Chinese researchers unveil phase imaging method with unlimited field of view

Scientists in China developed L2-CPI, a computational phase imaging technique that pairs diffraction-limited resolution with an arbitrarily large field of view. The method could improve defect detection in semiconductor wafers and expand high-throughput imaging in biology and digital pathology. Why it matters: - L2-CPI addresses a long-standing imaging trade-off between resolution and field of view. - The method could improve nanometer-scale defect detection in semiconductor manufacturing. - The same approach could support large-area biological imaging for digital pathology and automated screening. What happened: - A team led by Professor Jinlong Zhu at Huazhong University of Science and Technology in China developed Lateral Line-scan Computational Phase Imaging, or L2-CPI. - The work was published in Light: Advanced Manufacturing . - The system delivers diffraction-limited phase imaging across a field of view limited only by the scan stage travel range. - L2-CPI uses a Linnik-type interferometric setup and captures phase data while the sample moves laterally. The details: - The system removes the need for step-and-repeat stitching, which can introduce mechanical misalignment and phase discontinuities. - A Dynamic Compensation System suppresses mechanical vibrations in real time. - A three-parameter cosine-fitting algorithm reconstructs phase information with high precision. - The researchers say the lateral scanning architecture serves three functions at once: diffraction-limited imaging over an arbitrary field of view, elimination of proximal errors from step-scan systems, and high-fidelity 3D reconstruction under industrial vibration. - The team says the approach is more robust to noise and environmental disturbance because it retrieves phase from a large number of data points during the scan. - In tests on patterned defect-array wafers and microlens arrays, L2-CPI identified sub-wavelength bridge and cutting defects on a wafer with a 60 nm critical dimension. - Conventional intensity-based imaging systems typically cannot see those defects. - Simulations suggest the method could still detect defects when the critical dimension shrinks to 15 nm. - The DOI is 10.37188/lam.2026.020 . - The work was funded by China’s National Natural Science Foundation, National Key Research and Development Program, Shenzhen research programs, and the Innovation Project of Optics Valley Laboratory. Between the lines: - The technical advance is not just higher resolution; it is scalable inspection without the stitching artifacts that can undermine large-area microscopy. - That matters in semiconductor manufacturing, where inspection speed and data integrity both affect yield. - The biological imaging angle suggests the same hardware could move beyond cleanroom inspection into tissue and cell analysis. What’s next: - The researchers expect the technique to be useful for high-throughput, non-destructive nanometrology. - Future applications may include faster wafer inspection, digital pathology, and automated biological screening. - The team also expects the approach to remain useful as semiconductor features continue to shrink.