arXiv subjects

Bin Gao

Publications and source records attributed to Bin Gao.

At least 109 records · Page 6Linked to original sources

arXiv2019

Device and Circuit Interaction Analysis of Stochastic Behaviors in Cross-Point RRAM Arrays

Stochastic behaviors of resistive random access memory (RRAM) play an important role in the design of cross-point memory arrays. A Monte Carlo compact model of oxide RRAM is developed and calibrated with experiments on various device stack configurations. With Monte Carlo SPICE simulations, we show that an increase in array size and interconnect wire resistance will statistically deteriorate write functionality. Write failure probability (WFP) has an exponential dependency on device uniformity and supply voltage (VDD), and the array bias scheme is a key knob. Lowering array VDD leads to higher effective energy consumption (EEC) due to the increase in WFP when the variation statistics are included in the analysis. Random-access simulations indicate that data sparsity statistically benefits write functionality and energy consumption. Finally, we show that a pseudo-sub-array topology with uniformly distributed pre-forming cells in the pristine high resistance state is able to reduce both WFP and EEC, enabling higher net capacity for memory circuits due to improved variation tolerance.

cs.ET↗

arXiv2019

Record-High Proximity-Induced Anomalous Hall Effect in (Bi$_x$Sb$_{1-x}$)2Te$_3$ Thin Film Grown on CrGeTe$_3$ Substrate

Quantum anomalous Hall effect(QAHE) can only be realized at extremely low temperatures in magnetically doped topological insulators(TIs) due to limitations inherent with the doping precess. In an effort to boost the quantization temperature of QAHE, magnetic proximity effect in magnetic insulator/TI heterostructures has been extensively investigated. However, the observed anomalous Hall resistance has never been more than several Ohms, presumably owing to the interfacial disorders caused by the structural and chemical mismatch. Here, we show that, by growing (BixSb1-x)2Te3(BST) thin films on structurally and chemically well-matched, ferromagnetic-insulating CeGeTe3(CGT) substrates, the proximity-induced anomalous Hall resistance can be enhanced by more than an order of magnitude. This sheds light on the importance of structural and chemical match for magnetic insulator/TI proximity systems.

cond-mat.mes-hall↗

arXiv2019

Mapping the structural transitions controlled by the trilinear coupling in Ca3-xSrxTi2O7

We present the results of the high-temperature neutron and x-ray diffraction experiments on the Ca3-xSrxTi2O7 (x = 0.5, 0.8, 0.85, 0.9) compounds. The ferro- to paraelectric transition in these hybrid improper ferroelectric materials arises from the so-called trilinear coupling. Depending on the Strontium content, various structures and phase transitions, different from theoretical predictions, emerge. The in-situ x-ray powder diffraction indicates a direct ferro- to paraelectric transition between the orthorhombic A21am and tetragonal undistorted I4/mmm phase for x < 0.6. We identified a reduction in the trilinear coupling robustness by increasing the Sr-doping level to lead to the emergence of the intermediate tetragonal P42/mnm phase and the gradual suppression of the orthorhombic phase. The observed character of the structure transitions and the Ca3-xSrxTi2O7 phase diagram are discussed in the framework of theoretical models of other related hybrid improper ferroelectric systems.

cond-mat.mtrl-sci↗

arXiv2019

Magnetic properties of double perovskite $Ln_2$CoIrO$_6$ ($Ln$ = Eu, Tb, Ho): hetero-tri-spin $3d$-$5d$-$4f$ systems

The field of double perovskites is now advancing to three magnetic elements on the A, B and B$'$ sites. A series of iridium-based double perovskite compounds, $Ln_2$CoIrO$_6$ ($Ln$ = Eu, Tb, Ho) with three magnetic elements were synthesized as polycrystalline samples. The compounds crystalize in monoclinic structures with the space group $P2_1/n$. Magnetic properties of these hetero-tri-spin $3d$-$5d$-$4f$ systems were studied by magnetic susceptibility and field dependent magnetization in both DC and pulsed magnetic fields. All these compounds show ferrimagnetic transitions at temperatures $T_C$ above 100 K, which are attributable to antiferromagnetic coupling between Co$^{2+}$ and Ir$^{4+}$ spins. For Eu$_2$CoIrO$_6$, the magnetic properties are similar to those of La$_2$CoIrO$_6$. The Eu$^{3+}$ spins show Van Vleck paramagnetism that don't significantly interact with transition-metal cations. By contrast, Tb$_2$CoIrO$_6$ and Ho$_2$CoIrO$_6$ reveal a second transition to antiferromagnetic order below a lower temperature $T_N$. The temperature-induced ferrimagnetic-to-antiferromagnetic phase transition might be explained by a spin-reorientation transition. Moreover, a magnetic-field-induced spin-flop transition with a small hysteresis was observed below $T_N$ in these two compounds. The magnetic moment of all three compounds do not saturate up to 60 T at low temperatures. Moderate magnetocaloric effect was also observed in all three compounds. Our results should motivate further investigation of the spin configuration on single crystals of these iridium-based double perovskites.

cond-mat.mtrl-sci↗

arXiv2018

Imaging antiferromagnetic antiphase domain boundaries using magnetic Bragg diffraction phase contrast

Manipulating magnetic domains is essential for many technological applications. Recent breakthroughs in Antiferromagnetic Spintronics brought up novel concepts for electronic device development. Imaging antiferromagnetic domains is of key importance to this field. Unfortunately, some of the basic domain types, such as antiphase domains, cannot be imaged by conventional techniques. Herein, we present a new domain projection imaging technique based on the localization of domain boundaries by resonant magnetic diffraction of coherent x rays. Contrast arises from reduction of the scattered intensity at the domain boundaries due to destructive interference effects. We demonstrate this approach by imaging antiphase domains in a collinear antiferromagnet Fe2Mo3O8, and observe evidence of domain wall interaction with a structural defect. This technique does not involve any numerical algorithms. It is fast, sensitive, produces large-scale images in a single-exposure measurement, and is applicable to a variety of magnetic domain types.

cond-mat.mes-hall↗

arXiv2018

Topological spin excitations in honeycomb ferromagnet CrI$_3$

In two dimensional honeycomb ferromagnets, bosonic magnon quasiparticles (spin waves) may either behave as massless Dirac fermions or form topologically protected edge states. The key ingredient defining their nature is the next-nearest neighbor Dzyaloshinskii-Moriya (DM) interaction that breaks the inversion symmetry of the lattice and discriminates chirality of the associated spin-wave excitations. Using inelastic neutron scattering, we find that spin waves of the insulating honeycomb ferromagnet CrI$_3$ ($T_C=61$ K) have two distinctive bands of ferromagnetic excitations separated by a $\sim$4 meV gap at the Dirac points. These results can only be understood by considering a Heisenberg Hamiltonian with DM interaction, thus providing experimental evidence that spin waves in CrI$_3$ can have robust topological properties useful for dissipationless spintronic applications.

cond-mat.str-el↗

arXiv2018

Parallelizable Algorithms for Optimization Problems with Orthogonality Constraints

To construct a parallel approach for solving optimization problems with orthogonality constraints is usually regarded as an extremely difficult mission, due to the low scalability of the orthonormalization procedure. However, such demand is particularly huge in some application areas such as materials computation. In this paper, we propose a proximal linearized augmented Lagrangian algorithm (PLAM) for solving optimization problems with orthogonality constraints. Unlike the classical augmented Lagrangian methods, in our algorithm, the prime variables are updated by minimizing a proximal linearized approximation of the augmented Lagrangian function, meanwhile the dual variables are updated by a closed-form expression which holds at any first-order stationary point. The orthonormalization procedure is only invoked once at the last step of the above mentioned algorithm if high-precision feasibility is needed. Consequently, the main parts of the proposed algorithm can be parallelized naturally. We establish global subsequence convergence, worst-case complexity and local convergence rate for PLAM under some mild assumptions. To reduce the sensitivity of the penalty parameter, we put forward a modification of PLAM, which is called parallelizable column-wise block minimization of PLAM (PCAL). Numerical experiments in serial illustrate that the novel updating rule for the Lagrangian multipliers significantly accelerates the convergence of PLAM and makes it comparable with the existent feasible solvers for optimization problems with orthogonality constraints, and the performance of PCAL does not highly rely on the choice of the penalty parameter. Numerical experiments under parallel environment demonstrate that PCAL attains good performance and high scalability in solving discretized Kohn-Sham total energy minimization problems.

math.OC↗

arXiv2018

Vortex Ferroelectric Domains, Large-loop Weak Ferromagnetic Domains, and Their Decoupling in Hexagonal (Lu, Sc)FeO3

The direct domain coupling of spontaneous ferroelectric polarization and net magnetic moment can result in giant magnetoelectric (ME) coupling, which is essential to achieve mutual control and practical applications of multiferroics. Recently, the possible bulk domain coupling, the mutual control of ferroelectricity (FE) and weak ferromagnetism (WFM) have been theoretically predicted in hexagonal LuFeO3. Here, we report the first successful growth of highly-cleavable Sc-stabilized hexagonal Lu0.6Sc0.4FeO3 (h-LSFO) single crystals, as well as the first visualization of their intrinsic cloverleaf pattern of vortex FE domains and large-loop WFM domains. The vortex FE domains are on the order of 0.1-1 μm in size. On the other hand, the loop WFM domains are ~100 μm in size, and there exists no interlocking of FE and WFM domain walls. These strongly manifest the decoupling between FE and WFM in h-LSFO. The domain decoupling can be explained as the consequence of the structure-mediated coupling between polarization and dominant in-plane antiferromagnetic spins according to the theoretical prediction, which reveals intriguing interplays between FE, WFM, and antiferromagnetic orders in h-LSFO. Our results also indicate that the magnetic topological charge tends to be identical to the structural topological charge. This could provide new insights into the induction of direct coupling between magnetism and ferroelectricity mediated by structural distortions, which will be useful for the future applications of multiferroics.

cond-mat.mtrl-sci↗

arXiv2018

Revealing Controllable Anisotropic Magnetoresistance in Spin Orbit Coupled Antiferromagnet Sr2IrO4

Antiferromagnetic spintronics actively introduces new principles of magnetic memory, in which the most fundamental spin-dependent phenomena, i.e. anisotropic magnetoresistance effects, are governed by an antiferromagnet instead of a ferromagnet. A general scenario of the antiferromagnetic anisotropic magnetoresistance effects mainly stems from the magnetocrystalline anisotropy related to spin-orbit coupling. Here we demonstrate magnetic field driven contour rotation of the fourfold anisotropic magnetoresistance in bare antiferromagnetic Sr2IrO4/SrTiO3 (001) thin films hosting a strong spin-orbit coupling induced Jeff=1/2 Mott state. Concurrently, an intriguing minimal in the magnetoresistance emerges. Through first principles calculations, the band-gap engineering due to rotation of the Ir isospins is revealed to be responsible for these emergent phenomena, different from the traditional scenario where relatively more conductive state was obtained usually when magnetic field was applied along the magnetic easy axis. Our findings demonstrate a new efficient route, i.e. via the novel Jeff=1/2 state, to realize controllable anisotropic magnetoresistance in antiferromagnetic materials.

cond-mat.mtrl-sci↗

arXiv2018

Soft antiphase tilt of oxygen octahedra in the hybrid improper multiferroic Ca3Mn1.9Ti0.1O7

We report a single crystal neutron and x-ray diffraction study of the hybrid improper multiferroic Ca3Mn1.9Ti0.1O7 (CMTO), a prototypical system where the electric polarization arises from the condensation of two lattice distortion modes.With increasing temperature (T ), the out-of-plane, antiphase tilt of MnO6 decreases in amplitude while the in-plane, in-phase rotation remains robust and experiences abrupt changes across the first-order structural transition. Application of hydrostatic pressure (P) to CMTO at room temperature shows a similar effect. The consistent behavior under both T and P reveals the softness of antiphase tilt and highlights the role of the partially occupied d orbital of the transition-metal ions in determining the stability of the octahedral distortion. Polarized neutron analysis indicates the symmetry-allowed canted ferromagnetic moment is less than the 0.04 μB/Mn site, despite a substantial out-of-plane tilt of the MnO6 octahedra.

cond-mat.str-el↗

arXiv2018

High-temperature terahertz optical diode effect without magnetic order in polar FeZnMo$_3$O$_8$

We present a terahertz spectroscopic study of polar ferrimagnet FeZnMo$_3$O$_8$. Our main finding is a giant high-temperature optical diode effect, or nonreciprocal directional dichroism, where the transmitted light intensity in one direction is over 100 times lower than intensity transmitted in the opposite direction. The effect takes place in the paramagnetic phase with no long-range magnetic order in the crystal, which contrasts sharply with all existing reports of the terahertz optical diode effect in other magnetoelectric materials, where the long-range magnetic ordering is a necessary prerequisite. In \fzmo, the effect occurs resonantly with a strong magnetic dipole active transition centered at 1.27 THz and assigned as electron spin resonance between the eigenstates of the single-ion anisotropy Hamiltonian. We propose that the optical diode effect in paramagnetic FeZnMo$_3$O$_8$ is driven by signle-ion terms in magnetoelectric free energy.

cond-mat.str-el↗

arXiv2017

(LaTiO$_3$)$_n$/(LaVO$_3$)$_n$ as a model system for unconventional charge transfer and polar metallicity

At interfaces between oxide materials, lattice and electronic reconstructions always play important roles in exotic phenomena. In this study, the density functional theory and maximally localized Wannier functions are employed to investigate the (LaTiO$_3$)$_n$/(LaVO$_3$)$_n$ magnetic superlattices. The electron transfer from Ti$^{3+}$ to V$^{3+}$ is predicted, which violates the intuitive band alignment based on the electronic structures of LaTiO$_3$ and LaVO$_3$. Such unconventional charge transfer quenches the magnetism of LaTiO$_3$ layer mostly and leads to metal-insulator transition in the $n=1$ superlattice when the stacking orientation is altered. In addition, the compatibility among the polar structure, ferrimagnetism, and metallicity is predicted in the $n=2$ superlattice.

cond-mat.mtrl-sci↗

arXiv2017

Magnetic and electronic properties of La$_3M$O$_7$ and possible polaron formation in hole-doped La$_3M$O$_7$ ($M$=Ru and Os)

Oxides with $4d$/$5d$ transition metal ions are physically interesting for their particular crystalline structures as well as the spin-orbit coupled electronic structures. Recent experiments revealed a series of $4d$/$5d$ transition metal oxides $R_3M$O$_7$ ($R$: rare earth; $M$: $4d$/$5d$ transition metal) with unique quasi-one-dimensional $M$ chains. Here first-principles calculations have been performed to study the electronic structures of La$_3$OsO$_7$ and La$_3$RuO$_7$. Our study confirm both of them to be Mott insulating antiferromagnets with identical magnetic order. The reduced magnetic moments, which are much smaller than the expected value for ideal high-spin state ($3$ $t_{2g}$ orbitals occupied), are attributed to the strong $p-d$ hybridization with oxygen ions, instead of the spin-orbit coupling. The Ca-doping to La$_3$OsO$_7$ and La$_3$RuO$_7$ can not only modulate the nominal carrier density but also affect the orbital order as well as the local distortions. The Coulombic attraction and particular orbital order would prefer to form polarons, which might explain the puzzling insulating behavior of doped $5d$ transition metal oxides. In addition, our calculation predict that the Ca-doping can trigger ferromagnetism in La$_3$RuO$_7$ but not in La$_3$OsO$_7$.

cond-mat.mtrl-sci↗

arXiv2016

Solving Verbal Comprehension Questions in IQ Test by Knowledge-Powered Word Embedding

Intelligence Quotient (IQ) Test is a set of standardized questions designed to evaluate human intelligence. Verbal comprehension questions appear very frequently in IQ tests, which measure human's verbal ability including the understanding of the words with multiple senses, the synonyms and antonyms, and the analogies among words. In this work, we explore whether such tests can be solved automatically by artificial intelligence technologies, especially the deep learning technologies that are recently developed and successfully applied in a number of fields. However, we found that the task was quite challenging, and simply applying existing technologies (e.g., word embedding) could not achieve a good performance, mainly due to the multiple senses of words and the complex relations among words. To tackle these challenges, we propose a novel framework consisting of three components. First, we build a classifier to recognize the specific type of a verbal question (e.g., analogy, classification, synonym, or antonym). Second, we obtain distributed representations of words and relations by leveraging a novel word embedding method that considers the multi-sense nature of words and the relational knowledge among words (or their senses) contained in dictionaries. Third, for each type of questions, we propose a specific solver based on the obtained distributed word representations and relation representations. Experimental results have shown that the proposed framework can not only outperform existing methods for solving verbal comprehension questions but also exceed the average performance of the Amazon Mechanical Turk workers involved in the study. The results indicate that with appropriate uses of the deep learning technologies we might be a further step closer to the human intelligence.

cs.CL↗

arXiv2016

Strain enhanced superconductivity of Mo$X_2$ ($X$=S or Se) bilayers with Na intercalation

Mo$X_2$ ($X$=S or Se) is a semiconductor family with two-dimensional structure. And a recent calculation predicted the superconductivity in electron doped MoS$_2$ monolayer. In this work, the electronic structure and lattice dynamics of Mo$X_2$ bilayers with monolayer Na intercalated, have been calculated. According to the electron-phonon interaction, it is predicted that these bilayers can be transformed from indirect-gap semiconductors to a superconductors by Na intercalation. More interestingly, the biaxial tensile strain can significantly enhance the superconducting temperature up to $\sim10$ K in Na-intercalated MoS$_2$. In addition, the phonon mean free path at room-temperature is also greatly improved in Na intercalated MoSe$_2$, which is advantaged for related applications.

cond-mat.mtrl-sci↗

arXiv2016

Sentence Level Recurrent Topic Model: Letting Topics Speak for Themselves

We propose Sentence Level Recurrent Topic Model (SLRTM), a new topic model that assumes the generation of each word within a sentence to depend on both the topic of the sentence and the whole history of its preceding words in the sentence. Different from conventional topic models that largely ignore the sequential order of words or their topic coherence, SLRTM gives full characterization to them by using a Recurrent Neural Networks (RNN) based framework. Experimental results have shown that SLRTM outperforms several strong baselines on various tasks. Furthermore, SLRTM can automatically generate sentences given a topic (i.e., topics to sentences), which is a key technology for real world applications such as personalized short text conversation.

cs.LG↗

arXiv2015

Unveiling hidden ferrimagnetism and giant magnetoelectricity in polar magnet Fe2Mo3O8

Magnetoelectric (ME) effect is recognized for its utility for low-power electronic devices.Largest ME coefficients are often associated with phase transitions in which ferroelectricity is induced by magnetic order. Unfortunately, in these systems, large ME response is revealed only upon elaborate poling procedures. These procedures may become unnecessary in single-polar-domain crystals of polar magnets. Here we report giant ME effects in a polar magnet Fe2Mo3O8 at temperatures as high as 60 K.Polarization jumps of 0.3 μC/cm2, and repeated mutual control of ferroelectric and magnetic moments with differential ME coefficients on the order of 10$^4$ ps/m are achieved.Importantly, no electric or magnetic poling is needed, as necessary for applications.The sign of the ME coefficients can be switched by changing the applied "bias" magnetic field. The observed effects are associated with a hidden ferrimagnetic order unveiled by application of a magnetic field.

cond-mat.mtrl-sci↗

arXiv2015

Learning Better Word Embedding by Asymmetric Low-Rank Projection of Knowledge Graph

Word embedding, which refers to low-dimensional dense vector representations of natural words, has demonstrated its power in many natural language processing tasks. However, it may suffer from the inaccurate and incomplete information contained in the free text corpus as training data. To tackle this challenge, there have been quite a few works that leverage knowledge graphs as an additional information source to improve the quality of word embedding. Although these works have achieved certain success, they have neglected some important facts about knowledge graphs: (i) many relationships in knowledge graphs are \emph{many-to-one}, \emph{one-to-many} or even \emph{many-to-many}, rather than simply \emph{one-to-one}; (ii) most head entities and tail entities in knowledge graphs come from very different semantic spaces. To address these issues, in this paper, we propose a new algorithm named ProjectNet. ProjecNet models the relationships between head and tail entities after transforming them with different low-rank projection matrices. The low-rank projection can allow non \emph{one-to-one} relationships between entities, while different projection matrices for head and tail entities allow them to originate in different semantic spaces. The experimental results demonstrate that ProjectNet yields more accurate word embedding than previous works, thus leads to clear improvements in various natural language processing tasks.

cs.CL↗