Bevel-edge epitaxy of ferroelectric rhombohedral boron nitride single crystal.

Within the family of two-dimensional dielectrics, rhombohedral boron nitride (rBN) is considerably promising owing to having not only the superior properties of hexagonal boron nitride1-4-including low permittivity and dissipation, strong electrical insulation, good chemical stability, high thermal conductivity and atomic flatness without dangling bonds-but also useful optical nonlinearity and interfacial ferroelectricity originating from the broken in-plane and out-of-plane centrosymmetry5-23. However, the preparation of large-sized single-crystal rBN layers remains a challenge24-26, owing to the requisite unprecedented growth controls to coordinate the lattice orientation of each layer and the sliding vector of every interface. Here we report a facile methodology using bevel-edge epitaxy to prepare centimetre-sized single-crystal rBN layers with exact interlayer ABC stacking on a vicinal nickel surface. We realized successful accurate fabrication over a single-crystal nickel substrate with bunched step edges of the terrace facet (100) at the bevel facet (110), which simultaneously guided the consistent boron-nitrogen bond orientation in each BN layer and the rhombohedral stacking of BN layers via nucleation near each bevel facet. The pure rhombohedral phase of the as-grown BN layers was verified, and consequently showed robust, homogeneous and switchable ferroelectricity with a high Curie temperature. Our work provides an effective route for accurate stacking-controlled growth of single-crystal two-dimensional layers and presents a foundation for applicable multifunctional devices based on stacked two-dimensional materials.

Epitaxy of wafer-scale single-crystal MoS2 monolayer via buffer layer control.

Monolayer molybdenum disulfide (MoS2), an emergent two-dimensional (2D) semiconductor, holds great promise for transcending the fundamental limits of silicon electronics and continue the downscaling of field-effect transistors. To realize its full potential and high-end applications, controlled synthesis of wafer-scale monolayer MoS2 single crystals on general commercial substrates is highly desired yet challenging. Here, we demonstrate the successful epitaxial growth of 2-inch single-crystal MoS2 monolayers on industry-compatible substrates of c-plane sapphire by engineering the formation of a specific interfacial reconstructed layer through the S/MoO3 precursor ratio control. The unidirectional alignment and seamless stitching of MoS2 domains across the entire wafer are demonstrated through cross-dimensional characterizations ranging from atomic- to centimeter-scale. The epitaxial monolayer MoS2 single crystal shows good wafer-scale uniformity and state-of-the-art quality, as evidenced from the ~100% phonon circular dichroism, exciton valley polarization of ~70%, room-temperature mobility of ~140 cm2v-1s-1, and on/off ratio of ~109. Our work provides a simple strategy to produce wafer-scale single-crystal 2D semiconductors on commercial insulator substrates, paving the way towards the further extension of Moore's law and industrial applications of 2D electronic circuits.

Tracking cubic ice at molecular resolution.

Ice is present everywhere on Earth and has an essential role in several areas, such as cloud physics, climate change and cryopreservation. The role of ice is determined by its formation behaviour and associated structure. However, these are not fully understood1. In particular, there is a long-standing debate about whether water can freeze to form cubic ice-a currently undescribed phase in the phase space of ordinary hexagonal ice2-6. The mainstream view inferred from a collection of laboratory data attributes this divergence to the inability to discern cubic ice from stacking-disordered ice-a mixture of cubic and hexagonal sequences7-11. Using cryogenic transmission electron microscopy combined with low-dose imaging, we show here the preferential nucleation of cubic ice at low-temperature interfaces, resulting in two types of separate crystallization of cubic ice and hexagonal ice from water vapour deposition at 102 K. Moreover, we identify a series of cubic-ice defects, including two types of stacking disorder, revealing the structure evolution dynamics supported by molecular dynamics simulations. The realization of direct, real-space imaging of ice formation and its dynamic behaviour at the molecular level provides an opportunity for ice research at the molecular level using transmission electron microscopy, which may be extended to other hydrogen-bonding crystals.

3D multi-scale deep convolutional neural networks for pulmonary nodule detection.

With the rapid development of big data and artificial intelligence technology, computer-aided pulmonary nodule detection based on deep learning has achieved some successes. However, the sizes of pulmonary nodules vary greatly, and the pulmonary nodules have visual similarity with structures such as blood vessels and shadows around pulmonary nodules, which make the quick and accurate detection of pulmonary nodules in CT image still a challenging task. In this paper, we propose two kinds of 3D multi-scale deep convolution neural networks for nodule candidate detection and false positive reduction respectively. Among them, the nodule candidate detection network consists of two parts: 1) the backbone network part Res2SENet, which is used to extract multi-scale feature information of pulmonary nodules, it is composed of the multi-scale Res2Net modules of multiple available receptive fields at a granular level and the squeeze-and-excitation units; 2) the detection part, which uses a region proposal network structure to determine region candidates, and introduces context enhancement module and spatial attention module to improve detection performance. The false positive reduction network, also composed of the multi-scale Res2Net modules and the squeeze-and-excitation units, can further classify the nodule candidates generated by the nodule candidate detection network and screen out the ground truth positive nodules. Finally, the prediction probability generated by the nodule candidate detection network is weighted average with the prediction probability generated by the false positive reduction network to obtain the final results. The experimental results on the publicly available LUNA16 dataset showed that the proposed method has a superior ability to detect pulmonary nodules in CT images.

Development of in situ optical spectroscopy with high temporal resolution in an aberration-corrected transmission electron microscope.

Exploring the corresponding relation between structural and physical properties of materials at the atomic scale remains the fundamental problem in science. With the development of the aberration-corrected transmission electron microscopy (AC-TEM) and the ultrafast optical spectroscopy technique, sub-angstrom-scale spatial resolution and femtosecond-scale temporal resolution can be achieved, respectively. However, the attempt to combine both their advantages is still a great challenge. Here, we develop in situ optical spectroscopy with high temporal resolution in AC-TEM by utilizing a self-designed and manufactured TEM specimen holder, which has the capacity of sub-angstrom-scale spatial resolution and femtosecond-scale temporal resolution. The key and unique design of our apparatus is the use of the fiber bundle, which enables the delivery of focused pulse beams into TEM and collection of optical response simultaneously. The generated focused spot has a size less than 2 µm and can be scanned in plane with an area larger than 75 × 75 µm2. Most importantly, the positive group-velocity dispersion caused by glass fiber is compensated by a pair of diffraction gratings, thus resulting in the generation of pulse beams with a pulse width of about 300 fs (@ 3 mW) in TEM. The in situ experiment, observing the atomic structure of CdSe/ZnS quantum dots in AC-TEM and obtaining the photoluminescence lifetime (∼4.3 ns) in the meantime, has been realized. Further ultrafast optical spectroscopy with femtosecond-scale temporal resolution could be performed in TEM by utilizing this apparatus.