Quantitative electron phase imaging with high sensitivity and an unlimited field of view

Abstract

As it passes through a sample, an electron beam scatters, producing an exit wavefront rich in information. A range of material properties, from electric and magnetic field strengths to specimen thickness, strain maps and mean inner potentials, can be extrapolated from its phase and mapped at the nanoscale. Unfortunately, the phase signal is not straightforward to obtain. It is most commonly measured using off-axis electron holography, but this is experimentally challenging, places constraints on the sample and has a limited field of view. Here we report an alternative method that avoids these limitations and is easily implemented on an unmodified transmission electron microscope (TEM) operating in the familiar selected area diffraction mode. We use ptychography, an imaging technique popular amongst the X-ray microscopy community; recent advances in reconstruction algorithms now reveal its potential as a tool for highly sensitive, quantitative electron phase imaging.

Lensless LED matrix ptychographic microscope: problems and solutions

“…This flexibility holds promise for other forms of microscopy. In electron ptychography [26], for example, the temporal and spatial coherence are limited and the illumination variance also exists because of lens instabilities. For Fourier ptychography [27], a lens-aided imaging method that also uses LED matrix, can also make use of the proposed reconstruction strategy to accommodate the imperfections of the LEDs.…”

Ten implementations of ptychography

“…For electrons, the lensless setup can be implemented on a scanning electron microscope, but a practical difficulty is adjustment of the magnification over a good range with a fixed or marginally adjustable camera length (Humphry et al, 2012). This is avoided by the use of the intermediate lenses in the transmission electron microscope, where the available options are the microscope setup (Putkunz et al, 2012) and the selected area setup (Maiden et al, 2015). Resolution improvements are not apparent in these two cases, where the primary gain is an effective way to quantify the phase of the electron wave.…”

Modal decomposition of a propagating matter wave via electron ptychography

“…The longest exposure was 3 seconds (dataset 2) and the shortest was 1 second (dataset 1). Due to slowly varying diffraction lens current instability and a relatively long data collection time, the centre of the diffraction patterns drifted during our experiments by around 50 pixels -this drift was corrected algorithmically during the reconstruction process, as detailed in [27].…”

“…However, for the first 500 iterations we used an approximate method for accounting for the partial coherence using a Gaussian convolution of the diffraction pattern, as described elsewhere [27,28]. With reference to Figure 4, which shows the error metric as a function of iteration number, we see that when the multi-mode method is switched on, the error drops significantly after an initial instability, indicating that the modal decomposition is much more effective at handling the coherence properties of the beam.…”