We propose a simple, but efficient and accurate machine learning (ML) model for developing high-dimensional potential energy surface. This so-called embedded atom neural network (EANN) approach is inspired by the well-known empirical embedded atom method (EAM) model used in condensed phase. It simply replaces the scalar embedded atom density in EAM with a Gaussian-type orbital based density vector, and represents the complex relationship between the embedded density vector and atomic energy by neural networks. We demonstrate that the EANN approach is equally accurate as several established ML models in representing both big molecular and extended periodic systems, yet with much fewer parameters and configurations. It is highly efficient as it implicitly contains the three-body information without an explicit sum of the conventional costly angular descriptors. With high accuracy and efficiency, EANN potentials can vastly accelerate molecular dynamics and spectroscopic simulations in complex systems at ab initio level. TOC graphic3 Accurate and efficient interaction potential energy surfaces (PESs) are crucial for spectroscopic, molecular dynamics, and thermodynamics simulations 1 . Traditional empirical or semi-empirical force field models such as the embedded atom method (EAM) 2-3 , while physically meaningful and highly efficient, are severely limited by their accuracy. More recently, tremendous efforts have been devoted to developing machine learning (ML) based PESs 4 , which are capable of representing a large set of ab initio data more accurately on chemical properties 5 , molecules 6-13 , gas phase and surface reactions 14-18 , and materialsError! Bookmark not defined. 10,[19][20][21][22] .Since known ML techniques in computer science themselves do not recognize the intrinsic symmetry of a chemical system, it is essential to design a ML representation for a PES that preserves rotational, translational, and permutational symmetry in an accurate and efficient way 23 . In this regard, permutation invariant polynomials (PIPs) in terms of internuclear distances 24 were used as the input of neural networks (NNs) 14 and Gaussian process regression (GPR) 25 in low-dimensional systems. For highdimensional problems, Behler and Parrinello 26 first handcrafted a set of atom centered symmetry functions (ACSFs) 23 as the input of atomistic neural networks (AtNN), which were later improved in various ways 9, 27 . More recently, the deep learning molecular dynamics (DPMD) model 10 and its symmetrized edition (DeepPot-SE) 28 map the coordinate matrix to a multi-output NN making descriptors themselves self-adapted in training. Deep tensor NN (DTNN) model 11 utilizes a vector of nuclear charges and an inter-atomic distance matrix as descriptors and its variant SchNet model 12 further introduces a continuous-filter convolutional layer to extract features from these
Machine learning has revolutionized the highdimensional representations for molecular properties such as potential energy. However, there are scarce machine learning models targeting tensorial properties, which are rotationally covariant. Here, we propose tensorial neural network (NN) models to learn both tensorial response and transition properties in which atomic coordinate vectors are multiplied with scalar NN outputs or their derivatives to preserve the rotationally covariant symmetry. This strategy keeps structural descriptors symmetry invariant so that the resulting tensorial NN models are as efficient as their scalar counterparts. We validate the performance and universality of this approach by learning response properties of water oligomers and liquid water and transition dipole moment of a model structural unit of proteins. Machine-learned tensorial models have enabled efficient simulations of vibrational spectra of liquid water and ultraviolet spectra of realistic proteins, promising feasible and accurate spectroscopic simulations for biomolecules and materials.
The van der Waals (vdW) materials offer an opportunity to build all-two-dimensional (all-2D) spintronic devices with high-quality interfaces regardless of the lattice mismatch. Here, we report on an all-2D vertical spin valve that combines a typical layered semiconductor MoS 2 with vdW ferromagnetic metal Fe 3 GeTe 2 (FGT) flakes. The linear current−voltage curves illustrate that Ohmic contacts are formed in FGT/MoS 2 interfaces, while the temperature dependence of the junction resistance further demonstrates that the MoS 2 interlayer acts as a conducting layer instead of a tunneling layer. In addition, the magnitude of the magnetoresistance (MR) of 3.1% at 10 K is observed, which is around 8 times larger than that of the reported spin valves based on MoS 2 sandwiched by conventional ferromagnetic electrodes. The MR decreasing monotonically with increasing temperature follows the Bloch's law. As the bias current decreases exponentially, the MR increases linearly up to a maximum value of 4.1%. Our results reveal the potential opportunities of vdW heterostructures for developing novel spintronic devices.
The activator protein 1 (AP-1) transcription factor c-Jun is crucial for neuronal apoptosis. However, c-Jun dimerization partners and the regulation of these proteins in neuronal apoptosis remain unknown. Here we report that c-Jun-mediated neuronal apoptosis requires the concomitant activation of activating transcription factor-2 (ATF2) and downregulation of c-Fos. Furthermore, we have observed that c-Jun predominantly heterodimerizes with ATF2 and that the c-Jun/ATF2 complex promotes apoptosis by triggering ATF activity. Inhibition of c-Jun/ATF2 heterodimerization using dominant negative mutants, small hairpin RNAs, or decoy oligonucleotides was able to rescue neurons from apoptosis, whereas constitutively active ATF2 and c-Jun mutants were found to synergistically stimulate apoptosis.
2D layered chalcogenide semiconductors have been proposed as a promising class of materials for low‐dimensional electronic, optoelectronic, and spintronic devices. Here, all‐2D van der Waals vertical spin‐valve devices, that combine the 2D layered semiconductor InSe as a spacer with the 2D layered ferromagnetic metal Fe3GeTe2 as spin injection and detection electrodes, are reported. Two distinct transport behaviors are observed: tunneling and metallic, which are assigned to the formation of a pinhole‐free tunnel barrier at the Fe3GeTe2/InSe interface and pinholes in the InSe spacer layer, respectively. For the tunneling device, a large magnetoresistance (MR) of 41% is obtained under an applied bias current of 0.1 µA at 10 K, which is about three times larger than that of the metallic device. Moreover, the tunneling device exhibits a lower operating bias current but a more sensitive bias current dependence than the metallic device. The MR and spin polarization of both the metallic and tunneling devices decrease with increasing temperature, which can be fitted well by Bloch's law. These findings reveal the critical role of pinholes in the MR of all‐2D van der Waals ferromagnet/semiconductor heterojunction devices.
Atomically thin layers of van der Waals (vdW) crystals offer an ideal material platform to realize tunnel field-effect transistors (TFETs) that exploit the tunneling of charge carriers across the forbidden gap of a vdW heterojunction. This type of device requires a precise energy band alignment of the different layers of the junction to optimize the tunnel current. Among 2D vdW materials, black phosphorus (BP) and indium selenide (InSe) have a Brillouin zone-centered conduction and valence bands, and a type II band offset, both ideally suited for band-to-band tunneling. TFETs based on BP/InSe heterojunctions with diverse electrical transport characteristics are demonstrated: forward rectifying, Zener tunneling, and backward rectifying characteristics are realized in BP/InSe junctions with different thickness of the BP layer or by electrostatic gating of the junction. Electrostatic gating yields a large on/off current ratio of up to 10 8 and negative differential resistance at low applied voltages (V ≈ 0.2 V). These findings illustrate versatile functionalities of TFETs based on BP and InSe, offering opportunities for applications of these 2D materials beyond the device architectures reported in the current literature.challenges require a shift from traditional approaches toward transformative material systems and integration technologies. [1,2] Atomically thin layers of van der Waals (vdW) crystals and their heterostructures, [3][4][5] generally referred to as 2D materials, offer opportunities to study and exploit quantum phenomena for a wide range of applications. [6][7][8][9][10] These crystals have strong covalent atomic bonding in the 2D planes and weak vdW interaction between the layers, which enable the fabrication of stable thin films down to the atomic monolayer thickness and stack them into multilayered heterostructures. [11][12][13] The science of these 2D systems is developing rapidly with important technological breakthroughs emerging from recent studies.Among the extended family of 2D systems, the metal chalcogenide indium selenide (InSe) [14][15][16][17][18] and the elemental compound black phosphorous (BP) [19][20][21][22][23][24] have received increasing attention. These two semiconductors have electronic properties distinct from those of other 2D materials, such as transition metal dichalcogenides (TMDCs), including a higher electron mobility beneficial for field-effect transistors
tunable bandgap energy (from E g = 1.1 to 2.1 eV) and strong light absorption offer opportunities for a variety of optoelectronic devices. [2][3][4] Amongst the TMDs, MoTe 2 is an attractive semiconductor. In the monolayer form, it has a direct bandgap, E g = 1.10 eV at room temperature, larger than that of bulk MoTe 2 , which has an indirect bandgap (E g = 0.85 eV). [5][6][7] Thus, unlike other TMDs, such as MoS 2 and WS 2 , photodetectors based on MoTe 2 can have a broadband photoresponse that extends from the visible (VIS) to the near infrared (NIR) spectral range. [8][9][10] In particular, in MoTe 2 -based field effect transistors (FETs), the photoresponsivity (R) can be enhanced by a photogating effect and achieve values of up to R = 24 mA W −1 under illumination with NIR light. [9] Although Si has a similar bandgap to that of MoTe 2 , its absorption coefficient in the NIR spectral range is smaller than that of MoTe 2 : for Si, the absorption coefficient is 8.17 cm −1 at λ = 1064 nm, which is smaller than that for MoTe 2 (4.9 × 10 4 cm −1 ). [11] In contrast to traditional bulk semiconductors such as Si, Ge, or III-V compounds, 2D vdW crystals have surfaces that are free of dangling bonds. [2] This unique feature arises from their atomic structure: the atoms are arranged into layers that are held together by strong covalent in-plane bonds; in contrast, in the out-of-plane direction, the atomic layers interact with weak vdW interactions. This offers opportunities to combine them with other materials without the limitations of lattice mismatch that apply to covalent crystals. [2,12] For example, MoTe 2 has been used in different multilayer structures: in MoTe 2 /MoS 2 heterojunctions, the on/off photocurrent ratio can reach values of about 780; [13] also, the photoconductive gain in MoTe 2 /graphene heterostructures can be as large as 4.69 × 10 8 . [9] More generally, asymmetric contact barriers between two electrodes and a 2D vdW crystal can be exploited to construct high performance photodetectors: [14][15][16][17][18][19][20][21] Au and In Schottky contacts to a 2D material can be used to realize self-powered photodetectors with high photoresponsivity (R = 110 mA W −1 ). [14] Also, graphene can form a clean interface with 2D materials and its near perfect optical transparency makes it suitable for use as the top electrode of vertical heterostructure photodetectors. [22][23][24][25] Au/MoTe 2 /graphene vertical heterostructures have good photoresponsivity and photoresponse time of about 96 ms. [15] However, the photoresponse of 2D vdW heterostructure devices in the current literature remain still slow due to relatively long optically active Atomically thin 2D materials are promising candidates for miniaturized highperformance optoelectronic devices. This study reports on multilayer MoTe 2 photodetectors contacted with asymmetric electrodes based on n-and p-type graphene layers. The asymmetry in the graphene contacts creates a large (E bi ∼ 100 kV cm −1 ) built-in electric field across the short (l = 15 nm) M...
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.