A magnetic electron energy analyser for use in an SEM

Abstract

A new concept for an electron energy analyser is presented whereby a magnetic field is used to disperse the electrons on to an imaging detector. Although using magnetic fields to disperse electrons is not a new idea [1], we show that some of the previous difficulties with using magnetic fields as an electron energy analyser in a SEM can be overcome. Figure 1 shows a schematic of the basic principle. Electrons emitted from the sample pass through a narrow slit placed very close to the sample. Electrons are then dispersed on to an imaging detector. The principle is very close to the well known 180 degree magnetic spectrograph. However, in this case electrons are detected out of the dispersion plane. Using this very simple approach, one can perform Auger electron spectroscopy (AES) over a large energy range and a large angular range in parallel. The small red lines on the detector (see Fig. 1) show how electrons of the same energy but different azimuthal angle will land on the detector. Since the electrons on the red curves are spread quite far, this will create a poor energy resolution since different electrons of different energies and take off angles can land at the same location on the detector. We show that by increasing the angle of rotation (to greater than 180 degrees) that the electron undertake before striking the detector , we can improve the energy resolution and maintain a high transmission while retaining the parallel acquisition capability. The major advantage of the analyser is that it can acquire electron energy spectra in parallel and over a large angular range with a much greater transmission than another parallel acquisition analyser the Hyperbolic Field Analyser (HFA). An Active Pixel Sensor (APS) was used to detect the electrons directly. No use of microchannel plates was made to detect the electron energy spectrum. Since the APS is not very sensitive to low energy electrons, this resulted in rather poor statistics. However, considerable improvement would be expected if a microchannel plate (MCP) were to be used. Figure 2 shows a schematic of the actual realisation of the analyser. Due to space constraints (largely due to the size of PCB board and the narrow space between the Helmholtz coils), the APS had to be placed below the sample as indicated. Figure 3 shows an image acquired on the Active Pixel Sensor. The primary electrons were selected to have 900 eV electrons and the curved line shows elastic peak electrons acquired for many different polar take off angles. This is indicated by the line A. The image was integrated to form a spectrum. Lines B and C were used as the limits of integration. Figure 4 shows the resultant electron energy loss spectrum (EELS) after the process of integration. The energy resolution can be seen to be about 4eV. Further details of the analyser will be presented such as estimates of the energy reolution, an Auger electron spectrum and calculations of the Field of View. Advantages of the analyser are: (a) Acquisition of electron spectra over a large energy range and large angular range in parallel. (b) The detector and plate containing the sperture have a small size and could be fitted between the sample and objective lens (c) Simple construction Disadvantages of the analyser are: (a) Helmholtz coils needed to produce the magnetic field are quite bulky. (b) The spectrum is sensitive to the distance between the location of the electron source (i.e. where the primary beam strikes the sample) and the aperture slit. (c) The magnetic field of the Helmholtz coils changes the electron current needed to focus the objective lens (but does not affect the spatial resolution). (d) Current detectors are too large to fit between sample and objective lens. A bespoke electron sensor is required.