## Vol. 64, No. 20 (2015)

##### 2015-10-20

## INVITED REVIEW

2015, 64 (20): 205204.
doi: 10.7498/aps.64.205204

Abstract +

The processing morphology of cubic crystal LiF irradiated by femtosecond laser varies with the polarization direction. When the polarization direction is parallel to the crystal orientation 110, the distance between the starting point and the surface is 1.08 times that along 100 polarization, and the distance between the end point and the surface is 1.01 times. While the cubic crystal is irradiated by a femtosecond laser, self-focusing and inverse bremsstrahlung are two probable mechanisms dependent on polarization. In order to investigate the relation between the self-focusing and polarization, in this paper we report the nonlinear refractive index n2 of LiF crystal which is linear with respect to selffocusing coefficient. The Z-scan technique is used to measure the nonlinear refractive indexes at different polarizations. As the polarization direction is rotated from 110 to 100, the nonlinear refractive index decreases, and the self-focusing effect becomes weaker. If self-focusing leads to the dependence of morphology on polarization, the distance between the starting point and the surface for 100 polarization should be longer than that for 110 polarization. However, the experiment exhibits an opposite result that the distance between starting point and the surface for 100 polarization is shorter than that for 110 polarization. Therefore, the processing morphology which changes with polarization is not a consequence of the self-focusing. So in order to understand why the processing morphology varies with polarization, in this paper we present a model which combines inverse bremsstrahlung, avalanche ionization and radiationless transition. We believe that the recombination due to radiationless transition has a great effect on laser machining. The inverse bremsstrahlung coefficient of 110 polarization is less than that of 100 polarization, as a result, the density of free electrons which are produced by inverse bremsstrahlung and avalanche ionization at 110 polarization is less than that at 100 polarization. At first, the laser energy is transferred to the free electrons by inverse bremsstrahlung and avalanche ionization, which is described by the paraxial nonlinear Schrodinger equation and evolution equation of electron density. The density of free electrons is obtained by solving the equations. Then free electrons transfer the energy to the crystal lattice in the process of recombination through radiationless transition, which is depicted by energy conservation and gives the distribution of lattice temperature along the propagation direction. Finally, the area in LiF crystal of which the lattice temperature climbs up to above the melting point is processed. According to the simulation, the distance between the starting point and the surface at 110 polarization is 1.03 times that at 100 polarization, and the distance between the end point and the surface at 110 polarization is 0.981 times that at 100 polarization. These are consistent with the experimental results. Simulation and experimental results demonstrate that the inverse bremsstrahlung, which is dependent on polarization, is the main reason for morphology changing with the polarization of femtosecond laser. These research results may contribute to inducing microstructure in transparent dielectrics through femtosecond laser.

## INVITED REVIEW

2015, 64 (20): 207804.
doi: 10.7498/aps.64.207804

Abstract +

Bismuth telluride (Bi2Te3) and its alloys are regarded as the best thermoelectric materials available nowadays at room temperature and can be well prepared by using existing technology. In this paper, Bi2Te3 nanocrystals are prepared by hydrothermal method and then treated by a spark plasma sintering (SPS) process at five temperatures of 300, 350, 400, 450 and 500 ℃ each for 5 min under a pressure of 20 MPa. X-ray diffraction (XRD) and positron annihilation spectroscopy are used to study the microstructures of the samples after SPS treatment at different temperatures. According to the XRD patterns, the diffraction peaks of the as-grown powder are consistent with those indicated in the standard card for Bi2Te3, which confirms successful synthesis of Bi2Te3 powders. Scanning electron microscope images show that the particles of all the samples take on flake-like structures, and the particle sizes increase from about 100 nm to a few m with the sintering temperature increasing from 350 to 500 ℃. This suggests significant reorganization of nanograins in sintering process, and some grains are agglomerated into larger particles. However, the grain sizes estimated from the X-ray diffraction peaks show little change in all the samples sintered at temperatures between 300-500 ℃. And most of the grains have sizes around 30 nm. Positron lifetime spectra are measured for Bi2Te3 samples sintered at different temperatures. The measurements reveal vacancy defects existing in all the sintered samples. With the increase of sintering temperature, there appears no significant change in trapped positron lifetime (2). This suggests that the defect size has no change during sintering. However, intensity I2 decreases monotonically with increasing sintering temperature, which indicates the lowering of vacancy concentration. The average positron lifetime shows a monotonous decrease with increasing sintering temperature, which indicates the recovery of vacancy defects at higher sintering temperatures. The thermal conductivity of the sample increases from 0.3 Wm-1K-1 to about 2.4 Wm-1K-1 with the sintering temperature increasing from 300 to 500 ℃. Since the lattice thermal conductivity dominates the total thermal conductivity, it can be inferred that sintering at higher temperature leads to the increase of lattice thermal conductivity. According to the positron annihilation lifetime result, the vacancy defects in the interface region gradually recover after sintering at higher temperatures. This shows good correlation with the increase of lattice thermal conductivity, indicating that vacancy-type defects are effective phonon scattering centers for Bi2Te3.

## INVITED REVIEW

2015, 64 (20): 208801.
doi: 10.7498/aps.64.208801

Abstract +

Organic solar cells (OSCs) with the structure of ITO/MoO3(6 nm)/Rubrene(30 nm)/C70(30 nm)/PTCBI(x nm)/Al(150 nm) are fabricated. Role of perylenebisimide with extended pi system (PTCBI) modified cathode layer in Rubrene/C70 based organic solar cells is investigated. Experimental results show that the insertion of PTCBI between C70 and Al electrode can significantly improve the performance of the devices. PTCBI contributes to an Ohmic contact between the C70 layer and Al cathode, which enhances the built-in potential in OSCs. Furthermore, PTCBI avoids the contact between the excitons and the Al electrode, and reduces the damage of high energy Al ions to C70 in the cathode preparation process. The effect of PTCBI thickness on the performances of OSC is also studied. The results indicate that the optimized PTCBI thickness is 6 nm. Compared with the performances of OSC without PTCBI, the open circuit voltage (VOC), fill factor (FF), short current density (JSC), and power conservation efficiency (P) of the optimum device are ameliorated by 70.4%, 55.5%, 125.1%, 292.2%, respectively. The cause of S-shape J-V curve in organic solar cells with thick modified cathode layer is analyzed. The modified cathode layer can be divided into two regions: the PTCBI layer and the Al permeated PTCBI layer. The electron mobility of PTCBI layer is lower than the hole mobility of Rubrene layer, which results in the charge accumulation on the unaffected PTCBI layer. When the thickness value of PTCBI layer is small, the whole modified cathode layer is permeated by Al ions, and this layer has better electron mobility than the unaffected one. When the thickness of PTCBI layer is 6 nm or more, the series resistance of OSC will increase and the S-shape J-V curve appears.

2015, 64 (20): 204208.
doi: 10.7498/aps.64.204208

Abstract +

To meet the requirements for high speed, low cost, and more information capacity, silicon photonics has been booming in recent years. Silicon photonics covers a very wide field. For the silicon photonics, researchers have successfully achieved silicon-based optical waveguides, switches, modulators, and detectors. But the problem of silicon based light source has not been really resolved, which has become a primary bottleneck for further developing the silicon photonics. The momentum of a phonon is required to allow an electron to transit from the minimum of the conduction band to the maximum of the valence band in Si because of the indirect bandgap. This two-particle process with a low probability makes it difficult to achieve high-efficiency silicon-based light source by itself.
However, much effort has been made to characterize and understand the light-emission phenomena of silicon-based devices. Also, more attempts were made to enhance the emission efficiency of silicon. Practical silicon lasers are very important for silicon photonics and have been a long goal for semiconductor scientists. A number of important breakthroughs in the past decade have focused on silicon as a photonic platform thanks to the efforts of scientists.
In this review, we introduce the recent progress of silicon-based luminescence materials, silicon light emitting diodes and silicon lasers. In the first part of this paper, common types of silicon-based light emitting materials, including porous silicon, silicon nanocrystals, rare earth-doped silicon, silicon defect emission, germanium on silicon and semiconducting silicides are comprehensively reviewed. Among them, the quantum effects and surface effects of low-dimensional silicon can greatly enhance the light emission efficiency. The erbium atoms in silicon-based rare earth materials can produce the light emission at communication wavelength band independently of the host. The transition from the lowest excited state to the 4f ground state yields light at 1.54 m. Moreover the emission energy is independent of the temperature due to the inner atomic transition. Group IV materials grown on silicon such as Ge and GeSn alloy can change from indirect bandgap into direct bandgap by introducing mechanically strain and modifying the component. Strong enhancement of photoluminescence and net gain emerging from the direct transition are very significant for fabricating the devices.
In the second part, different light emitting diodes (LEDs) fabricated with above luminescent materials are introduced. The Si PN diodes were once popular at the earlier research stage. One approach was to modify the effective surface on high-purity single crystal silicon and the other idea was to use optically active defects in silicon. Ten years later, silicon LEDs in which the dressed-photon-phonons assisted method is used, made the Si PN diode rejuvenated. LEDs fabricated on nano-structured Si and silicon-based film were limited in the optoelectronic integration since the luminescence wavelength is not corresponding to the low-loss communication region. Although erbium-doped and Er silicate LEDs emit suitable light, their high turn-on voltage and low luminescence efficiency block the practical application. The researches of Ge-on-Si LED mainly focus on modifying the band structure by introducing strain and n-doping.
In the third part, firstly we summarize the basic rules of the silicon laser. Then, we review the most recent progress in the field. Nanometer Si with periodic array can only behave unambiguous laser action using optical pumping and at very low temperature. Low threshold silicon Raman lasers with racetrack ring resonator cavities can only stop on paper also due to the difficulty in electrical pumping. The Ge-on-Si lasers operating at room temperature by optical and electrical pumping were accomplished in the past 5 years. The GeSn laser that is CMOS-compatible also came into being this year. Although so far, lasing has been implemented only by using pulsed optical pumping and stopped working at 90 K, this first demonstration of lasing in a direct-gap group IV alloy grown on standard silicon substrates is potentially an important step on the way to a platform of fully integrated silicon-based photonics. Hybrid III-V-on-Si lasers are considered as one of the most practical means due to the excellent photoelectric properties and mature preparation technology.
Finally, current problems and future development direction in the silicon light source are also presented briefly.

###### SPECIAL ISSUE—Physics and devices of silicon photonics

2015, 64 (20): 204209.
doi: 10.7498/aps.64.204209

Abstract +

In this article, the recent progress of III-V quantum dot lasers on silicon substrates for silicon photonic integration is reviewed. By introducing various epitaxial techniques, room-temperature 1.3-m InAs/GaAs quantum dot laser on Si, Ge and SiGe substrates have been achieved respectively. Quantum dot lasers on Ge substrate has an ultra-low threshold current density of 55.2 A/cm2 at room temperature, which can operate over 60 ℃ in continuous-wave mode. Futhermore, by using the SiGe virtual substrate, at 30 ℃ and an output power of 16.6 mW, a laser lifetime of 4600 h has been reached, which indicates a bright future for the large-scale photonic integration.

2015, 64 (20): 206102.
doi: 10.7498/aps.64.206102

Abstract +

Si-based optical interconnection is expected to solve the problems caused by electric interconnection with increasing the density of integrated circuits, due to its merits of high speed, high bandwidth, and low consumption. So far, all of the key components except light source of Si-based optical interconnection have been demonstrated. Therefore, the light source has been considered as one of the most important components. Ge and GeSn based on Si have emerged as very promising candidates because of their high compatibility with Si CMOS processing, and the pseudo direct-bandgap characteristic. The energy difference between the direct and indirect bandgap of Ge is only 136 meV at room temperature. Under tensile strain or incorporation with Sn, the energy difference becomes smaller, and even less than zero, which means that Ge or GeSn changes into direct bandgap material. What is more, using large n-type doping to increase the fraction of electrons in valley, we can further increase the luminous efficiency of Ge or GeSn. In this paper, we briefly overview the recent progress that has been reported in the study of Ge and GeSn light emitters for silicon photonics, including theoretical models for calculating the optical gain and loss, several common methods of introducing tensile strain into Ge, methods of increasing the n-type doping density, and the method of fabricating luminescent devices of Ge and GeSn. Finally, we discuss the challenges facing us and the development prospects, in order to have a further understanding of Ge and GeSn light sources. Several breakthroughs have been made in past years, especially in the realizing of lasing from GeSn by optically pumping and Ge by optically and electrically pumping, which makes it possible to fabricate a practical laser used in silicon photonics and CMOS technology.

2015, 64 (20): 207803.
doi: 10.7498/aps.64.207803

Abstract +

The purpose of the semiconductor Materials Genome Initiative is to discover, develop, and deploy new materials in such a way that the research and development period is reduced to a half of original period, and the cost to a fraction of the present cost, thereby speeding up the advance of clean energy sourse, state security, and human welfare, through the organic integration of experiment, computation and theory. Semiconductors play a key role in developing technologies and industries relating to economy, state security, and human welfare. The implement of the semiconductor materials genome initiative will promote the development of semiconductor science and technology into a new era. In this paper, we present a demo of the semiconductor material genome project through introducing our early work on designing silicon-based light emission materials. We first briefly review the status of development of silicon-compatible light emission and challenges facing it. We then demonstrate the power and value of semiconductor materials genome initiative by presenting our recent work on the inverse design of strongly dipole-allowed direct bandgap two-dimensional Si/Ge superlattices and one-dimensional Si/Ge core/multi-shell nanowires, respectively, from two indirect-gap materials (Si and Ge). We use a combination of genetic algorithms with an atomistic pseudopotential Hamiltonian to search through the astronomic number of variants of Sin/Gem//Sip/Geq stacking sequences. We finally give a short perspective of semiconductor materials genome initiative.

2015, 64 (20): 208501.
doi: 10.7498/aps.64.208501

Abstract +

Germanium based metal oxide semiconductor (MOS) device has been a research hotspot and considered as a potential candidate for future complementary MOS (CMOS) technology due to its high and symmetric carrier mobility. However, the poor quality of gate dielectric/channel interface significantly restricts the performances of germanium based MOS devices. Besides, the solid-solubility and activation concentration of dopants in Ge are both quite low, and the dopants diffuse fast in Ge, which makes it difficult to achieve ultra-shallow junction with high dopant concentration, especially for Ge NMOS devices.To solve these problems, different techniques are proposed and overviewed. The proposed nitrogen-plasma-passivation method can effectively suppress the regrowth of germanium sub-oxide and reduce the interface state density. Thus the performance of the fabricated Ge NMOS device is significantly improved. To enhance the n-type dopant activation in Ge, the multiple implantation technique and the multiple annealing technique are proposed. High electrical activation over 1 1020 cm-3 is achieved, and the corresponding contact resistivity is reduced to 3.8 10-7 cm2. Besides, the implantation after germanide (IAG) technique is first proposed to modulate the Schottky barrier height (SBH). The record-low electron SBH of 0.10 eV is obtained by IAG technique, and the optimized process window is given. In addition, the poor thermal stability of NiGe restricts the further improvement in performance of Ge MOS device. P and Sb co-implantation technique and novel ammonium fluoride pretreatment method are proposed to improve the thermal stability of NiGe. The electrical characteristic of NiGe/Ge diode is also improved simultaneously. The results provide the guidelines for further enhancing the performances of germanium-based MOS devices.

###### GENERAL

2015, 64 (20): 200501.
doi: 10.7498/aps.64.200501

Abstract +

Evaluation of node importance is helpful to improve the invulnerability and robustness of complex networked systems. At present, the classic ranking methods of quantitatively analyzing node importance are based on the centrality measurements of network topology, such as degree, betweenness, closeness, eigenvector, etc. Therefore, they often restrict the unknown topological information and are not convenient to use in large-scale real networked systems. In this paper, according to the idea of self-avoiding random walking, we propose a novel and simplified ranking method integrated with label propagation and local topological information, in which the number of labels that node collects from propagating process quantitatively denotes the ranking order. Moreover, the proposed method is able to characterize the structural influence and importance of node in complex networked system because it comprehensively considers both the direct neighbors of node and the topological relation of node to other ones. Through performing the experiments on three benchmark networks, we obtain interesting results derived from four common evaluating indices, i. e., the coefficient of giant component, the spectral distance, the links of node, and the fragility, which indicate that the proposed method is much more efficient and effective for ranking influential nodes than the acquaintance algorithm.

2015, 64 (20): 200502.
doi: 10.7498/aps.64.200502

Abstract +

In the paper, we analyse the basic dynamical model of the chaotic oscillator by using the theory of intermittent chaos to construct the column of the intermittent chaos, and present an effective method to detect the weak underwater acoustic signal with an unknown frequency. Duffing oscillator is sensitive to phase transformation from chaos to intermittent chaos, whose frequency difference is slightly different from that between the driving force signal and the signal to be detected. By employing this theory, we detect the frequency of the ship signal in ocean background noise. For the detection by using the intermittent chaos, there exists no effective way to estimate the frequency of the signal to be detected, which can only be judged by empirical methods and therefore the man-made error will exist. All of these will affect the consequence of the intermittent chaos and make the practical application difficult. To solve this problem, in this paper, we first study the basic theory of the chaotic system, then construct the simulated signal to examine the system, and finally detect the ship signal. To make the detection feasible, a chaotic oscillator column is considered to sweep through the unknown frequency of the signal. By using this method, we can obtain the frequency range. Finally Hilbert transform is used to detect the envelope of the intermittent chaos followed by measuring the frequency of the envelope through using Fourier spectrum. Thus the frequency of the signal can be calculated by using the function describing the relationship among the driving force signal, ship signal and the envelope. The simulations and the detection processing of the measured acoustic signal are carried out by using the proposed method, which can effectively detect the frequency of the ship signal embedded within strong background noise and also the frequency of the signal to be detected can be calculated, which is conducive to solving the presently existing problem about frequency estimation. Signal-to-noise ratio can be enhanced by 4.4 dB based on the method by using the Hilbert transform compared with by using the method just through using an intermittent chaotic oscillator column, which verifies the effectiveness of the method in this paper.

2015, 64 (20): 200503.
doi: 10.7498/aps.64.200503

Abstract +

Stochastic resonance (SR) describes a nonlinear phenomenon in nature, of which the essential ingredients are a nonlinear system, a weak signal, and a source of noise. Using the nonlinear system, the signal-to-noise ratio (SNR) of the output signal of the system will peak at a certain value of noise intensity under a synergistic action of input signal and noise. Besides the traditional Langevin equation, the new SR models such as monostable oscillators, chaotic systems, time-delay systems and bistable Duffing systems, can also produce SR phenomena. In this paper, a normalized symmetrical tri-stable potential function is constructed by using equilibrium parameters p and q, and a tri-stable system model simultaneously driven by weak signal and noise is further proposed. The tri-stable system model can be understood through a cantilever beam structure with three magnets, and deduced from the Brownian motion equation. We study in-depth and summarize the influences of parameters p and q on the potential barrier heights ΔU1, ΔU2 and their difference value. By analyzing the steady-state solution of the tri-stable system under invariable input, the concept of system steady-state solution curve (SSS curve) is proposed, and is used to further study the system dynamic response under low-frequency harmonic signal input. In these situations, the system response can be obtained by combining the steady-state solutions of the system following time t under a group of tempolabile inputs. Moreover, with the noise injection, the tri-stable system can realize SR under appropriate parameter condition, which can be demonstrated by the output amplitude curve and also the output SNR curve of the system against noise intensity. The mechanism of noise-induced SR of tri-stable system can be analyzed from the perspective of SSS curve. Finally, we further study the influence of tri-stable SR against system parameters. The value of damping ratio k affects the value of damping force acting on the Brownian particle, thus the tri-stable system needs noise with larger intensity to produce SR under a larger k. The values of equilibrium parameters p and q both affect the shape of the SSS curve, a larger p or a smaller q may result in larger-intensity noise for the system to produce SR.

2015, 64 (20): 200504.
doi: 10.7498/aps.64.200504

Abstract +

Anti-control of bifurcation, as an inverse problem of conventional bifurcation analysis, is aimed at creating a certain bifurcation with desired dynamic properties at a pre-specified system parameter location via control. The main purpose of this paper is to address the problem of anti-control of Neimark-Sacker bifurcation of a three-degree-of-freedom vibro-impact system with clearance (i.e., the second Hopf bifurcation of the original system), which may be viewed as a design approach to creating a quasi-periodic impact motion (or torus solution) at a specified system parameter location via control. Firstly, in the premise of no change of periodic solutions of the original system, when the difficulties that are brought about by the implicit Poincaré map of the vibro-impact system are considered, a linear feedback controller is added to the original system and a six-dimensional Poincaré map of the close-loop control system is established. In order to design a desired bifurcation solution by control, the multiple control gains are used to tune the existence of this bifurcation based on the corresponding critical criterion. However, for six-dimensional map of the vibro-impact system in the paper, the analytical expressions of all eigenvalues of Jacobi matrix with respect to parameters are unavailable. This implies that when the classical critical criterion described by the properties of eigenvalues is used, we have to numerically compute eigenvalues point by point and check their properties to search for the control gains. So, the numerical calculation is a laborious job in the process of determining the control gains. To overcome the difficulty originating from the classical bifurcation criterion, the explicit critical criterion without using eigenvalue calculation of high-dimensional map is used to obtain the controlling parameters area when quasi-periodic impact motion occurs. Then, the stability of quasi-periodic bifurcation solution is analyzed by utilizing the center manifold and normal formal theory. Finally the numerical experiments verify that the stable quasi-periodic impact motion can be generated at a designated system parameter point by the proposed control.

2015, 64 (20): 200505.
doi: 10.7498/aps.64.200505

Abstract +

Starting from time evolution of wave function, quantum dynamics for a periodically kicked free top system is studied in this paper. For an initial spherical coherent state wave packet (localized) we find that 1) as the number of kicking is small, the speed and the direction of the diffusion for a time-evolving wave packet on a periodically kicked free top is related to the kicking strength: the stronger the kicking strength, the more chaotic for the diffusion (which means the more randomized in direction) is and the faster the speed of diffusion is, and then more quickly the full phase space is filled up; 2) as the kicking number is large, the time-evolving wave function will take on fine structure distribution in phase space, and the scope of the distribution for the fine structure will expand with the increase of the kicking strength, and the whole phase space will be filled up finally, and then the wave function will show multifractal property in phase space.#br#We study the multifractal behavior for a time-evolving wave function by partition function method: 1) for different kicking strengths and different q values, we study the scaling properties of partition function X(q), and find the power law relation between the partition function and the scaling L, i.e., X(q)-Lτ(q); 2) at different kicking strength, for a time-evolving wave function we calculate the singularity spectrum f(a)-a, and find that a maximum value of f(a) is 2.0 independent of the kicking strength, but the width of the singularity spectrum becomes narrow with the increase of the kicking strength, which means that the scope of the distribution for a is widest for regular state (localized), and is narrower for transition state from regular to chaotic, and is narrowest for chaotic state; 3) in the time-evolving process, the fluctuation for the width of the singular spectrum is smallest for chaotic state, intermediate for transition state of regular to chaotic, and the largest for regular state; 4) we calculate the generalized fractal dimension Dq-q for different kicking strengths, and find D0 = 2 independent of the kicking strength.#br#We study the mutifractal behaviors for the mean propbability amplitude distribution for a sequence of time-evolving wave functions and find that the result is similar to that of the single wave function type but has the difference: the width of the spectrum is reduced for each kicking strength.

2015, 64 (20): 200506.
doi: 10.7498/aps.64.200506

Abstract +

Chaos phenomenon which exists widely in nature and society affects people's production and life. It has great important significance to find out the regularity of chaotic time series from a chaotic system. Since chaotic system has extremely complex dynamic characteristics and unpredictability, and chaotic time series prediction through traditional methods has low prediction precision, slow convergence speed and complex model structure, a prediction model about Hermite orthogonal basis neural network based on improved teaching-learning-based optimization algorithm is proposed. Firstly, according to the chaotic time series, autocorrelation method and Cao method are used to determine the best delay time and the minimum embedding dimension respectively, then a phase space is reconstructed to obtain the refactoring delay time vector. Secondly, on the basis of phase space reconstruction and best square approximation theory, combined with the neural network topology, a prediction model about Hermite orthogonal basis neural network with excitation functions based on the Hermite orthogonal basis functions is put forward. Thirdly, in order to optimize the parameters of the prediction model, an improved teaching-learning-based optimization algorithm is proposed, where a feedback stage is introduced at the end of the learning stage based on the teaching-learning-based optimization algorithm. Finally, the parameter optimization problem is transformed into a function optimization problem in the multidimensional space, then the improved teaching-learning-based optimization algorithm is used for parameter optimization of the prediction model so as to establish it and analyze it. Lorenz and Liu chaotic systems are taken as models respectively, then the chaotic time series which will be used as simulation object is produced by the fourth order Runge-Kutta method. The comparison experiments with other prediction models are conducted on single-step and multi-step prediction for the chaotic time series. The simulation results and numerical analysis show that compared with radial basis function neural network, echo state network, least square support vector machine prediction model and Hermite orthogonal basis neural network based on teaching-learning-based optimization algorithm, the proposed prediction model has the mean absolute error and root mean square error reduced significantly, has a decision coefficient close to 1, meanwhile, has a mean modeling time shortened greatly. So the proposed prediction model can improve the prediction precision, accelerate the convergence speed and simplify the model structure, thus the prediction model is effective and feasible, which makes it promoted and applied easily.

2015, 64 (20): 200507.
doi: 10.7498/aps.64.200507

Abstract +

The given English phonemes, words and sentences are sampled and preprocessed. For these real measured speech signal series, time delay and embedding dimension are determined by using mutual information method and Cao's method, respectively, so as to perform phase space reconstruction of the speech signal series. By using small data set method, the largest Lyapunov exponent of the speech signal series is calculated and the fact that its value is greater than zero presents chaotic characteristics of the speech signal series. This, in fact, performs the chaotic characteristic identification of the speech signal series. By introducing second-order Volterra series, in this paper we put forward a type of nonlinear prediction model with an explicit structure. To overcome some intrinsic shortcomings caused by improper parameter selection when using the least mean square (LMS) algorithm to update Volterra model efficiency, by using a variable convergence factor technology based on a posteriori error assumption on the basis of LMS algorithm, a novel Davidon-Fletcher-Powell-based second of Volterra filter (DFPSOVF) is constructed and is performed to predict speech signal series of the given English phonemes, words and sentences with chaotic characteristics. Simulation results under MATLAB 7.0 environment show that the proposed nonlinear model DFPSOVF can guarantee its stability and convergence and there are no divergence problems in using LMS algorithm; for single-frame and multi-frame of the measured speech signals, when root mean square error (RMSE) is used as an evaluation criterion the prediction accuracy of the proposed nonlinear prediction model DFPSOVF in this paper is better than that of the linear prediction (LP) that is traditionally employed. The primary results of single-frame and multi-frame predictions are given. So, the proposed DFPSOVF model can substitute linear prediction model on certain conditions. Meanwhile, it can better reflect trends and regularity of the speech signal series and fully meet requirements for speech signal prediction. The memory length of the proposed prediction model may be selected by the embedding dimension of the speech signal series. The proposed model can present a nonlinear analysis and more valuable model structure for speech signal series, and opens up a new way to speech signal reconstruction and compression coding so as to improve complexity and process effect of speech signal processing method.

2015, 64 (20): 200508.
doi: 10.7498/aps.64.200508

Abstract +

Based on the parameter switching algorithm and the discrete chaotic system, a new chaotic system based parameter switching algorithm is proposed. The principles of parameter switching algorithm and chaotic system based parameter switching algorithm are presented in detail by means of flow chart and step description. By applying phase diagram observation method, chaotic attractor approximation of the unified chaotic system is investigated based on parameter switching algorithm and chaotic system based parameter switching algorithm. It shows that chaos can be obtained by switching two periodic parameters and periodic state can be observed by switching two chaotic parameters. Thus the formulas chaos+ chaos = periodic and period+ period = chaos are proved to be workable in this paper. Chaotic attractor approximation of Rössler chaotic system is also studied by employing the two switching methods. Two cases are investigated. Firstly, a chaotic switching system is obtained by switching a chaotic parameter and a periodic parameter. Then a more complex switching scheme is carried out. Periodic system is switched by two periodic parameters and a chaotic parameter. So, the formulas chaos+ periodic = chaos and periodic+ period+ chaos = periodic are proved to be workable. It shows that the switching system is the approximation of the original system under specified parameter, and the attractor is in accordance with the attractor of the targeting system. The outputs of the Logistic map based parameter switching algorithm are more complex than those of existing parameter switching algorithm. As the distribution of logistic map is not uniform, the approximate attractor does not consist of the targeting system and shows more complicated structure. But approximate attractors can be obtained when the distribution of discrete sequence is uniform. In addition, the chaotic map based parameter switching algorithm has larger secret key space since it has the initial values and parameter of the chaotic map. Finally, the parameter switching circuit of Rössler system is designed by introducing a square wave generator. Compared with the traditional switching chaotic circuit (switching between different systems), the design of parameter switch circuit is simpler as it does not need to change the original structure of the system. The output is affected by the frequency of the square wave. By adding an appropriate frequency square wave generator, the circuit simulation agrees with the numerical simulation. It presents a theoretical and experimental base for the practical application of the parameter switching chaotic systems.

2015, 64 (20): 200701.
doi: 10.7498/aps.64.200701

Abstract +

Diagnostic measurement of single picosecond event in high energy density physics, laser fusion, plasma radiation, and combustion, is of great importance. However, the measuring of the shape of the single X-ray pulse and the synchronization of X-ray and the laser pulse in picosecond resolution is still a great challenge. Restricted by the transit time of electrons, the time-resolution limit of a conventional framing camera based on the microchannel plate is 40 ps. Centered on the full-optical modulation effect of the light-probe, a novel method for X-ray detection of picoseconds temporal resolution based on low temperature GaAs is proposed in this work. The basic physical mechanism of the detector can be explained in both macroscopical and microcosmic ways. In the macroscopical way, the X-ray radiation absorption in the sensor material produces a transient, non-equilibrium electron-hole pair distribution that results in a transient differential change of the local refractive index, which is then sensed by the reflectivity changes of the optical probe beam. In the microcosmic way, X-ray absorption creates photoelectrons and the core level holes are subsequently filled through Auger or fluorescence processes. These excitations ultimately increase conduction and valence band carriers that perturb optical reflectivity.#br#To verify the proposed X-ray detection method, a Fabry-Perot detector is designed, which consists of a 5 μm thick GaAs layer surrounded by a GaAs/AlAs distributed Bragg reflector. The test is carried out on a femtosecond laser facility, where the X-ray source is produced by focusing the 56 fs Ti: Sapphire facility laser, with a central wavelength of 800 nm, onto an aluminum foil. Then the X-ray pulse induces a transient optical reflectivity change in GaAs, which is a powerful tool for establishing the high-speed X-ray detection.#br#The experimental results indicate that this technology can be used to provide X-ray detectors with a temporal resolution of tens of picoseconds. By optimizing the material, the temporal resolution can be enhanced to be less than 1 ps. Through further development, this X-ray detector could provide an insight into previously unmeasurable phenomena in many fields. Future work will focus on developing much faster devices characterizing both the rise and fall time and imaging array technology.

###### ATOMIC AND MOLECULAR PHYSICS

2015, 64 (20): 203101.
doi: 10.7498/aps.64.203101

Abstract +

We use the density functional theory (DFT) with dispersion correction to investigate the stability and electronic structure of hydrazine (N2H4) adsorpted on Ni8Fe8/Ni (111) alloy surface. The geometries and adsorption characteristics of the structure on the Ni8Fe8 alloy surface are presented. Results show that N2H4 bridging between two iron atoms gives the strongest adsorption with an adsorption energy of -1.578 eV/N2H4. Top modes turn out to be the local minima with adsorption energies of -1.346 eV/N2H4 (for the top site on a Fe atom) and -1.061 eV/N2H4 (for the top site on a Ni atom). It is demonstrated that the bridging mode is more favorable than the top mode on the NiFe alloy surface with a coverage of 1/16 ML, and Fe atom can provide stronger adsorption site than Ni atom. The van der Waals contribution is significant with a value of about 0.4 eV/N2H4. Meanwhile, the van der Waals contribution is larger for adsorption on Fe atom than on Ni atom, and for adsorption of the bridging mode than of the top mode. We also find that the structure of N2H4 in the anti molecule, rather than the gauche molecule, is bound on the top site of Fe atom on the NiFe alloy surface with a coverage of 1/16 ML, which demonstrates that the repulsive adsorbate-adsorbate interaction is weak on the surface with low coverage. The strong interaction between the surface atom and the adsorbate contributes to the result that the lone pair electrons of N2H4 in gauche conformer are attracted by the Fe atom. In addition, for the five adsorption structures of N2H4 on Ni8Fe8/Ni(111) alloy surface, we analyze the projected electronic density of states (DOS), induced charge density and electron localisation function (ELF) slices through the Fe-N or Ni-N bonds of the adsorbed molecule on the alloy surface. It shows that the electronic DOS presents the mixture between HOMO of N2H4 and the d orbital of the surface atom, which corresponds to charge transfer between the substrate and the adsorbate. The charges are transferred mainly from N2H4 to the surface atoms, and the extents of charge transfer are different for the bridging mode and the top one which is present in the induced charge density. Furthermore, the region of localisation in the ELF slices can be found for the adsorptions between the N atom of N2H4 and the Fe or Ni atom of surface, which gives a clear view of the coordination bonds for the interactions of N–Fe or N–Ni.

2015, 64 (20): 203301.
doi: 10.7498/aps.64.203301

Abstract +

For most diatomic electronic states, it is very difficult to obtain the accurate vibrational spectra of the highly-excited states directly by using the modern experimental techniques and quantum theories. Based on the general expression of diatomic molecular vibrational energy, the difference converging method (DCM) is used to give a new analytical expression in this paper. By using ten known vibrational energies, the full vibrational spectra, the vibrational spectroscopic constants of the highly-excited states, and the dissociation energy can be predicted for a diatomic electronic state. In this study, the full vibrational spectra of the electronic states 31Π, 41Π and A1Σ+ of NaLi molecule are studied with the DCM and the new formula. Results show that all the vibrational levels given in the experiments can be reproduced with an error rate less than 0.02 percent in our study. In addition, By comparing with the reported experimental results, we find 26, 45 and 31 new vibrational levels for 31Π, 41Π and A1Σ+ of NaLi molecule, respectively.

###### ELECTROMAGENTISM, OPTICS, ACOUSTICS, HEAT TRANSFER, CLASSICAL MECHANICS, AND FLUID DYNAMICS

2015, 64 (20): 204101.
doi: 10.7498/aps.64.204101

Abstract +

Based on time-reversal (TR) technique, the model of single frequency spatial power combining using sparse array is established. The efficiency function of spatial power combining is defined. The expression for the relationship of the statistical characteristics of combining efficiency at the time of maximum amplitude with the phase error and the number of array elements is derived. The analysis shows that when other parameters are determined, if the phase errors of the array nodes are mutually independent, and obey the uniform distribution to a certain extent, the combing efficiency's mean would not be related to the number of array elements N, but related to the statistic parameter of phase error. The combing efficiency's variance is related to not only the statistic parameter of phase errors, but also N. Once the statistic parameter of phase error is fixed, the greater the value of N, the smaller the variance is. So, in the engineering application, a large number of small power nodes could be used to reduce the phase error's effect. In addition, the influence of phase error on the combining efficiency is investigated by both theoretical analysis and the numerical simulation. The results show that when the array elements work at the same frequency, polarization and antenna type, the parameter of phase error would affect the combing result. The smaller the parameter of phase error, the larger the power of the effective point is, and the more concentrative the effective points' distribution is; the greater the parameter of phase error, the smaller the power of the effective point is, and the more dispersed the effective points' distribution is. It is also seen that even though the phase error occurs, the spatial power combining can still be realized with the time reversal technique. The determination of the phase control precision is the compromise between the requirements and the possibility. The results presented in this paper are useful for developing the microwave weapons with high power and electronic warfare.

2015, 64 (20): 204201.
doi: 10.7498/aps.64.204201

Abstract +

Based on the theory of paraxial approximation of beam propagation, the analytical expression of the intensity of the off axial radially polarized beam (OARPB) is derived and the effect of the off axial magnitude on the distribution of intensity of the OARPB is studied. Meanwhile, according to the definition of the first-order moment of centroid, the coordinate of centroid of the OARPB is derived and the variation of cenreoid of the OARPB is studied. Simulation result shows that the intensity distribution of the OARPB is different from that of the radially polarized beam. The intensity distribution of the OARPB is not uniform in the near-field. With increasing propagation distance, the beam spreads and the uniformity of intensity of the OARPB is improved gradually. However, the intensity distribution of the radially polarized beam keeps the form of symmetric doughnut spot during propagation all the time. When the off axial magnitude is small, the intensity distribution of the OARBP is obviously asymmetric in the near-field, and it becomes nearly symmetric while the beam propagates a certain distance. The smaller the off axial magnitude, the shorter the required propagation distance to become symmetric for the OARPB. When the off axial magnitude is larger, the hollow part of intensity distribution disappears, and the doughnut beam of the OARPB changes into a Gaussian beam spot gradually during propagation. On the other hand, the centroid of the OARPB does not change with increasing propagation distance. The value of the ordinate of centroid is equal to zero all the time. And the value of the abscissa of centroid is related to the beam size and the off axial magnitude. While the beam size increases, the abscissa of centroid increases linearly at the same time. When the off axial magnitude is small, the abscissa of the centroid of the OARPB increases with the increase of the off axial magnitude, nonlinearly and slightly; however, when the off axial magnitude is larger, the abscissa of centroid of the OARPB increases with the increase of the off axial magnitude, linearly and significantly.

2015, 64 (20): 204202.
doi: 10.7498/aps.64.204202

Abstract +

Image reconstruction from sparse data is one of the key technologies in physical imaging, and it can often be mathematically described as an underdetermined linear inverse problem. Mathematical models for sparse reconstruction often choose the sparseness constraints or data fidelity term directly as objective functions. However, physics concepts and laws for these modeling or solving processes have never been explored. In this paper, sparse reconstruction is investigated for the first time from the perspective of physical motion. Firstly, a physical model is created to describe a particle motion in viscous medium, in which the particle gravity potential energy function is the norm of l2-l1 after the relaxation transformation. In discrete calculations, the particle displacement is determined by the corresponding iterative result, and its velocity can be described as the change between two adjacent iterations. Then, a new mathematical model based on the physical motion model is studied for sparse reconstruction, in which the total energy of particle is chosen as a new objective function and nonnegative displacements as constraints. This new model preserves sparse constrains and fidelity term of original l2-l1 model, and adds the constrains of deviations between two adjacent iterations so as to avoid oscillations caused by large deviations. Furthermore, a targeted gradient projection technique is adopted to solve such a reconstruction model, and its convergence is discussed as well. Especially in this algorithm, the gradient of this new objective function contains the iterative step of previous iteration, and such iterative steps play the role of physical inertia property in iterative process, which can effectively enlarge the iterative steps to accelerate the convergence and avoid local optima. Finally, two sets of experimental results are presented, including natural grayscale image reconstruction and micro focus X-ray defect detection in precision electronic package. The results demonstrate that the proposed method outperforms its competitors distinctly in time efficiency on the basis of guaranteeing the reconstruction quality. Additionally, on detecting internal defects in solder joint of integrated circuit, the proposed method is well performed in retaining edge details of the reconstructed micro focus X-ray images. Therefore, the proposed method can identify the solder joint internal defects more accurately and is more suitable to rapid and precise micro focus X-ray defect detection in industry.

2015, 64 (20): 204203.
doi: 10.7498/aps.64.204203

Abstract +

Previous investigations demonstrated that a semiconductor laser subjected to optical injection can realize period-one (P1) oscillation output under suitable operational parameters, which can be used to obtain high quality photonic microwave. In this paper, we propose a scheme for simultaneously generating two channel photonic microwave based on the P1 oscillations of two orthogonally polarization modes in a vertical-cavity surface-emitting laser (VCSEL) subjected to an elliptical polarization optical injection, and the relevant characteristics of obtained photonic microwave are numerically simulated and analyzed. The results show that under suitable operational parameters, a free-running VCSEL (named master VCSEL, M-VCSEL) can output an elliptical polarization light in which both X and Y polarization components of the elliptical polarization light oscillate at the same frequency. By using the elliptical polarization light from the M-VCSEL as an injection light into another VCSEL (named slave VCSEL, S-VCSEL), both two polarization components of the S-VCSEL can be driven into P1 oscillation through selecting suitable injection strength under a fixed frequency detuning between the M-VCSEL and the S-VCSEL. Based on the P1 oscillation, two orthogonally photonic microwave signals can be obtained. With the increase of the injection strength from the M-VCSEL, the frequency of photonic microwave shows a gradually increasing trend while the power of photonic microwave displays an increasing process accompanied by slight ripples. Combining the distribution mappings of the frequency, the power, and the amplitude difference between the first sideband and the second sideband of the photonic microwave in the parameter space of the injection strength and the frequency detuning, certain regions with optimally operational parameters can be determined for acquiring high quality photonic microwave.

2015, 64 (20): 204204.
doi: 10.7498/aps.64.204204

Abstract +

Laser diode end-pumped Nd:YAG based multiple weak-lines laser from 4F3/2-4I13/2 translation is reported. Fluorescence spectra for both Nd:YAG crystal and ceramic are present. Simple two-mirror cavity with reasonable optical coating is used for multiple weak-line laser operation around 1.3 and 1.4 μm. The variations of laser output and spectra with the pump power are compared experimentally between 1.0 at.% Nd3+ doped Nd:YAG ceramic with 3 mm×3 mm×7.5 mm in size and 0.8 at.% Nd3+-doped Nd:YAG crystal with 3 mm×3 mm×8 mm in size. First, 1338 and 1356 nm dual-wavelength laser outputs are achieved using output coupler with transmittance values of 9.6%, 6.2%, 2.4% and 1.8% at 1338, 1356, 1414 and 1444 nm, respectively. Under low pump power around the threshold, the 1356 nm single wavelength laser is obtained. With increasing the pump power, the laser with a wavelength of 1338 nm appears first in the ceramic. At an incident pump power of 11.8 W, a 1338 and 1356 nm dual-wavelength laser with an output power of 3.7 W and an intensity ratio of 1 : 5 for crystal, and a 1338 and 1356 nm dual-wavelength laser with an output power of 3.5 W and an intensity ratio of 1:3 for ceramic are obtained. Replacing the output couplers with transmittance values of 15.4%, 6.5%, 1.1% and 0.8% at wavelengths of 1338, 1356, 1414 and 1444 nm, respectively, 1356 and 1414 nm dual-wavelength laser outputs are achieved. Under the low pump power, even triple-wavelength (1356, 1414 and 1444 nm) laser is obtained. With increasing the pump power, the intensities of 1414 and 1444 nm wavelengths turned down and the 1444 nm wavelength disappears first in ceramic. At an incident pump power of 11.8 W, a 1356 and 1414 nm dual-wavelength output power of 3.56 W with intensity ratio of 44 : 1 for crystal and a 1356 and 1414 nm dual-wavelength laser output power of 3.25 W with intensity ratio of 12 : 1 for ceramic are obtained. The results show that slight difference between fluorescence spectra results from the difference in laser spectrum between transparent ceramic and single crystal materials for multi-wavelength output. Thresholds of two different output couplers are both about 2 W. When the incident pump power increases to 13.5 W, a 1338 and 1356 nm dual-wavelength laser output power of 4.05 W and a 1356 and 1414 nm dual-wavelength laser output power of 3.65 W are achieved in Nd:YAG transparent ceramic. The corresponding slope efficiency values are 33.9% and 31.9%, respectively.

2015, 64 (20): 204205.
doi: 10.7498/aps.64.204205

Abstract +

We demonstrate an ultra-long cavity multi-wavelength Yb-doped fiber laser mode-locked by carbon nanotubes. The total length of the fiber laser is 1021.2 m. The different regimes of noise-like soliton and soliton rain mode-locking with the multi-wavelength operation are experimentally obtained with a repetition rate of 199.8 kHz. The higher output power and pulse energy from the soliton rain are measured to be 40.3 mW and 201.5 nJ, respectively, with a pulse width of about 102.5 ns.

2015, 64 (20): 204206.
doi: 10.7498/aps.64.204206

Abstract +

The analytical expressions for the average intensity and the centroid position of partially coherent decentred annular beams propagating through oceanic turbulence are derived, and the propagation equation of the position of the maximum intensity is also given. Changes of the average intensity, the centroid position and the position of the maximum intensity of partially coherent decentred annular beams during propagation are studied in detail. It is shown that both in free space and in oceanic turbulence, the position of the maximum intensity moves to the propagation z-axis with increasing the propagation distance, and is kept unchanged when the propagation distance is large enough. Furthermore, in free space the position of the maximum intensity is closer to the propagation z-axis than to the centroid position when the propagation distance is large enough. The position of the maximum intensity is closer to the propagation z-axis with increasing the correlation parameter, and far from the propagation z-axis with increasing the decentered parameter and the obscure ratio. However, in oceanic turbulence the position of the maximum intensity is close to the centroid position when the propagation distance is large enough, and the evolution is speeded with increasing the strength of oceanic turbulence. The influence of the beam coherence on propagation characteristics decreases due to oceanic turbulence. On the other hand, the centroid position is independent of the beam coherence, the propagation distance and the oceanic turbulence. The centroid position is far from the propagation z-axis with increasing the decentered parameter and the obscure ratio. In addition, the hollow core of partially coherent decentred annular beams is filled up as the propagation distance increases, and the evolution is speeded with increasing the strength of oceanic turbulence. The results obtained in this paper are very useful for applications of partially coherent decentred annular beams in oceanic turbulence.

Design and experimental study of mode selective all-fiber fused mode coupler based on few mode fiber

2015, 64 (20): 204207.
doi: 10.7498/aps.64.204207

Abstract +

Three types of all-fiber mode-selection couplers based on fused few-mode fibers (FMFs) are proposed and demonstrated. The specific mode conversions are achieved with appropriate parameters, keeping to the coupling mode theory. LP01 mode is selectively converted into the LP11, LP21, LP02 mode via a 2×2 fused fiber coupler composed of single-mode fiber (SMF) and FMF. By changing the preset parameters in fabrication, mode conversions are also realized between the LP01 and the mixed high order modes. Moreover, conversions from the LP01 mode to other higher order modes are implemented as well in a 3×3 fiber coupler comprising FMF-SMF-FMF structures. Besides, different modes are simultaneously obtained in separated channels to reduce model crosstalk. Distinguished from other techniques, symmetric and asymmetric fused biconical taper are employed in this paper. The 2×2 fiber coupler achieves the conversion from LP01 mode to a single higher order mode such as LP11 or LP21 mode over a broadband spectral range from 1530 nm to 1560 nm. Meanwhile, the mode conversion efficiency exceeding 80% is recorded in experiment, while the insertion loss remains as low as 0.8 dB. Through the comparison with all-fiber mode-selection couplers reported, the relationship between fusion-degree and conversion efficiency is further studied. The experimental results are consistent with the numerical simulations. In addition, the coupler based mode-selection with lower insertion loss and higher conversion efficiency shows potential applications in mode-division multiplexing and sensing systems.

2015, 64 (20): 204301.
doi: 10.7498/aps.64.204301

Abstract +

By solving the nonlinear wave equation coupled with the modified Rayleigh-Plesset equation, the characteristics of the acoustic field and bubble motion in cavitation environment can be described. In general, the cavitation cloud consists of many kinds of bubbles with different ambient radii. For simplicity, in this work the cavitation process of the mixture of two kinds of bubbles with different ambient radii is numerically simulated, and the ratio of the mixture is adjustable. Suppose that the cavitation in water contained in a cylindrical container is stimulated by ultrasonic horn. The dissipative absorption of the container wall is taken into account, which plays an important role in forming the stationary standing wave field, otherwise, the beat signal of acoustic pressure will appear which is absent in the observation. Based on the stationary acoustic wave field, for the case of the mixed-bubble cavitation, the interactions between bubbles and acoustic field, bubbles and bubbles, as well as the spectrum of acoustic signal are analyzed. We choose the cases that the ratio of two kinds of bubble species is varying, but the total density of bubble number is fixed to be 1/mm3, and find that those results are very different. For the case that the ambient radii of two bubble species are both a few micron, revealing that the interaction between bubbles and acoustic field is usually weak. As the proportion of bigger bubble increases, the change of the acoustic pressure and the averaged radius of bubble behave regularly; for the case that the ambient radius of one of bubble specie is relatively big, for example, the ambient radius is about a few tens of microns, the interactions between bubbles and acoustic field become stronger, and the nonlinearity is more apparent. We can observe the similar trends from the frequency spectrum. For the bubble of a few microns in size, the base frequency is dominant; in contrast, for the bubble of a few tens of microns in size, the components of harmonic frequencies are far beyond the base frequency component. The interesting phenomenon is that there is the cut off frequency and the cut of frequencies for different mixture of bubbles are almost the same.

2015, 64 (20): 204401.
doi: 10.7498/aps.64.204401

Abstract +

Constructal optimization of a rectangular fin heat sink with two-dimensional heat transfer model is carried out through using numerical simulation by finite element method, in which the minimized maximum thermal resistance and the minimized equivalent thermal resistance based on entransy dissipation are taken as the optimization objectives, respectively. The optimal constructs based on the two objectives are compared. The influences of a global parameter (a) which integrates convective heat transfer coefficient, overall area occupied by fin and its thermal conductivity, and the volume fraction (φ), on the minimized maximum thermal resistance, the minimized equivalent thermal resistances and their corresponding optimal constructs are analyzed. The results show that there does not exist optimal thickness of fins for the two objectives when the shape of the heat sink is fixed, and the optimal constructs based on the two objectives are different when the shape of the heat sinks can be changed freely. Besides, the global parameter has no influence on the optimal constructs based on the two objectives, but the volume fraction does. The increases of the global parameter and the volume fraction reduce the minimum values of the maximum thermal resistance and the equivalent thermal resistance, but the degrees are different. The reduce degree of the global parameter to the minimized equivalent thermal resistance is larger than that to the minimized maximum thermal resistance. The minimized equivalent thermal resistance and the minimized maximum thermal resistance are reduced by 40.03% and 41.42% for a= 0.5, respectively, compared with those for a = 0.3. However, the reduce degree of the volume fraction to the minimized maximum thermal resistance is larger than that to the minimized equivalent thermal resistance. The minimized equivalent thermal resistance and the minimized maximum thermal resistance are reduced by 59.69% and 32.80% for φ= 0.4, respectively, compared with those for φ= 0.3. As a whole, adjusting the parameters of the heat sink to make the equivalent thermal resistance minimum can make the local limit performance good enough at the same time; however, the overall average heat dissipation performance of the heat sink becomes worse when the parameters of the heat sink are adjusted to make the maximum thermal resistance minimum. Thus, it is more reasonable to take the equivalent thermal resistance minimization as the optimization objective when the heat sink is optimized.

###### PHYSICS OF GASES, PLASMAS, AND ELECTRIC DISCHARGES

2015, 64 (20): 205201.
doi: 10.7498/aps.64.205201

Abstract +

Comparing different collisional-radiative models is of great importance for validating the models for plasma spectroscopy and improving the diagnostic accuracy of plasma parameters. In this paper, the widely applied K-shell spectroscopic models, FAC and FLYCHK, are compared based on their calculation results of the aluminum K-shell emissivity and absorption coefficient. The state abundances, K-shell line ratios, K-shell emissivities and absorption coefficients in a wide range of plasma temperatures and densities are calculated and compared, and the reasons for the differences between these two models are discussed. In an electron temperature range from 200 to 800 eV, and an electron density range from 1017 to 1024 cm-3, the Al ions in the plasma are mainly composed of H-like and He-like ions. The ground-state populations of the H-like and He-like ions, calculated from FAC model, are in good agreement with the results from FLYCHK. Number densities of the excited states are two orders or more less than those of the ground states from both the models, and significant differences are observed in the number densities of n=2 and n=3 states of both the H-like and He-like ions. These differences will further result in the differences in spectral line emissivity and their line emissivity ratio, such as He-IC/He-αup and H-βup/He-βup, which are key parameters used to diagnose the electron temperature and density. The line emissivity ratio Ly-αup/(He-αup+He-IC) is less dependent on the electron density, and the difference in line emissivity ratio between the two models mainly lies in the parameter region where both the electron temperature and density are high. The ratio He-IC/He-αup is less dependent on the electron temperature when the electron density is more than 1019 cm-3 while significant differences are observed at a lower electron density.#br#The reason for the difference between the number densities of the low-energy excited states from FAC and FLYCHK models is analyzed by comparing the rate coefficients of various collisional and radiative processes in the rate equation of each state. The differences in the n=2 excited states of H-like ions come from the fact that FAC and FLYCHK models use the detailed-level model and the super-configuration model respectively to construct the rate equations of these states. The FAC model ignores the collisional excitation and de-excitation processes between the n=3 state and higher excitation states (e.g. n = 4) in H-like and He-like ions, which are responsible for the density difference in the n=3 excited state. Higher Rydberg states considered in FLYCHK model do not have any significant influence on the density of the ground-states. The difference in the absorption coefficient between the two models is smaller than that in the emissivity as discussed above, for the absorption coefficient mainly depends on the number density of the ions in ground state.

2015, 64 (20): 205202.
doi: 10.7498/aps.64.205202

Abstract +

LiF is often used as a window in laser-driven shock experiments, which can transmit and reflect visible probe laser. Researches of LiF transparency almost focus on its optical reflectivity compressed by strong shock, but there is almost no research on its optical transmissivity compressed by weak shock. In order to study the optical transmissivity of LiF, the quasi-isentropic compression experiment is carried out on the ShenGuang-III prototype laser facility, in which the velocity interferometer system for any reflector is used to diagnose the optical reflectivity of the quasi-isentropic compression sample CH/Al/LiF. The experimental results indicate that the velocity interferometer fringes are missing in the late stage of this experiment. The probe laser should penetrate LiF before it hits the rear surface of aluminum and the laser reflected by aluminum should penetrate LiF before it is collected by the velocity interferometer system for any reflector. Therefore, the reflectivity diagnosed by the velocity interferometer system for any reflector is the product of the optical reflectivity of aluminum and the optical transmissivity of LiF under the experimental condition. However, there is no research about the optical transmissivity model of thick LiF compressed by laser-driven shock. In this paper, we develop a transmissivity model for transparent window LiF and simulate the optical reflectivity of sample CH/Al/LiF. Firstly, we simulate the temperature and density of the sample by the code for one-dimensional multigroup radiation hydrodynamics (MULTI-1D). Then, based on the resulting temperature and density, we simulate the optical reflectivity of the sample by using the optical reflectivity model of aluminum and the optical transmissivity model of LiF. Without considering the transparency of LiF, the simulated result indicates that there is no signal missing in the late stage, which is different from the experimental result. By considering the transparency of LiF, the simulated result is in good agreement with the experimental result. The simulated result indicates that the formation of the strong shock, because of the later shock's catching up with the early one, obviously reduces the optical transparency of LiF and finally causes the velocity interferometer fringes to disappear. The simulated result also indicates that the energy gap of LiF calculated from density-functional theory is 1-2 eV. In this experiment, when LiF becomes opaque, its temperature is 1 eV and its pressure is 2-3 Mbar.

2015, 64 (20): 205203.
doi: 10.7498/aps.64.205203

Abstract +

A parallel intake diffusion combustion physical model is designed to study the influence of plasma on the secondary combustion of boron-based gas in the after-burning chamber, with excluding mixing effects of the intake air. The flame images of the diffusion combustion of the boron-based gas in the after-burning chamber are obtained by a high-speed photographic apparatus. The diffusion combustion characteristics of the physical model and the secondary ignition distance of boron particles are analyzed. The King ignition model, finite-rate/eddy-dissipation model, particle-trajectory model, RNG k-ε model, and plasma model are adopted to simulate the influence of plasma on the diffusion combustion of boron-based two-phase flow in a certain condition. The results show that the secondary ignition distance of boron particles, which is based on the boron-based flame image, is consistent well with the numerical simulation result, which verifies the accuracy of the boron-based two-phase flow diffusion combustion numerical model and the calculation method. When the boron-based gas passes through the plasma area, the temperature of the boron particles increases while the diameter decreases significantly on their trajectory. The distribution area of the B2O3 mass fraction increases significantly, and more than 70% boron particles reach a 100% combustion efficiency before they arrive at the area of the two-thirds after-burning chamber. More heat is released by fully burning the boron particles under the influence of plasma, which results in a half increase of the central area. It can be indicated that plasma can obviously enhance the combustion process of the boron-based gas, which improves the combustion efficiency of boron particles and releases more energy.

2015, 64 (20): 205205.
doi: 10.7498/aps.64.205205

Abstract +

When high-speed vehicles enter into the atmosphere, plasma sheath may be excited around due to aerodynamic heating, resulting in difficulties in communicating and changes of electromagnetic scattering properties. Those facts have received lots of attention due to their influences on the aerospace communication and radio telemetry applications. While analytic and numerical studies have been carried out by many native institutions on the electromagnetic radiation/scattering problems in the presence of plasma sheath, there remains the lack of measurement data to support and verify those researches. This work reports the backscattering measurements for the target surrounded by plasma sheath in the ground high-enthalpy shock tunnel facility. Using the step frequency sweeping mode of a commercial instrument, i.e., vector network analyzer, we conduct the experiments in the JF-10 high-enthalpy shock tunnel. The dynamic electromagnetic scattering measurement must be completed on a time scale of ms while the shock tunnel is running. The implementation details are demonstrated in this work, including the experimental configurations, data processing procedures, timing synchronization, and discussion on the relationships between the air flow status and measured target scattering signals. The influences of the plasma sheath on the target RCS (radar cross section) in the C band are successfully and clearly observed. The influence of the air flow status on the measured data can be concluded as follows: the front section of high-speed air flow lasting about 0.5-1 ms will change the measured signal dramatically, which should be avoided in observation due to its instability; the effective plasma sheath lasts only about 2 ms, resulting in an overall reduction on the target RCS by about 2 dB in the measurements. Afterwards, the effects by the plasma sheath on the target scattering vanish quickly.

2015, 64 (20): 205206.
doi: 10.7498/aps.64.205206

Abstract +

The stream formation in a 1-atm nitrogen gas switch is investigated by the two-dimensional and three-velocity (2D3V) particles through the cell-Monte Carlo collision (PIC-MCC) simulation and theoretical analysis. For simplicity, two parallel plane electrodes of 0.6 mm width are separated by a distance of 1.6 mm. It is found that the analytical solution of the electron density equation can be used to study the evolution of the plasma before the stream breaks down, for the ionization frequency, mean electron energy and electron drift velocity are all constant. After the breakdown of the stream, random collisions destroy the symmetry of the plasma region and cause plasma to branch. As plasma density increases, the electric field inside the plasma region decreases due to the shielding effect. However, charge densities at both ends of the plasma region increase and the density at the anode end is larger than that at the cathode end, for the plasma exponentially grows as electrons move from the cathode toward the anode. This causes the electric field at the end of plasma near the anode to be larger than that near the cathode. It is found that the electrons can achieve their stable mean energy in several picoseconds due to the high transfer frequency (1011-1012 Hz) of the electron energy in the nitrogen plasma. After the breakdown of the stream, the mean electron energy decreases due to the decrease of the electron energies inside the plasma. By increasing the electrode voltage, it is found that the mean electron energy increases, the electron drift velocity increases linearly, and the variation rate of ionization frequency with electric field is in a range between E4 and E5. Therefore, the time taking for breaking down the stream decreases with the increase of the electrode voltage.

###### CONDENSED MATTER: STRUCTURAL, MECHANICAL, AND THERMAL PROPERTIES

2015, 64 (20): 206101.
doi: 10.7498/aps.64.206101

Abstract +

In order to study the effects of working conditions and solid surface topography on the wettability of material, a series of No. 45 steel specimens with the same surface arithmetic average height Sa and different surface microstructures is designed and manufactured by laser surface texturing. All the surfaces are measured by a non-contact three-dimensional (3D) optical profiler Talysulf CCI Lite and characterized by the ISO25178. A series of wetting experiments is carried out with the No. 32 turbine oil on an optical contact angle and surface tension meter SL200 KS. The effects of temperature, droplet volume and surface structure on the wettability are analyzed. Meanwhile, quantitative research of the relationship between the random characteristics of topography and wettability of the solid surface is conducted with parameters obtained from the ISO25178. Based on the fact that the contact angle is an acute angle, the results show that the contact angle of the droplet on the solid surface decreases rapidly to a stable value in the wetting process. The stable value decreases with the increase of the temperature, while it first increases and then decreases with the increase of the droplet volume. The surface wettability can be affected by the laser micro-texturing. Surfaces with similar values of Sa show different wettabilities for different micro-textures with different shapes and directions. Textured surfaces with grooves along the spreading direction of the droplet perform the best wettability in our research. Results also predicate that the wettability of surface is greatly influenced by the amplitude parameters (Sku, Ssk), spatial parameters (Str, Sal), hybrid parameters (Sdq, Sdr), and feature parameters (Sda, Sdv), which are all obtained from the ISO25178. The wettability of hydrophilic surface becomes better with increasing Sku, Sal, and Sdr and reducing Ssk, Str, and Sdq.

2015, 64 (20): 206201.
doi: 10.7498/aps.64.206201

Abstract +

The effective modulus of transversely isotropic compound material containing aligned ellipsoidal inhomogeneity is derived using the method of sphere-equivalency of effective scattering. Based on this approach, we derive the integral solution of the Eshelby tensor for the anisotropic medium, allowing for numerically evaluating the effects of anisotropy for the solution. The numerical results show that the effective modulus of the medium decreases monotonically with increasing the concentration of the inhomogeneties. The anisotropy increases if the inhomogeneity alignment direction is perpendicular to the TI symmetry axis of the background medium. By reducing the numbers of matrix elastic modulus from 5 to 2, we calculate the slowness surfaces for the three modes of propagation in an isotropic medium containing aligned ellipsoidal inhomogeneity. The results are the same as the existing ones, which validates the exactness of our theory. The modeling results can be used to evaluate elastic property of an anisotropic medium with aligned inclusions, such as earth formation shale rocks containing aligned cracks.

###### CONDENSED MATTER: ELECTRONIC STRUCTURE, ELECTRICAL, MAGNETIC, AND OPTICAL PROPERTIES

2015, 64 (20): 207101.
doi: 10.7498/aps.64.207101

Abstract +

Layered transition metal dichalcogenides (LTMDs) have renewed interest as electronic materials, but the poor conductivities hinder their further development. Chemical doping can often significantly modify atomic structures and electronic functionalities of a wide range of materials and thus acts as one of the most effective ways to precisely tune material properties for technological application. Here, the geometries and band structures as well as the densities of states of pure NbSe2 and Ti-doped NbSe2 nanostructure are studied by employing the ab-initio plane-wave ultra-soft pseudo potential technique based on the density functional theory. We optimize the ground state of NbSe2 in the layered structure by using the generalized gradient approximation for the exchange-correlation potential. The computational structural parameters are in good agreement with experimental values within 2.5%. To investigate the stability of the doped system with changing the concentration of Ti atoms, 2×2×1 2H-NbSe2 supercells are taken into consideration. Meanwhile, we consider a total of three possible Ti-doping models: substitution, intercalation, and embedded model, and investigate the energy band diagrams, state densities and densities of partial wave state diagram before and after the doping. The results show that the energy electron density of states reaches a higher peak, and the band structure near Fermi level (EF) is changed obviously, resulting in the variations of the band gap and EF position and then the increase of electronic conductivity after doping. In addition, our calculations also predict that the electron transport properties can be enhanced by doping Ti and it can be regarded as a useful way to tailor electronic states so as to improve electron transport properties of 2H-NbSe2. Such a remarkable modification of electronic structure of 2H-NbSe2 by chemical doping offers an additional way of modulating performances of LTMDs and developing new electrical contact composite materials.

2015, 64 (20): 207102.
doi: 10.7498/aps.64.207102

Abstract +

In this paper, Ca/Sr atoms are confirmed to have symmetric distributions on 4c sites by using the minimum energy principle, and the stable crystal structure of Ca0.5Sr0.5TiO3 is built. The lattice parameters, elastic constants, bulk modulus, shear modulus, Young's modulus and Poisson's ratio of Ca0.5Sr0.5TiO3 (CST50) are investigated by the plane wave pseuedopotential method based on the first-principles density functional theory within the local density approximate (LDA) and generalized gradient approximation. The properties of planar acoustic velocity are studied by Christoffel equation, and the minimum thermal conductivity is investigated with Cahill and Cahill-Pohl models. The results show that the calculated lattice parameters are consistent with the corresponding experimental values. The larger calculated elastic constasnts C11, C22, and C33 suggest the incompressibility along the principle axes. The bulk modulus B is larger than the shear modulus G; G/BLDA = 0.5789 and G/BGGA = 0.5999, indicating that CST50 is a brittle material. The three-dimensional image of Young's modulus along [100], [010], and [001] crystal orientations shows the anisotropic elasticity of CST50. The planar projections of Young's modulus in (001) and (010) planes show the stronger anisotropy than in (100) plane and all the planar projections have two-fold symmetry. The Poisson's ratio exhibits the incompressbility of CST50. The universal elastic anisotropy indexes ALDAU = 0.0235 and AGGAU= 0.0341 indicate the weak anisotropy of CST50. The planar acoustic wave which has a branch of longitudinal wave and two branches of transverse wave is anisotropic along (010) and (001) planes and isotropic along (100) plane, and all the corresponding planar projections have two-fold symmetry. The minimum thermal conductivity calculated in Cahill model is isotropic in each plane, while the minimum thermal conductivity calculated in Cahill-Pohl model is proportional to the second power of T under low temperatures and reaches a constant at high temperatures. In the quasi harmonic Debye model, the molar heat capacity and thermal expansion coefficient of CST50 are close to those of calcium titanate, indicating that CST50 has the stable thermal expansion property at high temperatures. The direct band gap of CST50 is 2.19 eV and the bottom of the valence band is mainly determined by the electron orbitals of Ti-3d and O-2p. The analysis of the charge populations shows that the covalence of Ti–O is stronger than those of Sr–O and Ca–O, and the band length of Ti–O is shorter than those of Sr–O and Ca–O; (200), (110) and (002) planar contour charge densities indicate that Ti atoms interact strongly with O atoms. The charge population and contour charge density prove that CST50 has a stable Ti–O octahedral structure.

2015, 64 (20): 207103.
doi: 10.7498/aps.64.207103

Abstract +

Novel unconventional physical phenomena, such as metal-insulator transition, high temperature superconductivity, colossal magneto-resistance and quantum criticality, are usually found in transition metal oxides (TMOs) with layered perovskite structures. Great success has been achieved in 3d TMOs, in which the localized 3d states yield strongly correlated narrow bands with a large on-site Coulomb repulsion U and a small band width W. Anomalous insulating behaviors are reported in the 5d TMOs, such as Sr2IrO4 system, which is surprising since the 5d TMOs are usually considered as weakly correlated wide band systems with largely reduced on-site Coulomb repulsion U due to delocalized 5d states. The crystal structure of Sr2IrO4 consists of two-dimensional (2D) IrO2 layers, similar to the parent compound La2CuO4 of the cuprates. Theoretically, a variational Monte Carlo study of Sr2IrO4 suggests that d-wave like superconductivity may appear but only within a narrow region of electron doping. In contrast, an s±*-wave phase is established for hole doping deduced from functional renormalization group, and triggered by spin fluctuations within and across the two conduction bands. Moreover, triplet p-wave pairing state with relatively high transition temperature emerges on the hole-doped side when the Hund's coupling is comparable to spin-orbit coupling. Several experiments are tried to search for the predicted unconventional superconductivity due to both electron-and hole-doping. However, to the best of our knowledge, it has not been found yet in the carrier-doped Sr2IrO4 system. Hence, more detailed studies are needed to explore the potential superconductivity.#br#A series of La doped Sr2-xLaxIrO4 samples is synthesized based on solid state reaction method. The evolution of the crystal structure is studied by the X-ray diffraction, scanning electron microscopy, together with the Raman spectrum. It is found that the crystal constant of the c-axis decreases with increasing doping level as well as the apical Ir—O1 bond length, indicating the lattice construction. Moreover, the distortion of the IrO6 octahedron reduces with increasing doping level. Therefore, blue shift occurs of the Raman scattering peaks. The temperature dependence of the Raman spectrum is also studied. It is found that the frequencies of the A1g and B1g vibration modes increase with temperature decreasing and an abnormal jump occurs around 110 K, which is believed to be correlated with the structural change and the magnetic transition around this temperature.

2015, 64 (20): 207301.
doi: 10.7498/aps.64.207301

Abstract +

Plasmonics with subwavelength characteristics can break the diffraction limit of light and be used to produce the sub-wavelength optoelectronic device, thus it has aroused great interest for decades. Local surface plasmon resonance of metal nanoparticles has become one of the research hotspots due to the fact it can produce extinction and near-field enhancement effect. How to achieve controllable plasmon line shape and generate strong electromagnetic field enhancement is of great significance for improving the sensing performance, nonlinear effect and surface enhanced Raman factor of metallic nanostructures. The optical properties of plasmonic oligomer clusters composed of normal and L-shaped nanrod dimers are investigated by using the finite-difference time-domain method in this paper. There are two energy modes for an L-shaped nanorod due to its shaped anisotropy, where plasmons oscillate along the arms of the L-shaped nanorod or oscillate over the whole length of the L-shaped nanorod. Therefore, two bonding resonances appear in the spectrum of an L-shaped nanorod dimer, while only one bonding resonance exists for normal nanorod dimer. When a normal nanorod dimer and an L-shaped nanorod dimer are aligned together to form a quadrumer, the three bonding resonances can be excited simultaneously and radiative damping can be suppressed effectively around the dip spectral positions. It is shown that the optical responses of quadrumer can be strongly tuned by manipulating the geometry parameters. For example, the coupling between the two dimers can be modified by adjusting the separation, and the three resonances shift toward higher energies with the increasing of the separation. In addition, the optical responses of individual nanorod depend on the corresponding arm length. As a result, the three resonances of the quadrumer can also be well tuned by adjusting the arm length. Comparing the variation of resonance peak positions between L-shaped nanorod dimer and normal nanorod dimer, we can more intuitively understand spectral lineshape variation of quadrumer. These results can be used for guiding the design of nano-photonic devices for plasmonic oligomer clusters and also for developing the application of surface-enhanced Raman scattering and biological sensing.

2015, 64 (20): 207302.
doi: 10.7498/aps.64.207302

Abstract +

The origin of the resistance switching behavior in HfO2 is explained in terms of filament formation/rupture under an applied voltage. In order to investigate the position and process of conductive filament in resistive switching memory, the resistive switching and chemical structure of Ti/HfO2/Pt memory device are studied. Through current-voltage measurement, typical resistive switching behavior is observed in Ti/HfO2/Pt device cells; through detecting Hf 4f with different depths by using X-ray photoelectron spectroscopy. It is observed that the Hf4+ decreases monotonically with depth increasing towards HfO2/Pt interface in low resistance state, while a fluctuation distribution of Hf4+ is shown in high resistance state and in the pristine Ti/HfO2/Pt device. The concentration of Hf4+ in high resistance state is higher than that in low resistance state, which is confirmed by measuring the electron energy loss spectrum. Additionally, the O 1s spectrum shows a similar result consistent with the Hf 4f one. The above result is explained by the existence of locally accumulated oxygen vacancies in the oxide bulk layer in high resistance state and pristine states. It is proposed that the oxygen vacancy clusters dominantly determine the resistivity by the connecting/rupture between the neighbor cluster sites in the bulk. The cluster defects are the preexisting structural distortion/injure by charge trapping defects due to the fixed charge which could confine the nucleation of oxygen vacancies and bigger distortion could be enhanced or recovered via the transportation of oxygen vacancies under the external voltage. Oxygen vacancies are driven away from the clusters under SET electrical stimulus, and then recover back to original cluster sites under RESET process.#br#The previous presumption of the ideal evenly-distributed state for oxygen vacancies in the bulk of resistance random access memories (RRAMs) device leads to an issue about where the filaments occur/form first since the oxygen vacancy defects show uniform distribution in the active oxide bulk layer. Since the conductive filament is easily formed in the cluster region of oxygen vacancies, this study could provide a deep understanding of the formation of conductive filament in RRAMs device.

2015, 64 (20): 207303.
doi: 10.7498/aps.64.207303

Abstract +

La-doped BaSnO3 is regarded as a very essential material to construct transparent perovskite devices due to its super high electrical mobility in perovskite transparent conducting oxides. For understanding the high electrical mobility, the effective mass of the carrier in La-doped BaSnO3 is a critical factor and should be determined. In this work, the performances of epitaxial La-doped BaSnO3 thin films grown on (LaAlO3)0.3 (SrAl0.5Ta0.5O3)0.7 (001) substrate by radio-frequency (RF) magnetron sputtering technique are investigated. The electrical properties (resistivity, carrier density, mobility and Seebeck coefficient) and the optical transmittance are analyzed. In addition, it is proved from both the Hall effect and thermoelectric power measurements that the La-doped BaSnO3 thin films are n-type degenerate semiconductor. At 300 K, the resistivity, carrier density, mobility and Seebeck coefficient are 0.987 mΩ·cm, 2.584×1020 cm-3, 24.49 cm2·V-1·s-1 and 45.71 μV/K, respectively. The electron effective mass ～ 0.31m0 (m0, the free electron mass) is extracted by combining the Seebeck coefficient and carrier density. Ba0.99La0.01SnO3 (BLSO) thin film exhibits a high optical transmittance of 73% in the visible spectral region. In order to derive the band-gap energy, the complex dielectric constant, and the film thickness, the transmittance spectrum is simulated based on the dielectric model comprising the band-gap transition (O'Leary-Johnson-Lim model) and free electron excitation (Drude-Lorentz model). The band-gap energy, exponential band tail and thickness of the BLSO thin film are 3.43 eV, 0.27 eV and 781.2 nm, respectively. Wavelength-dependence of complex dielectric function of the BLSO thin film is also obtained from the fitted line. Additionally, the parameters (optical carrier density and mobility) resulting from the optical measurement are in agreement with the results from the electrical measurement, which supports the calculated electron effective mass aforementioned.

2015, 64 (20): 207304.
doi: 10.7498/aps.64.207304

Abstract +

A two-terminal Aharonov-Bohm (A-B) interferometer coupled with linear di-quantum dot molecules is presented. By employing Keldysh non-equilibrium Green's function technique, the conductance without introducing time-dependent external field and the average current with applying time-dependent external field are theoretically studied. In the absence of time-dependent external field, two identical linear diquantum dot molecules embedded respectively in the two arms of A-B interferometer lead to degeneracy energy levels. The central resonance peak at εd = 0 in the conductance spectrum splits into two resonance peaks as the inter-coupling strength of di-quamtum dot increases over a threshold. In the case that the two linear di-quantum dot molecules are different, three or four resonance peaks appear in the conductance spectrum. When tuning magnetic flux ψ= π, the destructive quantum interference of electron waves in the A-B interferometer takes place. The conversion between 0 and 1 for conductance is performed by switching on/off the magnetic flux, which suggests a new physical scheme of quantum switches. The effect of Rashba spin-orbit interaction on the conductance is discussed. The functionality of spin filter is suggested through adjusting the Rashba spin-orbit coupling strength and the external magnetic flux. When time-dependent external field is applied, the notable side-band effect appears in the average current curve. A series of resonance peaks is produced, with the peak-peak separation of ħω. Two main peaks become reduced as the amplitude of time-dependent external field increases, however, the sideband peaks grow gradually. This indicates that both the magnitude and the position of average current resonance peak are controllable by adjusting the amplitude of time-dependent external field. The sideband effect remains always in the average current curve no matter how much the frequency of time-dependent external field changes. But the increase in the frequency of external field leads to the growth of two main peaks at the bonding and anti-bonding energy respectively, and the decay of the corresponding sideband peaks as well. The conversion between the current peak and valley can be realized by tuning the frequency of time-dependent field. Moreover, the dependence of A-B effect of the average current on the magnetic flux is found. As the magnetic flux is ψ≠nπ, each peak in average current curves splits into two peaks. But under the condition of ψ=2nπ, the splitting phenomenon disappears. The spin-dependent average current shows effective controllability by tuning the magnetic flux and Rashba spin-orbit coupling. The results would be useful for gaining a physical insight into electron transport in the multi-quantum-dot molecules coupled A-B interferometer and for designing the quantum devices.

2015, 64 (20): 207801.
doi: 10.7498/aps.64.207801

Abstract +

Photoacoustic imaging has recently emerged as a promising imaging modality for prostate cancer. As ausual light illumination model in the previous studies, the external light illumination is difficult to obtain an accurate reconstructed photoacoustic image. It suffers a great deal of light absorption attenuation by the surrounding scattering tissue and cannot colletct sufficient ultrasound signals for image reconstruction. Some particular methods are required to be considered in the photoacoustic imaging technique for examining prostate, such as a light delivery to prostate with sufficient penetrating depth and minimal invasiveness. According to the structural characteristic of prostate tissue, a photoacoustic imaging system is built by using a novel technique for prostate in this paper. In our photoacoustic imaging system, a cylindrical diffusing source with a 2-cm-long diffuser tip is used for an internal light irradiation through a urethra, and a focused transducer with a 3.5 MHz central frequency and 30.3 mm extended focal zone is located in the rectum for scanning the photoacoustic signal. Phantom experimental imaging is carried out. In the experiment, a transverse resolution of 2.21 mm and an axial resolution of 0.39 mm are obtained. The results demonstrate that the system could achieve the accurate imaging position of the absorber in the tissue sample. Because of the symmetrical emitting of the cylindrical diffusing light source and a relatively better lateral uniformity of light absorption around the light source through the internal irradiation model via urethra, light absorption of the upper side of the light source is almost the same as that of the lower side. Therefore the lengthways and lateral imaging ranges can be improved. In addition, the laser energy is allowed to be increased appropriately to obtain a further imaging result without worrying about heat damages to normal tissues, for the light absorption is less around the cylindrical diffusing light source. In conclusion, the preliminary studies show that the new technique, where the internal light irradiation is implemented by using a cylindrical diffusing source and a focused transducer with extended focal zone, has a potential application in the early noninvasive diagnosis of prostate cancer.

2015, 64 (20): 207802.
doi: 10.7498/aps.64.207802

Abstract +

Unlike the general substrates such as SiO2, ITO, and AZO, the metal foil used as a substrate is rarely studied in application in the substrate, however, it has lots of advantages including cheapness, good conductivity and excellent scalability. In this paper, an acanthosphere-like structure named ZnO nanoflowers is successfully synthesized on Cu foil by using chemical vapor deposition method. The gas flows with oxygen-argon ratios ranging from 1 : 150, 1 : 200, 1 : 250 to 1 : 400, which impacted on Cu foil, and the property of the ZnO nanoflowers are carefully studied. The SEM images shown that there are lots of ZnO nanorods grown on the sphere cores, and look like flowers. The ZnO nanoflowers contains uniformly sized ZnO nanorods and morphology with best flower structure when the oxygen/argon gas flow ratio is 1 : 250. Furthermore, the length-diameter ratio of the ZnO nanorods on the ZnO nanoflowers decreases as the oxygen-argon gas flow ratio decreases. The ZnO is of hexagonal wurtzite structure indicated by XRD pattern and there exist no other diffraction peaks existence except those from the Cu foil. In addition, the photoluminescence of ZnO nanoflower changes from a wave packet into a broad peak in the visible region when the oxygen-argon gas flow ratio between decreases. Further study of the photoluminescence by fitting the peaks in visible region with gaussian function indicates that the photoluminescence relating to the oxygen vacancy defects increases, but that relating to the zinc vacancy defects decreases. Therefore, the white light emitting device may be constructed based on the ZnO nanoflowers studied shown above. Finally, a possible model of the ZnO nanoflowers grown on Cu foil is proposed based on the experimental results.

2015, 64 (20): 207901.
doi: 10.7498/aps.64.207901

Abstract +

An analytical model of secondary electron (SE) emission (SEE) from metal surface with regular structure is presented. In this model, the quantitative relationship between the SE emission yield (SEY) and surface topography is examined. Using the idea of multi-generation for SE emission, the first-generation of SEs is considered as being dominant in total SEs. The shielding effect of the surface structures on the SE is found to be the main factor influencing final SEY. On the basis of the cosine distribution of secondary electrons emission direction, the quantitative relationship between the SEY and surface topography parameters is revealed. Then taking the rectangular and triangular grooves for example, the analytical formulas of first-generation SEY are derived for both normal and oblique incidence. The analytical results are then verified with the Monte Carlo simulation results and experimental data. The results show that a rectangular groove with a bigger depth-to-width ratio can suppress the SEE more efficiently. For a triangular groove, owing to having both enhancing and suppressing effects on SEE, a small groove angle is required for effective SEE suppression. The present analytical model gives an insight into the relationship between the SEY and the surface topography parameters and is helpful for the structure design to modify SEY.

2015, 64 (20): 207902.
doi: 10.7498/aps.64.207902

Abstract +

Multipactor discharge always causes disastrous damage to a vacuum window in high power microwave system, which actually becomes a limiting factor for the output power of vacuum device. To explore the multipactor phenomenon of complicated pill-box window, the mulitpactor in normal field between the metal boundary and the window disk is studied. Through Monte Carlo (MC) simulations, the susceptive curve is fitted and analyzed. The secondary electrons' avalanche behavior under the normal RF field is discussed. It is noticed that the one-sided multipactor is excited within a very limited Vrf-fD region when two-sided multipactor is excited initially. The development condition from two-sided multipactor to the one-sided multipactor is proposed. Through analyzing and MC simulation, the condition is achieved. When the normal RF electric field can satisfy the phase focus conditions of one-sided multipactor, the two-sided multipactor will develop into one-sided multpactor and then reach a saturation value. Meanwhile, the initial effect of electrostatic field on one-sided multipactor is also discussed. On condition that two-sided multipactor can be excited, the number of secondary electrons can increase up to a saturation value when Edc0 is lower than the minimal saturate value of Edc. When Edc0 is lager than the minimal saturate value of Edc and in the Edc/Erf threshold of one-sided resonant multipactor, the number of secondary electrons can also increase to a saturate value. However, When Edc0 is lager than the minimal saturate value of Edc but beyond the Edc/Erf threshold of one-sided resonant multipactor, secondary electrons will be suppressed.

###### INTERDISCIPLINARY PHYSICS AND RELATED AREAS OF SCIENCE AND TECHNOLOGY

2015, 64 (20): 208101.
doi: 10.7498/aps.64.208101

Abstract +

The mechanical strengths of silicon wafers are crucial for the manufacturing yield of integrated circuits (ICs), which have received intensive attention over the years. With reducing the feature size of ICs, the mechanical strengths of silicon wafers become more significant. Actually, the gliding of indentation dislocations on single-crystalline silicon wafers at a given temperature reflects the mechanical strengths of silicon wafers. Since the gliding of indentation dislocations is driven by the residual stress around the indentation, the investigation on the correlation between the residual stress and dislocation gliding is of significance. In this paper, we first use micro-Raman microscopy to characterize the relief of stress around the indentation due to the annealling at 300 or 500 ℃. Then the effect of such a relief-stress on the gliding of indentation dislocations at 700-900 ℃ is investigated. In the case without the prior stress-relief, the indentation dislocations glide to the maximum distance after 2 h annealling at 700-900 ℃. With the prior stress-relief due to the annealling at 300 or 500 ℃, the indentation dislocations can still glide to the maximum distance after 2 h annealling at 900 ℃, however the gliding velocity significantly decreases and the gliding distance is remarkably reduced after 2 h annealling at 700 or 800 ℃. Such a reduction of gliding distance is most significant in the case of 700 ℃ annealling following the stress-relief with the 500 ℃/2 h annealling. Despite the prior stress-relief, as long as the annealing time at 700 or 800 ℃ is sufficiently extended, the indentation dislocations can glide to the maximum distance. In view of the above results, it is believed that the maximum gliding distance of indentation dislocations at a given temperature is independent of the values of residual stress around the indentation provided that the residual stresses are larger than the critical stress for driving the dislocation movement. Nevertheless, the annealing time for achieving the maximum gliding distance at a given temperature should be remarkably extended as the residual stresses around the indentation are relieved.

2015, 64 (20): 208102.
doi: 10.7498/aps.64.208102

Abstract +

In a crystal growth system, the crystal quality is greatly affected by the coupling properties between unsteady melt flow and thermal transfer. In this paper, an improved lattice Bolzmann method is proposed. This incompressible axisymmetric model based method transforms the fluid equations of cylindrical coordinate into those of the two-dimensional Cartesian coordinate and constructs the evolutionary relationship of the external force terms, such as rotational inertia force and the thermal buoyancy. In the unsteady melt, the temperature distribution and the rotational angular velocity are determined based on the D2Q4 model and the velocity of axisymmetric swirling fluid is calculated based on the D2Q9 model. The mirror bounce format is adopted as the boundary conditions of the free surface and the axis symmetry. For the remaining boundary conditions, the non-equilibrium extrapolation format is used. In the simulation, 12 sets of flow function results are obtained by choosing different sets of Grashof number and Reynolds number. By comparing with the finite crystal growth results, the effectiveness of the proposed method can be shown. Furthermore, by studying the convection shape and the temperature distribution of the melt under coupling between high Grashof number and high Reynolds number, it can be concluded that the thermal coupling properties and flow in the unsteady melt relate to Grashof number and Reynolds number. By adjusting the high Reynolds number generated by the crystal and crucible rotation, the strength of the forced convection in the melt can be changed. Therefore, the natural convection in the melt can be suppressed effectively and the temperature distribution results can be improved significantly. In addition, it is worth mentioning that the findings in this paper can be straightforwardly extended to the silicon single crystal growth experiment by turning the dimensionless crystal rotation Reynolds number and crucible rotation Reynolds number into the actual rotation speed.

2015, 64 (20): 208701.
doi: 10.7498/aps.64.208701

Abstract +

It has been widely observed that atherosclerotic diseases occur in regions with complex hemodynamics, such as artery bifurcations and regions of high curvature. These regions usually have low or oscillatory wall shear stress, which is a main factor that results in thrombus formation. In addition, after the thrombosis, the stenosis will in turn affect the hemodynamics. In the blood circulation, the abnormal substances that do not dissolve in the blood can block the vascular cavity, which is called embolism. These substances such as fat particles are called embolus. Embolism results in high velocity and wall shear stress (WSS), which is harmful to the vessel wall. Embolism leads to stroke easily, resulting in the death of the patient. Here, the authors focus on the formation process of fat embolism and its influence on hemodynamics. In order to investigate the influence of various factors on the movement of a fat particle, we carry out the single factor simulation. Fat particles do not dissolve in the blood and easily adhere to the vessel wall. We use a virtual fluid that represents the fat particles. In the present work, a two-dimensional (2D) carotid bifurcation is established, and the simulation is carried out by the computational fluid dynamics software. The movement of the fat particles relies on the thrust and surface friction of the blood, and the values of thrust and surface friction are governed by the blood velocity, viscosity and the diameter of the fat particle, which has little relationship with the density, especially for a blood vessel that is not too long. The fat particles can smoothly pass through the carotid sinus when the vessel is 0 or 25% stenosed, which indicates that the embolism does not occur and the fat particle does not adhere to the vessel wall. Small deformation occurrs when the vessel is 25% stenosed for 0.6 s. When the degree of stenosis increases to 50%, the fat particles partially blocks the vascular cavity. We give an experiment about the influences of the stenosis on the movement of two fat particles and thrombus. When the carotid sinus is 0 or 25% stenosed, the two fat particles adhere to the vessel wall at the end of internal carotid artery (ICA), resulting in fat embolism. One fat article is on the upper wall of ICA and the other is on the lower wall. At the end of ICA, the vascular diameter becomes smaller and the two fat particles cannot pass through it. When the carotid sinus is 50% stenosed, the two fat particles merge into a larger one and partially block the narrow vascular cavity, impeding blood circulation. The findings in this paper may help hematological experts to check the spread of atherosclerotic disease. The main conclusions drawn from the present study are a) the vascular stenosis has an important influence on the movement of fat particles and the formation of embolism as well; b) the fat particles may adhere to the vessel wall, and due to the flow of blood, the fat particles spread slightly on the wall; c) the region behind the stenosis might be the next site of thrombosis; d) with the movement of fat particles, the maximum negative WSS increases slightly, and its coordinate position moves to the right; e) when embolism occurs, the velocity and WSS distribution become very complex, which is harmful to the vessel.

2015, 64 (20): 208901.
doi: 10.7498/aps.64.208901

Abstract +

Extensive studies have shown that the fractal scaling exists widely in real complex systems, and the fractal structure significantly affects the spreading dynamics on the networks. Although node influence in spreading dynamics of complex networks has attracted more and more attention, systematical studies about the node influence of fractal networks are still lacking. Based on the flower model, node influences of the fractal scale-free structures are studied in this paper. Firstly, the node influences of different fractal dimensions are compared. The results indicate that when the fractal dimension is very low, the discriminability of node influences almost does not vary with node degree, thus it is difficult to distinguish the influences of different nodes. With the increase of fractal dimension, it is easy to recognize the super-spreader from both the global and local viewpoints. In addition, the network noise is introduced by randomly rewiring the links of the original fractal networks, and the effect of network noise on the discriminability of node influence is analyzed. The results show that in fractal network with low dimension, it becomes easier to distinguish the influences of different nodes after adding network noises. In the fractal networks of infinite dimensions, the existence of network noises makes it possible to recognize the influences of medium nodes. However it is difficult to recognize the influences of central nodes from either the global or local perspective.