Atomic-Level Structural Engineering of Graphene on a Mesoscopic Scale.

Structural engineering is the first step toward changing properties of materials. While this can be at relative ease done for bulk materials, for example, using ion irradiation, similar engineering of 2D materials and other low-dimensional structures remains a challenge. The difficulties range from the preparation of clean and uniform samples to the sensitivity of these structures to the overwhelming task of sample-wide characterization of the subjected modifications at the atomic scale. Here, we overcome these issues using a near ultrahigh vacuum system comprised of an aberration-corrected scanning transmission electron microscope and setups for sample cleaning and manipulation, which are combined with automated atomic-resolution imaging of large sample areas and a convolutional neural network approach for image analysis. This allows us to create and fully characterize atomically clean free-standing graphene with a controlled defect distribution, thus providing the important first step toward atomically tailored two-dimensional materials.

Hybrid pixel direct detector for electron energy loss spectroscopy.

We characterize a hybrid pixel direct detector and demonstrate its suitability for electron energy loss spectroscopy (EELS). The detector has a large dynamic range, narrow point spread function, detective quantum efficiency ≥ 0.8 even without single electron arrival discrimination, and it is resilient to radiation damage. It is capable of detecting ~5 × 106 electrons/pixel/second, allowing it to accommodate up to 0.8 pA per pixel and hence >100 pA EELS zero-loss peak (ZLP) without saturation, if the ZLP is spread over >125 pixels (in the non-dispersion direction). At the same time, it can reliably detect isolated single electrons in the high loss region of the spectrum. The detector uses a selectable threshold to exclude low energy events, and this results in essentially zero dark current and readout noise. Its maximum frame readout rate at 16-bit digitization is 2250 full frames per second, allowing for fast spectrum imaging. We show applications including EELS of boron nitride in which an unsaturated zero loss peak is recorded at the same time as inner shell loss edges, elemental mapping of an STO/BTO/LMSO multilayer, and efficient parallel acquisition of angle-resolved EEL spectra (S(q, ω)) of boron nitride.

Direct imaging of light-element impurities in graphene reveals triple-coordinated oxygen.

Along with hydrogen, carbon, nitrogen and oxygen are the arguably most important elements for organic chemistry. Due to their rich variety of possible bonding configurations, they can form a staggering number of compounds. Here, we present a detailed analysis of nitrogen and oxygen bonding configurations in a defective carbon (graphene) lattice. Using aberration-corrected scanning transmission electron microscopy and single-atom electron energy loss spectroscopy, we directly imaged oxygen atoms in graphene oxide, as well as nitrogen atoms implanted into graphene. The collected data allows us to compare nitrogen and oxygen bonding configurations, showing clear differences between the two elements. As expected, nitrogen forms either two or three bonds with neighboring carbon atoms, with three bonds being the preferred configuration. Oxygen, by contrast, tends to bind with only two carbon atoms. Remarkably, however, triple-coordinated oxygen with three carbon neighbors is also observed, a configuration that is exceedingly rare in organic compounds.

Electron-Beam Manipulation of Silicon Dopants in Graphene.

The direct manipulation of individual atoms in materials using scanning probe microscopy has been a seminal achievement of nanotechnology. Recent advances in imaging resolution and sample stability have made scanning transmission electron microscopy a promising alternative for single-atom manipulation of covalently bound materials. Pioneering experiments using an atomically focused electron beam have demonstrated the directed movement of silicon atoms over a handful of sites within the graphene lattice. Here, we achieve a much greater degree of control, allowing us to precisely move silicon impurities along an extended path, circulating a single hexagon, or back and forth between the two graphene sublattices. Even with manual operation, our manipulation rate is already comparable to the state-of-the-art in any atomically precise technique. We further explore the influence of electron energy on the manipulation rate, supported by improved theoretical modeling taking into account the vibrations of atoms near the impurities, and implement feedback to detect manipulation events in real time. In addition to atomic-level engineering of its structure and properties, graphene also provides an excellent platform for refining the accuracy of quantitative models and for the development of automated manipulation.

Insights into radiation damage from atomic resolution scanning transmission electron microscopy imaging of mono-layer CuPcCl16 films on graphene.

Atomically resolved images of monolayer organic crystals have only been obtained with scanning probe methods so far. On the one hand, they are usually prepared on surfaces of bulk materials, which are not accessible by (scanning) transmission electron microscopy. On the other hand, the critical electron dose of a monolayer organic crystal is orders of magnitudes lower than the one for bulk crystals, making (scanning) transmission electron microscopy characterization very challenging. In this work we present an atomically resolved study on the dynamics of a monolayer CuPcCl16 crystal under the electron beam as well as an image of the undamaged molecules obtained by low-dose electron microscopy. The results show the dynamics and the radiation damage mechanisms in the 2D layer of this material, complementing what has been found for bulk crystals in earlier studies. Furthermore, being able to image the undamaged molecular crystal allows the characterization of new composites consisting of 2D materials and organic molecules.

Software electron counting for low-dose scanning transmission electron microscopy.

The performance of the detector is of key importance for low-dose imaging in transmission electron microscopy, and counting every single electron can be considered as the ultimate goal. In scanning transmission electron microscopy, low-dose imaging can be realized by very fast scanning, however, this also introduces artifacts and a loss of resolution in the scan direction. We have developed a software approach to correct for artifacts introduced by fast scans, making use of a scintillator and photomultiplier response that extends over several pixels. The parameters for this correction can be directly extracted from the raw image. Finally, the images can be converted into electron counts. This approach enables low-dose imaging in the scanning transmission electron microscope via high scan speeds while retaining the image quality of artifact-free slower scans.

Computational insights and the observation of SiC nanograin assembly: towards 2D silicon carbide.

While an increasing number of two-dimensional (2D) materials, including graphene and silicene, have already been realized, others have only been predicted. An interesting example is the two-dimensional form of silicon carbide (2D-SiC). Here, we present an observation of atomically thin and hexagonally bonded nanosized grains of SiC assembling temporarily in graphene oxide pores during an atomic resolution scanning transmission electron microscopy experiment. Even though these small grains do not fully represent the bulk crystal, simulations indicate that their electronic structure already approaches that of 2D-SiC. This is predicted to be flat, but some doubts have remained regarding the preference of Si for sp 3 hybridization. Exploring a number of corrugated morphologies, we find completely flat 2D-SiC to have the lowest energy. We further compute its phonon dispersion, with a Raman-active transverse optical mode, and estimate the core level binding energies. Finally, we study the chemical reactivity of 2D-SiC, suggesting it is like silicene unstable against molecular absorption or interlayer linking. Nonetheless, it can form stable van der Waals-bonded bilayers with either graphene or hexagonal boron nitride, promising to further enrich the family of two-dimensional materials once bulk synthesis is achieved.

Buckyball sandwiches.

Two-dimensional (2D) materials have considerably expanded the field of materials science in the past decade. Even more recently, various 2D materials have been assembled into vertical van der Waals heterostacks, and it has been proposed to combine them with other low-dimensional structures to create new materials with hybridized properties. We demonstrate the first direct images of a suspended 0D/2D heterostructure that incorporates C60 molecules between two graphene layers in a buckyball sandwich structure. We find clean and ordered C60 islands with thicknesses down to one molecule, shielded by the graphene layers from the microscope vacuum and partially protected from radiation damage during scanning transmission electron microscopy imaging. The sandwich structure serves as a 2D nanoscale reaction chamber, allowing the analysis of the structure of the molecules and their dynamics at atomic resolution.

Automated Image Acquisition for Low-Dose STEM at Atomic Resolution.

Beam damage is a major limitation in electron microscopy that becomes increasingly severe at higher resolution. One possible route to circumvent radiation damage, which forms the basis for single-particle electron microscopy and related techniques, is to distribute the dose over many identical copies of an object. For the acquisition of low-dose data, ideally no dose should be applied to the region of interest before the acquisition of data. We present an automated approach that can collect large amounts of data efficiently by acquiring images in a user-defined area-of-interest with atomic resolution. We demonstrate that the stage mechanics of the Nion UltraSTEM, combined with an intelligent algorithm to move the sample, allow the automated acquisition of atomically resolved images from micron-sized areas of a graphene substrate. Moving the sample stage automatically in a regular pattern over the area-of-interest enables the collection of data from pristine sample regions without exposing them to the electron beam before recording an image. Therefore, it is possible to obtain data with minimal dose (no prior exposure during focusing), which is only limited by the minimum signal needed for data processing. This enables us to minimize beam-induced damage in the sample and to acquire large data sets within a reasonable amount of time.

Unraveling the 3D Atomic Structure of a Suspended Graphene/hBN van der Waals Heterostructure.

In this work we demonstrate that a free-standing van der Waals heterostructure, usually regarded as a flat object, can exhibit an intrinsic buckled atomic structure resulting from the interaction between two layers with a small lattice mismatch. We studied a freely suspended membrane of well-aligned graphene on a hexagonal boron nitride (hBN) monolayer by transmission electron microscopy (TEM) and scanning TEM (STEM). We developed a detection method in the STEM that is capable of recording the direction of the scattered electron beam and that is extremely sensitive to the local stacking of atoms. A comparison between experimental data and simulated models shows that the heterostructure effectively bends in the out-of-plane direction, producing an undulated structure having a periodicity that matches the moiré wavelength. We attribute this rippling to the interlayer interaction and also show how this affects the intralayer strain in each layer.