Maximum Efficiency STEM Phase Contrast Imaging
Timothy J. Pennycook
1,2
, Andrew R. Lupini
3
, Lewys Jones
1
and Peter D. Nellist
1,2
1.
Department of Materials, University of Oxford, Parks Road, Oxford OX1 3PH, UK
2.
SuperSTEM Laboratory, Daresbury, WA4 4AD, UK
3.
Oak Ridge National Laboratory, Oak Ridge, TN 37831, USA
Various bright-field (BF) imaging modes are available in scanning transmission electron microscopy
(STEM). The mode equivalent to phase contrast imaging in conventional high-resolution transmission
electron microscopy (TEM) involves a detector about an order of magnitude smaller than the
convergence angle of the probe. This means a large proportion of the incident electrons go to waste as
most of the bright field disk is left unused. The annular bright-field (ABF) mode of imaging uses a
larger area of the bright-field disk and has been successful in revealing columns of low-Z elements.
However complications in interpretation remain as contrast in ABF images can be generated through
both channelling and phase-contrast effects arrising from residual aberrations [1]. In the differental
phase-contrast (DPC) mode of imaging, almost the entire bright field disk can be used. In DPC multiple
detectors are placed in the BF disk, typically in a four quadrant configuration. To generate contrast, the
signal from oposing quadrants is subtracted, providing sensitivity to shifts in the beam caused by
electrostatic or magnetic fields in the sample [2].
An alternative approach that provides greater flexability is to record the entire BF disk with a pixelated
detector [3]. This allows one to choose the BF mode after acquiring the data, and even compare different
modes from the same scan. However a more sophisticated method can be used with a pixelated detector.
Rodenburg et al. used such a detector to allow STEM ptychography to retrieve the phase of the object
function [4]. Their work was focused on overcoming the diffraction limit to imaging resolution. Now with
aberration correction enabling routine atomic resolution imaging, the limiting parameter is often beam
induced damage rather than the spatial resolution of the microscope. With beam sensitive materials one
desires to extract the maximum amount of information per electron. We have therefore adapted the
technque developed by Rodenburg et al to provide maximum efficiency phase contrast imaging.
Figure 2 shows an example of a phase-contrast image created with STEM ptychography. A ronchigram
was recorded at each probe position simultaneously with the annular dark field (ADF) intensity. The
roughly 4000 Ronchigrams were then Fourier transformed with respect to probe position. By extracting
information only from regions of Fourier space in which the wavefunctions are transferred we reject noise
and obtain the best possible statistics on the phase and amplitude of each spacial frequency. We can then
interfere the spatial frequencies to form the highest efficiency phase contrast image possible in STEM.
Compared to a TEM the method has the advantage of offering phase contrast images at the full
resolution of the microscope without the need for aberrations. Through combining with simultaneous
ADF imaging we maximize the information by collecting signal over essentially all available scattering
angles. ADF imaging also facilitates the interpretation of the images and provides access to low-spatial
frequencies, which are not transferred in convention BF imaging. We will present further experimental
developments, data and theoretical calculations to investigate the optimal configuration for the
maximum phase sensitivity with the lowest dose [5].
382
doi:10.1017/S1431927614003638
Microsc. Microanal. 20 (Suppl 3), 2014
© Microscopy Society of America 2014
https://www.cambridge.org/core/terms. https://doi.org/10.1017/S1431927614003638
Downloaded from https://www.cambridge.org/core. IP address: 34.228.24.229, on 08 Jun 2020 at 15:26:55, subject to the Cambridge Core terms of use, available at