An energy discriminating direct detector for multi-color SEM

Project Summary / Abstract Understanding brain function and neurological disorder is predicated on mapping the connectivity among neurons, distinguishing various cellular and molecular populations, and elucidating the protein-protein interactions that drive neurological function. Such studies span a wide range of scales, requiring both a large field- of-view to map connectivity and high-resolution to visualize subcellular and intrasynaptic molecular details. Multi-color electron microscopy (EM) has shown promise in studying biological ultrastructure at nanometer resolution while also detecting specific molecular components of interest. The technique is analogous to multi- color fluorescence microscopy, but at about ~100× higher magnification. However, the current method for acquiring multi-color EM data is based on energy-filtered TEM (EFTEM), which significantly limits is usefulness in neurobiology due to its severely low throughput and limited field-of-view. We propose to develop a new ultra-fast direct detection camera for scanning electron microscopy (SEM) capable of operating at more than 100,000 frames per second (fps) and measuring the energy of detected electrons. Such a camera will be an astounding leap forward, dramatically improving throughput and enabling sophisticated multi- color EM techniques using serial block-face SEM (SBEM), so that small structures like synaptic vesicles, nucleosomes, nuclear pores, and viruses (all a few nanometers to 10-40 nm) can be identified and quantified. We have already developed a Phase I prototype of this new direct detection SEM camera, based on a low-energy- optimized version of Direct Electron’s current generation TEM direct detection cameras. Initial results have confirmed sensitivity to electrons down to 2 kV energy, showed far superior information content compared to current state-of-the-art scintillator-coupled SEM cameras, and most importantly, revealed that our new sensor design is capable of energy discrimination of detected electrons. These initial results were used to finalize the requirements for the new ultra-fast pixelated direct detector proposed here, the speed of which is required to make the technique useful for large field-of-view, high-resolution multi-color SBEM for imaging neurons. During Phase II we will advance the development and commercialization of this new ultra-fast SEM camera system, by fabricating and assembling the new ultra-fast SEM camera, further refining hardware and software to efficiently handle the enormous volumes of data produced and identify multi-color EM labels, and then demonstrating high-speed multi-color SBEM of neuronal tissue. The success of this project will create an analog of the ubiquitous fluorescence light microscopy technique, but at significantly higher resolution using serial block-face SEM. This will not only have wide ranging applications for neuroscience research but will also extend to cellular microscopy in a wide range of other biological fields. Additionally, the new camera will also enable energy-filtered electron backscattered diffraction (EBSD), which is widely used in materials science research and industrial quality control. Therefore, as a new enabling technology, we anticipate that the proposed detector will have broad impact across a variety of fields.