Diamond characterization

See the defect field, not just the defect.

Visible-light metrology for diamond substrates - mapping dislocation loci, surface variation, bending, roughness, stress signatures, and 3D topography from whole-wafer scale to a 20 micron field of view.

Non-destructive optical capture Macro-to-nano analysis AI-assisted defect classification
Research foundation

Built on tens of thousands of diamond defects.

From 2023 through 2026, NanoFraction developed this use case with GLCT using diamond wafer data across macro, micro, and nanoscale views. The resulting analytical pipeline was trained to identify patterns in defect geometry and gradient signatures.

The objective is practical: screen more of the wafer rapidly, localize regions that require attention, and use higher-cost methods where they add the most value.

GLCT research dataset

Measured outputs from one optical workflow.

NanoFraction combines real-time image capture with gradient analysis and AI-assisted interpretation to locate dislocation signatures and characterize the surrounding field.

99%+Analytical-model detection of dislocation loci in the evaluated research dataset.
80%+AI-model defect detection in the evaluated dataset, with additional training expected to improve classification.
5Defect categories identified from their gradient signatures.

Reported figures are research results from the GLCT collaboration dataset and should be validated against each customer's material, process, and reference method.

Gradient defect map showing diamond dislocation signatures in a 20 micron field of view
Gradient defect map / 20 micron field of view
Dislocations pinpointed

Map the locus and its surrounding stress signature.

Each feature exists within a wider field of tension, compression, topography, and surface variation. The NanoFraction pipeline is designed to preserve that context rather than reducing inspection to a binary defect marker.

  • Defect location
  • Type and geometry
  • Local and global fields
  • Whole-wafer bending
  • Surface roughness
  • 3D topography
Inspection workflow

Use high-cost tools where they matter most.

NanoFraction is designed as a screening and decision layer that can reduce unnecessary measurement cycles while directing confirmatory analysis toward the most informative locations.

01

Capture

Acquire whole-wafer and targeted optical views without destructive preparation.

02

Map

Generate defect, gradient, stress-signature, roughness, and topography outputs.

03

Prioritize

Identify the locations and signatures that warrant deeper investigation.

04

Confirm

Correlate selected regions with etch-pit, birefringence, cathodoluminescence, SEM, or other reference methods.

Model outputs

From image to actionable defect map.

The analytical pipeline detects candidate loci, measures relationships among features, and groups defects according to gradient signature. Outputs can be adapted to a customer's wafer geometry, process history, and existing inspection criteria.

Three-part NanoFraction model output showing a diamond defect image, detected feature relationships, and classified regions
Example analytical output / 20 micron field of view
12-week feasibility pilot

A/B test NanoFraction against your current diamond inspection process.

Define one high-value defect-detection question, compare outputs with your present workflow, and measure whether NanoFraction improves coverage, cycle time, or the selection of locations for confirmatory analysis.