Presentation Information
[9a-A23-8]Lensless Imaging for Characterization of Through Glass Vias
〇Christoph Lederbogen1,2, Stephan Reichelt1, Giancarlo Pedrini1, Yoshio Hayasaki2 (1.ITO, Univ. Stuttgart, 2.CORE, Utsunomiya Univ.)
Keywords:
Lensless Imaging,Through Glass Vias,Computational Imaging
Through Glass Vias (TGVs) are vertical electrical interconnects etched through glass substrates with diameters ranging from 10 to 100 µm. As a glass-based alternative to Through Silicon Vias (TSVs), they offer advantageous properties such as electrical isolation and thermal stability, making them a key enabling technology for advanced semiconductor packaging and optoelectronic integration. The production of TGVs requires precise and cost-effective characterization.
This work presents a lensless imaging approach for the characterization of TGVs. The vias are illuminated with a laser, and the diffraction patterns behind the sample are recorded by a camera at various distances. A modified Gerchberg-Saxton algorithm retrieves the phase information from these intensity measurements, enabling back-propagation of the reconstructed field to the TGV plane. The resulting amplitude and phase information at the TGV plane serve as the basis for quantitative geometric characterization, such as the assessment of via shape and roundness.
To validate and investigate the method's applicability to TGVs, a simulation framework based on the Wave Propagation Method (WPM) models various via cross-sectional shapes, enabling systematic study of the full imaging and characterization pipeline.
The simulated results demonstrate the feasibility of the proposed pipeline, showing that the reconstructed phase at the TGV plane can serve as a basis for geometric characterization. Experimental validation with real TGV samples is ongoing, and further characterization metrics beyond roundness are under investigation.
This work presents a lensless imaging approach for the characterization of TGVs. The vias are illuminated with a laser, and the diffraction patterns behind the sample are recorded by a camera at various distances. A modified Gerchberg-Saxton algorithm retrieves the phase information from these intensity measurements, enabling back-propagation of the reconstructed field to the TGV plane. The resulting amplitude and phase information at the TGV plane serve as the basis for quantitative geometric characterization, such as the assessment of via shape and roundness.
To validate and investigate the method's applicability to TGVs, a simulation framework based on the Wave Propagation Method (WPM) models various via cross-sectional shapes, enabling systematic study of the full imaging and characterization pipeline.
The simulated results demonstrate the feasibility of the proposed pipeline, showing that the reconstructed phase at the TGV plane can serve as a basis for geometric characterization. Experimental validation with real TGV samples is ongoing, and further characterization metrics beyond roundness are under investigation.