Search for: All records

Rapid identification of newly emerging or circulating viruses is an important first step toward managing the public health response to potential outbreaks. A portable virus capture device, coupled with label-free Raman spectroscopy, holds the promise of fast detection by rapidly obtaining the Raman signature of a virus followed by a machine learning (ML) approach applied to recognize the virus based on its Raman spectrum, which is used as a fingerprint. We present such an ML approach for analyzing Raman spectra of human and avian viruses. A convolutional neural network (CNN) classifier specifically designed for spectral data achieves very high accuracy for a variety of virus type or subtype identification tasks. In particular, it achieves 99% accuracy for classifying influenza virus type A versus type B, 96% accuracy for classifying four subtypes of influenza A, 95% accuracy for differentiating enveloped and nonenveloped viruses, and 99% accuracy for differentiating avian coronavirus (infectious bronchitis virus [IBV]) from other avian viruses. Furthermore, interpretation of neural net responses in the trained CNN model using a full-gradient algorithm highlights Raman spectral ranges that are most important to virus identification. By correlating ML-selected salient Raman ranges with the signature ranges of known biomolecules and chemical functional groups—formore »example, amide, amino acid, and carboxylic acid—we verify that our ML model effectively recognizes the Raman signatures of proteins, lipids, and other vital functional groups present in different viruses and uses a weighted combination of these signatures to identify viruses.« less

Individual atomic defects in 2D materials impact their macroscopic functionality. Correlating the interplay is challenging, however, intelligent hyperspectral scanning tunneling spectroscopy (STS) mapping provides a feasible solution to this technically difficult and time consuming problem. Here, dense spectroscopic volume is collected autonomously via Gaussian process regression, where convolutional neural networks are used in tandem for spectral identification. Acquired data enable defect segmentation, and a workflow is provided for machine-driven decision making during experimentation with capability for user customization. We provide a means towards autonomous experimentation for the benefit of both enhanced reproducibility and user-accessibility. Hyperspectral investigations on WS₂sulfur vacancy sites are explored, which is combined with local density of states confirmation on the Au{111} herringbone reconstruction. Chalcogen vacancies, pristine WS₂, Au face-centered cubic, and Au hexagonal close-packed regions are examined and detected by machine learning methods to demonstrate the potential of artificial intelligence for hyperspectral STS mapping.

Molecular insights into graphene-catalyst surface interactions can provide useful information for the efficient design of copper current collectors with graphitic anode interfaces. As graphene bending can affect the local electron density, it should reflect its local reactivity as well. Using ReaxFF reactive molecular simulations, we have investigated the possible bending of graphene in vacuum and near copper surfaces. We describe the energy cost for graphene bending and the binding energy with hydrogen and copper with two different ReaxFF parameter sets, demonstrating the relevance of using the more recently developed ReaxFF parameter sets for graphene properties. Moreover, the draping angle at copper step edges obtained from our atomistic simulations is in good agreement with the draping angle determined from experimental measurements, thus validating the ReaxFF results.

Realization of wafer-scale single-crystal films of transition metal dichalcogenides (TMDs) such as WS2 requires epitaxial growth and coalescence of oriented domains to form a continuous monolayer. The domains must be oriented in the same crystallographic direction on the substrate to inhibit the formation of inversion domain boundaries (IDBs), which are a common feature of layered chalcogenides. Here we demonstrate fully coalesced unidirectional WS2 monolayers on 2 in. diameter c-plane sapphire by metalorganic chemical vapor deposition using a multistep growth process to achieve epitaxial WS2 monolayers with low in-plane rotational twist (0.09°). Transmission electron microscopy analysis reveals that the WS2 monolayers are largely free of IDBs but instead have translational boundaries that arise when WS2 domains with slightly offset lattices merge together. By regulating the monolayer growth rate, the density of translational boundaries and bilayer coverage were significantly reduced. The unidirectional orientation of domains is attributed to the presence of steps on the sapphire surface coupled with growth conditions that promote surface diffusion, lateral domain growth, and coalescence while preserving the aligned domain structure. The transferred WS2 monolayers show neutral and charged exciton emission at 80 K with negligible defect-related luminescence. Back-gated WS2 field effect transistors exhibited an ION/OFF of ∼107more »and mobility of 16 cm2/(V s). The results demonstrate the potential of achieving wafer-scale TMD monolayers free of inversion domains with properties approaching those of exfoliated flakes.« less