In this thesis, the design and engineering considerations for a multi-beam scanning electron microscope (MBSEM) are discussed. This microscope can benefit biological research in various ways. It can give new insights into the inner workings of a multitude of biological systems that were hard to get using previously existing instrumentation. For instance, a higher throughput gives the option to do statistical analysis of multiple samples instead of drawing conclusions from only one. The goal of this thesis was to get from a proof of principle to a final system that can actually be used to do the research. It is divided into 5 chapters showing a step-by-step process of getting to the final system as it is now on the market. Chapter 1 is an introduction to the subject showing the current state of the art with respect to high throughput imaging. Chapter 2 Describes a novel imaging method in scanning electron microscopes. This chapter does not focus on the multi-beam application but shows it in the context of the often-used backscatter imaging. In this method, we place the tissue section directly on top of a thin scintillator screen (thinner than 200 μm) which is coated with a conductive layer. The light signal generated by the electrons transmitted through the sample is collected by a high NA objective lens and the light is imaged onto a photon detector outside of the vacuumchamber. A noise model is created to calculate the signal-to-noise ratio and the contrast-tonoise ratio of this imaging method. It shows that the best images are generated around a landing energy of about 5keV. There are some dependencies on sample thickness, staining level, and light collection efficiency which are also explored. This method does lower the resolution in the image to some extent (by a factor of 2 at low energies and thick sections), which is shown at the end of the chapter. Chapter 3 Goes into the considerations that have to be taken into account when dealing with the imaging method from chapter 2. This chapter is applicable to a single beam SEM as much as anMBSEM. A list of possible light detectors is given from which silicon photomultipliers are selected as the best candidate for the MBSEM. Combined with the light detector, multiple options for a scintillator were discussed, from which YAG:CE is selected. Organic scintillators are discarded due to their bleaching behavior due to electron beam irradiation. The surface of the scintillator and the coating layer are shown to have a large impact on image quality. For this reason, the scintillators are ion-beam polished and coated with a Boron layer. Unexpected behavior in the form of scintillator saturation is observed which is then described by a model and connected to the noise model fromchapter 2. Chapter 4 Gives an analysis of all the hardware requirements for a MBSEM. First a measurement of the crosstalk as a function of landing energy and pitch. It is found that a crosstalk of at least 10 % is to be expected in the system. Next, an overview is given for all the parameters that are related to the stage and the light optics. These are then related to the final throughput of the system. Two imaging strategies are described, in one the beam scans in one direction and the stage in the other. In the other strategy, the beams scan like in a regular SEM and are subsequently descanned in the light-optical system. It is found that with a step and scan approach in combination with planned beamshifts, the maximum throughput that can be achieved is around 420 mpix/s. Chapter 5 Shows results from the final prototype system. Alignments are of great importance in any SEMbut even more so in theMBSEM. Therefore a large part of this chapter is dedicated to describing this alignment. This starts with the electron optical alignment of the source and the beam through the column. The grid of beams has to be optimized to show as little as possible distortions to improve system throughput. The scan and descan have to be aligned to the grid axes and the amplitude has to be precisely correct. The beams have to be perfectly aligned to the detector array. On the processing side, a description of how can be compensated for varying dark and gain levels in the detector array. In the end, a final image is shown, consisting of 400 megapixels. Chapter 6 Describes the valorization of the project and all the challenges and choices that were involved.

The analytical theory of statistical Coulomb interactions allows to determine the trajectory displacement in a single rotationally symmetrical beam with well-behaved spatial and angular particle distributions. This can be used to estimate the trajectory displacement in a multi-beam system using the so called fully-filled segment approximation. This approach predicts full compensation of trajectory displacement for a specific setup of the system. We show that this prediction is not consistent with Monte Carlo simulations and we develop a new approach to the calculation, showing that two independent trajectory displacement contributions are present in a multi-beam system. We support this calculation with Monte Carlo simulations as well as with experimental data from a multi-beam system.

Large area electron microscopy (EM) imaging has long been difficult due to fundamental limits in throughput for conventional electron microscopes. New developments in transmission electron microscopy and multi-beam scanning electron microscopy (MBSEM) imaging have however made it possible to generate large EM datasets [1,2,3]. This article describes a transmission imaging technique that is suitable for a MBSEM as it allows for a relatively straightforward way of separating the signals generated by each beam. The technique places a thin (50nm-200nm) tissue section directly on top of a coated scintillator. The electrons that are transmitted through the section generate light in the scintillator which is collected by a high NA objective and imaged onto a photon detector. This article gives a model for the contrast-to-noise (CNR) and signal-to-noise (SNR) ratio that is to be expected for this imaging technique. These parameters were calculated using Monte-Carlo simulations. It was found that the CNR increases when decreasing landing energy and SNR increases with increasing landing energy. These two trends cause that there is an intermediate energy where imaging performance is best. The energy of this optimum was calculated for various levels of heavy metal staining, section thickness, coating material, coating thickness and light collection efficiency. The model was verified experimentally on a synthetic sample.

Nanodiamonds containing fluorescent nitrogen-vacancy centers are increasingly attracting interest for use as a probe in biological microscopy. This interest stems from (i) strong resistance to photobleaching allowing prolonged fluorescence observation times; (ii) the possibility to excite fluorescence using a focused electron beam (cathodoluminescence; CL) for high-resolution localization; and (iii) the potential use for nanoscale sensing. For all these schemes, the development of versatile molecular labeling using relatively small diamonds is essential. Here, we show the direct targeting of a biological molecule with nanodiamonds as small as 70 nm using a streptavidin conjugation and standard antibody labelling approach. We also show internalization of 40 nm sized nanodiamonds. The fluorescence from the nanodiamonds survives osmium-fixation and plastic embedding making them suited for correlative light and electron microscopy. We show that CL can be observed from epon-embedded nanodiamonds, while surface-exposed nanoparticles also stand out in secondary electron (SE) signal due to the exceptionally high diamond SE yield. Finally, we demonstrate the magnetic read-out using fluorescence from diamonds prior to embedding. Thus, our results firmly establish nanodiamonds containing nitrogen-vacancy centers as unique, versatile probes for combining and correlating different types of microscopy, from fluorescence imaging and magnetometry to ultrastructural investigation using electron microscopy.