Rad-Hard High-Dynamic-Range TEM Direct Detector

Although direct detection technology delivers substantially better image quality for transmission electron microscopy (TEM), its limited dynamic range and radiation sensitivity generally preclude its use for a variety of TEM applications. At present, direct detection is well-suited for biological cryo-EM, where illumination is relatively uniform and the electron exposure rate is limited. However, many structural biology experiments make use of lower magnifications to acquire large areas of multicellular tissues and/or specimen staining to reduce radiation sensitivity, resulting in high illumination on the camera. Other experiments, such as electron energy loss spectroscopy (EELS), also produce high illumination levels on some regions of the camera. In all these cases, the high illumination levels often exceed the capabilities of current direct detectors. We propose to develop a new large-format (4096 × 4096), high-speed direct detection camera with high dynamic range (HDR) to accommodate both high illumination and sparse single-electron detection. This new camera will make use of state-of-the-art fully rad-hard design principles that have been successfully developed in the high-energy physics community to prevent most radiation damage on the detector. During Phase I, we characterized the radiation hardness and lifetime of our latest pixel designs, as implemented in our ultra-fast 1k×1k camera for 4D STEM (Celeritas) and our event-based electron counting camera for cryo-EM. By comparing radiation sensitivity between several variants of our sensors, we were able to identify the pixel designs and architectures that provide the highest radiation hardness. Additionally, we characterized our simultaneous high-dynamic range (HDR) mode, completed an initial sensor architecture design study, and identified a solution for a data path that accommodates the new cameraĺs high data rate. During Phase II, we will develop and build an initial prototype of this new camera by completing the sensor fabrication, designing and fabricating the circuit boards necessary to operate the camera at the maximum speed of the new sensor, testing the various components, and assembling a prototype camera. In parallel, we will develop software to automatically and seamlessly transition between the various camera readout modes and process the output data to produce accurate and interpretable images. Finally, we will integrate the camera with the CEOS CEFID energy filter and its software. Successful development of a fully rad-hard high dynamic range direct detection camera will dramatically expand the capabilities of electron microscopy by significantly increasing both spatial and temporal resolution under high brightness conditions, including large-area tomography of multicellular biological specimens, high-speed electron diffraction, and EELS. Such a camera would obsolete the much lower-performance indirect scintillator-coupled cameras currently used for these experiments.