Imaging and Spectroscopy of Backscattered Electrons at Ultra-Low Energies—A New Characterization Approach for Beam Sensitive Organic Functional Materials

With the progress in research and development of carbon materials, organic materials and devices, as well as biological and bioinspired functional nano-materials, the need for novel in-situ characterization methods becomes paramount. Such materials are in general highly beam sensitive and conventional sample preparation and electron microscopic observation cannot sensibly be used in an in-situ setting. In particular metal sputtering, heavy metal staining and beam induced degradation of the sample are obstacles hampering the direct – and dynamic – observation of organic samples.<br/><br/>In our study we use ultra-low voltage Scanning Electron Microscopy (SEM) for direct imaging of organic and biological materials with high spatial resolution, surface sensitivity, and significantly reduced beam-damage. Applying electron spectroscopic imaging of secondary and backscattered electrons we can for the first time in a SEM identify excited states (e.g. plasmon excitations) and use them for low-dose material characterization. For this we utilize a novel kind of instrument, the DELTA SEM [1]. First studies on the functional characterization of organic thin film transistors exploiting dynamic charging by secondary electron (SE) spectroscopy have recently been published [2].<br/><br/>Next, we introduce Electron Energy Loss Spectroscopy of backscattered electrons (BSE) to scanning electron microscopy (bsEELS): As prerequisite for spectroscopic data collection the DELTA provides sub-nanometer imaging resolution in a wide energy range from a few keV (as in conventional SEM) down to ultra-low landing energies around 20 eV. Since for electron energies below 1 keV the backscattering coefficient for low-Z materials increases, SEM image contrast and imaging quality (SNR) in particular for low-dose imaging of organic materials is significantly improved.<br/><br/>As initial model system we study graphene, a sample for which we can correlate our spectral data with data from TEM-EELS measurements [3]. Although the energy resolution of this first DELTA prototype is limited (worse than about 5eV), we are able to correlate SEM spectral data to the known characteristic surface plasmon signal of free-standing graphene. This proves that backscattered electrons carry specific material information from their elastic and inelastic interaction with the sample surface.<br/><br/>Furthermore, we investigated improved contrast and minimized beam damage for imaging with ultra-low energy electrons on different organic samples. Such optimized imaging conditions are requirements for direct and dynamic imaging of organic materials. We show that the survival rate of Cy5- and Cy3- fluorescently labeled double-stranded DNA after electron irradiation increases drastically when decreasing the landing energy (90 % survival rate at 100 eV primary energy at an electron dose of 300 electrons per nm²). We are currently working on the prospect to image the excitation of organic fluorescence dyes within single layer DNA origami (~2 nm thick) directly via bsEELS at high resolution. For quantum dots in 3D-printed structures [4] we can already identify the fluorescence signal in the bsEELS spectrum.<br/><br/>In conclusion, SEM imaging with ultra-low electron energies provides a new, powerful characterization tool for the direct visualization and spectroscopy of organic materials. This is a prerequisite for further dynamic studies of such materials at nanometer resolution.<br/><br/>Acknowledgements: The authors thank Frederik Mayer and Martin Wegener (KIT, Karlsruhe, Germany) for Quantum Dot polymer samples [4]. Research funded by DFG (German Research Foundation) via the Excellence Cluster “3D Matter Made to Order” (EXC-2082/1-390761711).<br/><br/>[1] RR Schröder et al., Microsc. Microanal. 24 (Suppl 1), 2018<br/>[2] Zhang et al., Adv. Electron. Mater., 2100400, 2021<br/>[3] Wachsmuth et al, Physical Review B 90, 235434, 2014<br/>[4] Meyer et.al., Science Advances 5, 2, eaau9160, 2019