Seeing the Invisible: Ultrathin (UT) Membrane Chip for Fluidic-Cell Electron Microscopy

In the recent decade, <i>in-situ</i> or operando S/TEM utilizing SiNx membrane encapsulated chips to confine fluids for electron microscopy examination has become popular, A great number of prior innovators have shown this to be an effective approach for probing fluid-surface/nanostructure interactions and related phenomena or reactions. Such a “closed cell” chip based on silicon nitride (SiN<i>_x</i>) membranes as electron transparent encapsulation material, has many practical and technological advantages over the differential pumping environmental TEM (ETEM). Unfortunately, however, conventional fluid-cells suffer from additional and significant electron scattering from the top and the bottom membranes, which are typically 30-50nm thick to maintain integrity/stability during the operation. Thus, the total thickness of &gt;70-80 nm of the encapsulating membranes imposes many adverse effects on the post electron optics, such as increased chromatic aberrations. This naturally results in significant degeneration of signal quality and loss of spatiotemporal resolution, diffused interference in the electron diffraction, and plasmon-dominated electron energy loss spectra (EELS).<br/>Further, to implement advanced STEM techniques such as quantitative EELS analysis to resolve electronic structure and 4D-STEM for pixel-specific acquisition, one needs to significantly reduce electron scattering, which primarily stem from the thick SiN<i>_x</i> windows in gas-cells. For example, the log-ratio of scattered electron over zero-loss electron () already exceeds ~1 for two-50 nm SiN<i>_x</i> encapsulation (without any specimen). Howwever, low scattering “thin” membrane is risky since the mechanical robustness is compromised, resulting in potentially catastrophic failure. Thus, novel design strategies for fabrication of stable ultrathin SiN<i>_x</i> membrane are needed.<br/>We have recently developed a robust, functional and scalable backing support strategy to enable the thinnest possible (&lt;10 nm) SiN<i>_x</i> gas encapsulation material [2]. Inspired by the natural honeycomb geometry, our novel design provides for honeycomb backbone that can neatly anchor ultrathin (~&lt;10 nm) SiNx membrane with excellent stability and consistent performance. It can still withstand up to <i>6 Atm</i> pressure with ~50 % less bulging. Unlike graphene-based encapsulations [3], stability under the electron beam is comparable to a 50 nm SiN<i>_x</i> membrane, which is sufficient for most high-resolution S/TEM applications on non-electron sensitive materials.<br/>We show that our UT chip increases contrast of typical nanoparticles at <i>1 atm</i> Ar gas by ~70 % and the accessible information limit is enhanced by &gt;130 % compared to the conventional encapsulation. More importantly, the is reduced from nominally ~1.0 to 0.3 using a <i>1 Atm</i> gas cell. This greatly enhances spectral visibility and significantly improved S/N for EELS excitations. Thus, spatiotemporal detection of gas species, down to ~nanometer scale, which otherwise is unachievable with integrated residual gas analyser (RGA), is being achieved.<br/>The presentation will cover the design and implementation of UT membrane fluid-cell for in-situ gas-solid interactions. [4]<br/><br/>References: [1] K Koo, SM Ribet, C Zhang, PJM Smeets, Rd Reis, X Hu, and VP Dravid, Nano Lett <b>22</b> (2022), p. 4137. doi: 10.1021/acs.nanolett.2c00893; [2] VP Dravid, X Hu, and K Koo, US Provisional Patent, No. 63413097 (2022); [3] K Koo, J Park, S Ji, S Toleukhanova, and JM Yuk, Adv Mater <b>33</b> (2021), p. 2005468. doi: 10.1002/adma.202005468; [4] <i>Acknowledgement:</i> This work is made use of the EPIC facility of Northwestern University’s NU<i>ANCE</i> Center, which has received support from the SHyNE Resource (NSF ECCS-2025633), the IIN, and Northwestern’s MRSEC program (NSF DMR-1720139).<br/><i>Disclosure: </i>US Provisional Patent Application (No. 63413097) regarding this work is filed on 04-Oct-2022.