Substrate Glass Scratch Depth Testing: Methods, Standards, and Applications

Substrate glass scratch depth testing is a critical quality control process used to quantify linear surface defects on display, semiconductor, and optical glass substrates. By combining optical defect localization with high-resolution profilometry, manufacturers can measure scratch depth from the sub-micron to tens-of-microns range and verify compliance with product specifications. This article explains the sample types, measured parameters, testing methods, applicable standards, and typical applications of scratch depth evaluation for substrate glass.

Technical Overview

Substrate glass serves as the foundational material for thin-film-transistor liquid-crystal displays (TFT-LCD), organic light-emitting diode (OLED) panels, touch sensors, cover windows, and glass carriers used in advanced semiconductor packaging. Because subsequent thin-film deposition, photolithography, and cell assembly processes are extremely sensitive to surface condition, even shallow scratches can degrade coating uniformity, concentrate mechanical stress, reduce bending strength, and ultimately lower panel yield. Scratch depth testing therefore focuses on accurately quantifying the vertical dimension of linear surface defects—typically ranging from several tens of nanometers to tens of micrometers—so that materials can be graded, dispositioned, and released against defined acceptance criteria.

Testing generally follows a two-step strategy: scratches are first located and classified by bright-field, dark-field, or machine-vision imaging, and their depth is then measured quantitatively using stylus profilometry, white-light interferometry, confocal microscopy, or atomic force microscopy. Measurement and evaluation are aligned with international surface-texture standards such as ISO 4287, ISO 4288, ISO 25178, and ISO 10110-7, together with the corresponding national standards GB/T 1031, GB/T 3505, and JIS B 0651.

Testing Samples

TFT-LCD and OLED Display Substrate Glass

Alkali-free borosilicate or aluminosilicate glass sheets with thicknesses of 0.3 mm to 0.7 mm and generation sizes from Gen 4.5 to Gen 10.5. Samples may be bare polished substrates or substrates carrying array and color-filter films, evaluated in both as-received and post-cleaning conditions.

Chemically Strengthened Cover Glass

Aluminosilicate cover glass for smartphones, tablets, notebooks, wearables, and automotive displays, with thicknesses from 0.3 mm to 2.0 mm and compressive-stress surface layers produced by ion exchange. Scratch depth testing supports both manufacturing defect control and abrasion-resistance evaluation.

Glass Wafers and Carrier Substrates

Glass wafers of 100 mm to 300 mm diameter used as carriers for wafer-level packaging, fan-out packaging, and through-glass via (TGV) processes, where nanometer-level surface quality is required for bonding, thinning, and lithographic alignment.

Optical and Specialty Glass Sheets

Optical windows, filter glass, and specialty float or rolled glass products, for which scratch tolerances are commonly defined by optical drawing callouts such as those described in ISO 10110-7.

Testing Items

Depending on the product specification and the position in the value chain, scratch depth testing may include the following measurable parameters:

• Maximum scratch depth, referenced to the leveled undamaged reference surface
• Mean and residual scratch depth, and depth distribution along the scratch length
• Scratch width at the surface and at half-depth, scratch length, and orientation
• Cross-sectional profile geometry, including edge pile-up, chipping, and lateral cracking
• Scratch density (defect count per unit area) and size-class distribution for surface-quality grading
• Correlation of measured depth with optical visibility or dark-field contrast for cosmetic grading
• Statistical comparison of measured values with drawing limits for lot acceptance decisions

Testing Methods

Optical Microscopy and Machine-Vision Localization

Samples are inspected under bright-field, dark-field, or differential-interference-contrast illumination at magnifications from 50x to 1000x to locate scratches, distinguish them from digs, bubbles, and handling marks, and record coordinates for subsequent profiling. Automated machine-vision systems map the full surface and generate defect maps that guide depth measurement at representative or worst-case positions.

Stylus Profilometry

A diamond-tipped stylus (tip radius typically 2 um, measuring force of 1 mN or less) is traced perpendicular to the scratch axis across the defect. The profile is leveled using a least-squares reference line and, where required, filtered in accordance with ISO 16610. Depth is reported as the maximum profile valley depth relative to the undamaged reference flanks, following the profile-method vocabulary of ISO 4287 and the sampling rules of ISO 4288. Vertical resolution is typically a few nanometers, making the method suitable for scratches deeper than approximately 0.05 um.

White-Light Interferometry (Optical Profiling)

Non-contact vertical-scanning interferometry provides areal three-dimensional topography with sub-nanometer vertical resolution over fields ranging from tens of micrometers to several millimeters. Areal parameters such as maximum pit depth are extracted per ISO 25178, and stitched measurements can capture the entire scratch length. This method is preferred for coated, fragile, or chemically strengthened surfaces where stylus contact could cause additional damage.

Confocal Laser Scanning Microscopy

Confocal profiling combines high-contrast imaging with quantitative height measurement. It is particularly useful on steep scratch walls, rough ground surfaces, or samples where edge chipping must be resolved together with depth in a single dataset.

Atomic Force Microscopy

AFM in contact or tapping mode resolves ultra-shallow scratches and polished-in damage with lateral resolution below 10 nm and sub-nanometer vertical resolution, supporting advanced failure analysis, coating development, and polishing-process research.

Test Procedure and Data Evaluation

• Sample conditioning: cleaning with filtered isopropyl alcohol or deionized water, nitrogen drying, and equilibration at 23 plus or minus 2 degrees C and 50 plus or minus 5 percent relative humidity.
• Localization and marking of representative scratch positions, typically the visually deepest point plus at least two additional positions along the scratch.
• Profile acquisition perpendicular to the scratch axis, with at least three replicate scans per position.
• Data leveling, filtering, and calculation of depth, width, and profile parameters per ISO 4287 and ISO 25178.
• Reporting of maximum, mean, and standard deviation of depth, together with instrument settings, measurement uncertainty, and pass/fail judgment against specification.

Applications

Scratch depth data are used throughout the glass and electronics value chain:

• Incoming and outgoing inspection of display substrate glass at panel fabrication lines
• Process optimization of cutting, grinding, polishing, edge finishing, and cleaning operations
• Lot release, supplier qualification, and periodic surveillance testing by accredited third-party laboratories
• Failure analysis of cracked cells, coating delamination, and transport-induced handling damage
• Research and development of strengthened cover glass, anti-reflective and anti-fingerprint coatings, and abrasion-resistant surfaces
• Quality documentation for automotive, medical, and industrial display certification programs

FAQ

What depth resolution is achievable for substrate glass scratch measurement?

Stylus profilometers typically resolve about 1 to 10 nm, white-light interferometers reach sub-nanometer vertical resolution, and atomic force microscopy provides sub-nanometer resolution in both vertical and lateral directions. Practical detection limits also depend on surface roughness: on polished substrates with Ra below 1 nm, scratches as shallow as 20 to 50 nm can be reliably quantified.

Should a contact or non-contact method be chosen?

For bare, hard glass surfaces, stylus profilometry offers robust traceability and low cost. Non-contact optical methods are recommended for coated, ion-exchanged, or extremely thin and fragile samples, and whenever plastic deformation by the stylus must be excluded. Many accredited third-party laboratories apply both techniques and cross-validate the results.

How are results reported and judged?

A typical report lists scratch location, orientation, maximum and mean depth, width, cross-sectional profiles, instrument parameters, measurement uncertainty, and a comparison with the drawing or specification limit. Acceptance thresholds are product-specific: active areas of display substrates commonly require scratch depths below 0.1 to 0.5 um, while cosmetic or non-active zones may permit deeper defects.

What sample preparation is required?

Samples should be cleaned to remove dust, fingerprints, and particles without introducing new scratches, then conditioned in a standard laboratory atmosphere of 23 plus or minus 2 degrees C and 50 plus or minus 5 percent relative humidity. Full sheets, coupons cut from sheets, and finished parts can all be tested with appropriate fixturing.

Which standards govern scratch depth testing?

Depth measurement relies on surface-texture standards including ISO 4287, ISO 4288, ISO 25178, and ISO 16610, while optical-component imperfection classes are defined in ISO 10110-7 and general flat-glass quality requirements in ASTM C1036. Corresponding national standards include GB/T 1031, GB/T 3505, and JIS B 0651. Product-level depth limits are normally specified in customer drawings or industry material specifications rather than in the measurement standards themselves.