The Study of the Etching Resistance of YOF Coating Deposited by Atmospheric Plasma Spraying in HBr/O₂ Plasma

Abstract

Yttrium oxyfluoride (YOF) coatings with different oxygen content were prepared using atmospheric plasma spraying (APS) technology. The etching resistance of the coatings in HBr/O₂ plasma was investigated. Shifts in diffraction peaks of the X-ray diffraction, along with XPS analysis conducted before and after etching, demonstrated that Br ions could replace O and F ions and fill the oxygen vacancies after exposure to HBr/O₂ plasma, which is supported by the first-principles calculations. Br ions formed a protective layer on the surface of the YOF coating, slowing down further etching by Br ions. By adjusting the oxygen mass fraction in YOF powder, the oxygen vacancy concentration and Br ion filling were regulated to enhance etching resistance. YOF coatings with 6% oxygen content exhibited improved etching resistance compared to YOF coatings with 3% and 9% oxygen content. This improvement was primarily due to the increased Br ion concentration. These findings provide a new approach for developing coatings with enhanced etching resistance.

1. Introduction

With the rapid development of the semiconductor industry, the size of semiconductor components is decreasing, and the accuracy requirements of the corresponding circuit diagrams are becoming higher and higher. This trend has encouraged the use of plasma etching technology in semiconductor device preparation, micro-nano processing technology, and micro-electronic manufacturing technology [1,2,3]. However, during the plasma etching process, the plasma erodes not only silicon wafers but also parts in the plasma etch machine’s reaction chamber. This process generates pollution particles in the reaction chamber, and as the particle size increases, the weaker corrosion particles fall off the surface. As the process repeats, the plasma corrodes the surface, and the contaminated particles fall off, contaminating the wafer and reducing the semiconductor product’s yield [4,5,6,7,8,9]. Therefore, it is essential to take reasonable measures to slow the plasma etching of the internal parts of the plasma etching machine for the development of the semiconductor industry. Generally, a protective coating resistant to plasma etching can be prepared on the surface of parts to effectively mitigate the etching damage [2,10,11,12].

Yttrium-based ceramic coatings produced by APS technology are widely used in the semiconductor industry, owing to their exceptional hardness, superior wear resistance, elevated dielectric strength, remarkable corrosion resistance, and chemical stability [13,14,15,16]. The YOF coatings produced by APS received significant attention for their superior etching resistance [14,17,18,19,20]. The concentration of F-free radicals generated in fluorinated plasma is elevated, facilitating the formation of fluoride pollution particles. In comparison to fluorine- and chlorine-based plasmons, HBr/O₂ plasma has a diminished pure chemical etching component and does not generate a polymer, rendering it less corrosive. Consequently, it exhibits superior anisotropy in etched components and an exceptional bombardment effect [21,22]. YOF exhibits commendable etching resistance in NF₃ and CF₄ plasma environments, attributed to its dense crystal structure, elevated hardness, and corrosion resistance [23,24]. However, its etching resistance in HBr plasma atmospheres remains unexamined. Furthermore, the plasma etching process will involve both physical and chemical etching [25,26]. For example, during silicon etching, an electric field induces charged particles in the plasma to continuously strike the silicon wafer, disrupting the bonds within the silicon atoms to facilitate physical etching and eliminate the etched regions [27]. In YOF coating, the lower electronegativity of Br compared to F and O results in atomic bonds that are not readily disrupted by bromine ions in HBr/O₂ plasma. The defects on the material surface also influence the etching resistance of the coating. The fewer the defects, the better the etching resistance [28,29].

In this study, X-ray diffraction (XRD) was employed to identify the phase composition of YOF powders and coatings. Scanning electron microscopy (SEM) was used to examine the surface morphology and microstructure of YOF coatings before and after etching. X-ray photoelectron spectroscopy (XPS) was used to analyze the elemental composition of the coatings before and after plasma etching. Simulations were conducted to confirm the ability of Br ions to replace O and F atoms in the coating lattice.

2. Materials and Methods

2.1. Experimental Methods

YOF coatings were prepared utilising America Oerlikon Metco’s APS system, which generates a high-temperature plasma jet by ionising a mixture of argon and hydrogen gas to melt the raw materials and pushes the molten particles onto the anodised aluminium substrate (50 mm × 50 mm × 3 mm), making a 200 μm thick coating. The spraying method relies on carrier gas. The detailed thermal spraying deposition parameters are shown in

Table 1. Thermal spraying deposition parameters.

Distance (mm)	Pitch (mm)	Robot Speed (mm/s)	Ar Gas	H2 Gas	Carrier Gas	Voltage (V)	Current (A)	Power (Kw)	Feeding
130	4	750	45	11.2	3.2	74	650	48.1	20%

. The durability of the produced coatings under plasma etching was evaluated utilising the Kiyo EX inductively coupled plasma (ICP) etching system from Lam Research Company, using HBr/O₂ and He gases as plasma sources. Before the APS process, sandblasting was used to roughen the substrate’s surface to Ra = 4–6 μm. A 200 μm thick YOF coating was formed on the surface of a smaller aluminium substrate (20 mm × 20 mm × 3 mm) using the APS method, which was the cover material used in the experiment. Figure 1 provides a schematic illustration of the etching chamber system. The etching chamber received HBr, O₂, and He gases at flow rates of 300, 5, and 15 standard cubic centimetres per minute (sccm), respectively, while maintaining a constant pressure of 6 mTorr. The transformer coupled plasma (TCP) RF power was set to 400 W, and the RF bias voltage was set to 250 V. The plasma exposure time of the coated substrate was 300 h, and the detailed process parameters are shown in

Table 2. HBr/O₂ plasma etching parameters.

Pressure (mTorr)	TCP RF Power (W)	Bias RF Voltage (V)	O2 Flow Rate (sccm)	HBr Flow Rate (sccm)	He Flow Rate (sccm)
6	400	250	5	300	15

. The cross-sectional and surface morphology as well as the microstructure of coated samples and HBr/O₂ plasma-etched samples were analysed using field emission scanning electron microscopy (CIQTEK SEM5000, Hefei, China). An Xplore 30 EDS detector (Oxford Instruments, UK) was used, and the software AZtecOne (Version 6.0 SP2) was employed for quantification. X-ray diffraction (XRD, Rigaku Ultima IV, Japan) analysis was performed using Cu K$\mathsf{\alpha }$ ($\mathsf{\lambda }$ = 1.5418 Å) radiation, with a scanning range of 10–90° (2$\mathsf{\theta }$) and a scanning rate of 2°/min. X-ray photoelectron spectroscopy (XPS, K-Alpha, Thermo Scientific, Waltham, MA, USA) used a 100 V monochromatic Al-K$\mathsf{\alpha }$ X-ray source with a spot size of 450 μm, a passing energy of 50 eV, and a detection angle of 45°. The subsurface elemental compositions were assessed using depth profiling with Ar⁺ ions on the etched surface for 300 s at 30 s intervals. The average particle diameter was measured by BT-9300ST (Dandong Bettersize Instruments, China) laser particle size distributor. The coating thickness was measured by a gauge Elcometer model A456 (UK). The simulation calculations involved in this study were performed using the CASTEP module in Materials Studio [30]. The K-point of the model in the first Brillouin zone was set to fine precision. The exchange–correlation function was adjusted by GGA-PBE (Perdew–Burke–Ernzerhof). The plane cutoff energy used in the calculation was 750 eV, and the k-point was set to 4 × 4 × 4 in the Monkhorst–Pack scheme. In order to ensure the accuracy of the calculation, the self-consistent field (SCF) convergence threshold was set to ${10}^{-6}$, the stress convergence deviation was less than 0.05 GPa, the maximum interatomic force was less than 0.03 eV/Å, and the interatomic tolerance offset was less than 0.001 Å.

2.2. Experimental Materials

Commercially purchased YOF/YF₃ powder mixtures (Nippon Yttrium Co., Ltd., Omuta, Japan) with different YOF content (represented by oxygen weight% of 3%, 6%, and 9%), referred to as YOF 3%, YOF 6%, and YOF 9%, were used as a feedstock coating material. Figure 2 presents the SEM image of YOF series powders, revealing that all raw powder materials consist of condensed spherical particles. The particle size of powder materials is a critical physical property in the production of powder coatings, significantly influencing the spray performance of the final coatings. Spherical powders with a narrow particle size distribution and a small angle of repose typically exhibit favourable fluidity [15].

Table 3. Parameters of YOF powders.

Parameter	YOF 3%	YOF 6%	YOF 9%
Average particle size (μm)	20.5	22.4	22.2
Angle of Repose (°)	34.7	31.5	30.7

presents the average particle size and the angle of repose of YOF 3%, YOF 6%, and YOF 9% powders. The average particle size of YOF 3%, YOF 6%, and YOF 9% powders was 20.5 μm, 22.4 μm, and 22.2 μm, respectively. The angles of repose for YOF powders at concentrations of 3%, 6%, and 9% were 34.7°, 31.5°, and 30.7°, respectively. Typically, the angle of repose represents the maximum angle that forms between the free slope of a powder accumulation layer and the horizontal plane. Reduced friction and improved fluidity are correlated with a smaller angle of repose.

3. Results and Discussion

Figure 3a presents the XRD pattern of the YOF series powder, indicating that the powder consists of several compositions, including Y₅O₄F₇, Y₆O₅F₈, Y₇O₆F₉, and YF₃. This result is consistent with standard data for YOF (JCPDS No. 71-2100) and YF₃ (JCPDS No. 74-0911). Figure 3b–d show the XRD pattern of the YOF coatings before and after etching. After etching, the entire YOF coating’s diffraction peak shifts to the left. Using the highest peak of each coating as an example, the diffraction peak of YOF₃% shifts from 28.04° to 27.61°. The YOF 6% diffraction peak shifts from 28.04° to 27.62°. The YOF 9% diffraction peak varies from 28.64° to 28.24°. According to the Bragg diffraction equation, the shift of the diffraction peak to the left indicates that the lattice constant increases. The reason for the diffraction peak deviation may be that some Br ions (Br ion radius of 0.196 nm) replaced F and O ions in the coating during the etching process (F and O ion radius of 0.133 nm and 0.14 nm, respectively).

The SEM images of YOF 3%, YOF 6%, and YOF 9% in Figure 4 illustrate the microstructure’s composition of smooth and rough regions before and after etching on the surface. Molten particles diffuse to form the smooth area. Rough areas are formed by particles that are not fully melted. Figure 3a–c show the surface topography of YOF 3%, YOF 6%, and YOF 9% coatings prior to etching, respectively. The figure illustrates that the surface of the YOF 6% coating (Figure 3b) is smoother than that of the YOF 3% (Figure 3a) and YOF 9% Figure 3c) coatings. Comparing the morphology of the coating before and after etching reveals different depths of depression in the surface morphology of YOF 3% and YOF 9% after etching (Figure 4d,f), a result of the coating’s exposure to Br ions. In contrast, the YOF 6% coating (Figure 4e) has a relatively flat surface after etching. The results show that YOF 6% has better etching resistance than the other two coatings.

To further assess the corrosion resistance of the coating, a baffle is employed to block a portion of the coating’s surface. The depth of the unshielded area is measured post-etching. As shown in Figure 5, the etching depths of YOF 3%, YOF 6%, and YOF 9% coatings are recorded as 5.672 μm, 3.003 μm, and 8.677 μm, respectively. The results show that the YOF 6% coating’s etching depth is the least profound, indicating superior etching resistance. Furthermore, examination of the coating cross section reveals that the cracks and pores in the YOF 6% coating (Figure 5b) post-etching are fewer and comparatively flat, indicating that the YOF 6% coating has superior etching resistance.

The XPS results of the YOF coatings before and after HBr/O₂ plasma exposure are displayed in Figure 6. It can be seen that the YOF coatings have four elements (Y, O, F, and C) before etching and five elements (C, O, F, Br, and Y) after etching. This suggests that bromine ions remain on the coatings’ surface after etching. The atomic percentage of the elements in the YOF coating was calculated using Advantage software (Thermo Scientific, USA), and the results are shown in

Table 4. The relative atomic percentages of Y, O, F, and Br elements in the coatings before and after plasma exposure (obtained via EDS quantification).

Atomic Percentage (%)	YOF 3%		YOF 6%		YOF 9%
	Pre- Etching	Post- Etching	Pre- Etching	Post- Etching	Pre- Etching	Post- Etching
Y3d	17.2%	29.4%	25.6%	28.9%	32.6%	33.1%
O1s	56.4%	55.0%	55.3%	52.9%	59.4%	43.3%
F1s	27.8%	13.3%	19.1%	17.1%	24.1%	7.3%
Br3d	/	0.9%	/	1.0%	/	0.3%

. It can be seen that the atomic percentage of Y3d in the three coatings of YOF 3%, YOF 6%, and YOF 9% increased after etching, from 17.2%, 25.6%, and 32.6% to 29.4%, 28.9%, and 33.1%, respectively. After etching, the atomic percentages of O1s and F1s decreased. The atomic percentages of O1s decreased from 56.4%, 55.3%, and 59.4% to 55.0%, 52.9%, and 43.3%, respectively. The atomic percentages of F1s decreased from 27.8%, 19.1%, and 24.1% to 13.3%, 17.1%, and 7.3%, respectively. After bromine plasma bombardment, O and F in the YOF series coatings were sputtered out, resulting in a decrease in O and F. Before etching, the coatings did not contain Br elements. After etching, the atomic percentages of Br3d in the three coatings of YOF 3%, YOF 6%, and YOF 9% were 0.9%, 1.0%, and 0.3%, respectively. The improvement of the etching resistance of the coatings may be related to the increase in the concentration of Br content.

XPS was used to fit oxygen vacancies to study the reasons for the change in Br content in YOF coatings before and after etching. The fitting results are shown in Figure 7. The light blue peaks indicate the binding energies of Y3$d_{5/2}$ and Y3$d_{3/2}$, which correspond to Y-O bonding, while the dark blue peaks indicate the binding energies of oxygen vacancies. The YOF series coatings have oxygen vacancies before etching. The oxygen vacancy concentrations of YOF 3%, YOF 6%, and YOF 9% coatings are 4.56%, 7.5%, and 33.55%, respectively. After etching, the oxygen vacancy concentrations of YOF 3%, YOF 6%, and YOF 9% coatings are reduced to 1.76%, 3.21%, and 14.07%, respectively. The reduction in oxygen vacancy concentration may be related to the filling of oxygen vacancies by Br ions.

In order to study the change in bromine content in the YOF series coatings before and after etching, the first principle simulation was used to calculate whether bromine can replace O and F as well as oxygen vacancies in the YOF coatings, and the lattice parameters after substitution were calculated. The simulation results are shown in Figure 8. Y₅O₄F₇ has been selected as the model because of its comparative stability. Figure 8 displays (a) the Y₅O₄F₇ cell, (b) the aerobic vacancy within the cell, (c) the cell following the replacement of F ion by Br ion, and (d) the cell following the replacement of O ion by Br ion. The unit cell system energy after Br atoms replace O vacancies, F atoms, and O atoms is close to the Y₅O₄F₇ system energy before substitution, indicating that after etching, B atoms can replace O atoms, F atoms, and oxygen vacancies in the Y₅O₄F₇ unit cell. Figure 8b–d show that the unit cell constants increase following the substitution of Br atoms for O vacancies, F atoms, and O atoms, in comparison to the Y₅O₄F₇ unit cell. The outcomes agree with the XRD results shown in Figure 3. The simulation results show that after etching, although most Br atoms do not enter the coating, a few Br atoms can replace O or F atoms and fill in O vacancies, slowing down the effect of etching on the coating. In summary, the Br atomic percentage of the YOF 6% coating post-etching exceeds that of the YOF 3% and YOF 9% coatings, which is 1.0%. Consequently, the incorporation of Br atoms into the coating mitigates further etching by Br plasma, resulting in the YOF 6% coating exhibiting superior Br plasma etching resistance.

4. Conclusions

YOF coatings with varying oxygen content on aluminum substrates were prepared using APS technology, and the etching resistance of these coatings in HBr/O₂ plasma was investigated. Plasma etching caused a leftward shift in the diffraction peaks of YOF coatings, suggesting an increase in the lattice constant due to Br ions replacing F and O ions in the lattice. SEM observations revealed distinct microstructural changes before and after etching. The YOF 6% coating exhibited superior etching resistance compared to YOF 3% and YOF 9%, with less surface roughness and fewer pores after etching. XPS analysis demonstrated the changes in elemental composition before and after etching. The atomic percentage of Y increased while F and O decreased, which can be attributed to bromine plasma bombardment. The presence of Br on the surface post-etching suggests that Br ions play a role in filling oxygen vacancies, corroborated by first-principles simulations. The simulation showed that Br ions could replace F and O atoms in the YOF lattice, contributing to the reduced oxygen vacancy concentration observed after etching. Compared to YOF 3% and YOF 9% coatings, YOF 6% has superior etching resistance. It is attributed to the replacement of O and F ions by Br ions and the filling of oxygen vacancies. This research indicates that YOF film is expected to provide a good protective barrier against damage caused by the HBr/O₂ plasma etching process.