SGH Imaging Defect Detection: Principles, Methods, and Industrial Applications

SGH imaging defect detection is an advanced, non-destructive optical inspection technique that combines structured grating projection with holographic phase reconstruction to identify surface and subsurface defects at micro- and nano-scale resolution. Widely adopted in semiconductor, display, optical, and precision manufacturing industries, SGH imaging delivers full-field, high-contrast defect visualization that conventional microscopy often cannot achieve. This article provides a comprehensive technical overview of SGH imaging defect detection, including its working principles, applicable sample types, testing items, standard methods, industrial applications, and frequently asked questions.

Technical Overview

SGH (Structured Grating Holography) imaging defect detection is a full-field optical metrology technique that projects a periodic grating pattern onto a sample surface and captures the deformed fringes reflected from the surface. Because the projected grating is modulated by the topography, texture, and reflectivity variations of the sample, any defect — whether a scratch, particle, pit, void, or subsurface anomaly — introduces measurable distortion in the fringe pattern. By reconstructing the phase information contained in the deformed fringes, SGH systems generate quantitative surface maps and high-contrast defect images in a single acquisition.

Unlike traditional bright-field or dark-field microscopy, which relies on intensity contrast and can miss low-contrast defects, SGH imaging converts subtle geometric and optical discontinuities into strong phase signals. This makes the technique particularly powerful for detecting shallow scratches, faint mura, micro-cracks, and buried interface anomalies on specular or semi-specular surfaces such as polished wafers, glass substrates, and optical coatings.

Key performance characteristics of SGH imaging defect detection include:

• Non-contact and non-destructive measurement with no risk of sample contamination or damage
• Sub-micrometer lateral resolution and nanometer-level height sensitivity
• Full-field acquisition covering large areas within seconds, enabling high-throughput inspection
• Simultaneous detection of surface topography defects and reflectivity-related anomalies
• Quantitative 3D reconstruction suitable for defect dimensioning and severity grading
• Strong suppression of background noise through phase-based filtering algorithms

Modern SGH systems typically integrate automated stage control, multi-wavelength or multi-angle illumination, and AI-assisted defect classification software, allowing unattended inspection of full wafers, panels, or component batches with detection sensitivities down to the sub-micron scale.

Testing Samples

Semiconductor Wafers and Dies

Polished silicon wafers, SOI substrates, patterned wafers, and singulated dies are among the most common samples. SGH imaging detects CMP scratches, slurry residues, handling marks, edge chipping, and shallow dicing damage that can compromise yield and reliability.

Display Panels and Cover Glass

LCD and OLED panel substrates, cover glass, polarizers, and touch-sensor films are inspected for mura, streaks, bubbles, inclusions, coating non-uniformity, and micro-scratches. The phase-sensitive nature of SGH imaging reveals low-contrast luminance and geometry defects invisible to standard visual inspection.

Optical Components

Lenses, mirrors, prisms, filters, diffusers, and coated optical flats benefit from SGH detection of digs, sleeks, coating pinholes, bubbles, and figure errors. The technique is especially effective on high-reflectivity and anti-reflection-coated surfaces.

Precision Machined Metal Parts

Machined, ground, and polished metal components such as bearings, seals, molds, and medical device surfaces are inspected for tool marks, corrosion pitting, debris embedding, and geometric deviations.

Photovoltaic and Ceramic Substrates

Solar cells, thin-film photovoltaic stacks, ceramic substrates, and sapphire wafers are analyzed for micro-cracks, material inclusions, printing defects, and texturing irregularities that affect performance and mechanical strength.

PCB and Advanced Packaging Substrates

Rigid and flexible circuit boards, IC substrates, and redistributed interposers are examined for pad defects, surface contamination, warpage, and laser-drilling anomalies.

Testing Items

SGH imaging defect detection covers a broad portfolio of measurable defect categories and geometric parameters:

• Surface defects: scratches, digs, pits, bumps, particles, fibers, fingerprints, stains, and residues
• Subsurface and interface defects: voids, delamination, inclusions, and buried micro-cracks
• Coating and film defects: pinholes, uncoated regions, thickness non-uniformity, peeling, and blistering
• Pattern defects: line breaks, bridges, residues, CD irregularities, and pattern displacement
• Geometry and topography: warpage, bow, step height, surface roughness, and local height deviation
• Edge and perimeter defects: chipping, edge cracks, and bevel damage
• Defect quantification: lateral size, height or depth, area, position coordinates, density, and severity classification

Each detected defect is typically documented with coordinates, dimensional measurements, grayscale or phase-contrast images, and a classified defect code, supporting statistical process control and root-cause investigation.

Testing Methods

The SGH imaging defect detection workflow follows a standardized procedure designed to ensure repeatability, traceability, and measurement integrity:

• Sample intake and preparation: samples are cleaned with dry nitrogen or lint-free wipes when appropriate, mounted on vacuum or edge-clamp stages, and registered in the laboratory information system.
• Region definition and scan planning: the area of interest, scan resolution, and grating parameters (pitch, direction, wavelength) are configured according to the material, reflectivity, and expected defect size.
• Grating projection and acquisition: structured fringe patterns, often with phase-shifting sequences, are projected onto the sample while a high-dynamic-range camera records the deformed fringes from one or more viewing angles.
• Phase reconstruction: wrapped phase maps are computed from the fringe images and unwrapped using robust algorithms to reconstruct surface topography and reflectivity-modulated phase images.
• Defect extraction and classification: differential comparison against reference maps or golden-sample databases, combined with machine-learning classifiers, isolates and categorizes defects while suppressing false alarms from design features.
• Verification and cross-analysis: flagged defects can be verified with complementary techniques such as SEM, AFM, white-light interferometry, or optical profilometry for confirmatory dimensioning when required.
• Reporting: results are delivered with defect maps, dimensional statistics, severity grading, and pass/fail judgment against client-specified acceptance criteria or relevant industry standards.

Typical measurement capabilities include lateral detection limits below one micrometer, vertical sensitivity at the nanometer level, and inspection throughput sufficient for full 300 mm wafer or Gen-6 panel mapping within practical cycle times.

Applications

SGH imaging defect detection serves critical quality and reliability functions across multiple high-technology industries:

• Semiconductor manufacturing: incoming wafer qualification, CMP process monitoring, post-dicing inspection, and yield improvement through systematic defect signature analysis.
• Display production: mura detection, cover glass inspection, film quality control, and new product introduction validation for mobile, automotive, and TV panels.
• Optics and photonics: incoming inspection of optical components, coating quality assurance, and R&D evaluation of prototype surfaces.
• Consumer electronics: glass and metal enclosure inspection, camera module component screening, and cosmetic quality grading of finished parts.
• Automotive: safety-critical component inspection for sensors, displays, mirror systems, and polished powertrain parts, supporting IATF 16949 quality workflows.
• Photovoltaics: micro-crack screening, textured surface evaluation, and metallization print quality control in cell and module production.
• Failure analysis and R&D: defect localization for subsequent SEM or FIB analysis, process development feedback, and material research involving surface evolution studies.

FAQ

What is the smallest defect SGH imaging can detect?

Depending on the optical configuration and surface condition, lateral detection limits typically reach the sub-micrometer level, while vertical sensitivity can approach nanometer-scale height differences. Practical detectability also depends on defect contrast relative to the surrounding surface.

Is SGH imaging destructive to the sample?

No. SGH imaging is a completely non-contact optical method. The sample is illuminated with low-power structured light, so there is no physical probing, coating, or sectioning, and the sample remains fully functional after inspection.

What sample sizes and materials can be tested?

Systems accommodate samples ranging from small diced components to 300 mm wafers and large display panels. Reflective, semi-reflective, and transparent materials including silicon, glass, sapphire, polymers, metals, and ceramics can all be inspected with appropriately configured illumination.

How does SGH imaging compare with conventional AOI or microscopy?

Conventional AOI relies mainly on intensity contrast, which can miss shallow or low-contrast defects. SGH imaging extracts phase information that is inherently sensitive to geometric discontinuities, enabling detection of defects such as faint scratches, shallow pits, and subsurface anomalies with significantly higher contrast and quantitative depth information.

How long does a typical inspection take?

A single full-field measurement usually completes within seconds to minutes depending on the area, required resolution, and phase-shifting sequence. Full-wafer or full-panel automated mapping is typically completed within one to several hours.

Can SGH results be combined with other analysis techniques?

Yes. SGH imaging is frequently used as a defect localization step prior to SEM, EDX, AFM, FIB cross-sectioning, or white-light interferometry, providing a powerful integrated workflow for defect characterization and root-cause analysis.

What reporting is provided after testing?

Standard deliverables include defect distribution maps, individual defect images with coordinates and dimensions, statistical summaries such as defect density and size distribution, severity classification, and conclusions against the specified acceptance criteria.