Hengyang Normal University

About: Hengyang Normal University is a education organization based out in Hengyang, China. It is known for research contribution in the topics: Graphene & Adsorption. The organization has 1087 authors who have published 1280 publications receiving 13850 citations. The organization is also known as: Hengyang Teachers' College & Héngyáng Shīfàn Xuéyuàn.

Fabrication of Amine-Modified Magnetite-Electrochemically Reduced Graphene Oxide Nanocomposite Modified Glassy Carbon Electrode for Sensitive Dopamine Determination

Abstract: Amine-modified magnetite (NH2–Fe3O4)/reduced graphene oxide nanocomposite modified glassy carbon electrodes (NH2–Fe3O4/RGO/GCEs) were developed for the sensitive detection of dopamine (DA). The NH2-Fe3O4/RGO/GCEs were fabricated using a drop-casting method followed by an electrochemical reduction process. The surface morphologies, microstructure and chemical compositions of the NH2–Fe3O4 nanoparticles (NPs), reduced graphene oxide (RGO) sheets and NH2–Fe3O4/RGO nanocomposites were characterized by scanning electron microscopy (SEM), transmission electron microscopy (TEM), X-Ray diffraction (XRD) and Fourier-transform infrared (FTIR) spectroscopy. The electrochemical behaviors of DA on the bare and modified GCEs were investigated in phosphate buffer solution (PBS) by cyclic voltammetry (CV). Compared with bare electrode and RGO/GCE, the oxidation peak current (ipa) on the NH2–Fe3O4/RGO/GCE increase significantly, owing to the synergistic effect between NH2–Fe3O4 NPs and RGO sheets. The oxidation peak currents (ipa) increase linearly with the concentrations of DA in the range of 1 × 10−8 mol/L – 1 × 10−7 mol/L, 1 × 10−7 mol/L – 1 × 10−6 mol/L and 1 × 10−6 mol/L – 1 × 10−5 mol/L. The detection limit is (4.0 ± 0.36) ×10−9 mol/L (S/N = 3). Moreover, the response peak currents of DA were hardly interfered with the coexistence of ascorbic acid (AA) and uric acid (UA). The proposed NH2–Fe3O4/RGO/GCE is successfully applied to the detection of dopamine hydrochloride injections with satisfactory results. Together with low cost, facile operation, good selectivity and high sensitivity, the NH2–Fe3O4/RGO/GCEs have tremendous prospects for the detection of DA in various real samples.

Photonic spin Hall effect enabled refractive index sensor using weak measurements.

Abstract: In this work, we theoretically propose an optical biosensor (consists of a BK7 glass, a metal film, and a graphene sheet) based on photonic spin Hall effect (SHE). We establish a quantitative relationship between the spin-dependent shift in photonic SHE and the refractive index of sensing medium. It is found that, by considering the surface plasmon resonance effect, the refractive index variations owing to the adsorption of biomolecules in sensing medium can effectively change the spin-dependent displacements. Remarkably, using the weak measurement method, this tiny spin-dependent shifts can be detected with a desirable accuracy so that the corresponding biomolecules concentration can be determined.

Facile synthesis of Ag@Cu2O heterogeneous nanocrystals decorated N-doped reduced graphene oxide with enhanced electrocatalytic activity for ultrasensitive detection of H2O2

Abstract: A nanohybrid composed of Ag@Cu2O heterogeneous nanocrystals supported on N-doped reduced graphene oxide (Ag@Cu2O/N-RGO) has been synthesized by a simple wet-chemical method. The resultant composite consists of N-RGO sheets fully and homogeneously coated with a dense layer of Ag@Cu2O nanocrystals. Both Ag and N-RGO are in direct contact with Cu2O, and Ag nanoparticles with sizes of 2–5 nm are mainly deposited on the surface of Cu2O cubes (edge length of 500 nm). The electrochemical studies reveal that the ternary Ag@Cu2O/N-RGO composite exhibit significantly enhanced electrocatalytic activity for H2O2 sensing compared with either the single component (N-RGO) or two component systems (Cu2O/N-RGO and Ag/N-RGO), which is mainly due to the synergetic catalysis of the ternary system. The nonenzymatic sensor based on Ag@Cu2O/N-RGO composite shows overwhelmingly superior comprehensive performance for the H2O2 detection over the documented Ag-based sensors. More specifically, it displays a rapid response (10 s) to H2O2 over a wide linear range of 54–700 nM with a high sensitivity of 1298.3 μA mM−1 cm−2 and a low detection limit of 10 nM. Moreover, the sensor also exhibits the preferable selectivity in the presence of biologically coactive compounds accompanied with long-term stability and good reproducibility.

Two-dimensional van der Waals electrical contact to monolayer MoSi2N4

Abstract: A two-dimensional (2D) MoSi2N4 monolayer is an emerging class of air-stable 2D semiconductors possessing exceptional electrical and mechanical properties. Despite intensive recent research effort devoted to uncover the material properties of MoSi2N4, the physics of electrical contacts to MoSi2N4 remains largely unexplored thus far. In this work, we study van der Waals heterostructures composed of MoSi2N4 contacted by graphene and NbS2 monolayers using first-principles density functional theory calculations. We show that the MoSi2N4/NbS2 contact exhibits an ultralow Schottky barrier height (SBH), which is beneficial for nanoelectronics applications. For the MoSi2N4/graphene contact, the SBH can be modulated via the interlayer distance or via external electric fields, thus opening up an opportunity for reconfigurable and tunable nanoelectronic devices. Our findings provide insights into the physics of 2D electrical contacts to MoSi2N4 and shall offer a critical first step toward the design of high-performance electrical contacts to MoSi2N4-based 2D nanodevices.

Frankenstein: Learning Deep Face Representations Using Small Data

Abstract: Deep convolutional neural networks have recently proven extremely effective for difficult face recognition problems in uncontrolled settings. To train such networks, very large training sets are needed with millions of labeled images. For some applications, such as near-infrared (NIR) face recognition, such large training data sets are not publicly available and difficult to collect. In this paper, we propose a method to generate very large training data sets of synthetic images by compositing real face images in a given data set. We show that this method enables to learn models from as few as 10 000 training images, which perform on par with models trained from 500 000 images. Using our approach, we also obtain state-of-the-art results on the CASIA NIR-VIS2.0 heterogeneous face recognition data set.