Surface Finish Standards Explained: Ra Values, Grit Chart, ASME BPE, and Semiconductor Metrics
A complete technical reference for surface roughness parameters — Ra, Rq, Rz, WIWNU, TTV — with grit-to-Ra cross-reference tables, ASME BPE SF classification breakdown, and industry-specific surface finish requirements for semiconductor and pharmaceutical applications.
Surface finish standards are the common language between design engineers who specify polished surfaces and the process teams and suppliers who produce and verify them. Without a shared, unambiguous definition of “smooth,” specifications are open to interpretation and disputes. This reference article defines all major surface finish parameters used in semiconductor manufacturing, pharmaceutical equipment, and precision industrial fabrication, provides a comprehensive grit-to-Ra cross-reference, and explains the ASME BPE surface finish classification system in full. For the polishing process context, see our complete mechanical polishing guide.
1. Surface Roughness Parameters Defined
Surface roughness is measured by tracing a stylus (contact profilometer) or optical beam (non-contact profilometer) across the surface and recording the height profile z(x). From this profile, multiple statistical parameters are calculated over a defined evaluation length (typically 5 × the cut-off wavelength λc per ISO 4288):
| 参数 | Symbol | Definition | Typical Use |
|---|---|---|---|
| Arithmetic Mean Roughness | Ra | Average absolute deviation of the profile from the mean line over the evaluation length | Most widely used; stainless steel finishing, CMP process control |
| Root Mean Square Roughness | Rq (or RMS) | Square root of the mean of the squared deviations; more sensitive to peaks and valleys than Ra | Optical surface specifications; CMP wafer metrology |
| Maximum Peak-to-Valley Height | Rz | Average of the five largest peak-to-valley heights within the evaluation length | ASME BPE; DIN standards; corrosion and cleanability assessment |
| Maximum Profile Height | Rmax (Ry) | Single largest peak-to-valley height across the full evaluation length | Sealing surface specifications; bearing applications |
| Skewness | Rsk | Asymmetry of the profile height distribution; negative Rsk = more valleys (plateau surface) | Bearing surface characterization; tribological performance |
| Kurtosis | Rku | Sharpness of the profile height distribution; high Rku = spiky peaks | Wear prediction; sealing performance |
For a Gaussian (random) surface profile, Rq ≈ 1.25 × Ra. However, real polished surfaces are often non-Gaussian — a freshly belt-ground surface with regular scratch marks has a more periodic profile where Rq/Ra can be significantly different. Always specify which parameter is required; do not assume Ra and Rq specifications are interchangeable.
2. Semiconductor-Specific Surface Metrics
In semiconductor manufacturing, wafer-level surface characterization uses additional metrics not commonly found in industrial surface finishing standards:
- WIWNU (Within-Wafer Non-Uniformity): Standard deviation of post-CMP film thickness across the 300 mm wafer, expressed as a percentage (1σ%) of the mean thickness. Target: ≤ 2.5% for advanced node oxide and metal CMP.
- TTV (Total Thickness Variation): Difference between the maximum and minimum wafer thickness values across all measured sites. Specified for silicon wafer starting material; typically ≤ 0.5 µm for advanced logic prime wafers.
- Nanotopography: Spatial wavelength surface height variation in the 0.2–20 mm range, distinct from roughness (shorter wavelengths) and flatness (longer wavelengths). Nanotopography of the incoming silicon wafer influences STI CMP performance — high nanotopography can cause step-height variation that exceeds the CMP planarization length, leading to residual oxide non-uniformity.
- SFQR (Site Flatness Front Reference): Site-by-site flatness metric used for silicon wafer qualification; maximum deviation from a best-fit plane within a defined site window (e.g., 26 × 33 mm EUV exposure field). Typical SFQR specification: ≤ 40 nm for EUV-node wafers.
3. Grit-to-Ra Cross-Reference Table
The relationship between abrasive grit number and achievable Ra is material-dependent and method-dependent. The values below represent typical achievable Ra on 316L stainless steel using aluminum oxide abrasives under normal polishing conditions:
| Grit No. | Avg. Particle Size (µm) | Typical Ra on SS (µm) | Finish Class | 应用 |
|---|---|---|---|---|
| 36 | 530 | 3.2 – 6.3 | As-ground rough | Heavy stock removal, weld cleanup |
| 60 | 270 | 1.6 – 3.2 | Coarse grind | Remove deep pits and mill scale |
| 80 | 200 | 1.0 – 2.0 | #3 Grinding | General fabrication cleanup |
| 120 | 125 | 0.6 – 1.2 | #3–4 transition | Pre-#4 preparation |
| 150 | 100 | 0.4 – 0.8 | #4 Brushed | 3-A Sanitary, food contact surfaces |
| 180 | 80 | 0.3 – 0.6 | #4 fine | ASME BPE SF1 |
| 220 | 65 | 0.2 – 0.4 | #6 Satin | ASME BPE SF1–SF2 |
| 320 | 46 | 0.1 – 0.2 | #6–7 transition | Pre-buffing step |
| 400 | 35 | 0.07 – 0.15 | #7 Buffed | ASME BPE SF2–SF3; pre-electropolish |
| 600 | 26 | 0.05 – 0.10 | #7 fine | Recommended pre-EP condition |
| 800 | 21 | 0.03 – 0.06 | #7–8 transition | Pre-mirror buffing |
| 1200 | 15 | 0.01 – 0.03 | #8 Mirror (pre-compound) | Optical and decorative mirror |
| Compound | 0.5 – 5 | < 0.025 | #8 Mirror | Final mirror finish, mold cavities |
| CMP 泥浆 | 12 – 200 nm | < 0.001 (1 Å) | Semiconductor | Wafer planarization |
4. Industry Finish Designations
Multiple designation systems exist in parallel for stainless steel surface finishes. Engineers working across industries frequently need to translate between them:
| ASTM Designation | EN 10088 (European) | Common Name | Ra (µm) | 说明 |
|---|---|---|---|---|
| #1 | 1D | Hot rolled, annealed | 3.2 – 12.5 | Rough, dull; structural use only |
| #2B | 2B | Cold rolled, bright | 0.1 – 0.5 | Standard mill finish; most common |
| #3 | 1G | Grinding / rough polish | 0.5 – 1.6 | Unidirectional scratch pattern, 80–120 grit |
| #4 | 1J | Brushed finish | 0.2 – 0.8 | Standard sanitary finish; 150–180 grit |
| #6 | — | Satin (Tampico) | 0.1 – 0.2 | Dull sheen; less directional than #4 |
| #7 | — | Buffed | 0.025 – 0.1 | High reflectance; pre-EP condition |
| #8 | 2P | Mirror | < 0.025 | True mirror; no visible grain |
5. ASME BPE Surface Finish Classification
The ASME BPE (Bioprocessing Equipment) standard defines a hierarchical Surface Finish (SF) classification specifically for stainless steel equipment used in biopharmaceutical and high-purity process industries. This system is also widely adopted in semiconductor equipment specification. The full SF classification:
| ASME BPE SF | 过程 | Max Ra (µm) | Max Rz (µm) | Additional Requirements |
|---|---|---|---|---|
| SF1 | Mechanically polished | 0.84 | — | Minimum: #4 finish; visual inspection |
| SF2 | Mechanically polished | 0.51 | — | Profilometer Ra verification required |
| SF3 | Mechanically polished | 0.25 | — | Ra ≤ 0.25 µm; must be verified |
| SF4 | Mechanically polished + electropolished | 0.25 | — | EP after mechanical; Cr:Fe ratio verification |
| SF5 | Electropolished | 0.25 | — | EP only; enhanced Cr:Fe ratio; rouge resistance test |
| SF6 | Electropolished + tested | 0.25 | — | Full documentation; Cr:Fe ≥ 1.5; rouging qualification |
Both SF3 and SF4 specify a maximum Ra of 0.25 µm — but they are fundamentally different surfaces. SF3 is a mechanically polished surface: chromium-depleted, with a cold-worked layer and potentially embedded abrasive. SF4 is mechanically polished followed by electropolishing: the cold-worked layer is removed, the Cr:Fe ratio at the surface is verified to exceed 1.5, and the surface exhibits significantly superior corrosion resistance and cleanability. For semiconductor process chemical systems, SF4 is the minimum specification for wetted surfaces.
6. Industry-Specific Requirements by Sector
| Industry / Application | Surface Material | Specification | Primary Standard |
|---|---|---|---|
| Biopharmaceutical vessels | 316L SS | Ra ≤ 0.25 µm; EP preferred | ASME BPE SF3–SF4 |
| Semiconductor CMP slurry systems | 316L SS | Ra ≤ 0.25 µm; EP + passivation | ASME BPE SF4+ |
| Semiconductor process gas delivery | 316L SS tubing | Ra ≤ 0.25 µm ID; EP; no rouge | SEMI F20; ASME BPE SF4–SF5 |
| Food contact surfaces | 304 / 316 SS | Ra ≤ 0.8 µm; #4 minimum | 3-A Sanitary Standard 68; FDA 21 CFR |
| Semiconductor Si wafer | 硅 | Ra < 0.1 nm; SFQR ≤ 40 nm | SEMI M1; JEIDA |
| Optical components (lithography) | Fused silica / ULE | Rq < 0.05 nm; flatness < λ/20 | ISO 10110; customer spec |
| Injection mold (optical grade) | Hardened tool steel | Ra < 0.01 µm | SPI A1; customer spec |
7. Measurement Methods and Instrumentation
Surface finish measurement method selection must match the surface material and roughness range. Using the wrong instrument or settings is a common source of measurement errors in production environments:
- Contact profilometer (stylus): Gold standard for Ra measurement on metals. Stylus tip radius (typically 2 µm) limits the ability to resolve very fine features. Cutoff wavelength λc must be selected appropriately: ISO 4288 specifies λc based on the expected Ra range (e.g., λc = 0.8 mm for Ra 0.1–2 µm; λc = 0.25 mm for Ra 0.02–0.1 µm). Incorrect λc selection is a systematic measurement error.
- Optical profilometer (white-light interferometry / confocal): Non-contact; preferred for soft surfaces (Cu, Al), delicate components, and very smooth surfaces (Ra < 10 nm) where stylus contact would damage the surface. Provides full 3D topographic maps (Sa, Sq, Sz — area-based equivalents of Ra, Rq, Rz).
- Atomic Force Microscope (AFM): Used for semiconductor wafer roughness measurement at the Å level. Provides Ra, Rq, and power spectral density (PSD) data at spatial wavelengths from 1 nm to 100 µm. Standard metrology tool for CMP wafer characterization in production and R&D.
Ra values measured under different conditions (different λc, different evaluation length, different stylus radius) are not directly comparable. When specifying or reporting Ra, always document: instrument type, stylus tip radius, cutoff wavelength λc, evaluation length, number of measurement traces, and trace locations on the part. This is required by ASME BPE Section SF for documentation compliance.
8. Frequently Asked Questions
ASME BPE specifies Ra ≤ 0.84 µm for SF1 (the minimum mechanical finish) through Ra ≤ 0.25 µm for SF3–SF6. In practice, biopharmaceutical process equipment wetted surfaces are typically specified at ASME BPE SF3 (Ra ≤ 0.25 µm, mechanically polished) or SF4 (Ra ≤ 0.25 µm with electropolishing), depending on the product contact and cleaning validation requirements. Regulatory guidance from the FDA and EMA does not prescribe a specific Ra value but requires that surfaces can be validated as clean — electropolished surfaces at SF4 or above are far easier to validate due to their superior chemical inertness and reduced particle shedding.
Ra is the most commonly used parameter for cleanability specifications, but it is not always the most informative. Ra measures average roughness amplitude but does not capture surface valley depth (Rz) or the lateral frequency of surface features — both of which influence bacterial adhesion and cleaning efficacy. For critical biopharmaceutical and semiconductor applications, specifying both Ra and Rz (or using 3D area-based parameters Sa and Sz from optical profilometry) provides a more complete surface characterization. ASME BPE SF4–SF6 also requires Cr:Fe ratio verification by XPS or AES, recognizing that surface chemistry — not just roughness — governs corrosion resistance and cleanability.
Related Technical Articles
Precision CMP Consumables Backed by Technical Expertise
JEEZ supplies CMP slurries, polishing pads, and backing films with full lot documentation and applications engineering support. Contact us to discuss surface finish requirements for your process.
