Project Details
Projekt
GeSn Based Infrared Detection and Imaging
Subject Area
Electronic Semiconductors, Components and Circuits, Integrated Systems, Sensor Technology, Theoretical Electrical Engineering
Term
since 2025
Project identifier
Deutsche Forschungsgemeinschaft (DFG) - Project number 568675133
The project's ambition is to provide technological building blocks for AI infrared detection and imaging based on GeSn alloy, a semiconductor material of Group IV elements that is compatible with the low-cost processes of the silicon industry. The spectral detection range of GeSn is expanded to longer wavelengths to enable demonstrator devices operating in the mid infrared (MidIR) range of 2-5 µm, unattainable to date CMOS-compatible material platforms. GERALD develops and demonstrates different classes of devices, with particular focus on quantum well (QW) heterostructure diodes and phototransistors, proposing different design solutions that will be optimized using AI-driven evolutionary algorithm procedures. After a stand-alone device demonstrator, the realization of photodetector arrays foreseeing external read-out integrated circuits (ROIC) follows. The phototransistor approach is the step into a new field of in-sensing computing, where light detection is coupled with electrical charge storage (memory) and read by a transistor. This is evaluated for their possible use in neuromorphic networks for on-chip, edge computing. The proposed GeSn-based technology, that is intended to be seamlessly integrated in standard microelectronic manufacturing, represents a major breakthrough and a key enabler for expanding the realm of applications of silicon photonics. The possibility of low-cost detection and imaging, given by the proposed technology, would add functionalities to portable or diffused objects, manufacturable in large volumes, potentially addressing mass-scale markets of e.g. lab-on-a-chip, biosensors, imager for assisted driving (e.g. “see-through fog”), gas and liquid detectors, pollution monitors, and security scanners. Importantly, these devices could be realized together with edge computing components, such as CMOS-realized neuromorphic networks, that could allow machine learning or/and AI-driven on- chip spectral analysis and image recognition. The consortium is made up of world-renowned European experts in the field of growth of innovative (Si)GeSn semiconductors on silicon, of design, processing, and testing of photonic devices. The partners of the consortium have already been involved together into collaborative projects through common publications or European and national funded projects.
DFG Programme
Research Grants
International Connection
France
Cooperation Partners
Dr. Patrick Bouchon; Professor Dr. Moustafa El Kurdi; Dr. Yanko Todorov
