arXiv subjects

Kai Xiao

Publications and source records attributed to Kai Xiao.

84 records · Page 5Linked to original sources

arXiv2018

Deep Learning Analysis of Defect and Phase Evolution During Electron Beam Induced Transformations in WS2

Understanding elementary mechanisms behind solid-state phase transformations and reactions is the key to optimizing desired functional properties of many technologically relevant materials. Recent advances in scanning transmission electron microscopy (STEM) allow the real-time visualization of solid-state transformations in materials, including those induced by an electron beam and temperature, with atomic resolution. However, despite the ever-expanding capabilities for high-resolution data acquisition, the inferred information about kinetics and thermodynamics of the process and single defect dynamics and interactions is minima, due to the inherent limitations of manual ex-situ analysis of the collected volumes of data. To circumvent this problem, we developed a deep learning framework for dynamic STEM imaging that is trained to find the structures (defects) that break a crystal lattice periodicity and apply it for mapping solid state reactions and transformations in layered WS2 doped with Mo. This framework allows extracting thousands of lattice defects from raw STEM data (single images and movies) in a matter of seconds, which are then classified into different categories using unsupervised clustering methods. We further expanded our framework to extract parameters of diffusion for the sulfur vacancies and analyzed transition probabilities associated with switching between different configurations of defect complexes consisting of Mo dopant and sulfur vacancy, providing insight into point defect dynamics and reactions. This approach is universal and its application to beam induced reactions allows mapping chemical transformation pathways in solids at the atomic level.

cond-mat.mtrl-sci↗

arXiv2018

Dynamics of liquid crystal on hexagonal lattice

Nematic liquid crystal (LC) molecules adsorbed on two dimensional materials are aligned along the crystal directions of the hexagonal lattice. It was demonstrated that short electric pulses can reorient the aligned LC molecules in the preferred armchair direction of hexagonal boron nitride (h-BN). Several states with a variety of colors were obtained by changing the direction and strength of the electric pulses. The ab initio calculations based on density functional theory was carried out to determine the favorable adsorption configurations of the LC molecules on the h-BN surface. A non-volatile display, in which pixel resolution can be determined by grains of hexagonal surface, is proposed, which can offer a pathway towards dynamic high-quality pixels with low power consumption, and could define a new paradigm for all non-volatile display applications.

cond-mat.soft↗

arXiv2018

Mapping mesoscopic phase evolution during e-beam induced transformations via deep learning of atomically resolved images

Understanding transformations under electron beam irradiation requires mapping the structural phases and their evolution in real time. To date, this has mostly been a manual endeavor comprising of difficult frame-by-frame analysis that is simultaneously tedious and prone to error. Here, we turn towards the use of deep convolutional neural networks (DCNN) to automatically determine the Bravais lattice symmetry present in atomically-resolved images. A DCNN is trained to identify the Bravais lattice class given a 2D fast Fourier transform of the input image. Monte-Carlo dropout is used for determining the prediction probability, and results are shown for both simulated and real atomically-resolved images from scanning tunneling microscopy and scanning transmission electron microscopy. A reduced representation of the final layer output allows to visualize the separation of classes in the DCNN and agrees with physical intuition. We then apply the trained network to electron beam-induced transformations in WS2, which allows tracking and determination of growth rate of voids. These results are novel in two ways: (1) It shows that DCNNs can be trained to recognize diffraction patterns, which is markedly different from the typical "real image" cases, and (2) it provides a method with in-built uncertainty quantification, allowing the real-time analysis of phases present in atomically resolved images.

cond-mat.mtrl-sci↗

arXiv2018

Deep Learning of Atomically Resolved Scanning Transmission Electron Microscopy Images: Chemical Identification and Tracking Local Transformations

Recent advances in scanning transmission electron and scanning probe microscopies have opened exciting opportunities in probing the materials structural parameters and various functional properties in real space with angstrom-level precision. This progress has been accompanied by an exponential increase in the size and quality of datasets produced by microscopic and spectroscopic experimental techniques. These developments necessitate adequate methods for extracting relevant physical and chemical information from the large datasets, for which a priori information on the structures of various atomic configurations and lattice defects is limited or absent. Here we demonstrate an application of deep neural networks to extract information from atomically resolved images including location of the atomic species and type of defects. We develop a 'weakly-supervised' approach that uses information on the coordinates of all atomic species in the image, extracted via a deep neural network, to identify a rich variety of defects that are not part of an initial training set. We further apply our approach to interpret complex atomic and defect transformation, including switching between different coordination of silicon dopants in graphene as a function of time, formation of peculiar silicon dimer with mixed 3-fold and 4-fold coordination, and the motion of molecular 'rotor'. This deep learning based approach resembles logic of a human operator, but can be scaled leading to significant shift in the way of extracting and analyzing information from raw experimental data.

cond-mat.mtrl-sci↗

arXiv2017

High conduction hopping behavior induced in transition metal dichalcogenides by percolating defect networks: toward atomically thin circuits

Atomically thin circuits have recently been explored for applications in next-generation electronics and optoelectronics and have been demonstrated with two-dimensional lateral heterojunctions. In order to form true 2D circuitry from a single material, electronic properties must be spatially tunable. Here, we report tunable transport behavior which was introduced into single layer tungsten diselenide and tungsten disulfide by focused He$^+$ irradiation. Pseudo-metallic behavior was induced by irradiating the materials with a dose of ~1x10$^{16} He^+/cm^2$ to introduce defect states, and subsequent temperature-dependent transport measurements suggest a nearest neighbor hopping mechanism is operative. Scanning transmission electron microscopy and electron energy loss spectroscopy reveal that Se is sputtered preferentially, and extended percolating networks of edge states form within WSe$_2$ at a critical dose of 1x10$^{16} He^+/cm^2$. First-principles calculations confirm the semiconductor-to-metallic transition of WSe$_2$ after pore and edge defects were introduced by He$^+$ irradiation. The hopping conduction was utilized to direct-write resistor loaded logic circuits in WSe$_2$ and WS$_2$ with a voltage gain of greater than 5. Edge contacted thin film transistors were also fabricated with a high on/off ratio (> 10$^6$), demonstrating potential for the formation of atomically thin circuits.

cond-mat.mtrl-sci↗

arXiv2017

Tunable Quasiparticle Band Gap in Few Layer GaSe/graphene Van der Waals Heterostructures

Two-dimensional (2D) materials have recently been the focus of extensive research. By following a similar trend as graphene, other 2D materials including transition metal dichalcogenides (MX2) and metal mono-chalcogenides (MX) show great potential for ultrathin nanoelectronic and optoelectronic devices. Despite the weak nature of interlayer forces in semiconducting MX materials, their electronic properties are highly dependent on the number of layers. Using scanning tunneling microscopy and spectroscopy (STM/STS), we demonstrate the tunability of the quasiparticle energy gap of few layered gallium selenide (GaSe) directly grown on a bilayer graphene substrate by molecular beam epitaxy (MBE). Our results show that the band gap is about 3.50 +/-0.05 eV for single-tetralayer (1TL), 3.00 +/-0.05 eV for bi-tetralayer (2TL) and 2.30 +/-0.05 eV for tri-tetralayer (3TL). This band gap evolution of GaSe, in particularly the shift of the valence band with respect to the Fermi level, was confirmed by angle-resolved photoemission spectroscopy (ARPES) measurements and our theoretical calculations. Moreover, we observed a charge transfer in GaSe/graphene van der Waals (vdW) heterostructure using ARPES. These findings demonstrate the high impact on the GaSe electronic band structure and electronic properties that can be obtained by the control of 2D materials layer thickness and the graphene induced doping.

cond-mat.mtrl-sci↗

arXiv2016

Anisotropic distributions in a multi-phase transport model

With A Multi-Phase Transport (AMPT) model we investigate the relation between the magnitude, fluctuations and correlations of the initial state spatial anisotropy $\varepsilon_{n}$ and the final state anisotropic flow coefficients $v_{n}$ in Au+Au collisions at $\sqrt{s_{_{\rm NN}}}=$ 200 GeV. It is found that the relative eccentricity fluctuations in AMPT account for the observed elliptic flow fluctuations, in agreement with measurements of the STAR collaboration. In addition, the studies based on 2- and multi-particle correlations and event-by-event distributions of the anisotropies suggest that the Elliptic-Power function is a promising candidate of the underlying probability density function of the event-by-event distributions of $\varepsilon_{n}$ as well as $v_{n}$. Furthermore, the correlations between different order symmetry planes and harmonics in the initial coordinate space and final state momentum space are presented. Non-zero values of these correlations have been observed. The comparison between our calculations and data will, in the future, shed new insight into the nature of the fluctuations of the Quark-Gluon Plasma produced in heavy ion collisions.

nucl-ex↗

arXiv2015

Factorization of event-plane correlations over transverse momentum in relativistic heavy ion collisions in a multi-phase transport model

Momentum-space azimuthal harmonic event planes (EP) are constructed from final-state midrapidity particles binned in transverse momentum (pT ) in sqrt(s_NN) = 200 GeV Au+Au collisions in a multi-phase transport (AMPT) model. The EP correlations between pT bins, corrected by EP resolutions, are smaller than unity. This indicates that the EP's decorrelate over pT in AMPT, qualitatively consistent with data and hydrodynamic calculations. It is further found that the EP correlations approximately factorize into single pT-bin EP correlations to a common plane. This common plane appears to be the momentum-space EP integrated over all pT, not the configuration space participant plane (PP).

nucl-th↗

arXiv2015

Online Algorithms Modeled After Mousehunt

In this paper we study a variety of novel online algorithm problems inspired by the game Mousehunt. We consider a number of basic models that approximate the game, and we provide solutions to these models using Markov Decision Processes, deterministic online algorithms, and randomized online algorithms. We analyze these solutions' performance by deriving results on their competitive ratios.

cs.DS↗

arXiv2013

Eccentricity is not the only source of elliptic flow fluctuations

Sources of event-by-event elliptic flow fluctuations in relativistic heavy-ion collisions are investigated in a multiphase transport model. Besides the well-known initial eccentricity fluctuations, several other sources of dynamical fluctuations are identified. One is fluctuations in initial parton configurations at a given eccentricity. Second is quantum fluctuations in parton interactions during system evolution. Third is fluctuations caused by hadronization and final-state hadronic scatterings. The magnitudes of these fluctuations are investigated relative to eccentricity fluctuations and average flow magnitude. The fluctuations from the latter two sources are found to be negative. The results may have important implications to the interpretation of elliptic flow data.

nucl-th↗

arXiv2012

Event-plane decorrelation over pseudo-rapidity and its effect on azimuthal anisotropy measurement in relativistic heavy-ion collisions

Within A Multi-Phase Transport model, we investigate decorrelation of event planes over pseudorapidity and its effect on azimuthal anisotropy measurements in relativistic heavy-ion collisions. The decorrelation increases with increasing η gap between particles used to reconstruct the event planes. The third harmonic event planes are found even anticorrelated between forward and backward rapidities, the source of which may root in the opposite orientation of the collision geometry triangularities. The decorrelation may call into question the anisotropic flow measurements with pseudorapidity gap designed to reduce nonflow contributions, hence the hydrodynamic properties of the quark-gluon plasma extracted from those measurements.

nucl-th↗

arXiv2012

Study of triangular flow $v_3$ in Au+Au and Cu+Cu collisions with a multiphase transport model

We studied the relation between the initial geometry anisotropy and the anisotropic flow in a multiphase transport model (AMPT) for both Au+Au and Cu+Cu collisions at $\sqrt{s_{NN}}$ = 200 GeV. It is found that unlike the elliptic flow $v_2$, little centrality dependence of the triangular flow $v_3$ is observed. After removing the initial geometry effect, $v_3/ε_3$ increases with the transverse particle density, which is similar to $v_2/ε_2$. The transverse momentum ($p_T$) dependence of $v_3$ from identified particles is qualitatively similar to the $p_T$ dependence of $v_2$.

nucl-th↗