Exploring the role of −NH2 functional groups of ethylenediamine in chemical mechanical polishing of GCr15 bearing steel

Ethylenediamine with two −NH2 functional groups was used as a critical complexing agent in chemical mechanical polishing (CMP) slurries for a high carbon chromium GCr15 bearing steel (equivalent to AISI 52100). The polishing performance and corresponding mechanism of −NH2 functional groups were thoroughly investigated as a function of pH. It is revealed that, when polished with ethylenediamine and H2O2-based slurries, the material removal rate (MRR) and surface roughness Ra of GCr15 steel gradually decrease as pH increases. Compared with acidic pH of 4.0, at alkaline pH of 10.0, the surface film of GCr15 steel has much higher corrosion resistance and wear resistance, and thus the material removal caused by the pure corrosion and corrosion-enhanced wear are greatly inhibited, resulting in much lower MRR and Ra. Moreover, it is confirmed that a more protective composite film, consisting of more Fe3+ hydroxides/oxyhydroxides and complex compounds with −NH2 functional groups of ethylenediamine, can be formed at pH of 10.0. Additionally, the polishing performance of pure iron and a medium carbon 45 steel exhibits a similar trend as GCr15 steel. The findings suggest that acidic pH could be feasible for amine groups-based complexing agents to achieve efficient CMP of iron-based metals.


Introduction
Bearing steels, such as a high carbon chromium GCr15 bearing steel (also named AISI 52,100, 100Cr6 and SUJ2. The chemical composition as follows: 0.95-1.05 wt% C, 1.30-1.65 wt% Cr, 0.15-0.35 wt% Si, 0.2-0.4 wt% Mn, < 0.027 wt% P, < 0.02 wt% S, and balance Fe [1]), have been widely used in industrial key basic parts such as bearings due to their superior properties [2]. It was found that, under the harsh working conditions, smooth surfaces of the contact elements of bearings can result in a decrease in the heat generation rate and a significant improvement of the lubrication performance between the rollers and the raceways [3,4]. Moreover, the decrease in the surface roughness of the contact elements can improve the ratio between the thickness of the lubricating film and the composite roughness, which can effectively extend the service life of bearings [5]. Therefore, ultra-smooth surfaces are required to improve the service life and operation performance of ultraprecise devices which rely on bearings and other key basic parts made of bearing steels. However, traditional machining techniques, such as grinding and lapping, might not be able to meet such stringent requirements [6]. Therefore, it is urgent to develop an efficient ultra-precision machining technique to achieve an ultra-smooth surface.
Chemical mechanical polishing (CMP) technique has been intensively used in the semiconductor industry for manufacturing ultra-large scale integrated circuits [7][8][9]. CMP can achieve an ultra-smooth and damage-free surface by the synergistic effect of chemical and mechanical interactions [7,8]. For this reason, CMP has been gradually applied in processing key basic parts with bearing steels as the main material, such as GCr15 steel [10][11][12][13][14]. The complexing agent is one of the most critical additives for metals CMP, and it is used to enhance the material removal rate (MRR) as well as to reduce defects by chelating metal ions to form soluble complexes. Complexing agents containing -NH 2 functional groups are widely used in H 2 O 2 -based slurries for CMP of metals, such as copper [15][16][17]. Liu et al. [15] reported that the copper MRR increased with the concentration of ethylenediamine (containing two -NH2 functional groups) increasing, and a rather high copper MRR of 1,899 nm/min could be achieved with the slurry containing 0.1 M ethylenediamine and 0.6 wt% H 2 O 2 . The research group of Babu et al. found that, when polished with the slurry containing 0.13 M ethylenediamine and 5 wt% H 2 O 2 , the dissolution rate and polishing rate of copper increased as pH increased, indicating that -NH 2 functional groups are more active to bind cooper ions in alkaline environment [16,17]. In short, pH plays an important role in the polishing performance of -NH 2 functional groups of complexing agents. However, the investigation of the effect of -NH 2 functional groups of complexing agents on the CMP of GCr15 steel, a typical bearing steel, is still lacking. In our previous work, it was found that the organic base, highly likely a type of amine, in TiSol-NH 4 dispersion could serve as a complex agent to achieve efficient CMP of GCr15 steel [18]. Moreover, our preliminary experimental results reveal that, the MRR of GCr15 steel gradually decreases rather than increases when it is polished with TiSol-NH 4 dispersion and pH value increases from 5.0 to 7.0, which is apparently different from that of copper [16,17]. Therefore, it is significant to figure out the role of -NH 2 functional groups of complexing agents in GCr15 steel CMP.
In this study, ethylenediamine (H 2 NCH 2 CH 2 NH 2 ) with merely two -NH 2 functional groups was used as the complexing agent [16,17] for GCr15 steel CMP to exclude the interference from other functional groups such as -COOH functional groups. In most cases, a proper oxidizer is essential to obtain a high metal MRR by the synergistic effect of oxidation and complexation since complexing agents cannot directly dissolve metal. Conventionally, H 2 O 2 has been used as a strong oxidizer for metals CMP since it is readily available and free of contamination [19]. It was found that, when GCr15 steel was polished with the slurries containing H 2 O 2 and glycine, with the increase in the H 2 O 2 content, the MRR of GCr15 steel reached the peak value at 0.01 wt% H 2 O 2 and nearly the valley value at 0.1 wt% H 2 O 2 , indicating that 0.01 wt% and 0.1 wt% H 2 O 2 can provide two typical chemical-mechanical states, respectively [13]. In light of these, 0.01 wt% and 0.1 wt% H 2 O 2 as the proper oxidation conditions together with ethylenediamine were used to elucidate the role of -NH 2 functional groups of complexing agents for GCr15 steel CMP. The effect of -NH 2 functional groups of ethylenediamine on the polishing performance of GCr15 steel was studied as a function of pH value. The corresponding surface film of GCr15 steel was characterized by various experiments including static etching, electrochemical and atomic force microscopy (AFM) wear experiments as well as X-ray photoelectron spectroscopy (XPS) measurements. Afterward, a preliminary polishing mechanism of GCr15 steel was put forward. Additionally, the effect of -NH 2 functional groups of ethylenediamine was further assessed on other iron-based metals, including pure iron and a medium carbon 45 steel (also named AISI 1045 steel).

Material and methods
Polishing slurries were prepared with deionized (DI) water, colloidal silica (YZ8040 with an average particle size of approximately 80 nm, purchased from Shanghai YZ-Lapping Material Co., Ltd.), and chemical reagents including ethylenediamine sulfate (purchased from TCI (Shanghai) Development Co., Ltd.) and H 2 O 2 (purchased from Sinopharm Chemical Reagent Co., Ltd.). After all the components were added and mixed homogenously, the pH was adjusted with dilute H 2 SO 4 and KOH. In this study, ethylenediamine sulfate was used to provide ethylenediamine since ethylenediamine has an ultra-strong alkalinity. The ultra-strong alkalinity will bring about several issues, such as dissolution of colloidal silica [20], and the addition of a large quantity of H 2 SO 4 for adjusting pH to neutral and acidic range.
Polishing experiments were performed on a benchtop polisher (UNIPOL-1200S, Shengyang Kejing Auto-instrument Co., Ltd, China). A GCr15 steel disk with diameter of 50.8 mm and thickness of 1.5 mm was polished under the following polishing conditions: set-pressure of 5.0 kg, carrier speed/ platen speed of 60 rpm/60 rpm, slurry flow rate of and polishing time of 1 min. The polishing pad used in the experiments was an IC1010/Suba-IV composite pad with K-type groove. In between each polishing step, a pad ex-situ conditioning was carried out with a diamond conditioning disk. Before the polishing, the disk was pre-polished to obtain a surface roughness R a of approximately 2.7 nm. After polishing, the MRR was calculated by measuring the weight loss with a Sartorius ME36S microbalance (1 μg readability). The post-CMP surface morphology was measured by using an optical 3D surface profiler (Chotest SuperView W1, Chotest Technology Inc., China), and the surface roughness R a was calculated with the analysis software coming with the instrument. The corresponding measured surface area was 97.9 μm × 97.9 μm. If no otherwise specified, each experimental condition was repeated at least four times for all the polishing and characterization experiments.
In order to investigate the pure corrosion of GCr15 steel caused by the slurries, static etching experiments were carried out at room temperature. A GCr15 steel disk with diameter of 25.4 mm and thickness of 1.5 mm was first mechanically polished with an abrasive paper made of silicon carbide (P2000), and then was degreased with ethyl alcohol and cleaned with DI water. Afterwards, it was immersed in the designated solution (without adding abrasive) for 3 min, during which the solution was gently stirred to ensure the uniform dispersion. The static etching rate (SER) was calculated by measuring the weight loss. Given the fact that the mechanically polished surface of GCr15 steel is so rough that it is practically impossible to characterize the post-etching surface morphology, a GCr15 steel sample of 7 mm by 7 mm was first pre-polished with the slurry (pH of 10.0) containing 2 wt% colloidal silica, 0.1 M ethylenediamine sulfate, 0.01 wt% H 2 O 2 , and DI water to obtain a smooth surface with 1.9 nm R a , and then it was immersed in the designated solution (without adding abrasive) for 1 min. Afterward, the post-etching surface morphology was measured by using the optical 3D surface profiler, and an AFM (SPA-300HV/SPI3800N Probe Station, Seiko, Japan) in much smaller areas of 10 μm × 10 μm and 1 μm × 1 μm with a silicon nitride AFM probe (MSCT, Bruker, USA, nominal radius of curvature less than 20 nm).
In order to characterize the electrochemical properties of GCr15 steel in the designated solutions (without adding abrasive), electrochemical experiments, including open-circuit potential (OCP), electrochemical impedance spectroscopy (EIS) and potentiodynamic polarization, were performed by using an electrochemical workstation (Autolab PGSTAT 302N, Metrohm Autolab B.V., Switzerland) with a 200 mL three-electrode cell. A platinum electrode was employed as the counter electrode, an Ag/AgCl electrode with a saturated KCl reference solution was employed as the reference electrode, and a cylindrical GCr15 steel sample of 5 mm diameter was employed as the working electrode, which was encased in epoxy resin and electrically connected through a copper wire. Prior to each measurement, the working electrode was first mechanically polished with an abrasive paper made of silicon carbide (P2000), and then was degreased with ethyl alcohol and cleaned with DI water. For each experimental condition, the OCP Friction 9(6): 1673-1687 (2021) | https://mc03.manuscriptcentral.com/friction measurement was first carried out for 30 min to obtain a steady state. Then, the EIS measurement was performed in the frequency range from 100 kHz to 0.1 Hz with a sinusoidal AC perturbation of 5 mV amplitude. The disturbance electrical signal was applied on the basis of the corresponding opencircuit potential. Afterward, the potentiodynamic polarization measurement was conducted with a scan rate of 1 mV/s and a step of 0.45 mV. The EIS data was analyzed by equivalent electrical circuit modeling with the NOVA analysis software.
In order to further assess the corrosionenhanced wear of GCr15 steel caused by the slurries, wear tests of a pretreated GCr15 steel sample rubbed against a chemically inert cone-shaped diamond tip (NC-LC, Adama, Ireland, spring constant of 94.25 N/m, and nominal radius of curvature of 20 nm) were conducted on an AFM (SPA-300HV/SPI3800N Probe Station, Seiko, Japan) in vacuum. The experimental conditions were set as follows: applied load F n of 2-10 μN, relative sliding velocity v of 0.5 μm/s, relative sliding length L of 1 μm, and reciprocating sliding cycle N of 1. Prior to each test, a GCr15 steel sample of 7 mm by 7 mm was first pre-polished with the slurry (pH of 10) containing 2 wt% colloidal silica, 0.1 M ethylenediamine sulfate, 0.01 wt% H 2 O 2 and DI water, and a smooth surface with 1.9 nm R a was acquired. Then, it was immersed in the designated solution (without adding abrasive) for 1 min. Afterward, it was rinsed with DI water, dried, and then immediately placed in a vacuum chamber for the wear test. After the wear test, a silicon nitride AFM probe (MSCT, Bruker, USA) was used to scan the surface topography of the wear area in vacuum.
In order to characterize the chemical composition of the post-CMP and post-etching GCr15 steel surfaces, XPS measurements were performed on an X-ray photoelectron spectrometer (AXIS Ultra DLD, Kratos Analytical, United Kingdom) with a hemispherical energy analyzer and 128 channel detectors. Prior to the measurement, the GCr15 steel sample was either pre-polished or pre-etched with the designated slurries (the etching solutions did not contain the abrasive). Afterward, it was rinsed with DI water, dried, and then immediately placed in a vacuum chamber. The XPS vacuum chamber pressure was less than 7 × 10 10 mbar. All the spectra were obtained at 90° photoelectron take-off angle from the surface. The high-resolution spectra of Fe, Cr, O, and N were measured. Afterward, CasaXPS software was employed to analyze the data. In addition, at pH of 2.0, 0.1 wt% H 2 O 2 leads to a higher MRR of GCr15 steel than 0.01 wt% H 2 O 2 , which is different from other pH values from 4.0 to 10.0 [13,21]. As reported, at pH of 2.0, the direct dissolution of copper could be the main reaction, and the influence of -NH 2 functional groups of complexing agents could be insignificant [22].  [13]. Based on the polishing results, in both of the two distinct chemical-mechanical states with the addition of either 0.01 wt% or 0.1 wt% H 2 O 2 and 0.1 M ethylenediamine, the MRR and surface roughness R a of GCr15 steel gradually decrease when pH increases, indicating pH has a significant impact on the performance of -NH 2 functional groups of ethylenediamine for GCr15 steel CMP. Moreover, it should be pointed out particularly that, the MRR trend of GCr15 steel as pH changes is entirely different from that of copper as reported, where the copper MRR gradually increased as pH increased with ethylenediamine as the complexing agent [16,17].  The total MRR in CMP can be decomposed into the following four parts [26][27][28][29]: the MRR by the pure wear, the MRR by the corrosion-enhanced wear, the MRR by the pure corrosion, the MRR by the wear-enhanced corrosion. Specifically, the MRR by the corrosion-enhanced wear is mainly attributed to the modification of the mechanical strength of the surface film by slurries, especially by the complexing agent [29][30][31]. In this study, the pure corrosion and the corrosion-enhanced wear caused by -NH 2 functional groups of ethylenediamine were emphatically investigated by using static etching, electrochemical and AFM wear experiments collectively.
The results of the static etching experiments of GCr15 steel are shown in Fig. 3. The etching solutions were composed of 0.1 M ethylenediamine sulfate, 0.01 wt% H 2 O 2 , DI water, and with pH of 4.0 or pH of 10.0. It can be seen from Fig. 3(a) that, the SER of GCr15 steel at pH of 4.0 is 23.1 nm/min, and in striking contrast, the SER at pH of 10.0 is nearly undetectable. As shown in Fig. 3(b) and 3(c), after being statically etched at pH of 4.0, the surface of GCr15 steel becomes quite porous and rough; In contrast, after being statically etched at pH of 10.0, the surface remains compact and smooth. As shown in Fig. 3(d), the corresponding surface roughness R a of GCr15 steel with 97.9 μm × 97.9 μm, 10 μm × 10 μm, and 1 μm ×1 μm measured surface areas all substantially confirms the improvement of the post-etching surface quality as pH rises from 4.0 to 10.0.
The results of the electrochemical experiments of GCr15 steel, including OCP and potentiodynamic polarization plots, are shown in Fig. 4. The solutions were composed of 25 mM Na 2 SO 4 , 0.1 M ethylenediamine sulfate, 0.01 wt% H 2 O 2 , DI water, and with pH of 4.0 or pH of 10.0. As shown in Fig. 4(a), after 30-min immersion, both of the two OCP plots at pH of 4.0 and pH of 10.0 become extremely stable, which provides a favorable surface condition for the following electrochemical measurements. Moreover, the OCP value at pH of 4.0 is higher than that at pH of 10.0. The difference in the OCP value should be associated with the change in the anodic and cathodic reactions at different pH [32]. As shown in Fig. 4(b), in comparison with the potentiodynamic polarization plot at pH of 4.0, both the anodic and cathodic reactions of the plot at pH of 10.0 are suppressed pronouncedly, especially the anodic reactions, which is consistent with the trend of the SER results shown in Fig. 3(a). Moreover, at pH of 10.0, an apparent passivation region emerges at an overpotential of approximately 0.4 V (vs. E Ag/AgCl ) in the anodic curve. More specifically, at pH of 4.0 and with a low concentration of H 2 O 2 (0.01 wt%), the main anodic reaction can be depicted as follows [33]: The cathodic reaction can be depicted as follows [29]: (2) Then, a majority of the reaction product Fe 2+ are complexed by -NH 2 functional groups of ethylenediamine. Specifically, for the acidic form of ethylenediamine, pK a1 is 6.85 and pK a2 is 9.93 [17,34]  and dissolved, and then the fresh surface of the GCr15 steel working electrode underneath is exposed continuously, further promoting the anodic and cathodic reactions.
While at pH of 10.0, the anodic reaction is the same as that at pH of 4.0 shown in Eq. (1). The cathodic reactions can be depicted as follows [39]: However, unlike the scenario at pH of 4.0, at pH of 10.0, the reaction product Fe 2+ can then react with OH  in the solution as the following reactions [33]: (7) and (9) could be either from the dissolved oxygen in the solution or from the decomposition product of H 2 O 2 [40]. The oxidation function of O 2 in Eq. (9) can also be achieved by H 2 O 2 . As a result, insoluble iron hydroxides/ oxyhydroxides, such as Fe(OH) 3 and FeOOH, can be formed on the GCr15 steel electrode surface, which should primarily account for the suppression of the anodic dissolution. Additionally, as revealed by the following XPS results, ethylenediamine molecules can be adsorbed onto the oxide film as a supplementary protection layer, resulting in the further enhancement of the corrosion resistance [34,41].
EIS experiments were further carried out to quantitatively investigate the electrochemical properties of the surface film of GCr15 steel formed by -NH 2 functional groups of ethylenediamine at different pH. The Nyquist plots and the corresponding equivalent electrical circuits are shown in Fig. 6. The corresponding quantitative fitting results are listed in Table 1. There is merely one semicircle for both pH of 4.0 and pH of 10.0. In the equivalent electrical circuits, R sol represents the solution resistance, R ct represents the charge transfer resistance between the surface film of GCr15 steel and the solution, and CPE dl represents the double layer capacitance between the surface film of GCr15 steel and the solution. According to the fitting results shown in Table 1, when pH changes from 4.0 to 10.0, R ct rises from 77.8 to 921.2 ·cm 2 , indicating that the surface film gradually provides effective protection for the substrate of GCr15 steel by restricting the mass transfer of reactants and products between GCr15 steel and the electrolyte [42]. Meanwhile, the admittance Y 0dl of CPE dl decreases as pH increases, which could be caused by the increase in the thickness of the electrical double layer [43,44]. In addition, at pH of 10.0, there is an extra Warburg impedance W associated with the diffusion process of oxygen through the solution to the surface film of GCr15 steel [45]. The above results of the static etching and electrochemical experiments substantially demonstrate that, in the presence of ethylenediamine and H 2 O 2 , the surface film of GCr15 steel formed at pH of 10.0 can provide a much higher resistance to the pure corrosion than that at pH of 4.0.  Then, the corrosion-enhanced wear caused by -NH 2 functional groups of ethylenediamine was assessed by using AFM wear tests, where the surface film of GCr15 steel was rubbed against a chemically inert diamond tip in vacuum [46]. The immersion solutions for the GCr15 steel pretreatment were composed of 0.1 M ethylenediamine sulfate, 0.01 wt% H 2 O 2 , DI water, and with pH of 4.0 or pH 10.0. The typical AFM surface topographical images and average cross-section profiles of the wear traces after the wear tests are shown in Figs. 7(a) and 7(b). As displayed in Fig. 7(a), after being pretreated at pH of 4.0, wide and deep wear scars are formed on the GCr15 steel surface by scratching with the diamond tip for one reciprocating sliding cycle; In comparison, after being pretreated at pH of 10.0, narrow and shallow wear scars along with plenty of pileups are formed on the GCr15 steel surface. The width difference in the wear scars could be related to the surface structure. In particular, as shown in Fig. 3(b), at pH of 4.0, the surface of GCr15 steel is porous and rough, and thus the mechanical strength is relatively low. The diamond tip penetrates deeply into the surface and push the material surrounding the tip readily as it scratches, and thereby the wear scar is wide. As show in Fig. 3(c), at pH of 10.0, the surface of GCr15 steel becomes compact and smooth, and thus the mechanical strength grows high. It becomes difficult for the diamond tip to push the surrounding material, and thereby the wear scar is narrow. The corresponding statistical data of the wear depth and volume are shown in Fig. 7(c) and 7(d), respectively. It can be seen that, with the increase in the applied load, both the wear depth and volume almost linearly increase, which could be explained by Archard model [47]. Moreover, both the wear depth and volume at pH of 4.0 are much higher than those at pH of 10.0, suggesting that the surface film of GCr15 steel formed at pH of 10.0 can provide a much higher resistance to the  XPS technique was used to further identify the chemical composition of the surface film of GCr15 steel. The XPS spectra of the GCr15 steel surfaces after being either polished or etched are shown in Fig. 8. The polishing slurries were composed of 2 wt% colloidal silica, 0.1 M ethylenediamine sulfate, 0.01 wt% H 2 O 2 , DI water, and with pH of 4.0 or pH of 10.0. The etching solutions were the same as the polishing slurries except for removing the abrasive. As shown in Fig. 8(a), the deconvolution of Fe(2p 3/2 ) spectra of both pH of 4.0 and pH of 10.0 reveals three peaks. The left peak with 706.5 eV binding energy should be assigned to metallic Fe [48], the medium peak with 709.6 or 709.3 eV binding energy should be assigned to Fe 2+ compounds [48,49], and the right peak with 711.1 or 711.0 eV should be assigned to Fe 3+ compounds [50,51]. Moreover, for the polishing preparation method, when pH increases from 4.0 to 10.0, the proportion of metallic Fe keeps approximately 26.0 at%, the proportion of Fe 2+ compounds decreases from 48.5 at% to 41.0 at%, and the proportion of Fe 3+ compounds increases from 24.2 at% to 33.3 at%. The etching preparation method shares a similar trend of the proportion change to the polishing one. In combination with the results of static etching and electrochemical experiments, it can be concluded that a more protective layer with a higher Fe 3+ content can be formed at a higher pH value. As shown in Fig. 8(c), the deconvolution of O(1s) spectra of both pH of 4.0 and pH of 10.0 reveals two peaks, of which that with the lower binding energy at 529.6 eV should be assigned to oxygen as a form of O  [52], and that with the higher binding energy at 531.3 or 531.1 eV should be assigned to oxygen as a form of OH  [53]. As shown in Fig. 8(d), the deconvolution of N(1s) spectra of both pH of 4.0 and pH of 10.0 reveals one peak, which should be assigned to nitrogen atoms of the complex compounds formed by the complexing reactions between iron ions and of -NH 2 functional groups of ethylenediamine [36]. Moreover, after being either polished or etched, the area of the N(1s) peak increases when pH changes from 4.0 to 10.0, indicating more ethylenediamine molecules are adsorbed onto the GCr15 steel surface probably due to the complexing reaction between ethylenediamine and Fe 3+ [41].

Polishing mechanism of GCr15 steel by -NH 2 functional groups at different pH
Summarizing up all the above polishing and characterization results, a preliminary polishing mechanism of GCr15 steel by -NH 2 functional groups of ethylenediamine at different pH value can be proposed as follows, and a schematic diagram of the polishing mechanism is depicted in Fig. 9. As shown in Fig. 9 (5)), and severe corrosion occurs on the surface, resulting in a high SER and the formation of a porous and nonuniform surface film with many deep corrosion pits (Fig. 3), which is of a low wear resistance to the mechanical abrasion ( Fig. 7). Hence, under such corrosion condition, the MRR and surface roughness R a is high (Fig. 1); As pH gradually increases to alkaline pH of 10.0, as shown in Fig. 9(b), insoluble Fe 3+ hydroxides/ oxyhydroxides are mainly formed (Eqs. (8) -(10)) on the GCr15 steel surface, then -NH 2 functional groups of ethylenediamine are absorbed onto the oxide film to further enhance the corrosion resistance (confirmed by the XPS results shown in Fig. 8), resulting in an extremely low SER and a compact and protective surface film (Fig. 3), which is of a high wear resistance to the mechanical abrasion (Fig. 7). Hence, under such increasingly protective condition, the MRR and surface roughness R a decrease accordingly.

Extension to other iron-based metals
To and surface roughness R a . Specifically, the MRR and surface roughness R a both decrease when pH changes from 4.0 to 10.0. Therefore, it can be concluded that the CMP performance of -NH 2 functional groups of ethylenediamine at different pH can be potentially applied to other iron-based steels, such as die steel and tool steel. In summary, the findings of the study provide beneficial guide for screening appropriate amine groups-based complexing agents for CMP of iron-based metals in acidic pH range. More complexing agents, such as polyamines [54], can be evaluated to achieve a high MRR in an acidic environment. Such a high MRR can be used for the initial step of the finishing process of rough surfaces of iron-based metals. More investigation on introducing appropriate corrosion inhibitors as counterparts of complexing agents is needed to balance the corrosion and the protection, so as to achieve a high MRR and a satisfactory surface quality simultaneously.

Conclusions
Ethylenediamine (obtained from ethylenediamine sulfate to avoid the strong alkalinity) with two -NH 2 functional groups was introduced into H 2 O 2 -based slurries as a key complexing agent for GCr15 steel CMP. The role of -NH 2 functional groups of ethylenediamine in GCr15 steel CMP and the underlying polishing mechanism were investigated. Additionally, the effect of -NH 2 functional groups of ethylenediamine on CMP of other iron-based metals was assessed. Based on the above results and discussion, the conclusions can be reached as follows: 1) In the presence of 0.1 M ethylenediamine and either 0.01 wt% or 0.1 wt% H 2 O 2 , the MRR and surface roughness R a of GCr15 steel gradually decrease as pH increases from 2.0 to 10.0, which clearly shows that pH has an effective influence on the polishing performance of -NH 2 functional groups of ethylenediamine for GCr15 steel CMP.
2) At pH of 4.0, with the addition of ethylenediamine and a low concentration of H 2 O 2 , severe corrosion occurs on the surface, which is attributed to the complexation reactions between -NH2 functional groups of ethylenediamine and Fe 2+ ions to form soluble [Fe(H 2 O) m (En) n ] 2+ complex compounds, and a porous and nonuniform surface film with a low wear resistance to the mechanical abrasion is formed, and thus the MRR and R a of GCr15 steel is high. When pH increases to 10.0, a compact and protective surface film with a high wear resistance to the mechanical abrasion, comprising more Fe 3+ hydroxides/oxyhydroxides and complex compounds with -NH 2 functional groups of ethylenediamine, is formed, and thus the MRR and R a of GCr15 steel decreases.
3) In ethylenediamine and H 2 O 2 -based slurries, the MRR and surface roughness R a of the three types of iron-based metals, i.e. GCr15 steel, pure iron and 45 steel, perform the same trend when pH changes from 4.0 to 10.0. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.