Mechanical and electrical properties of a nano-gap or how to play the nano-accordion

TL;DR

This study addresses how microstructure, particularly grain boundaries and crystallographic orientation, governs electrical and thermoelectric transport in a misfit-layered oxide, Ca2.93Sr0.07Co4O9, by introducing a nano-gap at a grain boundary that enables controlled mechanical bending via a differential heating current $I_H$. Using in-situ TEM with a custom thermoelectric chip, the authors elastically modulate the gap to form a nano-accordion and monitor both structural evolution and electrical response. They find a large intrinsic voltage offset associated with the gap that is strongly modulated by adsorbed molecules (notably water) in the gap region, with desorption under high vacuum heating reducing the offset; full-gap configurations show even larger offsets and sensitivity to vacuum history. The results highlight the importance of interior-gap surfaces and adsorbates in nanoscale conduction studies and offer a route to engineer nano-gaps for studying gas- or interfacial conduction, while cautioning about potential artifacts in in-situ TEM experiments due to mechanical stresses and surface chemistry.

Abstract

In-situ transmission electron microscopy (TEM) has become an important technique to study dynamic processes at highest spatial resolution and one branch is the investigation of phenomena related with electrical currents. Here, we present experimental results obtained from a peculiar in-situ TEM device, which was prepared with the aim to analyze the relationship between (thermo)electric properties and specific crystal orientations of a misfit layered compound. The formation of a nano-sized gap at a grain boundary facilitated a precisely controllable mechanical bending of the device by application of differential heating currents. The devices' electrical properties were found to be substantially influenced by the gap, leading to a high intrinsic voltage. This voltage additionally depends on the vacuum environment and on the history of applied heating currents. These findings are largely attributed to the presence of adsorbed molecules within the gap region. The electrical in-situ TEM studies of this work illustrate that interior surfaces can strongly influence electrical properties even under high vacuum conditions.

Figures (6)

• Figure 1: Preparation of the nano-accordion. (a) SEM image of the membrane area with the differential heating element and the contact pads of the in-situ chip, see (d) for a zoomed-in view of the pads. (b) SEM image of the CCO material reveals its sheet-like structure, the predominant c axis direction of the material has been indicated. A sketch of the MLC structure with Co (dark blue), Ca/Sr (light blue/green) and O (red) atoms has been added. The deposited Pt, highlighted in yellow, indicates the place from where the lamella was taken and its orientation with respect to the CCO sheets. Stage tilt is 52°. (c) SEM image of the thinned device on the in-situ chip with an incident angle of the electron beam of 5° with respect to the surface of the chip. The two Pt contact pads, the thick and thin parts of the lamella as well as the long hole in the membrane are visible. The inset shows the bridge-shaped form of the lamella after transfer to a Cu TEM grid. (d) Top view of the final in-situ device. Scale bars are (a) 200 $\upmu$m, (b) 10 $\upmu$m, (c) 2 $\upmu$m, width of inset SEM image is 15 $\upmu$m and (d) 5 $\upmu$m.
• Figure 2: (a) HAADF-STEM image of the nano-accordion at a stage tilt of 11.5° revealing the presence of a gap in the device. The inset STEM image shows the entire region of the thin lamella. The selected area for the SAED pattern shown in (b) and the acquisition position of the HRTEM image shown in (d) are indicated. (b) SAED pattern taken from the area marked in (a) revealing the horizontal orientation of the c axis of the material in the left part of the lamella. The sets of [00l] and [13l] reflections are named and the placement position of the objective aperture for acquisition of the BF and DF-TEM images shown in (e,f) is indicated. (c) HRSTEM image of the central gap region revealing the stack along the c-axis on its left side. A line profile taken from the marked area is shown in the inset revealing a stack periodicity of 1.1 nm and a gap width of approximately 5.5 nm. (d) HRTEM image of the end of the gap with grain boundary positions (blue lines) and the direction of the respective c-axis orientation (green arrows) marked. (e) BF- and (f) DF-TEM images of the thin part of the lamella, description see text. Scale bars are (a) 100 nm, inset image width is 2 $\upmu$m, (b) 2 nm-1. (c) 9 nm, (d) 10 nm, (e) 200 nm and (f) 300 nm.
• Figure 3: (a-c) BF-TEM images of the thin part of the device for different applied $I_H$ of (a) 0 mA, (b) 3 mA and (c) 5 mA. The movement of the bending contours due to temperature-induced bending is clearly visible. (d) Plot of the measured width of the lamella as indicated by the blue line in (b) over $I_H$. The non-linearly decreasing width due to thermal expansion and bending is clearly visible. Black crosses correspond to the rise of the current and blue boxes to the descent. Error bars correspond to the pixel size. (e) HAADF-STEM image of the device with $I_H$ = 5 mA. The two parts of the lamella overlap in the upper part due to temperature-induced bending, which leads to the increased contrast in the overlapping region. (f) SAED pattern acquired from the area marked by a blue circle in (c) for $I_H$ = 5 mA. Due to bending, the electron-beam direction is now far from the initial proximity to the [310] orientation. Scale bars are (a-c) 300 nm. (e) 100 nm and (f) 3 nm-1.
• Figure 4: (a) Comparison of three I-V curves for $I_H$ = 0 mA (blue solid line), $I_H$ = 5 mA (red line) and again at 0 mA after having heated with 5 mA (blue dashed line). The inset shows the curves at low voltages and the black squares mark the origin and the voltages where blue and red curves pass zero measured current. (b) Plot of the voltage offset (as marked by black squares in the inset in (a)) for three different conditions of the device: Directly after insertion in the microscope (black crosses and circles), after 4h in the microscope (green) and after retraction to the loadlock (yellow). (c) TEM image of the gap region directly after insertion showing a bright contrast in the gap region. Scale bar is 20 nm.
• Figure 5: (a) SEM image of the device with full gap. The circle indicates the position with minimum distance. (b) BF-TEM images of the device with fully broken gap, circle indicates position for SAED pattern acquisition. (c) HRTEM image of the upper gap revealing amorphous material in the gap region. (d) HRTEM image of the triple-phase region showing that in this area, the gap only exhibits minimum width. (e) BF-TEM image of the device at an applied heating current of $I_H$ = 5 mA. (f) SAED pattern of the unheated device. Scale bars are (a,b,e) 300 nm, (c) 5 nm, (d) 10 nm, (f) 4 nm-1.