Microscopy Conference 2023

Session chairs

This session dial with new developments in SEM and FIB technology and methods. Especially new detector principals and designs for electrons, ions and photons locally emitted after interaction with different types of ion or electron beams. Techniques using the angular distribution of the scattered ions and electrons like EBSD and ECCI will be covered by the contributions. New strategies in beam control and data analysis will also be discussed.

Prof.

Hyperspectral visualization of picometric motion

Abstract text (incl. figure legends and references) The motion of nanostructures can be measured with picometric resolution using scattering of free electrons at sharp edges of the structures. Motion at the nano- to atomic scale is of growing technological importance and fundamental interest, in nano-electro-mechanical systems (NEMS), advanced materials (e.g. nanowires, 2D materials), mechanically reconfigurable photonic metamaterials; and in the study of systems governed by Van der Waals and Casimir forces, and quantum phenomena. However, there are no routinely available technologies for quantifying and spatially mapping fast, complex movements of picometric amplitude in nanostructures. We show how the spectrally resolved detection of scattering from a tightly-focused free-electron beam incident on the sharp edges of a nano-object can provide for quantitative 3D visualization of motion at the picoscale. For a range of nano/microstructures, from simple cantilevers to photonic metamaterials and MEMS comb-drive actuators, we demonstrate measurements of thermal (cf. Brownian) motion amplitudes down to a noise-equivalent displacement level of 1 pm/Hz1/2, and the mapping of driven-motion oscillatory &quot;mode shapes&quot; with spatial (SEM imaging) resolution far beyond the diffraction limit applicable to optical vibrometry techniques. We also report on the first observation of short-timescale &quot;ballistic&quot; thermal motion in the flexural mode of a nanomembrane cantilever, driven by thermal fluctuation of flexural phonon numbers in the membrane: over intervals &lt;10 µs, the membrane is found to move ballistically, at an average constant velocity of ~300 µm/s, while Brownian-like dynamics emerge for longer observation times. Access to the ballistic regime provides verification of the equipartition theorem and Maxwell-Boltzmann statistics for flexural modes and presents opportunities in fast thermometry and mass sensing.

IM 1: Progress in instrumentation and ultrafast EM

IM 3: SEM and FIB developments

Invited talk

Prof. Dr. rer. nat. habil.

Simultaneous acquisition of real and reciprocal space information in transmission SEM

Abstract text (incl. figure legends and references) In the recent years, a variety of transmission diffraction setups entered the field of scanning electron microscopy (SEM) to prove the worthwhile implementation of diffraction analysis in combination with lower energetic electrons and the analytical capabilities of a SEM [1-3]. The broad accessibility and simpler setup facilitate such developments compared to the more complex transmission electron microscopy (TEM). Moreover, the large microscope chamber offers a straightforward implementation and enough space for unique in situ systems [4-5]. Even state-of-the-art direct electron detection cameras combined with the popular 4D-STEM technique are available for SEM´s with the benefits of reduced knock-on damage and increased contrast [2]. However, in relation to in situ experiments, where a continuous observation of the occurring phenomena is indispensable, these cameras are not suitable for such an experiment, even with the fastest acquisition rate. Based on the limitations regarding parallel acquisition of real and reciprocal space information in situ, we introduce a unique setup in SEM, which solves the current restrictions. Figure 1a shows schematically the concept of our developed system, which is similar to our existing diffraction setup (LEND) [3]. The basis is a fluorescent screen positioned below the sample and an in-chamber mounted CMOS [WJ1] camera responsible for recording of the generated diffraction patterns on the screen. To enable the additional acquisition of the real space signal, a flexible STEM detector is installed below the fluorescent screen, which contains a concentric hole allowing the electrons to pass through and contribute to image formation. Figure 1b depicts the setup installed in our SEM. The combination of a simple integrating detector (STEM) with the parallel acquisition of the averaged diffraction patterns offers high frame rates linked to the corresponding reciprocal space information. In addition, all occurring signals in SEM can be collected. Here, we investigate a thin MoS2 flake covered with gold nanoparticles as model system for acquiring the four simultaneously available signals in one run (Figure 2). The sample topography is depicted by secondary electrons (SE), where all small particles and surface contamination are visible. Back-scattered electrons (BSE) offer elemental contrast, which facilitates to distinguish between the gold particles and MoS2. The inner sample structure can be investigated by bright-field (BF) STEM and the acquired diffraction pattern reveal the present crystal structure of MoS2. Besides BF STEM imaging, the fluorescent screen, together with the STEM detector, can be moved to switch to dark-field (DF) imaging by selecting a suitable diffraction spot. In the future, we are planning to perform dedicated heating experiments while acquiring real and reciprocal space information in situ. An upgrade of the current setup to an in situ chip-based heating stage will expand strongly the analytical potential of this concept. The real time monitoring of an ongoing process by STEM and diffraction signal will help to reveal appearing phenomena and mechanism in several material aspects [6]. [1] 10.1016/j.ultramic.2018.09.006 [2] 10.1016/j.ultramic.2020.113137 [3] 10.1016/j.ultramic.2020.112956 [4] 10.1017/S143192762100951X [5] 10.1017/S1431927621003974 [6] The authors acknowledge funding by the German Research Foundation (DFG) via the Research Training Group GRK 1896

Abstract talk

Prof. Dr.

Cryo focussed ion beam secondary ion mass spectrometry analysis of Galinstan

Abstract text (incl. figure legends and references) Introduction The combination of a focused ion beam and scanning electron microscope (FIB/SEM) with a time of flight secondary ion mass spectrometer (ToF-SIMS) [1] offers new possibilities for materials samples investigation. The combination of different information, such as highly sensitive 3D chemical mapping, in-situ sample preparation and high resolution imaging yield valuable insights into the samples. When adding cryo sample temperatures by using a cryo stage a number of experiments are possible for e.g. liquid metals, such as Galinstan, a Ga, In, Sn alloy. Galinstan can be printed into 3D shapes at room temperature [2]. As it solidifies only at about -19°C [2], it is the general expectation that printed structures are stabilized by a thin galliumoxide layer, which forms around the alloy by passivation. As Galinstan is liquid at room temperature, it is generally difficult to investigate the oxide layer with conventional spectroscopic methods. Objectives In our study we want to detect the stabilizing oxide layer with ToF-SIMS facilitated by the Gallium ion beam in a FIB/SEM. To prevent material mixing within the liquid metal we solidified the printed Galinstan by cooling it down on a cryo stage. This allows also the study of the alloy demixing of Galinstan when cooled down, which can have an influence of Galinstan&quot;s electrical properties at low temperatures. Materials &amp; Methods Printed Galinstan lines are cooled down below the solidification temperature in-situ. Afterwards ToF-SIMS analyses of the sample surface can be conducted (Figure 1a). For the cross section analyses, the Galinstan is cut with the FIB and then tilted to an angle of 54° (Figure 1d). Either negative ions (GaO) or positive ions (Ga, In, Sn) can be detected by ToF-SIMS. Results We can visualize the enclosing galliumoxide layer in two different detection geometries: First, on the surface of the lines (Figure 1b,c) and second, along the line surface adjacent to a cross section (Figure 1e,f). As additional finding we see that the alloy separates into a Ga and In-Sn phase at temperatures lower than -25°C (Figure 2). However, we found a number of experimental sample and instrument parameters influencing the secondary ion signal, such as sample geometry, the presence of oxygen (matrix effect [3]) and the charging of the sample. Moreover, water freezing out on the sample surface during the cooling process can strongly influence the ion signal. Conclusion In our study we showed a first application of our cryo FIBSIMS instrumental combination yielding valuable results for the investigation of Galinstan. It was possible to visualize the galliumoxide layer and phase separation of the cooled down alloy. The secondary ion detection shows strong influence of the detection geometry and the investigated sample. In future, we want to expand this novel analysis tool to investigate fully printed, flexible electronics. This might yield valuable information for both the design of new devices and their quality control. Acknowledgements: The author thanks Navid Hussain, Jasmin Aghassi-Hagmann, and Michael Hirtz (Karlsruher Institut für Technologie) for the samples. Research funded by DFG via the Excellence Cluster &quot;3D Matter Made to Order&quot; (EXC-2082/1-390761711) [1] Pillatsch et al., Progr. Cryst. Growth Charact. Mater., 65(1):1–19, 2019 [2] Hussain et al., Adv. Mater. Technol., 29:2100650, 2021 [3] Priebe et al., J. Anal. Atom. Spectrom., 35(6):1156–1166, 2020

IM 4: Development of cryo-EM instrumentation and techniques

References
1)	Cloots et al. (2016) Mat. Sc. Eng. A658, 68-76, doi:10.1016/j.msea.2016.01.058 
2)	Kunze et al. (2014) Mat. Sc. Eng. A620, 213-222, doi:10.1016/j.msea.2014.10.003 
3)	Geiger et al. (2016) Mat. Sc. Eng. A661, 240-246, doi:10.1016/j.msea.2016.03.036 
4)	Etter et al. (2015) IOP Conf. Series: Mat. Sc. Eng. 82, 012097, doi:10.1088/1757-899X/82/1/012097 
5)	Niu et al. (2022) Energy Environ. Sci.15, 2002-2010, doi:10.1039/d1ee03696c 
6)	Barroo et al. (2020) Nature Cat. 3, 30-39, doi:10.1038/s41929-019-0395-3 
7)	Basu et al. (2022) Acta Mater. 229, 117746, doi:10.1016/j.actamat.2022.117746 
8)	Prieto et al. (2017) Nanotechnology 28,135701 doi:10.1088/1361-6528/aa5ec4 
9)	Wang & Zaefferer (2017) Mat. Charact. 130, 33-38, doi:10.1088/1361-6528/aa5ec4 
10)	Becker et al. (2022) Crystals 12, 813, doi:10.3390/cryst12060813 
11)	Seita et al. (2016) npj Comp. Mat. 2, 16016, doi:10.1038/npjcompumats.2016.16 
12)	Laleh et al. (2021) Scrpta Mat 192, 115-119, doi:10.1016/j.scriptamat.2020.10.018 
13)	Saylor et al. (2004) Metall. Mater. Trans. A35, 1981-1989, doi:10.1007/s11661-004-0147-z 
14)	Rohrer et al. (2004) Z. Met., 95, 197-214, doi:10.3139/146.017934 
15)	Zaefferer et al. (2008) Metall. Mater. Trans. A39, 374-389, doi:10.1007/s11661-007-9418-9 
16)	Echlin et al. (2015) Mat. Charact.100, 1-12, doi: 10.1016/j.matchar.2014.10.023 
17)	Kelly et al. (2016) Acta Mater. 111, 22-30, doi:10.1016/j.actamat.2016.03.029 
18)	Tsai et al. (2022) Rev.Sc. Instr. 93, 093707, doi:10.1063/5.0087945 
19)	Winkelmann et al (2021) J. Micr. 284, 157-184, doi:10.1111/jmi.13051 
20)	Lenthe et al. (2019). Ultramicroscopy 207, 112841, doi:10.1016/j.ultramic.2019.112841 
21)	Kunze et al. (1994) MSF 157-162, 1243-1248, doi:10.4028/www.scientific.net/MSF.157-162.1243 
22)	Heidelbach et al. (2000) JSG 22, 91-104, doi:10.1016/S0191-8141(99)00125-X 
23)	Winkelmann & Nolze (2015) Ultramicroscopy 149, 58-63, doi:10.1063/1.4907938

Crystals and grains – EBSD in 2+ dimensions

Abstract text (incl. figure legends and references) Electron backscatter diffraction (EBSD) is a widespread microanalytical method to characterize crystal orientations, phases and defects in typically polycrystalline materials. At the core facility ScopeM, we are supporting user requests in a diversity of applications, also involving characterization by EBSD. Thus, in the first part of this presentation, a number of projects are presented, with emphasis on the links between microstructures and textures to the anisotropy of physical properties, and the effects of different processing routes for tuning them to the most desired properties. These in include the microstructural characterization and elastic anisotropy of additively manufactured Ni-based superalloys (1-4) produced by powder bed laser fusion (PBLF or SLM); crystal orientation dependent photoelectrochemical performance of thermally-oxidised Cu2O photocathodes (5); grain orientation dependent catalyzed surface reactions observed by in-situ experiments (6); twinning mediated plasticity in lean Mg alloys (7); bi-modal nanoheteroepitaxy of GaAs on Si (8). In the second part, a few recent developments are highlighted. As one of the general bottlenecks, EBSD is typically performed on 2D surfaces of bulk materials. Thus, planar features in three-dimensional microstructures, like interface planes or grain boundaries, are only accessible by their traces on the planar sample surface. A number of approaches has been applied to reveal the missing degree of freedom in order to obtain the complete 5-parameter grain boundary character in a statistically reliable population. These could be pseudo-3D EBSD (9); habit plane analysis (10); hybrid analysis by EBSD and optical reflectance on top and bottom sides (11); large number statistics on 2D mappings (12-14); and serial sectioning by (P)FIB, fs-laser or polishing robots (15-18). Finally, not only speed and sensitivity of EBSD detectors continue to see major developments. Also, the processing of EBSD patterns has advanced from conventional Hough transform based band detection and indexing to pattern matching approaches, where the recorded patterns are compared to the best-fit from a dictionary of simulated patterns (19-20). In the case of quartz microstructures, these simulations demonstrate the potential of EBSD to distinguish not only Dauphiné twin orientations (21-22), but even the handedness of left-handed and right-handed crystals (23), and polarity in other non-centrosymmetric structures.

Dr.-Ing.

Prof. Dr.-Ing.

3D-reconstruction of hierarchical nanoporous structures using multi-modal deep learning models trained on synthetic multi-voltage FIB tomography data

Abstract text (incl. figure legends and references) 3D reconstruction of hierarchical nanoporous structures requires high-resolution imaging data in the range of a few nanometers. Focused ion beam (FIB) tomography can be combined with a scanning electron microscope (SEM) to collect high-resolution images of nanostructures in the range of 1 nm in the SEM plane and 10 nm in the out-of-plane direction [1]. FIB tomography collects data layer-wise, where a FIB milling process follows each imaging process by an SEM. However, these consecutive series of cross-sectional images have artefacts, such as the "shine-through effect" [2]. Due to this effect, some structures become visible from posterior regions of the currently milled plane, and it is not straightforward to map pixel intensities with the material phase uniquely. Therefore, such images cannot be segmented accurately using conventional methods like thresholding or the k-means algorithm. Earlier, [3] presented a machine learning-based approach to reconstruct hierarchical nanoporous materials accurately. [3] also demonstrated a method to generate synthetic FIB-SEM images of any complex nanoporous structures to overcome the lack of training data. In this study, we present two ways to improve 3D reconstruction results. The first approach is to use a multi-modal machine learning method to take advantage of more information from the same region. In this case, we train a machine learning model using synthetic multi-voltage FIB tomography data generated by changing the accelerating voltage and simulating images of the same regions (Figure 1). Figure 2 shows pixel intensity profiles of images simulated using different accelerating voltages and the actual structure. The second approach is to make synthetic FIB-SEM data more similar to the real electron microscopy images using machine learning-based domain adaptation techniques. We show that using these approaches, our segmentation methods can suppress the "shine-through effect" and improve upon the methods presented in [3]. [1] Knott et al., Journal of Neuroscience, 28(12):2959–2964 (2008) [2] Prill et al., J Microsc., 250(2):77-87 (2013) [3] Sardhara et al., Frontiers in Materials, 9 2296-8016 (2022)

IM 5: Quantitative image and diffraction data analysis

FIB-notch: a novel method towards ultra-high throughput TEM sample preparation

Abstract text (incl. figure legends and references) Transmission electron microscopy (TEM) has become a routinely used technique in a wide range of industries. With an ever-increasing number of analyses performed, lowest time-to-sample is now a central demand on sample preparation. Here, conventional focused ion beam (FIB) preparation routines are facing a fundamental physical limitation: milling is either performed rapidly at high currents or accurately at very low currents [1]. The final thinning and polishing steps thus become a bottleneck for the whole process. FIB-Notch is a novel, FIB-based preparation approach designed to overcome the dichotomy between accurate and fast milling. It employs a special sample geometry, where a notch is cut into a thick lamella [2]. Upon glancing-angle milling of the sample, enhanced erosion occurs at the elevated notch faces, leading to the formation of a receded, electron-transparent terrace region (Fig. 1). The key advantage of this patented technique is that spatial localization is governed by the notch geometry rather than the ion beam profile, making the erosion process essentially self-aligned. This allows for rapid thinning to be performed at high beam currents in a quasi-broad ion beam setting, speeding up the preparation process significantly (Fig. 2). In order to control the final thickness of the lamella in the terrace region, the notch geometry needs to be matched to the initial thickness of the lamella. Accurate determination of this parameter is thus required, preferably performed in-situ. A quantitative thickness measurement technique based on electron beam absorbed current (EBAC) [3] was developed to for this purpose. First experiments performed on a VION plasma FIB and a customized Zeiss Auriga 40 instrument will be presented, where electron-transparent samples could directly be created with nA-range ion beam currents. The physical principles governing the erosion of a notched lamella will be discussed, which were studied analytically and numerically using a two-dimensional, kinetic model of sputter erosion. Finally, an outlook on further potential applications of the technique to tasks such as sample sectioning will be given. Future work will focus on validation of the thickness determination method and automation of the preparation process. Figure 1. Erosion of a notched lamella (schematic depiction). Figure 2. SEM micrograph of a TEM lamella prepared with the FIB-Notch technique, using a Xe plasma FIB at 180 nA. Financial support by the European Union via the Europäischer Fonds für Regionale Entwicklung (EFRE) and the Investitionsbank Sachsen-Anhalt (Germany) through the project FIBNotch under the grant number 2004/00071 is gratefully acknowledged. [1] Bassim N, Scott K. and Giannuzzi LA, MRS Bulletin 39, 317-325 (2014) [2] Busch R, Krause M, Coyle S, and Höche T, Micron 107, 35-42 (2018) [3] Moldovan, G. Microscopy and Microanalysis, 25(S2), 532-533. (2019)

IM 3
SEM and FIB developments

Appointment

Session chairs

Description

Programme

IM 3SEM and FIB developments

Appointment

Session chairs

Description

Programme

IM 3
SEM and FIB developments