Oral Presentation 24th Australian Conference on Microscopy and Microanalysis 2016

Microscopy with source and detector arrays (#13)

Colin J.R. Sheppard 1 , Marco Castello 1 2 , Giuseppe Vicidomini 1 , Alberto Diaspro 1
  1. Italian Institute of Technology, Genova, LIGURIA, Italy
  2. DIBRIS, University of Genoa, Genoa, Liguria, Italy

Electron microscopes are usually of the conventional type (as in TEM), or scanning type (SEM or STEM). In a conventional microscope an image is formed with a lens after the sample, but in the scanning type the sample is illuminated with a small focused scanning probe, generated before the sample. In optical microscopy both these types also exist, but for many years now confocal microscopes, which combine a scanning probe with a lens and single element detector after the sample, have been used to great benefit.

The system can also be parallelized, by using arrays of illuminating spots and detector elements, to speed up the imaging process. The arrays are thus situated in image planes before and after the sample, unlike techniques such as differential phase contrast, selected area diffraction or ptychography, where detectors are placed in the Fourier plane.In the general system with a source and detector array, effectively we generate a four-dimensional (4D) signal from a 2D object. This image is thus highly redundant, and many approaches can be considered for reconstructing the object. In general, the final result is an image with twice the spatial frequency bandwidth of either the conventional or scanning types.

If we illuminate with a focused spot, and detect with a detector array, one algorithm to process the resulting 4D signal is to assume that the imaged object point for a particular detector element is not the point illuminated, but the combination of illumination and detection, corresponding to a point midway between these two. This approach is called pixel reassignment. By summing over the pixel elements, an improved resolution is obtained, but more importantly, a much stronger signal can be recorded.  The resulting system is called image scanning microscopy. Further improvement in resolution can be obtained by using multi-image deconvolution techniques.