Low-Energy Secondary Electron Filtering with Immersion Lens SEM

Abstract

The scanning electron microscope (SEM) offers a rich array of different electron-beam induced signals with which to create the final micrograph. So in parallel with efforts to improve the beam resolution [1], it is also important to find ways to make the best use of these induced signals to bring out the desired information about the sample. As an example of this effort, this paper will describe the use of energy filtering of the secondary electron (SE) signal using an immersion lens SEM. Applications of this technique include visualizing dopant contrast [2] and improving the sensitivity of passive voltage contrast, Figure 1, as well as the potential for increasing the contrast between key materials of interest by exploiting the different SE energy distributions from different materials (see for example, [3] for experimental measurements of SE spectra).To better understand the complete imaging process Monte-Carlo simulations have been undertaken that include a 3D modeling of the SEM column, including the electromagnetic fields, so that the trajectories of the emitted SE and back-scattered electrons (BSE) from the beam-sample interaction can be accurately followed back into the SEM detection system [4]. Once such a model is in place virtual experiments can be carried to determine the optimum beam and detector conditions for a particular sample type. Such simulations can investigate the signal based on particular SE energies or trajectory angles from the sample, and also to simulate the actual image expected.This modeling has been applied to the energy filtering of the SE signal. Low-pass, high-pass and band-pass filtering can be carried out depending upon the beam, sample and detector conditions. In each case, the magnetic field of the immersion lens SEM combined with an in-lens detector play the crucial role in enabling the filtering. Figure 2 shows modeled results for a low pass filter where with in-lens detector parameter M=-2V the electrons up to ~5 eV are transmitted, while a value of M=-16V allows a wider range (~16 eV). The modeling showed that by adjustment of M it was possible to go from zero to full transmission of the SE range, with Fig.2 just being representative examples. Figure 1 shows experimental results for a high pass filter on a voltage contrast sampleby adjusting the starting point of the range from 1 eV to 2 eV it was possible to change which of the contacts appear bright or dark in the image.

Anisotropic Approach for Simulating Electron Transport in Layered Materials: Computational and Experimental Study of Highly Oriented Pyrolitic Graphite

“…SE spectra were collated by differentiating the mean intensity of each image from each individual M step. SE energy calibration method for this system can be found in Young et al 7…”

“…First, we investigated the dependence of the REEL spectra on the parameter f by spanning a range of possible values in Eqs. (6) and (7). .…”

“…SE spectra were collated by differentiating the mean intensity of each image from each individual M step. The SE energy calibration method for this system can be found in Young et al and the Supporting Information of Wan et al, while the absolute energy value was checked by fine structures reported experimentally in literature (3, 4, and 7.5 eV) for HOPG and fine structure for diamond (6 eV). Detection artifacts within a certain M range were identified by evaluating the average intensity of a reference gold sample with the smallest possible filter parameter step difference (0.1 V).…”

Nanoscale Mapping of Semi‐Crystalline Polypropylene

“…The design of this detector allows for the different SE energy ranges to be collected by changing a mirror voltage. [13,14] An iFast automatic collection recipe (iFast developers kit version 3.0.16.1738) was utilized to step the mirror voltage between À15 and 15 V and an image was collected with each successive 0.5 V. The mean intensity of each micrograph was plotted for each mirror voltage and the final SE spectra were generated by differentiating the intensity of the micrographs with respect to the mirror voltage step. The resulting SE spectra are characteristic of the surface of the material.…”