Carbon nanotube film-type broadband imager-based multi-wavelength computer tomography system for 3D composite structure reconstruction

Non-destructive testing techniques play important roles; optical imaging measurements in infrared (IR)–terahertz (THz) bands have great potential for defective detection of industrial products. In the above bands, photo-irradiation shows characteristic transmittance values to various non-metallic materials per composition and wavelength. This advantage leads to material identification through multi-wavelength transmissive imaging. Employing these optical properties, previous work demonstrated structure reconstruction of composite 3D objects with computer vision (CV)-driven measurements. Specifically, carbon nanotube (CNT)-based imagers installed in previous systems facilitate broadband operation, which leads to detailed material identification. CNT film exhibits highly efficient photo-absorptance in ultrabroadband containing IR–THz, which enables broadband detection via photo-thermoelectric (PTE) effect. However, longer spatial scanning measurements due to insufficient CNT pixel integration structure prevent further inspection applications. To this end, this work performed a high-speed operation CV system via pixel integration of CNT film PTE sensors.

As a preparation step, this work employed dispenser printing equipment for the fabrication of a CNT film PTE camera. Precision mechanical alignment of this facilitates high-yield fabrications compared to manual processes. This work realized a 2D integrated device as follows: 20 mm-15 mm-sq. for pixel area, 1 mm pitch for the x-axis, 5 mm pitch for the y-axis, 60 pixels total. Then, this work performed IR–THz tomography, one of the typical CV measurements. This system consists of multi-wavelength photo-sources (λ: 976 nm, 1,310 nm, 10.3 μm), the CNT film PTE camera, and a digital stepping motor stage for rotation scanning. Under this system, this work demonstrated material identification and structure restoration of composite 3D objects based on multi-wavelength images. Specifically, 2D integration of CNT film PTE camera facilitates reduced spital scanning, which shortens its measurement time at a magnitude of 1/2.14 compared to the existing 1D array-pixel sensor.

Mr. Kubota is a 2^nd year master’s course student at Department of Electrical, Electronic, and Communication Engineering, Faculty of Science and Engineering, Chuo University. His research focuses on the development of a non-destructive inspection system based on broadband infrared–terahertz imaging with carbon nanotube-based photo-detector.