Reduced Damage Electron Microscopy with Conditional Sample Re-illumination

We propose and theoretically analyze an electron microscopy scheme based on multiple sample illuminations with a very low current beam, in combination with Elitzur and Vaidman’s interaction-free measurement scheme. Our analysis starts with a prior guess for the characteristic transparency of each pixel on the sample. We update these priors after each round of illumination based on the statistics at the imaging detectors. The re-illumination for a pixel ceases once a stopping criterion is met. This criterion could be based on either a minimum required signal-to-noise-ratio, or a maximum imaging time. We show that this scheme reduces the damage suffered by the sample during imaging, compared to a conventional imaging scheme without conditional re-illumination. We calculated the maximum resolution at a given sample damage and signal-to-noise ratio that could be obtained with this scheme. This resolution shows an improvement over that from conventional imaging. We are working towards implementing this scheme on a scanning electron microscope, modified to include bright- and dark-field transmission electron detectors, and a fast beam blanker to obtain extremely low-dose electron pulses.