Catálogo de publicaciones - revistas

<jats:p>The influences of reducing the supply voltage on single event upset (SEU) and multiple-cell upset (MCU) in two kinds of 65-nm static random access memories (SRAMs) are characterized across a wide linear energy transfer (LET) range. The results show that the influence of the voltage variation on SEU cross section clearly depends on the LET value which is above heavy ion LET threshold no matter whether the SRAM is non-hardened 6T SRAM or radiation-hardened double dual interlocked cells (DICE) SRAM. When the LET value is lower than the LET threshold of MCU, the SEU only manifests single cell upset, the SEU cross section increases with the decrease of voltage. The lower the LET value, the higher the SEU sensitivity to the voltage variation is. Lowering the voltage has no evident influence on SEU cross section while the LET value is above the LET threshold of MCU. Moreover, the reduction of the voltage can result in a decrease in the highest-order MCU event cross section due to the decrease of charge collection efficiency of the outer sub-sensitive volume within a certain voltage range. With further scaling the feature size of devices down, it is suggested that the dependence of SEU on voltage variation should be paid special attention to for heavy ions with very low LET or the other particles with very low energy for nanometer commercial off-the-shelf (COTS) SRAM.</jats:p>

<jats:p>In order to investigate the characteristics and mechanisms of subthreshold voltage hysteresis (Δ<jats:italic>V</jats:italic> <jats:sub>th,sub</jats:sub>) of 4H-SiC metal–oxide–semiconductor field-effect transistors (MOSFETs), 4H-SiC planar and trench MOSFETs and corresponding P-type planar and trench metal–oxide-semiconductor (MOS) capacitors are fabricated and characterized. Compared with planar MOSFEF, the trench MOSFET shows hardly larger Δ<jats:italic>V</jats:italic> <jats:sub>th,sub</jats:sub> in wide temperature range from 25 °C to 300 °C. When operating temperature range is from 25 °C to 300 °C, the off-state negative <jats:italic>V</jats:italic> <jats:sub>gs</jats:sub> of planar and trench MOSFETs should be safely above –4 V and –2 V, respectively, to alleviate the effect of Δ<jats:italic>V</jats:italic> <jats:sub>th,sub</jats:sub> on the normal operation. With the help of P-type planar and trench MOS capacitors, it is confirmed that the obvious Δ<jats:italic>V</jats:italic> <jats:sub>th,sub</jats:sub> of 4H-SiC MOSFET originates from the high density of the hole interface traps between intrinsic Fermi energy level (<jats:italic>E</jats:italic> <jats:sub>i</jats:sub>) and valence band (<jats:italic>E</jats:italic> <jats:sub>v</jats:sub>). The maximum Δ<jats:italic>V</jats:italic> <jats:sub>th,sub</jats:sub> of trench MOSFET is about twelve times larger than that of planar MOSFET, owing to higher density of interface states (<jats:italic>D</jats:italic> <jats:sub>it</jats:sub>) between <jats:italic>E</jats:italic> <jats:sub>i</jats:sub> and <jats:italic>E</jats:italic> <jats:sub>v</jats:sub>. These research results will be very helpful for the application of 4H-SiC MOSFET and the improvement of Δ<jats:italic>V</jats:italic> <jats:sub>th,sub</jats:sub> of 4H-SiC MOSFET, especially in 4H-SiC trench MOSFET.</jats:p>

<jats:p>Significantly enhanced electroluminescence performance and stability of all-inorganic perovskite light-emitting devices (PeLEDs) have been achieved by adding triton X-100 into the perovskite precursors. The small perovskite grains arranged tightly and formed large grains as the triton X-100 were introduced. Thus the nonradiative defects originated from Pb atoms at the grain boundaries were highly passivated by triton X-100 and resulted in the promotion of PeLED performance, including a turn-on voltage of 3.2 V, a brightness of 63500 cd/m<jats:sup>2</jats:sup>, a current efficiency of 17.4 cd/A, and a prolonged lifetime of 2 h in air.</jats:p>

<jats:p>Coherent photon source is an important element that has been widely used in spectroscopy, imaging, detection, and teleportation in quantum optics. However, it is still a challenge to realize micro-scale coherent emitters in semiconductor systems. We report the observation of gain in a cavity-coupled GaAs double quantum dot system with a voltage bias across the device. By characterizing and analyzing the cavity responses to different quantum dot behaviors, we distinguish the microwave photon emission from the signal gain. This study provides a possibility to realize micro-scale amplifiers or coherent microwave photon sources in circuit quantum electrodynamics (cQED) hybrid systems.</jats:p>

<jats:p>The purpose of this article is to explore the cause of the over-response phenomenon of fiber x-ray sensor. The sensor is based on a length of PMMA fiber, whose end is filled with the scintillation material Gd<jats:sub>2</jats:sub>O<jats:sub>2</jats:sub>S:Tb. The Monte Carlo simulation software GEANT4 uses the phase space file provided by the International Atomic Energy Agency (IAEA), by irradiating the fiber x-ray sensor in the water phantom, counting the fluorescence signal of the optical fiber x-ray sensor after propagation through the fiber. In addition, the number of Cerenkov photons propagating through the fiber is also counted. Comparing this article with previous research, we believe that one of the reasons for the over-response of the fiber x-ray sensor is the non-linear response of the deposition energy of the scintillator to the fluorescence. By establishing a region of interest and counting the x-rays in this region, the simulation results show that the counted number of x-rays that may affect the fiber x-ray sensor is the biggest in the area of interest at a water depth of 5 cm. This result is close to the maximum dose point of the experimental and simulated percentage depth dose (PDD) curve of fiber x-ray sensor. Therefore, the second reason of the over-response phenomenon is believed to be fact that the inorganic materials such as Gd<jats:sub>2</jats:sub>O<jats:sub>2</jats:sub>S:Tb have larger effective atomic numbers, so the fiber x-ray sensors will cause more collisions with x-ray in a low energy region of 0.1 MeV–1.5 MeV.</jats:p>

<jats:p>Artificial neural networks can achieve impressive performances, and even outperform humans in some specific tasks. Nevertheless, unlike biological brains, the artificial neural networks suffer from tiny perturbations in sensory input, under various kinds of adversarial attacks. It is therefore necessary to study the origin of the adversarial vulnerability. Here, we establish a fundamental relationship between geometry of hidden representations (manifold perspective) and the generalization capability of the deep networks. For this purpose, we choose a deep neural network trained by local errors, and then analyze emergent properties of the trained networks through the manifold dimensionality, manifold smoothness, and the generalization capability. To explore effects of adversarial examples, we consider independent Gaussian noise attacks and fast-gradient-sign-method (FGSM) attacks. Our study reveals that a high generalization accuracy requires a relatively fast power-law decay of the eigen-spectrum of hidden representations. Under Gaussian attacks, the relationship between generalization accuracy and power-law exponent is monotonic, while a non-monotonic behavior is observed for FGSM attacks. Our empirical study provides a route towards a final mechanistic interpretation of adversarial vulnerability under adversarial attacks.</jats:p>

<jats:p>On the basis of the general potential vorticity theorem (GPV), a new wave–activity relation is derived in an ageostrophic and nonhydrostatic dynamic framework. When the Reynolds average is taken, the wave–activity relation shares an exchange term with the equation of the basic-state GPV. Thus, the two equations are capable of presenting the dynamic process of the wave–basic flow interaction. Unlike the E–P flux theory which can only be used in large-scale atmosphere, the corresponding derivation provides a useful tool to analyze the feedback of waves to basic states and the forcing of basic states to waves simultaneously, and it can be used in mesoscale systems, such as heavy rainfall processes. The theory was initially applied to the landfalling Typhoon Mujigae (2015) by assigning the scalar <jats:italic>φ</jats:italic> to the generalized potential temperature (GPT). The results showed that the newly-derived wave–activity density is able to describe the wave activities associated with strong precipitation in Typhoon Mujigae (2015), including the eyewall and spiral rainbands. However, the interaction between the basic-state vortex and mesoscale waves denoted by the exchange term between basic-state GPV and wave–activity density mainly occurs in the eyewall in Typhoon Mujigae (2015). A comparison of the exchange term with other forcing terms in the newly-derived wave–activity relation indicates that the basic state–wave interaction plays a significant role in enhancing wave activities in the high-precipitation eyewall. By a magnitude analysis of the interaction term, it is found that the strong interaction between basic-state vortex and mesoscale waves is mainly attributed to two factors: the vertical vorticity intensity of the basic-state vortex and the averaged perturbation advection of perturbation GPT which is an exchange between the basic-state GPT and perturbation GPT.</jats:p>

<jats:p>We study the exact solution of the Gaudin model with Dzyaloshinsky–Moriya and Kaplan–Shekhtman–Entin–Wohlman–Aharony interactions. The energy and Bethe ansatz equations of the Gaudin model can be obtained via the off-diagonal Bethe ansatz method. Based on the off-diagonal Bethe ansatz solutions, we construct the Bethe states of the inhomogeneous <jats:italic>XXX</jats:italic> Heisenberg spin chain with the generic open boundaries. By taking a quasi-classical limit, we give explicit closed-form expression of the Bethe states of the Gaudin model. From the numerical simulations for the small-size system, it is shown that some Bethe roots go to infinity when the Gaudin model recovers the <jats:italic>U</jats:italic>(1) symmetry. Furthermore, it is found that the contribution of those Bethe roots to the Bethe states is a nonzero constant. This fact enables us to recover the Bethe states of the Gaudin model with the <jats:italic>U</jats:italic>(1) symmetry. These results provide a basis for the further study of the thermodynamic limit, correlation functions, and quantum dynamics of the Gaudin model.</jats:p>

<jats:p>To study the incentive mechanisms of cooperation, we propose a preference rewarding mechanism in the spatial prisoner’s dilemma game, which simultaneously considers reputational preference, other-regarding preference and the dynamic adjustment of vertex weight. The vertex weight of a player is adaptively adjusted according to the comparison result of his own reputation and the average reputation value of his immediate neighbors. Players are inclined to pay a personal cost to reward the cooperative neighbor with the greatest vertex weight. The vertex weight of a player is proportional to the preference rewards he can obtain from direct neighbors. We find that the preference rewarding mechanism significantly facilitates the evolution of cooperation, and the dynamic adjustment of vertex weight has powerful effect on the emergence of cooperative behavior. To validate multiple effects, strategy distribution and the average payoff and fitness of players are discussed in a microcosmic view.</jats:p>

<jats:p>Rydberg atoms have been widely investigated due to their large size, long radiative lifetime, huge polarizability and strong dipole-dipole interactions. The position information of Rydberg atoms provides more possibilities for quantum optics research, which can be obtained under the localization method. We study the behavior of three-dimensional (3D) Rydberg atom localization in a four-level configuration with the measurement of the spatial optical absorption. The atomic localization precision depends strongly on the detuning and Rabi frequency of the involved laser fields. A 100% probability of finding the Rydberg atom at a specific 3D position is achieved with precision of ∼0.031<jats:italic>λ</jats:italic>. This work demonstrates the possibility for achieving the 3D atom localization of the Rydberg atom in the experiment.</jats:p>