Previous single-factor experiments have clarified the independent influence mechanisms of three process gases (Ar, CF₄, and O₂) on SiC plasma machining performance. However, actual mass production of SiC substrates involves strong coupling effects among five critical parameters, including gas flow rates, RF machining power, and torch-to-wafer distance. Single-factor analysis can only characterize the individual effect of a single variable and fails to adapt to the complex multi-parameter working conditions in industrial mass production.
Accordingly, this study systematically investigates the intrinsic mechanisms of two core electromechanical parameters, namely RF machining power and plasma torch-to-wafer distance. Based on L16(4⁵) orthogonal experiments, the coupling interactions among the five process parameters are comprehensively explored. The optimal mass-production-oriented parameter combination is screened out to simultaneously achieve supreme material removal efficiency, superior surface quality, and zero substrate damage.

1. Machining Power: The Core Balance Factor Between Machining Efficiency and Thermal Damage

RF machining power serves as the core energy input of the plasma system, which fundamentally determines plasma electron density, ionization intensity, jet temperature, and the total yield of active particles. It is the dominant parameter governing SiC etching efficiency. The magnitude of machining power directly modulates the CF₄ pyrolysis efficiency, fluorine radical (F*) concentration, and particle diffusion range of the plasma jet.
In this experiment, gas ratio, machining duration, and machining distance were fixed, and a power gradient ranging from 350 W to 650 W with an interval of 50 W was set for comparative controlled tests.

Figure 1: Influence curves of machining power on plasma machining performance
 
The experimental results indicate that under sufficient gas supply conditions, the peak removal rate, full width at half maximum (FWHM), and volumetric removal rate of SiC etching exhibit a significant linear increasing trend with the rise of machining power. Elevated power substantially enhances the electron collision frequency and energy density inside the plasma, enabling efficient pyrolysis of massive CF₄ molecules and a sharp increase in reactive F* radical concentration. Meanwhile, the diffusion range of high-energy active particles is expanded, broadening the effective machining width of the plasma jet and realizing dual improvement in etching depth and machining coverage.
Nevertheless, machining power has a strict safety threshold and acts as a critical risk control indicator for industrial mass production. The plasma jet temperature rises linearly with increasing power, leading to intensified thermal effects. Experimental verification demonstrates that when the power is elevated to 700 W, excessive local temperature and concentrated thermal stress induce obvious deformation or even fracture of SiC wafers, resulting in irreversible substrate scrapping.
Engineering Conclusion: Machining power can effectively improve SiC etching efficiency, whereas a rigorous balance between machining performance and thermal damage risk must be maintained. Over-power operation is prohibited in industrial production. Orthogonal experiments are essential to determine the optimal power balance point for comprehensive process optimization.

2. Machining Distance: Key Parameter Determining Particle Utilization Efficiency and Machining Uniformity

Machining distance is defined as the vertical spacing between the plasma torch nozzle and the SiC wafer surface. It directly dominates the divergence degree of plasma jets, the transmission distance of active particles, and energy loss magnitude, serving as a vital spatial parameter affecting etching uniformity and material utilization efficiency. Under atmospheric environments, plasma jets diverge rapidly after exiting the nozzle. Active particles collide with neutral air molecules during transmission, generating continuous energy loss, which becomes more severe as the machining distance increases.
With the machining power fixed at 500 W and standard gas ratio maintained, machining distance gradients ranging from 3 mm to 7 mm were set to systematically verify the influence law of spatial parameters on machining performance.

Figure 2: Influence curves of machining distance on plasma machining performance
 

Figure 3: Material removal contour at the machining distance of 3 mm
 

Figure 4: Material removal contour at the machining distance of 7 mm
 
The experimental results present a distinct staged attenuation law:
(1) Optimal interval (3–5 mm): The plasma jet exhibits low divergence, with negligible energy loss of active F* radicals and high particle utilization efficiency. The peak removal rate remains stable with minor FWHM fluctuation. The etching contour fits the standard Gaussian curve perfectly, delivering excellent machining uniformity.
(2) Attenuation interval (5–6 mm): The plasma jet is severely disturbed by ambient air, causing dramatic energy loss of active particles and a decline in effective radical concentration. Both volumetric removal rate and FWHM decrease sharply, resulting in degraded machining performance.
(3) Failure interval (>7 mm): Severe jet divergence and intense ambient airflow interference lead to obvious fluctuations in etching contours, which deviate significantly from the standard Gaussian morphology. The machining uniformity fails completely, and the material removal efficiency drops to a low level.
Engineering Conclusion: To ensure high machining efficiency and forming precision of SiC substrates in mass production, the machining distance must be strictly controlled within 5 mm, and the optimal process window is 3–3.5 mm.

3. Five-Parameter Orthogonal Experiment: Global Optimal Process Screening Under Multi-Parameter Coupling Effects

To address the limitations of single-factor experiments and adapt to complex industrial mass-production conditions, five core process parameters, including machining power, machining distance, Ar flow rate, O₂ flow rate, and CF₄ flow rate, were selected for multi-objective optimization. Each parameter was set at four levels, and 16 groups of comparative tests were designed based on the L16(4⁵) orthogonal experimental table. Taking volumetric removal rate (the core evaluation indicator for mass production) as the sole criterion, range analysis was adopted to clarify the influence weight of each parameter and screen out the optimal parameter combination.
 

Table 1: Orthogonal experimental design matrix

Table 2: Range analysis of volumetric removal rate
 
According to the range (R) value analysis, the influence weight of the five parameters on SiC plasma etching efficiency is ranked as follows: machining power > Ar flow rate > O₂ flow rate > machining distance > CF₄ flow rate. Among them, machining power and carrier gas flow rate are the dominant parameters determining overall machining efficiency, while fluorine source gas and auxiliary gas act as fine-tuning parameters for performance optimization.
A core innovation of this orthogonal experiment is that the optimal parameter combination obtained does not appear in the 16 experimental sample groups. This fully reflects the unique advantage of orthogonal experiments in mining global optimal solutions under multi-parameter coupling conditions and effectively avoids the local optimal trap of traditional single-factor optimization methods.

4. Verification of Mass-Production-Level Optimal Process Parameter Combination

Combining the results of range analysis, parameter coupling laws, surface quality requirements, and zero-damage processing criteria, the global optimal process parameters for atmospheric-pressure plasma etching of SiC substrates are ultimately determined as follows: 550 W machining power, 3.5 mm machining distance, 15 SLM Ar flow rate, 70 SCCM CF₄ flow rate, and 20 SCCM O₂ flow rate.
 

Figure 5: Removal function contour plot under optimal process parameters
 
Parameter verification experimental results show that under this optimized parameter combination, the peak plasma etching removal rate reaches 0.5736 μm, the FWHM is 3.456 μm, and the volumetric removal rate is as high as 9.119 μm³/s, which is the maximum value among all tested working conditions in this study. Meanwhile, the processed contour presents uniform morphology without obvious deposition defects or thermal damage. This parameter combination perfectly meets four core mass-production requirements: high machining efficiency, superior surface quality, reliable process stability, and zero substrate damage.

5. Core Engineering Conclusions

(1) Carrier gas (Ar): Low-flow operation is recommended under stable discharge conditions to prevent active particle dilution and improve etching efficiency.
(2) Fluorine source gas (CF₄): The optimal flow threshold must be strictly controlled; excessive gas supply will inhibit plasma ionization and reduce machining efficiency.
(3) Auxiliary gas (O₂): The optimal process window of 20–40 SCCM can completely eliminate surface deposition defects and synergistically enhance the etching reaction efficiency.
(4) Machining power: Etching efficiency increases linearly with power. Strict threshold control is required to avoid thermal deformation and fracture of SiC wafers, and 550 W is verified as the optimal power balance point.
(5) Machining distance: The machining distance should be strictly controlled within 5 mm for mass production, and 3.5 mm maximizes particle utilization efficiency and machining uniformity.
(6) Optimal mass-production parameter combination: The combination of 550 W power, 3.5 mm machining distance, 15 SLM Ar, 70 SCCM CF₄, and 20 SCCM O₂ enables high-efficiency, high-precision, and damage-free plasma modification of SiC substrates.

6. Industry Summary

Atmospheric-pressure plasma etching technology effectively solves the inherent industrial pain points of traditional SiC substrate processing, including inevitable machining damage, low efficiency, and high manufacturing cost. Through precise multi-parameter coupling optimization, this technology fully satisfies the ultra-precision modification requirements of high-end SiC substrates, possessing high industrialization and engineering application value. It provides a novel technical route for low-cost, high-yield mass production of third-generation semiconductor substrates.
 
JXT stably supplies full-specification silicon carbide substrates, covering 2–12 inch standard wafers and customized small-size square substrates. The product portfolio includes conductive, semi-insulating, and special crystal-type SiC substrates with customizable thickness, crystal orientation, and dimensional specifications. The products are widely applicable to full-process scenarios such as plasma modification, epitaxial growth, and device fabrication, supporting the efficient mass production of the third-generation semiconductor industry.