Development of Highly Sensitive Raman Spectroscopy for Subnano and Single-Atom Detection

Direct detection and characterisation of small materials are fundamental challenges in analytical chemistry. A particle composed of dozens of metallic atoms, a so-called subnano-particle (SNP), and a single-atom catalyst (SAC) are ultimate analysis targets in terms of size, and the topic is now attracting increasing attention as innovative frontier materials in catalysis science. However, characterisation techniques for the SNP and SAC adsorbed on substrates requires sophisticated and large-scale analytical facilities. Here we demonstrate the development of an ultrasensitive, laboratory-scale, vibrational spectroscopic technique to characterise SNPs and SACs. The fine design of nano-spatial local enhancement fields generated by the introduction of anisotropic stellate-shaped signal amplifiers expands the accessibility of small targets on substrates into evanescent electromagnetic fields, achieving not only the detection of isolated small targets but also revealing the effects of intermolecular/interatomic interactions within the subnano configuration under actual experimental conditions. Such a development of “in situ subnano spectroscopy” will facilitate a comprehensive understanding of subnano and SAC science.

TEMImageNet training library and AtomSegNet deep-learning models for high-precision atom segmentation, localization, denoising, and deblurring of atomic-resolution images

Atom segmentation and localization, noise reduction and deblurring of atomic-resolution scanning transmission electron microscopy (STEM) images with high precision and robustness is a challenging task. Although several conventional algorithms, such has thresholding, edge detection and clustering, can achieve reasonable performance in some predefined sceneries, they tend to fail when interferences from the background are strong and unpredictable. Particularly, for atomic-resolution STEM images, so far there is no well-established algorithm that is robust enough to segment or detect all atomic columns when there is large thickness variation in a recorded image. Herein, we report the development of a training library and a deep learning method that can perform robust and precise atom segmentation, localization, denoising, and super-resolution processing of experimental images. Despite using simulated images as training datasets, the deep-learning model can self-adapt to experimental STEM images and shows outstanding performance in atom detection and localization in challenging contrast conditions and the precision consistently outperforms the state-of-the-art two-dimensional Gaussian fit method. Taking a step further, we have deployed our deep-learning models to a desktop app with a graphical user interface and the app is free and open-source. We have also built a TEM ImageNet project website for easy browsing and downloading of the training data.

Nondestructive microwave detection of a coherent quantum dynamics in cold atoms

Cold atom quantum sensors based on atom interferometry are among the most accurate instruments used in fundamental physics, metrology, and foreseen for autonomous inertial navigation. However, they typically have optically complex, cumbersome, and low-bandwidth atom detection systems, limiting their practical applications. Here, we demonstrate an enabling technology for high-bandwidth, compact, and nondestructive detection of cold atoms, using microwave radiation. We measure the reflected microwave signal to coherently and distinctly detect the population of single quantum states with a bandwidth close to 30 kHz and a design destructivity that we set to 0.04%. We use a horn antenna and free-falling molasses cooled atoms in order to demonstrate the feasibility of this technique in conventional cold atom interferometers. This technology, combined with coplanar waveguides used as microwave sources, provides a basic design building block for detection in future atom chip-based compact quantum inertial sensors.