电子游戏

Southeast University Team Led by Xiaobao Xu and Wei Lei Reports New Progress in On-Chip Spectral Imaging

发布者:刘鹃发布时间:2026-08-25浏览次数:10


Recently, a research team led by Prof. Xiaobao Xu and Prof. Wei Lei at Southeast University developed a vertically bandgap-cascaded perovskite single-crystal film heterojunction. By integrating wavelength-dependent absorption depth with a stepwise band alignment within a monolithic device, the architecture enables bias-polarity-switchable, depth-selective carrier collection while combining fast photoresponse, broadband detection, and high-accuracy spectral reconstruction. The work, entitled“A Vertically Bandgap-Cascaded Perovskite Single-Crystal Film Heterojunction for On-Chip Spectral Imaging”, has been published inAdvanced Materials.

Figure 1. Spectral detection and hyperspectral imaging enabled by the vertically bandgap-cascaded perovskite single-crystal heterojunction: (a) device architecture, depth distribution of photocarriers, and spectral reconstruction principle; (b) cross-sectional morphology and elemental distribution of the heterojunction; (c, d) transient photoresponse and spectral reconstruction results; (e-h) hyperspectral imaging demonstration using a TFT-integrated device.

Spectral imaging simultaneously captures spatial and spectral information from a target, making it an important enabling technology for precision inspection, environmental sensing, biomedical analysis, and intelligent vision. For portable and on-chip applications, spectral systems must not only be miniaturized but also obtain sufficiently rich and mutually independent spectral information from a limited number of pixels. Conventional approaches typically rely on filter arrays, dispersive structures, or tunable optical components for spectral encoding, often at the cost of lower photon utilization, increased lateral pixel footprint, or limited modulation speed. Therefore, improving the spectral encoding capability and response speed of an individual detector unit without introducing complex optical paths remains a key challenge in on-chip hyperspectral imaging.

The team employed sequential solution epitaxy to grow three perovskite single-crystal films - MAPbCl3, MAPbBr3, and MAPbI3 - along the device thickness, forming a vertically cascaded heterojunction with progressively decreasing bandgaps. Because incident light of different wavelengths penetrates to different depths in the three layers, the combination of a stepwise type-I band alignment and a nonuniform built-in electric field gives rise to distinct transport and collection behaviors for photogenerated carriers generated at different depths. As the magnitude and polarity of the applied bias are varied, the dominant carrier-transport pathways within the device switch accordingly, producing multiple complementary and weakly correlated spectral responses.

By exploiting wavelength-dependent absorption depth in the vertically cascaded heterojunction and electrically regulating the transport and collection of photocarriers generated at different depths, the design reduces redundancy in spectral encoding and improves the stability and accuracy of spectral reconstruction. The device achieved a maximum external quantum efficiency of approximately 82%, a minimum rise time of 8 μs, and a spectral operating range of approximately 400-820 nm. The reconstructed spectra exhibited a peak-wavelength error of about 1.4 nm, while two near-infrared peaks could be resolved with a separation of approximately 2.7 nm.

The researchers further integrated the heterojunction detector onto a thin-film-transistor (TFT) substrate, realizing 64 × 64-pixel hyperspectral imaging and reconstructing spatial intensity distributions at different wavelengths. By combining vertical materials engineering with band-structure engineering, the work endows a single pixel with programmable spectral encoding capability and offers a new route toward compact devices that simultaneously support fast response and high-accuracy spectral reconstruction. The approach may find applications in portable sensing, machine vision, biomedical analysis, and other intelligent optoelectronic systems.

Prof. Xiaobao Xu and Prof. Wei Lei from the School of Electronic Science and Engineering, Southeast University, are the co-corresponding authors of the paper, and Shilin Liu, a doctoral student in the same school, is the first author. The work was supported by the National Natural Science Foundation of China, the National Key Research and Development Program of China, the Jiangsu International Science and Technology Cooperation Program, the Fundamental Research Funds for the Central Universities, and the Aeronautical Science Foundation, among other funding sources.

Paper://doi.org/10.1002/adma.74281



网站地图