A close atomic surface of stainless steel produced by novel green chemical mechanical polishing using silica and lanthana mixed abrasives

Abstract

Stainless steel poses machining challenges due to its corrosion and wear resistance. Traditional chemical mechanical polishing (CMP) often involves using noxious slurries, resulting in environmental risks. Moreover, surface roughness Sa is normally higher than 0.5 nm on stainless steel using conventional CMP, and the material removal rate (MRR) is extremely low. To address these challenges, a novel green CMP method was developed, which uses silica, lanthana, malic acid, γ-aminobutyric acid and hydrogen peroxide. Following this environmentally friendly CMP process, the surface roughness Sa of stainless steel was reduced to 0.286 nm, with an MRR of 82.14 nm min⁻¹. These results represent significant advancements compared to existing studies. X-ray photoelectron spectroscopy analysis demonstrates that hydrogen peroxide oxidized the surface of stainless steel, forming oxides. Malic acid then dissolved these oxides by releasing hydrogen ions. Meanwhile, γ-aminobutyric acid through its –COOH functional groups chelated Fe²⁺, Fe³⁺, and Cr³⁺ ions. This innovatively green CMP process suggests a fresh perspective on achieving fine surface roughness on stainless steel, simultaneously enhancing its wear and corrosion assistance. Its potential application in high-performance stainless steel devices is promising.

---

1. Introduction

Stainless steel^1,2 is widely used in the aerospace industry, precision instrumentation, medical devices, and power generation due to its exceptional corrosion resistance, high-temperature stability, superior mechanical properties, and proven durability.^3–5 Its key applications include precision gears, display substrates, solar cell substrates, and turbine blades. However, the overall performance of this component is directly determined by its surface quality.^6,7 Yu et al.⁸ investigated the effect of surface roughness on blade performance^9,10 with surface roughness Ra increasing from 3.1 to 18.8 μm. Liu et al.¹¹ researched the influence of surface roughness on the erosion wear of compressor blades. Their results showed an increase in surface roughness value from 1 μm to 60 μm for the main blade, with the wear region shifting from covering 80% to 30% of the blade height, whereas, for the splitter blade, it changes from 80% to 50%. Bae et al.¹² conducted a study on the impact of surface roughness on the fatigue limit of martensitic stainless steel. They observed a substantial decrease in the fatigue limit of 7.66% as the surface roughness increased from 0.226 μm to 2.053 μm, with the fatigue ratio dropping from 54.9% to 50.5%. This emphasizes the critical importance of reducing surface roughness to minimize wear and enhance the fatigue limit.

Polishing^13–15 is the final step in the precision manufacturing^16–18 of turbine blades and plays a key role in determining their dimensional accuracy. Therefore, both the polishing process¹⁹ and the reduction of blade surface roughness^20,21 must be thoroughly investigated to improve the thermal efficiency of heavy-duty gas turbines. Various methods such as robot-assisted abrasive cloth wheel polishing,²² belt wheel polishing^23–26 and multi-spindle machine tool polishing²⁷ are frequently utilized for blade polishing. Although these methods offer advantages, they can also pose challenges like tool wear and poor polishing consistency. In contrast, chemical mechanical polishing (CMP)^28–30 is a widely adopted technique for achieving surface flatness, with advanced processes capable of producing atomic-level surfaces.³¹ Achieving an atomic-level surface on stainless steel can significantly enhance its performance in various critical applications by leveraging the unique properties of ultra-smooth, defect-free surfaces. Liu et al.³² explored the application of N,N′-1,2-ethanediylbis-1-aspartic acid and 1,2,4-triazole in the CMP of GCr15 steel,³³ achieving a 1.8 nm surface roughness. Peng et al.³⁴ evaluated the use of CMP on 18CrNiMo7-6 steel, attaining a surface roughness Sa of 0.85 nm. The resultant smooth surface facilitated gear lubrication. In addition, Peng et al.³⁵ studied the CMP of 9Cr18Mo stainless bearing steel, where hydrogen peroxide facilitated the formation of a protective surface film caused by the reaction of Fe and Cr, resulting in a flat surface with a Sa of 0.63 nm. Evidently, CMP is capable of attaining a close atomic surface.³⁶ Despite these advancements, there is still significant room for improvement in CMP slurries to enhance surface roughness and improve material removal efficiency. Thus, there is a pressing need to delve into innovative formulations and processing parameters to unlock the full potential of CMP and address any existing performance gaps.

Various types of abrasives³⁷ play a crucial role in determining the surface quality and material removal rate (MRR). Luo et al.³⁸ opted for SiO₂–ZnO hybrid soft abrasives to polish sapphire wafers. The use of these hybrid abrasives resulted in a remarkable 91.68% increase in MRR compared to using SiO₂ abrasives alone. Similarly, in a separate investigation, He³⁹ investigated the CMP performance on K9 glass by employing a mixed abrasive of cerium oxide and silicon oxide. Notably, the composite abrasive containing 0.5% silicon oxide yielded the most effective polishing results, achieving a glass surface roughness (RMS) of 1.3157 nm and an MRR of 22.6 nm min⁻¹. These findings underscore the substantial enhancements in material removal efficiency and surface quality achievable through the use of mixed abrasives over singular abrasive particles.^40–42 Furthermore, the inclusion of rare earth elements has been identified to further boost the polishing performance.^43,44

To sum up, in this research, an innovative polishing solution containing rare earth abrasive particles, including SiO₂, La₂O₃, L-malic acid, H₂O₂, γ-aminobutyric acid (GABA), and deionized water (DI water), was developed. SiO₂ is a naturally occurring material and La₂O₃ is an environmentally friendly rare earth oxide.^39,45–48L-Malic acid, featuring two carboxyl groups, acts as the pH regulator, and GABA, possessing –COOH and –NH₂ groups as active sites, serves as the complexing agent. This study delves into the roles of rare earth elements in the polishing solution through a combination of experimental methods and characterization techniques. Furthermore, the study evaluates the impact of process parameters on the polishing performance, resulting in a Sa of 0.286 (±0.02) nm (50 × 50 μm²) and an MRR of 82.14 (±0.41) nm min⁻¹. These insights contribute significantly to enhancing the performance and quality of materials like 12Cr stainless steel and similar alloys.

2. Experimental design

12Cr stainless steel was cut into samples measuring 10 × 10 × 3 mm³. Three of these samples were evenly affixed to the edge of the counterweight plate using paraffin wax, as illustrated in Fig. 1. A flat polishing machine was employed for both the rough and fine polishing processes. During rough polishing, a polyurethane polishing pad was utilized with a CMP slurry. Silica abrasives averaging 1 μm in diameter were employed as coarse polishing particles. First, 10 wt% SiO₂ particles were dispersed in deionized water to form a suspension. Second, the polishing slurry was ultrasonically dispersed in deionized water for 30 minutes. Finally, the prepared slurry was applied to rough-polish 12Cr samples. Continuous stirring of the slurry was done throughout the CMP rough polishing process. Prior to each fine polishing step, the samples were polished to an average Sa of 300 (±10) nm (50 × 50 μm²) to ensure consistent thickness and surface roughness, as shown in Fig. 2. Subsequently, a black frosted leather polishing pad was used for precision polishing.

Fig. 1 Schematic diagram of the CMP process on 12Cr stainless steel.

Fig. 2 White light interferometer and microscope images after rough polishing.

To explore the impact of diverse factors on surface roughness and MRR, a series of 25 orthogonal experimental (L₂₅(5⁶)) setups were meticulously devised. The factors under study encompassed SiO₂ (characterized by an average particle size of 10 nm), pH (adjusted with L-malic acid), H₂O₂, GABA, counterweight plate pressure (hereafter referred to as “Pressure”), and polishing linear velocity (hereafter referred to as “Velocity”). Each of these six variables was examined at five distinct levels. The detailed parameter settings for the orthogonal test can be found in Table 1. Following the execution of the orthogonal test, the outcomes were methodically analyzed to pinpoint the optimal polishing process and formulation for the 12Cr steel samples.

Table 1 Configuration of experimental parameters L₂₅(5⁶)

	Factors
Levels	A	B	C	D	E	F
	SiO2 (wt%)	pH	H2O2 (wt%)	GABA (wt%)	Pressure (kPa)	Velocity (m min^−1)
1	5%	3.5	0.50%	0.5%	10	60
2	6%	4.0	0.75%	1.0%	20	75
3	7%	4.5	1.00%	1.5%	30	90
4	8%	5.0	1.25%	2.0%	40	105
5	9%	5.5	1.50%	2.5%	50	120

---

After the polishing process, the residual slurry was meticulously eliminated using deionized water, and the surface was dried utilizing nitrogen gas. The surface roughness was gauged employing a white light interferometer, and then the sample mass was accurately recorded using a precision balance. The MRR (nm min⁻¹) was calculated using the formula in eqn (1),⁴⁹ where Δm (g) represents the mass change of the 12Cr steel sample pre- and post-polishing, ρ (g nm⁻³) denotes the density ρ of the steel, s (nm²) is the polishing area, and t (min) denotes the polishing time.

The mechanism of CMP was further analyzed using an electrochemical workstation, Fourier transform infrared spectroscopy (FTIR), and X-ray photoelectron spectroscopy (XPS). For the measurement of surface roughness values, a five-point sampling approach was adopted. Precisely, roughness values were obtained from the central region as well as the four corners of the sample, subsequently culminating in an averaged value.

In this study, the UNIPOL-1200S surface polishing machine (manufactured by Shenyang Kejing Automation Equipment Co., Ltd) was utilized. The 12Cr steel samples were supplied by Dongfang Electric Corporation, while the silica and rare earth abrasive grains were sourced from Zhongye New Materials Co., Ltd. L-Malic acid (purity: ≥99.0%), hydrogen peroxide (30% purity), and GABA (purity ≥99.5%) were procured from Sinopharm Chemical Reagent Co., Ltd. Additionally, the following analytical instruments were employed: a white light interferometer (SuperView W1, Chotest Technology Inc., China), an optical microscope (BX53MRF-S, Olympus, Japan), an electrochemical workstation (CHI760E, Shanghai Chenhua Instrument Co., Ltd), an FTIR system (Thermo Scientific™ Nicolet™ iS50), an XPS system (Thermo Scientific K-Alpha, USA), and a TEM (JEOL JEM-F200, Japan).

3. Experimental results and analysis

3.1. Orthogonal experimental results and analysis

To explore the influence of six factors (SiO₂ (A), pH (B), H₂O₂ (C), GABA (D), Pressure (E), and Velocity (F)) on the surface roughness and MRR of 12Cr steel specimens, a comprehensive set of 25 orthogonal experiments was conducted across five distinct levels. Each experimental group was replicated three times, maintaining a consistent polishing duration of 45 minutes and a slurry flow rate of 8 mL min⁻¹. The average values for surface roughness and MRR can be found in Table 2. For each experimental group, a suite of white light interferometer images was carefully chosen, as shown in Fig. 3, with the test area measuring 50 × 50 μm².

Fig. 3 Sa values of all groups in 25 orthogonal experiments.

Table 2 Orthogonal experimental design and results

No.	A	B	C	D	E	F
	SiO2 (wt%)	pH	H2O2 (wt%)	GABA (wt%)	Pressure (kPa)	Velocity (m min^−1)	Sa (nm)	MRR (nm min^−1)
1	5%	3.5	0.50%	0.5%	10	60	5.294	95.73
2	5%	4.0	0.75%	1.0%	20	75	4.041	96.89
3	5%	4.5	1.00%	1.5%	30	90	1.898	89.48
4	5%	5.0	1.25%	2.0%	40	105	1.320	74.37
5	5%	5.5	1.50%	2.5%	50	120	3.109	70.22
6	6%	3.5	0.75%	1.5%	40	120	3.232	121.48
7	6%	4.0	1.00%	2.0%	50	60	4.111	103.85
8	6%	4.5	1.25%	2.5%	10	75	2.008	25.48
9	6%	5.0	1.50%	0.5%	20	90	3.043	29.04
10	6%	5.5	0.50%	1.0%	30	105	1.969	143.11
11	7%	3.5	1.00%	2.5%	20	105	2.921	150.52
12	7%	4.0	1.25%	0.5%	30	120	2.468	108.74
13	7%	4.5	1.50%	1.0%	40	60	1.291	78.22
14	7%	5.0	0.50%	1.5%	50	75	2.383	136.00
15	7%	5.5	0.75%	2.0%	10	90	3.451	29.63
16	8%	3.5	1.25%	1.0%	50	90	4.155	196.30
17	8%	4.0	1.50%	1.5%	10	105	1.435	32.89
18	8%	4.5	0.50%	2.0%	20	120	2.908	143.70
19	8%	5.0	0.75%	2.5%	30	60	1.643	55.41
20	8%	5.5	1.00%	0.5%	40	75	1.724	70.52
21	9%	3.5	1.50%	2.0%	30	75	4.605	148.15
22	9%	4.0	0.50%	2.5%	40	90	4.549	189.19
23	9%	4.5	0.75%	0.5%	50	105	2.155	203.56
24	9%	5.0	1.00%	1.0%	10	120	2.090	53.63
25	9%	5.5	1.25%	1.5%	20	60	5.503	37.63

---

Table 2 and Fig. 3 show that the no. 13 group exhibited the most favorable surface roughness, boasting a Sa-value of 0.822 nm. The corresponding MRR stood at 78.22 nm min⁻¹. In contrast, the no. 23 group exhibited the highest MRR of 203.56 nm min⁻¹, along with a Sa-value of 1.582 nm. When it comes to the polishing of 12Cr steel, surface roughness takes precedence over MRR in terms of importance. Therefore, Minitab software was used to analyze the main effect diagram of Sa. From Fig. 4, it can be deduced that with the increment of SiO₂ and pH levels, the overall surface roughness value exhibited a trend of initially decreasing and then increasing. Specifically, when the SiO₂ concentration was fixed at 8 wt% and the pH value was set at 4.5, the Sa-value hit its lowest point. Moreover, as the amounts of H₂O₂ and GABA were elevated, the surface roughness initially declined, then rose, and ultimately descended once more. Interestingly, when employing 1 wt% H₂O₂ and 1 wt% GABA, the Sa-value was minimized. Regarding the impact of pressure, as it was elevated, the surface roughness initially surged, then abated, and subsequently rebounded. Notably, at a pressure of 40 kPa, the Sa-value dropped to its lowest point. Likewise, in the case of velocity, as it was augmented, the surface roughness initially lessened, then grew, followed by another decrease and subsequent rise. Particularly noteworthy is that at a velocity of 105 m min⁻¹ the Sa-value was optimized. Among these variables, SiO₂, pH, and Velocity were demonstrated to exert a pronounced influence on surface roughness. From Table 3, the order of the influence of the variables on surface roughness was determined as follows: Velocity > pH > SiO₂ > Pressure > GABA > H₂O₂.

Fig. 4 Effects of different ingredients on the surface roughness Sa.

Table 3 Mean response

	Factors
Levels	A	B	C	D	E	F
	SiO2 (wt%)	pH	H2O2 (wt%)	GABA (wt%)	Pressure (kPa)	Velocity (m min^−1)
1	2.215	3.729	2.953	2.348	2.246	3.137
2	2.873	3.141	2.621	2.377	3.594	2.680
3	2.503	2.091	2.712	3.139	2.672	3.800
4	2.374	2.096	3.091	3.279	2.424	1.960
5	3.781	3.152	2.697	2.846	3.183	2.762
Delta	1.565	1.638	0.470	0.931	1.348	1.839
Rank	3	2	6	5	4	1

---

Within the CMP process, linear velocity emerges as a crucial parameter that holds substantial sway over the polishing results. Functioning dynamically, the linear velocity directly regulates the speed of relative motion between the abrasive particles and the workpiece surface. An escalation in linear velocity heightens the frequency at which abrasive particles sweep across the surface, thus bolstering material removal. Nevertheless, excessively high rotational speeds can disrupt the stability of the polishing process, amplifying the impact force of the abrasive particles on the workpiece and adversely affecting surface roughness.

In acidic environments, particularly at lower pH values,⁵⁰ the polishing slurry tends to become more corrosive. Under such conditions, the relatively mild chemical corrosion paves the way for the predominance of mechanical action by the abrasive particles. Nonetheless, an excess of mechanical action could potentially introduce scratches and surface imperfections, leading to heightened surface roughness and inconsistencies in surface quality. Furthermore, when the concentration of SiO₂ abrasive particles is decreased, as evidenced by a low quantity of abrasive particles in the slurry,⁵¹ the interaction between abrasive particles and the surface of 12Cr steel per unit time decreases. This reduction inevitably results in a lower MRR.

At relatively low pressures, the interplay between abrasive particles and the steel surface lacks sufficient force, hindering both chemical reactions and mechanical grinding. This inadequate force proves ineffective in removing surface protrusions and defects, resulting in higher surface roughness. As the pressure increases, the contact force and material removal efficiency improve, leading to reduced surface roughness. However, surpassing 40 kPa in pressure can prompt abrasive particles to embed into the steel surface, causing surface damage. Excessive pressure can also intensify chemical reactions on the surface, making it challenging to regulate them and further elevating surface roughness. In terms of H₂O₂ content, lower levels lead to the formation of a thin and discontinuous oxide film, hampering comprehensive coverage of the surface. In such scenarios, abrasive particles rely more on mechanical action, making it arduous to precisely remove small protrusions and defects, thereby increasing surface roughness.^52,53 As the H₂O₂ content increases, the oxide film thickens and becomes more continuous, playing a supportive role in the polishing process and enhancing surface quality.

Based on the above analysis, the optimal process parameters have been identified as A4B3C3D2E4F4, comprising 8 wt% SiO₂, pH 4.5, 1 wt% H₂O₂, 1 wt% GABA, a pressure of 40 kPa, and a velocity of 105 m min⁻¹. Using these process parameters for polishing yields a Sa of 0.491 (±0.03) nm within a 50 × 50 μm² area and an MRR of 56.32 (±2.32) nm min⁻¹.

3.2. Polishing results and analysis after adding La₂O₃

Lanthana^54,55 possesses distinctive physical and chemical attributes, including a stable crystal structure and high hardness,^56,57 enabling it to effectively aid mechanical grinding during polishing, thereby removing minor protrusions and surface defects.^58,59 Incorporating La₂O₃ with an average particle size of 20 nm based on the optimal process parameters (A4B3C3D2E4F4) had a significant impact. The relationship between surface roughness and MRR is shown in Fig. 5(b). As the quantity of La₂O₃ augments, the surface roughness experiences an initial upward trend, followed by a downward turn and then a renewed ascent. Correspondingly, the MRR first dips, then climbs, and finally drops again. Upon adding 0.75 wt% La₂O₃, the surface roughness decreased to 0.286 (±0.02) nm (50 × 50 μm²), and the MRR reached 82.14 (±0.41) nm min⁻¹. The corresponding white light interferometer and optical microscope images are shown in Fig. 6. Compared to the scenario without La₂O₃ addition, the surface roughness decreased by 41.75% and the MRR increased by 45.84%.

Fig. 5 Single factor tests of La₂O₃ and morphologies of abrasives: (a) the influence of La₂O₃ on surface roughness and MRR; (b) the influence of SiO₂ and La₂O₃ mixed abrasives on surface roughness and MRR; (c) TEM image of SiO₂; (d) TEM image of La₂O₃; and (e) TEM image of SiO₂ and La₂O₃ mixed abrasives.

Fig. 6 White light interferometer and microscope images after polishing with La₂O₃.

A comparative experiment using only La₂O₃ abrasive grains was also conducted, with the results shown in Fig. 5(a and b). As the lanthanum oxide concentration increases, the surface roughness gradually increases. The lowest Sa recorded was 2.419 (±0.04) nm at a concentration of 0.25 wt%. Across the five experimental groups, the minimum surface roughness appeared at the lowest lanthanum oxide concentration, although scratches were observed in every trial. Additionally, the MRR reached a maximum of 26.6 (±1.82) nm min⁻¹. The morphology of the abrasive grains was captured using a transmission electron microscope (TEM), as shown in Fig. 5(c), (d), and (e). Specifically, Fig. 5(c) shows the TEM image of the silicon oxide abrasive grains, Fig. 5(d) shows the TEM image of the lanthanum oxide abrasive grains, and Fig. 5(e) presents the TEM image after the mixing of silicon oxide and lanthanum oxide. The silicon oxide grains are quasi-spherical, whereas the lanthanum oxide grains exhibit an irregular shape. When only silicon oxide abrasives were used for CMP, the Sa was 0.491 (±0.03) nm within the area of 50 × 50 μm² and the MRR was 56.32 (±2.32) nm min⁻¹. The SiO₂ particles used in this study have an average particle size of 10 nm, in contrast, to the 20 nm size of the La₂O₃ particles. This uniform particle size contributes to the consistent scratch patterns observed during CMP. In addition, the La₂O₃ powder exhibits excellent dispersion in the polishing solution without agglomerating,^57,60,61 which ensures uniform interaction with the polished surface, prevents localized polishing irregularities, and ultimately leads to consistent quality.

In the synergistic polishing of stainless steel using SiO₂ and La₂O₃,^61,62 the complementary particle size and hardness properties are key to enhancing efficiency and surface quality. The larger, relatively softer La₂O₃ (Mohs hardness 6–7) acts as an abrasive for initial coarse grinding.⁵⁹ Its size provides high impact energy and cutting depth, efficiently removing thick oxide layers, macroscopic scratches, and work-hardened protrusions. Meanwhile, the smaller, harder SiO₂ (Mohs hardness 7) acts as the fine finishing agent. Its finer particles penetrate microscopic grooves left behind by La₂O₃, eliminating defects through uniform micro-cutting, significantly reducing roughness and enhancing glossiness. Combining these abrasives allows smaller SiO₂ particles to fill gaps between larger La₂O₃ particles. Experimentally, La₂O₃ first removes major protrusions and high-roughness areas, followed by SiO₂ refining the surface and reducing scratch depths.^43,45,63 Crucially, SiO₂ grains encapsulate the sharp edges of La₂O₃ particles. This dual-abrasive interaction produces a synergistic effect, significantly improving both material removal efficiency and final surface smoothness.

3.3. Electrochemical analysis

Using the optimal conditions, tests were carried out to evaluate the Tafel curves, corrosion potential, and corrosion current. Fig. 7 shows these results. In Fig. 7(a) and (b), the Tafel curves for varying H₂O₂ concentrations are presented alongside the corresponding corrosion potential (E_corr) and corrosion current density (I_corr). As the H₂O₂ concentration goes up, the open-circuit potential remains nearly constant. Meanwhile, the corrosion potential initially decreases before increasing, and the corrosion current first increases and then decreases. This behavior suggests that at higher H₂O₂ levels, an oxide layer develops on the material surface, thereby enhancing its resistance to corrosion. Notably, when the H₂O₂ content reaches 1 wt%, E_corr attains its minimum value while I_corr peaks, which aligns with the optimal mass fraction of hydrogen peroxide obtained from the orthogonal experiments. Fig. 7(c) and (d) show the Tafel curves of pH, E_corr, and I_corr. With increasing pH, the open-circuit potential exhibits more prominent shifts. Here, E_corr first climbs and then falls, whereas I_corr continuously decreases. These trends indicate that as the environment approaches neutrality, the corrosive impact on the material diminishes. The highest corrosion rate is observed at pH 3.5, where the corrosion potential is at its lowest and I_corr at its highest, as shown in Fig. 7(d). When the pH exceeds 4.5, the decline rate of I_corr becomes markedly slower and the increase in E_corr is gradual; however, at pH 5.0, E_corr begins to drop significantly. From the orthogonal test results, it is clear that pH 4.5 achieves a balanced compromise between the corrosion rate and material removal rate. Finally, Fig. 7(e) and (f) compare the Tafel curves, E_corr, and I_corr for five formulations: S1. DI water; S2. DI water + SiO₂; S3. DI water + La₂O₃; S4. DI water + pH + H₂O₂ + GABA; and S5. CMP slurry. The findings reveal that mixtures containing DI water, SiO₂, and La₂O₃ exhibit relatively mild corrosive effects. However, when L-malic acid (a pH modifier) is added along with H₂O₂ and GABA, the I_corr increases, suggesting that the polishing solution becomes more corrosive and accelerates material removal.⁵²

Fig. 7 Tafel curves, E_corr and I_corr of 12Cr steel samples after immersion: (a) and (b) H₂O₂ solution; (c) and (d) pH solution; and (e) and (f) five formulations (S1. DI water, S2. DI water + SiO₂, S3. DI water + La₂O₃, S4. DI water + pH + H₂O₂ + GABA, S5. CMP slurry).

3.4. FTIR analysis

To investigate the impact of L-malic acid and its associated chemical reactions on the 12Cr steel surface, two experimental setups were devised. In the first setup, pure L-malic acid crystals were used, as illustrated by the “L-malic acid crystals” curve. In the second, an L-malic acid—which had been used to soak the 12Cr steel for 12 hours and subsequently allowed to crystallize—was examined. The resulting spectrum from this treatment is presented in Fig. 8(a), labeled “After soaking”.

Fig. 8 FTIR spectra and complexation reaction: (a) FTIR spectrum of L-malic acid and (b) combinations of metal ions and L-malic acid.

L-Malic acid comprises two carboxyl (–COOH) groups and one hydroxyl (–OH) functional group. In aqueous solutions, the carboxyl groups dissociate, releasing hydrogen ions (H⁺). These ions, in conjunction with hydrogen peroxide, participate in redox reactions with metal atoms, oxidizing the metal and liberating electrons as metal ions form and dissolve into the solution. When the –COOH coordinates with metal ions such as Fe²⁺, Fe³⁺ or Cr³⁺, it induces a shift in the C=O stretching vibration band.⁴⁹ As shown in Fig. 8(a), the C=O stretching peak of free L-malic acid at approximately 1689 cm⁻¹ shifts to around 1683 cm⁻¹, confirming the coordination between the oxygen atom in the carboxyl group and the metal ions.⁵² Furthermore, both the carboxyl and hydroxyl groups in L-malic acid can complex with metal ions present in 12Cr steel. This complex formation increases the stability of the metal ions, thereby promoting their dissolution and exacerbating the corrosion of 12Cr steel.⁵⁰ In this context, divalent metal ions in the slurry are abbreviated as M²⁺ and trivalent metal ions as M³⁺. The forms of their complexation with L-malic acid are shown in Fig. 8(b); the reaction equation for this process is as follows, where n represents the coordination number.^{32,42,49,64,65}

Fe + H₂O₂ + H⁺ → Fe²⁺ (2)

Fe²⁺ + H₂O₂ + H⁺ → Fe³⁺ (3)

Cr + H₂O₂ + H⁺ → Cr³⁺ (4)

FeO + H⁺ → Fe²⁺ (5)

Fe₂O₃ + H⁺ → 2Fe³⁺ (6)

Cr₂O₃ + H⁺ → Cr³⁺ (7)

Fe²⁺ + nC₄H₆O₅ → [Fe(C₄H₆O₅)_n]²⁺ (8)

Fe³⁺ + nC₄H₆O₅ → [Fe(C₄H₆O₅)_n]³⁺ (9)

Cr³⁺ + nC₄H₆O₅ → [Cr(C₄H₆O₅)_n]³⁺ (10)

3.5. XPS analysis

To investigate the function of each component in the polishing liquid and the chemical reactions occurring on the workpiece surface, six groups of comparative experiments were designed. These experiments aimed to analyze both the broad (full spectrum) and narrow (fine spectrum) ranges of the primary elements (Fe and Cr) in 12Cr steel under various conditions. The first group featured the untreated 12Cr steel sample, acting as the baseline for subsequent experiments. This group was instrumental in discerning the element distribution on the original surface. Following this, the second group involved immersing the sample surface in a pH 4.5 L-malic acid solution for a duration of 12 hours. Contrasting the results of this group with the first one allowed for the evaluation of the effect of L-malic acid in the CMP process. Subsequently, the third group consisted of samples immersed in a 1 wt% H₂O₂ solution for 12 hours, facilitating an examination of the role of H₂O₂. In the fourth group, the sample surface was immersed in a 1 wt% GABA solution for the same duration to examine the impact of GABA. The fifth group involved immersing the sample in a 0.75 wt% La₂O₃ solution for 12 hours, aiming to assess the effect of rare earth elements. In the final group, the sample was exposed to the complete set of conditions of the polishing liquid for 12 hours. By comparing the results from this group with those from the previous five groups, the interplay of the components in the polishing liquid was investigated. The XPS test findings are presented in Fig. 9.

Fig. 9 XPS fine spectra and atomic percentages: (a) spectral diagrams of Fe 2p in different valence states; (b) atomic percentages of Fe metal, Fe²⁺, and Fe³⁺; (c) spectral diagrams of Cr 2p in different valence states; and (d) atomic percentages of Cr metal, Cr₂O₃, and Cr(OH)₃.

Fig. 9(a) and (b) show the fine spectra of Fe 2p and the atomic percentages of Fe metal, Fe²⁺, and Fe³⁺. The peaks identified at 706 eV, 709 eV, and 711 eV correspond to Fe metal, Fe²⁺ and Fe³⁺, respectively.⁶⁶ In Fig. 9(b), it can be seen that the original 12Cr steel sample has atomic percentages of approximately 12% Fe, 68% Fe²⁺, and 20% Fe³⁺. Subsequent immersion in a pH 4.5 L-malic acid solution resulted in a decrease in Fe³⁺ alongside an increase in Fe and Fe²⁺ concentrations. Exposure to 1 wt% H₂O₂ and 1 wt% GABA solutions led to a decline in Fe³⁺ concentrations and an increase in both Fe and Fe²⁺, with Fe²⁺ peaking at about 86%. In the case of immersion in a 0.75 wt% La₂O₃ solution, Fe³⁺ dwindled to 1%; conversely, both Fe and Fe²⁺ surged, with Fe surpassing its original content at 18%. Following exposure to the complete polishing solution, the atomic percentages of Fe, Fe²⁺, and Fe³⁺ were measured at 13%, 81%, and 6%, respectively.

Fig. 9(c) and (d) show the fine spectra of Cr 2p and the atomic percentages of Cr metal, Cr₂O₃ and Cr(OH)₃. The peaks at 573 eV, 575–578 eV and 576 eV correspond to Cr metal, Cr₂O₃ and Cr(OH)₃, respectively. From Fig. 9(d), the atomic percentages for the untreated 12Cr steel sample are found to be approximately 19% Cr, 50% Cr₂O₃ and 31% Cr(OH)₃. Treatment with a pH 4.5 L-malic acid solution led to a reduction in Cr content as the Cr₂O₃ and Cr(OH)₃ levels increased. Similar trends were observed upon exposure to 1 wt% H₂O₂ and 1 wt% GABA solutions, with decreasing Cr levels and increasing Cr₂O₃ and Cr(OH)₃ contents. Immersion in a 0.75 wt% La₂O₃ solution resulted in a 1% drop in Cr content, an increase in Cr₂O₃, and a decrease in Cr(OH)₃. Post-exposure to the complete formula, the Cr content decreased to 4%, the Cr₂O₃ content increased to 67%, and Cr(OH)₃ decreased to 29%.

The data in Fig. 9 reveal that H₂O₂ aids in the oxidation of Fe and Cr in 12Cr steel, transforming them into Fe²⁺, Fe³⁺, and Cr³⁺. L-Malic acid and GABA interacted with Fe³⁺ and Cr³⁺ to form complexes. The addition of La₂O₃ promoted a chemical reaction, resulting in a 19.18% increase in Fe³⁺ and a 20.32% increase in Cr³⁺. Notably, the metallic iron content on the surface post immersion in the complete formulation remained nearly unchanged, suggesting that GABA impedes both the oxidation and dissolution of metallic iron. Furthermore, after exposure to the total formula, the Cr metal content decreased, while Cr³⁺ oxide levels increased, indicating that the formulation aids in the removal of Cr from the surface.^67,68

Based on the above analysis, the CMP mechanism for 12Cr steel can be elucidated, as illustrated in Fig. 10. Initially, hydrogen peroxide in the CMP slurry exhibits potent oxidizing properties, leading to the oxidation of surface constituents of the 12Cr steel. This primarily involves the oxidation of metal elements (Fe, Cr) and low-valent metal oxides (Fe²⁺) to high-valent oxides (Fe³⁺, Cr³⁺), forming an oxide layer on the surface. In addition, L-malic acid and GABA in the slurry have a distinct corrosive effect on the 12Cr steel.⁶⁹ These reagents ionize to release H⁺ ions, which dissolve and react with both the metal and its oxide, thereby degrading the oxide layer. The metal ions (Fe²⁺, Fe³⁺, Cr³⁺, etc.) dissolved in the slurry subsequently interact with functional groups such as –COOH, –OH, and –NH₂ present in the L-malic acid and GABA molecules, leading to the formation of chelate precipitates.^70–75 Upon deprotonation of the carboxyl group (–COOH) in L-malic acid, malate ions are generated. These negatively charged ions coordinate with iron ions to form complexes.^42,66,76 Eventually, as the oxide film undergoes partial dissolution, softening, and degradation, the mixed abrasives of SiO₂ and La₂O₃ enhance material removal through micro-cutting. The interplay of chemical and mechanical actions occurs in a continuous cycle, ultimately reaching an equilibrium state, resulting in a finely polished atomic surface of the 12Cr steel.

Fig. 10 Atomic model of the CMP mechanism on 12Cr stainless steel.

4. Conclusions

In summary, a sustainable polishing solution incorporating rare earth La₂O₃ was developed through systematic experimentation and theoretical analysis to meet the exacting polishing standards of 12Cr steel. By meticulously probing into the polishing mechanisms and fine-tuning the process parameters, a surface on 12Cr steel with near-atomic smoothness was achieved post CMP. The incorporation of rare earth La₂O₃ facilitated the chemical reactions and enhanced material removal efficiency, leading to a 41.75% reduction in surface roughness and a substantial 45.84% increase in MRR. These results hold significant implications for advancing the use of 12Cr steel in high-end manufacturing applications and offer innovative technical approaches for the precise processing of similar materials.

Author contributions

Yaowen Wu: investigation, formal analysis, data curation, and visualization. Dong Wang: formal analysis and investigation. Zhenyu Zhang: funding acquisition, project administration, methodology, supervision, and conceptualization. Feng Zhao: formal analysis and investigation. Hongxiu Zhou: formal analysis, data curation, and visualization. Xiuqing Liu: formal analysis and data curation. Xiaofei Yang: formal analysis and data curation.

Data availability

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

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.

Acknowledgements

The authors acknowledge the financial support from the Hainan Provincial Natural Science Foundation of China (525QN264), the Key Research and Development Program of Yunnan Province (202402AB080009), the Natural Science Foundation of Jiangsu Province (BK20221361), the Natural Science Foundation of Liaoning Province (2023JH2/101700163), the State Key Laboratory of Clean and Efficient Turbomachinery Power Equipment Open Project (DEC8300CG202318698EE280503), the Fundamental Research Funds for the Central Universities (DUT24YG209), the Science and Technology Special Fund of Hainan Province (ZDYF2024GXJS026 and ZDYF2023GXJS148), and the Changjiang Scholars Program of the Chinese Ministry of Education.

References

1. S. Jiang, H. Wang, Y. Wu, X. Liu, H. Chen, M. Yao, B. Gault, D. Ponge, D. Raabe, A. Hirata, M. Chen, Y. Wang and Z. Lu, Nature, 2017, 544, 460–464.
2. Y. M. Wang, T. Voisin, J. T. McKeown, J. Ye, N. P. Calta, Z. Li, Z. Zeng, Y. Zhang, W. Chen, T. T. Roehling, R. T. Ott, M. K. Santala, P. J. Depond, M. J. Matthews, A. V. Hamza and T. Zhu, Nat. Mater., 2018, 17, 63–71.
3. S. Zhang, H. Feng, H. Li, Z. Jiang, T. Zhang, H. Zhu, Y. Lin, W. Zhang and G. Li, Nat. Commun., 2023, 14, 7869.
4. X. X. Wei, B. Zhang, B. Wu, Y. J. Wang, X. H. Tian, L. X. Yang, E. E. Oguzie and X. L. Ma, Nat. Commun., 2022, 13, 726.
5. J. Biehler, H. Hoche and M. Oechsner, Surf. Coat. Technol., 2017, 313, 40–46.
6. B. Blakey-Milner, P. Gradl, G. Snedden, M. Brooks, J. Pitot, E. Lopez, M. Leary, F. Berto and A. Du Plessis, Mater. Des., 2021, 209, 110008.
7. E. L. Dalibón, A. Dalke, H. Biermann and S. P. Brühl, Surf. Coat. Technol., 2024, 485, 130931.
8. X. Yu, S. Zhao, G. An, Y. Xu and X. Xu, J. Turbomach., 2024, 146, 31007.
9. D. Wu, H. Wang, K. Zhang and X. Lin, J. Manuf. Process., 2019, 39, 305–326.
10. Q. Miao, W. Ding, J. Xu, L. Cao, H. Wang, Z. Yin, C. Dai and W. Kuang, Int. J. Extreme Manuf., 2021, 3, 045102.
11. R. Liu, Y. Pan, A. Chen, G. Bin and H. Li, Powder Technol., 2023, 430, 119037.
12. D.-S. Bae and J.-K. Lee, Int. J. Precis. Eng. Manuf., 2024, 25, 2125–2131.
13. B. Yu, Y. Gu, J. Lin, S. Liu, S. Zhang, M. Kang, Y. Xi, Y. Gao, H. Zhao and Q. Ye, Surf. Coat. Technol., 2023, 475, 130162.
14. A. Temmler, I. Ross, J. Luo, G. Jacobs and J. H. Schleifenbaum, Surf. Coat. Technol., 2020, 403, 126401.
15. W. Tillmann, L. Hagen, D. Stangier, N. F. Lopes Dias, J. Görtz and M. D. Kensy, Surf. Coat. Technol., 2021, 428, 127905.
16. X. Wang, R. Chen and S. Sun, Int. J. Extreme Manuf., 2023, 5, 43001.
17. L. Liu, Z. Xu, Y. Hao and Y. Teng, Int. J. Extreme Manuf., 2025, 7, 015104.
18. B. Zhao, Y. Wang, J. Peng, X. Wang, W. Ding, X. Lei, B. Wu, M. Zhang, J. Xu, L. Zhang and R. Das, Int. J. Extreme Manuf., 2024, 6, 062012.
19. Y. Mei, W. Chen and X. Chen, Materials, 2023, 16, 3393.
20. G. Zhu, X. Zeng, Z. Gao, Z. Gong, W. Duangmu, Y. Zeng and C. Lu, J. Manuf. Process., 2023, 99, 636–651.
21. S. Luo, L. Liao and Y. Wang, J. Manuf. Process., 2024, 131, 494–506.
22. J. Zhang, J. Liu and S. Yang, Int. J. Adv. Manuf. Technol., 2022, 119, 8211–8225.
23. Z. Chen, Y. Shi, X. Lin, T. Yu, P. Zhao, C. Kang, X. He and H. Li, Results Phys., 2019, 12, 870–877.
24. J. Zhang, Y. Shi, X. Lin and Z. Li, Int. J. Adv. Manuf. Technol., 2017, 93, 3383–3393.
25. P. Zhsao and Y. Shi, Chin. J. Mech. Eng., 2013, 26, 988–996.
26. G. Xiao and Y. Huang, Int. J. Adv. Manuf. Technol., 2015, 78, 1473–1484.
27. Z. Yun, X. Wang, Z. Chen, Z. Zhengqing and Y. Huan, Int. J. Adv. Manuf. Technol., 2022, 123, 1669–1678.
28. H. Yan, X. Niu, M. Qu, F. Luo, N. Zhan, J. Liu and Y. Zou, Int. J. Adv. Manuf. Technol., 2023, 125, 47–71.
29. X. Shi, G. Pan, Y. Zhou, L. Xu, C. Zou and H. Gong, Surf. Coat. Technol., 2015, 270, 206–220.
30. Z. Liu, Z. Zhang, J. Feng, X. Yi, C. Shi, Y. Gu, F. Zhao, S. Liu and J. Li, Nanoscale, 2024, 16, 85–96.
31. Z. Luo, Z. Zhang, F. Zhao, C. Fan, J. Feng, H. Zhou, F. Meng, X. Zhuang and J. Wang, Mater. Today Sustainability, 2024, 27, 100841.
32. J. Liu, P. Hao, L. Jiang and L. Qian, Tribol. Lett., 2022, 70, 67.
33. G. Ji, H. Sun, H. Duan, D. Yang and J. Sun, Surf. Coat. Technol., 2021, 420, 127330.
34. W. Peng, Y. Gao, L. Jiang, J. Liu and L. Qian, Lubricants, 2022, 10, 199.
35. W. Peng, C. Huang, S. Zhang, Y. Chen, Y. Han, L. Jiang and L. Qian, J. Solid State Electrochem., 2023, 27, 467–477.
36. B. Pan, R. Kang, J. Guo, H. Fu, D. Du and J. Kong, J. Manuf. Process., 2019, 44, 47–54.
37. W. Xie, Z. Zhang, L. Liao, J. Liu, H. Su, S. Wang and D. Guo, Nanoscale, 2020, 12, 22518–22526.
38. Z. Luo, J. Lu, Q. Yan, D. Hu and Y. Zhou, Mater. Sci. Semicond. Process., 2024, 176, 108318.
39. Q. He, Appl. Nanosci., 2018, 8, 163–171.
40. L. Liao, Z. Zhang, F. Meng, D. Liu, J. Liu, Y. Li and X. Cui, J. Manuf. Process., 2021, 66, 198–210.
41. H. Wu, L. Jiang, J. Liu, C. Deng, H. Huang and L. Qian, Tribol. Lett., 2020, 68, 34.
42. H. Wu, L. Jiang, X. Zhong, J. Liu, N. Qin and L. Qian, Friction, 2021, 9, 1673–1687.
43. Z. Wang, Z. Zhang, H. Zhou, D. Han, C. Shi, L. Chen, J. Yao, S. Yu and J. Xu, Appl. Surf. Sci., 2025, 681, 161586.
44. E. Kim, J. Lee, C. Bae, H. Seok, H.-U. Kim and T. Kim, Powder Technol., 2022, 397, 117025.
45. H. Lei and K. Tong, Precis. Eng., 2016, 44, 124–130.
46. M. Moothedan and K. B. Sherly, AIP Conf. Proc., 2011, 1391, 549–551.
47. A. Trunschke, D. L. Hoang, J. Radnik and H. Lieske, J. Catal., 2000, 191, 456–466.
48. A. Kumar, B. Jayabalan, C. Singh, J. Jain, S. Mukherjee, K. Biswas and S. S. Singh, Met. Mater. Int., 2023, 29, 1067–1078.
49. H. Li, Z. Zhang, C. Shi, H. Zhou, J. Feng, D. Tong and F. Meng, Appl. Surf. Sci., 2024, 657, 159787.
50. M. Naznin, J. Choi, W. S. Shin and J. Choi, Sep. Sci. Technol., 2017, 52, 2888–2898.
51. W. Choi, J. Abiade, S.-M. Lee and R. K. Singh, J. Electrochem. Soc., 2004, 151, G512.
52. Z. Zhang, L. Liao, X. Wang, W. Xie and D. Guo, Appl. Surf. Sci., 2020, 506, 144670.
53. H. Li, Z. Zhang, C. Shi, H. Zhou, J. Feng, D. Tong and F. Meng, Appl. Surf. Sci., 2024, 657, 159787.
54. C. Yang, X. Hou, Z. Li, X. Li, L. Yu and Z. Zhang, Appl. Surf. Sci., 2016, 388, 497–502.
55. J. Liu, L. Jiang, H. Wu, X. Zhong and L. Qian, Tribol. Lett., 2021, 69, 161.
56. A. Behatha, V. K. Sharma, S. Gummula and K. Venkatakrishnan, Mater. Today Commun., 2021, 26, 101830.
57. Y. Chen, Y. Fang, P. Cheng, X. Ke, M. Zhang, J. Zou, J. Ding, B. Zhang, L. Gu, Q. Zhang, G. Liu and Q. Yu, Nat. Commun., 2024, 15, 4105.
58. H. Liu, S. Wang, J. Liang, H. Hu, Q. Li and H. Chen, Crystals, 2023, 13, 515.
59. K. Huang, S. Lai, M. Guo, X. Zhu, J. Yuan, Z. Liu, G. Hu and Y. Gao, J. Rare Earths, 2024, 42, 424–430.
60. A. Huminic, G. Huminic, C. Fleacă, F. Dumitrache and I. Morjan, J. Mol. Liq., 2019, 287, 111013.
61. Z. Kou, C. Wang, W. Zhou, A. Chen and Y. Chen, Appl. Surf. Sci., 2024, 657, 159733.
62. P. Li, X. Guo, S. Yuan, M. Li, R. Kang and D. Guo, Appl. Surf. Sci., 2021, 554, 149668.
63. J. Cheng, S. Huang, Y. Li, T. Wang, L. Xie and X. Lu, Appl. Surf. Sci., 2020, 506, 144668.
64. B. Ensing, F. Buda and E. J. Baerends, J. Phys. Chem. A, 2003, 107, 5722–5731.
65. Z.-L. Xie and Z.-H. Zhou, ACS Appl. Mater. Interfaces, 2023, 15, 35710–35719.
66. T. Marshall-Roth, N. J. Libretto, A. T. Wrobel, K. J. Anderton, M. L. Pegis, N. D. Ricke, T. V. Voorhis, J. T. Miller and Y. Surendranath, Nat. Commun., 2020, 11, 5283.
67. X. Qi, H. Cai, X. Zhang, J. Ouyang, D. Lu, X. Guo and S. Jia, Chem. Eng. J., 2023, 475, 146320.
68. W. Xie, Z. Sun, Z. Bian, H. Liu and J. Hu, Chem. Eng. J., 2023, 465, 142989.
69. P. S. Bagus, C. J. Nelin, C. R. Brundle, B. V. Crist, N. Lahiri and K. M. Rosso, J. Chem. Phys., 2021, 154, 094709.
70. J. Liu, L. Jiang, H. Wu, T. Zhao and L. Qian, J. Electrochem. Soc., 2020, 167, 131502.
71. L. Büker, R. Dickbreder, R. Böttcher, S. Sadowski and A. Bund, J. Electrochem. Soc., 2020, 167, 162509.
72. Y. Hao, H. Ma, Q. Wang, C. Zhu and A. He, Ecotoxicol. Environ. Saf., 2022, 240, 113676.
73. A. Dasque, M. Gressier, P.-L. Taberna and M.-J. Menu, Results Chem., 2021, 3, 100207.
74. J. A. Mahmud, M. Hasanuzzaman, K. Nahar, A. Rahman, Md. S. Hossain and M. Fujita, Ecotoxicology, 2017, 26, 675–690.
75. F. Li, X. Duan, H. Li, L. Zou, G. Liu, F. Liu, G. Zhang and J. Xu, Microchem. J., 2022, 178, 107426.
76. X. Guo, S. Yuan, J. Huang, C. Chen, R. Kang, Z. Jin and D. Guo, Appl. Surf. Sci., 2020, 505, 144610.

---