Ultra-Fast Electron Counting Through Edge Computing

Advanced techniques in transmission electron microscopy TEM) such as high-speed continuous tilt tomography, in-situ microscopy, and S)TEM at liquid helium temperature that are being developed to address important problems in materials and life sciences are hampered by the lack of a large field-of-view electron counting camera with sufficiently high frame rate. Currently available systems are an order of magnitude too slow and present huge challenges with the massive amounts of data they produce. We propose developing an edge-computing based system that will take advantage of a novel sparse, binary- readout direct electron sensor that the company has developed to provide a camera system that will allow 5 times the dose-per-pixel on a 16-megapixel sensor while performing all of the electron counting operations in hardware located in the camera head, thereby relieving associated computer resources of the enormous electron counting burden. During Phase I, we will demonstrate a centroiding algorithm in field-programmable gate arrays FPGAs) located in the camera head to identify, to sub-pixel accuracy, the incidence position of primary beam electrons. We will then determine the relationship between coincidence loss and exposure rate to establish the linearity performance of the system as a function of dose rate. We will also determine optimal ranges of camera frame rates for accumulated frames taking into account information contained in each frame and bandwidth to a host computer system. Finally, we will demonstrate the optimized system for continuous- rotation tomography and cryo-temperature and/or in situ TEM movies. Successful development of an ultra-fast electron counting direct detection camera with edge-computing enabled data reduction would be a tremendous benefit for multi-dimensional electron microscopy in private, public, and national laboratories working in a wide variety of fields. Microscopists will no longer need to worry about significant non-linearity or about fitting their experiments into a very narrow window of exposure conditions. Removing the camera as a limiting factor for TEM imaging will make it an indispensable tool and drive its adoption in all fields.