pu, Enhancement mechanism of Y-doped Ce_1−xY_xO₂ for photocatalytic-assisted chemical-mechanical polishing, Mater. Today Commun., 2024, 38, 107791.
42. D. Wang, L. Liu, Z. Zhang, Q. Peng, C. Shi and X. Liu, et al., Atomic-scale planarization surface of quartz glass induced by novel green chemical mechanical polishing using three ingredients, Mater. Today Sustain., 2024, 25, 100669.
43. T. D. Kuehne, M. Iannuzzi, M. Del Ben, V. V. Rybkin, P. Seewald and F. Stein, et al., CP2K: An electronic structure and molecular dynamics software package - Quickstep: Efficient and accurate electronic structure calculations, J. Chem. Phys., 2020, 152, 194103.
44. W. Humphrey, A. Dalke and K. Schulten, VMD: visual molecular dynamics, J. Mol. Graphics Modell., 1996, 14, 33–38.
45. A. Stukowski, Visualization and analysis of atomistic simulation data with OVITO-the Open Visualization Tool, Modell. Simul. Mater. Sci. Eng., 2010, 18, 015012.
46. S. Grimme, Semiempirical GGA-type density functional constructed with a long-range dispersion correction, J. Comput. Chem., 2006, 27, 1787–1799.
47. J. Perdew, K. Burke and M. Ernzerhof, Generalized gradient approximation made simple, Phys. Rev. Lett., 1996, 77, 3865–3868.
48. S. Goedecker, M. Teter and J. Hutter, Separable dual-space gaussian pseudopotentials, Phys. Rev. B: Condens. Matter Mater. Phys., 1996, 54, 1703–1710.
49. Y. Sun, X. Liu, L. Wang, L. Xu, K. Liu and L. Xu, et al., High-performance SOD mimetic enzyme Au@Ce for arresting cell cycle and proliferation of acute myeloid leukemia, Bioact. Mater., 2022, 10, 117–130.
50. W. Zhao, W. Jia, J. Zhou, T. Zhai, Y. Wu and Z. Rana, et al., Nanoscale grain boundary-weakened Ce-O covalency and surface confinement intrinsically boosting ceria surface oxygen reactivity, J. Am. Chem. Soc., 2025, 147, 13050–13058.
51. J. Seo, J. Moon, J. Kim, K. Lee, J. Hwang and H. Yoon, et al., Role of the oxidation state of cerium on the ceria surfaces for silicate adsorption, Appl. Surf. Sci., 2016, 389, 311–315.
52. S. Xie, Z. Wang, F. Cheng, P. Zhang, W. Mai and Y. Tong, Ceria and ceria-based nanostructured materials for photoenergy applications, Nano Energy, 2017, 34, 313–337.
53. P. Das, S. Dash, R. Ganguly, A. Santra, E. Venkatesan and A. Rajhi, et al., Effect of particle loading and temperature on the rheological behavior of Al₂O₃ and TiO₂ nanofluids, Energy Sources, Part A, 2022, 44, 7062–7079.
54. Y. He, G. Xiao, S. Zhu, G. Liu, Z. Liu and Z. Deng, Surface formation in laser-assisted grinding high-strength alloys, Int. J. Mach. Tools Manuf., 2023, 186, 104002.
55. S. Kim, K. Lee, H. Mun, H. Kim, S. Hwang and J. Roh, et al., Dense dislocation arrays embedded in grain boundaries for high-performance bulk thermoelectrics, Science, 2015, 348, 109–114.
56. M. Li, F. Song and Z. Huang, Control strategy of machining efficiency and accuracy in weak-chemical-coordinated-thickening polishing (WCCTP) process on spherical curved 9Cr18 components, J. Manuf. Process., 2022, 74, 266–282.
57. M. Li, B. Karpuschewski, H. Ohmori, O. Riemer, Y. Wang and T. Dong, Adaptive shearing-gradient thickening polishing (AS-GTP) and subsurface damage inhibition, Int. J. Mach. Tools Manuf., 2021, 160, 103651.
58. W. Yao, B. Lyu, T. Zhang, L. Guo and Y. Song, Effect of elastohydrodynamic characteristics on surface roughness in cylindrical shear thickening polishing process, Wear, 2023, 530–531, 205026.
59. K. Kawaguchi, H. Ito, T. Kuwahara, Y. Higuchi, N. Ozawa and M. Kubo, Atomistic mechanisms of chemical mechanical polishing of a Cu surface in aqueous H₂O₂: tight-binding quantum chemical molecular dynamics simulations, ACS Appl. Mater. Interfaces, 2016, 8, 11830–11841.
60. S. Utara, S. Hunpratub, S. Pinitsoontorn and S. Phokha, Characterization and magnetic performance of pure CeO₂ nanoparticles via an ozonolysis reaction, Results Phys., 2021, 30, 104890.
61. B. Choudhury, P. Chetri and A. Choudhury, Oxygen defects and formation of Ce³⁺ affecting the photocatalytic performance of CeO₂ nanoparticles, RSC Adv., 2014, 4, 4663–4671.
62. T. Onodera, H. Takahashi and S. Nomura, First-principles molecular dynamics investigation of ceria/silica sliding interface toward functional materials design for chemical mechanical polishing process, Appl. Surf. Sci., 2020, 530, 147259.
63. A. A. Isari, F. Hayati, B. Kakavandi, M. Rostami, M. Motevassel and E. Dehghanifard, et al., N, Cu co-doped TiO₂@functionalized SWCNT photocatalyst coupled with ultrasound and visible-light: An effective sono-photocatalysis process for pharmaceutical wastewaters treatment, Chem. Eng. J., 2020, 392, 123685.
64. G. Nelson, J. Nychka and A. McDonald, Flame spray deposition of titanium alloy-bioactive glass composite coatings, J. Therm. Spray Technol., 2011, 20, 1339–1351.
65. Y. Li, Z. Zhang, X. Deng, H. Zhou, Y. Gu and C. Shi, et al., Novel green chemical mechanical polishing for an aluminum alloy and mechanisms interpreted by molecular dynamics simulations and measurements, Surf. Interfaces, 2023, 42, 103493.
66. N. Bahadur, F. Chowdhury, M. Obaidullah, M. Hossain, R. Rashid and Y. Akter, et al., Ultrasonic-assisted synthesis, characterization, and photocatalytic application of SiO₂@TiO₂ core-shell nanocomposite particles, J. Nanomater., 2019, 2019, 6368789.
67. Y. Xingwen and L. Guoqiang, XPS study cerium conversion coating on the anodized 2024 aluminum alloy, J. Alloys Compd., 2004, 364, 193–198.
68. M. Shirkhanzadeh, XRD and XPS characterization of superplastic TiO₂ coatings prepared on Ti6Al4V surgical alloy by an electrochemical method, J. Mater. Sci.: Mater. Med., 1995, 6, 206–210.
69. Y. Zhang, J. Li, S. Che and Y. Tian, Electrochemical polishing of additively manufactured Ti-6Al-4V alloy, Met. Mater. Int., 2020, 26, 783–792.
70. T. Kim and J. Lim, Surface characterization of Ti and Ti-6Al-4V alloy by Ca vapor deoxidation: Relation between surface morphology and deoxidation reaction, Appl. Surf. Sci., 2022, 590, 153098.
71. T. Yu, X. Tan, L. Zhao, Y. Yin, P. Chen and J. Wei, Characterization, activity and kinetics of a visible light driven photocatalyst: Cerium and nitrogen co-doped TiO₂ nanoparticles, Chem. Eng. J., 2010, 157, 86–92.

---