High-SNR Dark-Field Photothermal Confocal Microscopy for Multimodal Characterization of Subsurface Defects

Nanoscale surface and subsurface defects in optical components act as deterministic failure sites in high-energy laser facilities, constraining operational fluence and system longevity. Nondestructive diagnostics simultaneously achieving high sensitivity for nanoscale defect detection and specificity for damage-relevant defect identification remain imperative for advancing laser resilience. We introduce dual-modal dark-field photothermal confocal microscopy enabling concurrent 3D tomography of defect geometry and absorption properties. A spatiotemporal double-filtering strategy isolates photothermal signals, enhancing signal-to-noise ratios by 7.4× over conventional photothermal imaging and 5.5× versus baseline dark-field microscopy. This sensitivity breakthrough reveals subsurface and phase defects that are inaccessible to established optical methods. Critically, the pixel-level fusion of scattering (structural) and photothermal (absorption-specific) modalities enables defect hazard grading, delivering essential specificity. By resolving the sensitivity-specificity trade-off, our approach establishes a predictive framework for laser damage resistance and preemptive mitigation in next-generation high-power optics.