The geometric phase has become a fundamental concept in many fields of physics since it was revealed. Recently, the study of the geometric phase has attracted considerable attention in the context of quantum phase transition, where the ground state properties of the system experience a dramatic change induced by a variation of an external parameter. In this work, we experimentally measure the ground-state geometric phase of the three-spin $XY$ model by utilizing the nuclear magnetic resonance technique. The experimental results indicate that the geometric phase could be used as a fingerprint of the ground-state quantum phase transition of many-body systems.

A Schottky barrier diode with low-barrier is presented, based on which a terahertz waveguide detector working at 500–600 GHz is designed and fabricated. By using the InGaAs/InP material system, the feature of the low barrier is obtained which greatly improves the performance of the detector. The measured typical voltage responsivity is about 900 V/W at 500–560 GHz and is about 400 V/W at 560–600 GHz. The proposed broadband waveguide detector has the characteristics of simple structure, compact size, low cost and high performance, and can be used in a variety of applications such as imaging, molecular spectroscopy and atmospheric remote sensing.

It is crucial to determine the spin and parity ($J^{P}$) of ${Z_{\rm c}(3900)^\pm}$ for understanding its structure. We perform a helicity amplitude analysis on ${Z_{\rm c}(3900)^\pm}$ in the process ${e^+e^-}\to\pi^+\pi^-{J/\psi}$ with the hypotheses $J^{P}=0^-$, $1^+$, $1^-$, $2^+$ and $2^-$. To estimate the significance of $J^{P}=1^+$ over other hypotheses, we perform a Monte Carlo simulation study, and we also discuss the prospect of measuring the spin parity in the future experiment with a large data sample.

An accurate theoretical study on the MgH radical is reported by adopting the high-level relativistic MRCI+Q method with a quintuple-zeta quality basis set. The reliable potential energy curves of the five ${\it \Lambda}$-S states of MgH are derived. Then the associated spectroscopic parameters are determined and found to be in good accordance with the available experimental results. The permanent dipole moments (PDMs) and the spin–orbit (SO) matrix elements of ${\it \Lambda}$-S states are computed. The results show that the abrupt changes of PDMs and SO matrix elements are attributed to the variations of electronic configurations at the avoided crossing point. The SOC effect leads to the five ${\it \Lambda}$-S states split into ten ${\it \Omega}$ states and results in the double potential well of (2)1/2 state. Finally, the transition properties from the (2)1/2, (1)3/2 and (3)1/2 states to the ground state $X^{2}{\it \Sigma}$+1/2 transitions are obtained, including the transition dipole moments, Franck–Condon factors and radiative lifetimes.

We report the initial test results of a rubidium ($^{87}$Rb) space cold atom clock (SCAC). The space-qualified $^{87}$Rb SCAC is composed of the physical package, the optical bench, the microwave synthesizer and the control electronics. After the system is integrated, about 10$^{8}$ $^{87}$Rb cold atoms are captured by magneto-optical trap. The linewidth of the Ramsey fringe is about 10 Hz for the free evolution time of 50 ms on the ground, and the signal-to-noise ratio is measured to be larger than 300. We demonstrate a good medium-term fractional frequency stability of $1.5\times10^{-14}$@1000 s in the closed-loop operation on the ground. The main effects of the noise on the stability are also presented, and the optimized operating parameter is analyzed for the operation of SCAC in the microgravity environment.

We propose a method to determine the optimal power of the microwave resonance transition that simultaneously improves the signal-to-noise ratio and reduces line width based on saturation broadening theory and experiment. Saturation broadening spectra of the ground state hyperfine transition of trapped $^{199}$Hg$^{+}$ ions are measured and analyzed. The value of the optimal microwave power is obtained by using the proposed method and is verified. Rabi oscillations decay spectra of trapped $^{199}$Hg$^{+}$ ions are observed and the optimal microwave irradiation time for the maximum transition signal intensity is determined. This work will help to improve the short-term frequency stability of the mercury ion microwave frequency standard.

A seven-core photonic crystal fiber (PCF) is fabricated and shown to possess a Gaussian-like far-field-intensity distribution. The seven-core PCF, designed with double-cladding structure and zero dispersion wavelength at 927 nm, is utilized to build up a 104 W all-fiber-integrated supercontinuum (SC) source with total conversion efficiency up to 74.3%. The average output power of SC can be further scaled based on this multi-core PCF.

The intensity distribution in the focal region of a high-NA lens for the incident azimuthally polarized multi Gaussian beam transmitted through a multi belt spiral phase hologram is studied on the basis of the vector diffraction theory. Here we report a new method used to generate a needle of transversely polarized light beam with sub diffraction beam size of 0.366$\lambda$ that propagates without divergence over a long distance of about 22$\lambda$ in free space. We also expect that such a light needle of transversely polarized beam may find its applications in utilizing optical materials or instruments responsive to the transversal field only.

We present a bidirectional reflection distribution function (BRDF) model for thermal coating surfaces based on a three-component reflection assumption, in which the specular reflection is given according to the microfacet theory and Snell's law, the multiple reflection is considered $N$th cosine distributed, and the volume scattering is uniformly distributed in reflection angles according to the experimental results. This model describes the reflection characteristics of thermal coating surfaces more completely and reasonably. Simulation and measurement results of two thermal coating samples SR107 and S781 are given to validate that this three-component model significantly improves the modeling accuracy for thermal coating surfaces compared with the existing BRDF models.

A monolithically active–passive integrated colliding pulse mode-locked semiconductor laser is demonstrated in the InGaAsP/InP material system. The device is mode locked at the second harmonic passive mode-locking regime with a wide mode-locking range. Pulse trains with the repetition rate of 40 GHz, 3-dB rf line width of 25 kHz, the pulse width of 2.5 ps, and a nearly transform-limited time–bandwidth product of 0.53 are obtained.

With spin-polarized-dependent band gap renormalization effect taken into account, the energy-dependent evolution of electron spin polarization in GaAs is calculated at room temperature and at a low temperature of 10 K. We consider the exciting light with right-handed circular polarization, and the calculation results show that the degree of electron spin polarization is dependent strongly on the quasi-Fermi levels of $|1/2\rangle$ and $|-1/2\rangle$ spin conduction bands. At room temperature, the degree of electron spin polarization decreases sharply from 1 near the bottom of the conduction band, and then increases to a stable value above the quasi-Fermi level of the $|-1/2\rangle$ band. The greater the quasi-Fermi level is, the higher the degree of electron spin polarization with excessive energy above the quasi-Fermi level of the $|-1/2\rangle$ band can be achieved. At low temperature, the degree of electron spin polarization decreases from 1 sharply near the bottom of the conduction band, and then increases with the excessive energy, and in particular, up to a maximum of 1 above the quasi-Fermi level of the $|1/2\rangle$ band.

We demonstrate a cw and actively Q-switched Er:LuAG laser resonantly dual-end-pumped by 1532 nm fibre-coupled laser diodes. A maximum cw output power of 1.9 W at 1650.3 nm is obtained at a pump power of 25.5 W, corresponding to a slope efficiency of 43.3%. In the Q-switched regime, the maximum pulse energy of 3.51 mJ is reached at a pulse repetition rate of 100 Hz, a pulse duration of 90.5 ns and a pump power of 25.5 W. At the repetition rate of 400 Hz, the output energy is 2.12 mJ, corresponding to a pulse duration of 125.4 ns.

La-doped and undoped $x$BiInO$_{3}$-$(1-x)$PbTiO$_{3}$ (BI-PT) thin films are deposited on (101)SrRuO$_{3}$/(100)Pt/(100) MgO substrates by the rf-magnetron sputtering method. The structures of the films are characterized by XRD and SEM, and the results indicate that the thin films are grown with mainly (100) oriented and columnar structures. The ferroelectricity and piezoelectricity of the BI-PT films are also measured, and the measured results illustrate that both performances are effectively improved by the La-doping with suitable concentrations. These results will open up wide potential applications of the films in electronic devices.

The onset of solutal convection during the directional solidification of Bridgman type of liquid Al-3.5 wt%Li is studied. Based on the analysis of a liquid-inhomogeneous-porous-double-layer system, a bimodal feature of neutral stability curve is found. The pulling rate is ascertained as the governing parameter for the mode transition, i.e., it determines whether the microstructure in the mushy layer is related to convection after the system destabilizes.

A Monte Carlo code is developed to study mega-electron-volt (MeV) electron scattering and transport in plasma based on multiple scattering. A scaling law relating the angular width of a scattered beam to the incident electron energy and the areal density of plasma is found, which may provide a method of MeV electron radiography for diagnosing the areal density of high-temperature, dense plasma under fusion conditions. The study on the MeV electron beam radiography also shows that plasma density interfaces could be discriminated by electron scattering.

Reflections of a Korteweg–de Vries (KdV) solitary wave and an envelope solitary wave are studied by using the particle-in-cell simulation method. Defining the phase shift of the reflected solitary wave, we notice that there is a phase shift of the reflected KdV solitary wave, while there is no phase shift for an envelope solitary wave. It is also noted that the reflection of a KdV solitary wave at a solid boundary is equivalent to the head-on collision between two identical amplitude solitary waves.

Cylindrical and spherical dust-electron-acoustic (DEA) shock waves and double layers in an unmagnetized, collisionless, complex or dusty plasma system are carried out. The plasma system is assumed to be composed of inertial and viscous cold electron fluids, nonextensive distributed hot electrons, Maxwellian ions, and negatively charged stationary dust grains. The standard reductive perturbation technique is used to derive the nonlinear dynamical equations, that is, the nonplanar Burgers equation and the nonplanar further Burgers equation. They are also numerically analyzed to investigate the basic features of shock waves and double layers (DLs). It is observed that the roles of the viscous cold electron fluids, nonextensivity of hot electrons, and other plasma parameters in this investigation have significantly modified the basic features (such as, polarity, amplitude and width) of the nonplanar DEA shock waves and DLs. It is also observed that the strength of the shock is maximal for the spherical geometry, intermediate for cylindrical geometry, while it is minimal for the planar geometry. The findings of our results obtained from this theoretical investigation may be useful in understanding the nonlinear phenomena associated with the nonplanar DEA waves in both space and laboratory plasmas.

The solutions of incompressible ideal Hall magnetohydrodynamics are obtained by using the traveling wave method. It is shown that the velocity and magnetic field parallel to the wave vector can be arbitrary constants. The velocity and magnetic field perpendicular to the wave vector are both helical waves. Moreover, the amplitude of the velocity perpendicular to the wave vector is related to the wave number and the circular frequency. In addition, further studies indicate that, no matter whether the uniform ambient magnetic field exists or not, the forms of the travelling wave solutions do not change.

We observe the spectra of molybdenum for the first time since the first wall of our experimental advanced superconducting tokamak (EAST) was changed mainly to molybdenum tiles. A large amount of molybdenum accumulated in the central plasma where the long-lived $m=1$ mode instability bursts is shown. Molybdenum is proved to be the main impurity species observed during the formation and lifetime of impurity-induced long-lived $m=1$ mode. This may indicate that a close relationship exists between the high-$Z$ impurity accumulation and the occurrence of long-lived $m=1$ mode in EAST plasmas.

Zinc oxide (ZnO) is one of the most promising and frequently used semiconductor materials. In-doped nanostructure ZnO thin films are grown on p-type gallium nitride substrates by employing the simultaneous rf and dc magnetron co-sputtering technique. The effect of In-doping on structural, morphological and electrical properties is studied. The different dopant concentrations are accomplished by varying the direct current power of the In target while keeping the fixed radio frequency power of the ZnO target through the co-sputtering deposition technique by using argon as the sputtering gas at ambient temperature. The structural analysis confirms that all the grown thin films preferentially orientate along the $c$-axis with the wurtzite hexagonal crystal structure without having any kind of In oxide phases. The presenting Zn, O and In elements' chemical compositions are identified with EDX mapping analysis of the deposited thin films and the calculated $M$ ratio has been found to decrease with the increasing In power. The surface topographies of the grown thin films are examined with the atomic force microscope technique. The obtained results reveal that the grown film roughness increases with the In power. The Hall measurements ascertain that all the grown films have n-type conductivity and also the other electrical parameters such as resistivity, mobility and carrier concentration are analyzed.

Substantial defects are produced in Al$_{2}$O$_{3}$ by 4 MeV Au ion irradiation with a fluence of $4.4\times10^{15}$ cm$^{-2}$. Rutherford backscattering spectrometry/channeling and cross-sectional transmission electron microscopy methods are used to investigate the irradiation damage. The 190 keV H ions with a fluence of $1\times10^{17}$ cm$^{-2}$ are used for implanting pristine and Au ion irradiated Al$_{2}$O$_{3}$ to explore the irradiation damage effects on the hydrogen retention in Al$_{2}$O$_{3}$. The time-of-flight secondary ion mass spectrometry method is used to obtain the single hydrogen depth profile and ions mass spectra (IMS), in which we find that implanted hydrogens interacted with defects produced by Au ion irradiation. In IMS, we also obtain the hydrogen retention at a certain depth. Comparing the hydrogen retention in different Al$_{2}$O$_{3}$ samples, it is concluded that the irradiation damage improves the tritium permeation resistance property of Al$_{2}$O$_{3}$ under given conditions. This result means that Al$_{2}$O$_{3}$ may strengthen its property of reducing tritium permeation under the harsh irradiation environment in fusion reactors.

As a potential flexible substrate for flexible electronics, a polymer-sandwiched ultra-thin silicon platform is studied. SU-8 photoresist coated on the silicon membrane improves its flexibility as shown by an ANSYS simulation. Using the plasma enhanced chemical vapor deposited SiO$_{2}$/Si$_{3}$N$_{4}$ composite film as an etching mask, a $4''$ silicon-(100) wafer is thinned to 26 μm without rupture in a 30 wt.% KOH solution. The thinned wafer is coated on both sides with 20 μm of SU-8 photoresist and is cut into strips. Then the strips are bent by a caliper to measure its bending radius. A sector model of bending deformation is adopted to estimate the radius of curvature. The determined minimal bending radius of the polymer-sandwiched ultra-thin silicon layer is no more than 3.3 mm. The fabrication process of this sandwich structure can be used as a post-fabrication process for high performance flexible electronics.

The recently discovered tetragonal, monoclinic and orthorhombic polymorphs of $M_{3}$N$_{4}$ ($M$=C, Si, Sn) are investigated by using first-principles calculations. A set of anisotropic elastic quantities, i.e., the bulk and shear moduli, Young's modulus, Poisson ratio, $B/G$ ratio and Vickers hardness of $M_{3}$N$_{4}$ ($M$=C, Si, Sn) are predicted. The quasi-harmonic Debye model, assuming that the solids are isotopic, may lead to large errors for the non-cubic crystals. The thermal effects are obtained by the traditional quasi-harmonic approach. The dependences of heat capacity, thermal expansion coefficient and Debye temperature on temperature and pressure are systematically discussed in the pressure range of 0–10 GPa and in the temperature range of 0–1100 K. More importantly, o-C$_{3}$N$_{4}$ is a negative thermal expansion material. Our results may have important consequences in shaping the understanding of the fundamental properties of these binary nitrides.

Cu/W multilayer nanofilms are prepared in pure Ar and He/Ar mixing atmosphere by the rf magnetron sputtering method. The cross-sectional morphology and the defect distribution of the Cu/W multilayer nanofilms are characterized by scanning electron microscopy and Doppler broadening positron annihilation spectroscopy. The results show that plenty of point defects can be produced by introducing He during the growth of the multilayer nanofilms. With the increasing natural storage time, He located in the near surface of the Cu/W multilayer nanofilm at room temperature could be released gradually and induce the segregation of He-related defects due to the diffusion of He and defects. However, more He in the deep region spread along the interface of the Cu/W multilayer nanofilm. Meanwhile, the layer interfaces can still maintain their stability.

Tin oxide (SnO$_{2}$) is one of the most promising transparent conducting oxide materials, which is widely used in thin film gas sensors. We investigate the dependence of the deposition time on structural, morphological and hydrogen gas sensing properties of SnO$_{2}$ thin films synthesized by dc magnetron sputtering. The deposited samples are characterized by XRD, SEM, AFM, surface area measurements and surface profiler. Also the H$_{2}$ gas sensing properties of SnO$_{2}$ deposited samples are performed against a wide range of operating temperature. The XRD analysis demonstrates that the degree of crystallinity of the deposited SnO$_{2}$ films strongly depends on the deposition time. SEM and AFM analyses reveal that the size of nanoparticles or agglomerates, and both average and rms surface roughness is enhanced with the increasing deposition time. Also gas sensors based on these SnO$_{2}$ nanolayers show an acceptable response to hydrogen at various operating temperatures.

The searches for large-gap quantum spin Hall insulators are important for both practical and fundamental interests. In this work, we present a theoretical observation of the two-dimensional fully fluorinated stanene (SnF) by means of density functional theory. Remarkably, a significant spin-orbit coupling is observed for the SnF monolayer in the valence band at the ${\it \Gamma}$ point, with a considerable indirect band gap of 278 meV. The direct gap of the SnF monolayer is at the ${\it \Gamma}$ point, which is slightly larger by as much as 381 meV. In addition, the elastic modulus of the SnF monolayer is about 20 J/m$^{2}$, which is comparable with the in-plane stiffness of black phosphorus monolayer along the $x$-direction ($\sim$28.94 J/m$^{2})$. Finally, the optical properties of stanene, SnF monolayer and stanene/SnF bilayer are calculated, in which the stanene/SnF bilayer is supposed to be an attractive sunlight absorber.

Surface leakage currents of AlGaN/GaN high electron mobility transistors are investigated by utilizing a circular double-gate structure to eliminate the influence of mesa leakage current. Different mechanisms are found under various passivation conditions. The mechanism of the surface leakage current with Al$_{2}$O$_{3}$ passivation follows the two-dimensional variable range hopping model, while the mechanism of the surface leakage current with SiN passivation follows the Frenkel–Poole trap assisted emission. Two trap levels are found in the trap-assisted emission. One trap level has a barrier height of 0.22 eV for the high electric field, and the other trap level has a barrier height of 0.12 eV for the low electric field.

Spinel (001)-orientated Mn$_{3}$O$_{4}$ thin films on Nb-doped SrTiO$_{3}$ (001) substrates are fabricated via the pulsed laser deposition method. X-ray diffraction and high-resolution transmission electron microscopy indicate that the as-prepared epitaxial film is well crystallized. In the film plane the orientation relationship between the film and the substrate is [100]Mn$_{3}$O$_{4}$$\parallel$[110] Nb-doped SrTiO$_{3}$. After an electroforming process, the film shows bipolar nonvolatile resistance switching behavior. The positive voltage bias drives the sample into a low resistance state, while the negative voltage switches it back to a high resistance state. The switching polarity is different from the previous studies. The complex impedance measurement suggests that the resistance switching behavior is of filament type. Due to the performance reproducibility and state stability, Mn$_{3}$O$_{4}$ might be a promising candidate for the resistive random access memory devices.

We study the control of gate voltage over the magnetization of a single-molecule magnet (SMM) weakly coupled to a ferromagnetic and a normal metal electrode in the presence of the temperature gradient between two electrodes. It is demonstrated that the SMM's magnetization can change periodically with periodic gate voltage due to the driving of the temperature gradient. Under an appropriate matching of the electrode polarization, the temperature difference and the pulse width of gate voltage, the SMM's magnetization can be completely reversed in a period of gate voltage. The corresponding flipping time can be controlled by the system parameters. In addition, we also investigate the tunneling anisotropic magnetoresistance (TAMR) of the device in the steady state when the ferromagnetic electrode is noncollinear with the easy axis of the SMM, and show the jump characteristic of the TAMR.

A novel AlGaN/GaN high electron mobility transistor (HEMT) with double buried p-type layers (DBPLs) in the GaN buffer layer and its mechanism are studied. The DBPL AlGaN/GaN HEMT is characterized by two equi-long p-type GaN layers which are buried in the GaN buffer layer under the source side. Under the condition of high-voltage blocking state, two reverse p-n junctions introduced by the buried p-type layers will effectively modulate the surface and bulk electric fields. Meanwhile, the buffer leakage is well suppressed in this structure and both lead to a high breakdown voltage. The simulations show that the breakdown voltage of the DBPL structure can reach above 2000 V from 467 V of the conventional structure with the same gate–drain length of 8 μm.

In CaFe$_{2}$As$_{2}$, superconductivity can be achieved by applying a modest $c$-axis pressure of several kbar. Here we use scanning tunneling microscopy/spectroscopy (STM/S) to explore the STM tip pressure effect on single crystals of CaFe$_{2}$As$_{2}$. When performing STM/S measurements, the tip-sample interaction can be controlled to act repulsive with reduction of the junction resistance, thus to apply a tip pressure on the sample. We find that an incoherent energy gap emerges at the Fermi level in the differential conductance spectrum when the tip pressure is increased. This energy gap is of the similar order of magnitude as the superconducting gap in the chemical doped compound Ca$_{0.4}$Na$_{0.6}$Fe$_{2}$As$_{2}$ and disappears at the temperature well below that of the bulk magnetic ordering. Moreover, we also observe the rhombic distortion of the As lattice, which agrees with the orthorhombic distortion of the underlying Fe lattice. These findings suggest that the STM tip pressure can induce the local Cooper pairing in the orthorhombic phase of CaFe$_{2}$As$_{2}$.

Single crystals of undoped CaFe$_{2}$As$_{2}$ are grown by an FeAs self-flux method, and the crystals are quenched in ice-water rapidly after high-temperature growth. The quenched crystal undergoes a collapsed tetragonal structural phase transition around 80 K revealed by the temperature-dependent x-ray diffraction measurements. Superconductivity below 25 K is observed in the collapsed phase by resistivity and magnetization measurements. The isothermal magnetization curve measured at 2 K indicates that this is a typical type-II superconductor. For comparison, we systematically characterize the properties of the furnace-cooled, quenched, and post-annealed single crystals, and find strong internal crystallographic strain existing in the quenched samples, which is the key for the occurrence of superconductivity in the undoped CaFe$_{2}$As$_{2}$ single crystals.

We report a reproducible approach in preparing high-quality overdoped Bi$_2$Sr$_2$CaCu$_2$O$_{8+\delta}$ (Bi2212) single crystals by annealing Bi2212 crystals in high oxygen pressure followed by a fast quenching. In this way, high-quality overdoped and heavily overdoped Bi2212 single crystals are obtained by controlling the annealing oxygen pressure. We find that, beyond a limit of oxygen pressure that can achieve most heavily overdoped Bi2212 with a $T_{\rm c}\sim$63 K, the annealed Bi2212 begins to decompose. This accounts for the existence of the hole-doping limit and thus the $T_{\rm c}$ limit in the heavily overdoped region of Bi2212 by the oxygen annealing process. These results provide a reliable way in preparing high-quality overdoped and heavily overdoped Bi2212 crystals that are important for studies of the physical properties, electronic structure and superconductivity mechanism of the cuprate superconductors.

HgCr$_{2}$S$_{4}$ is a typical compound manifesting competing ferromagnetic (FM) and antiferromagnetic (AFM) exchanges as well as strong spin–lattice coupling. Here we study these effects by intentionally choosing a combination of magnetization under external hydrostatic pressure and thermal conductivity at various magnetic fields. Upon applying pressure up to 10 kbar at 1 kOe, while the magnitude of magnetization reduces progressively, the AFM ordering temperature $T_{\rm N}$ enhances concomitantly at a rate of about 1.5 K/kbar. Strikingly, at 10 kOe the field polarized FM state is found to be driven readily back to an AFM one even at only 5 kbar. In addition, the thermal conductivity exhibits drastic increments at various fields in the temperature range with strong spin fluctuations, reaching about 30% at 50 kOe. Consequently, the results give new experimental evidence of spin–lattice coupling. Apart from the colossal magnetocapacitance and colossal magnetoresistance reported previously, the findings here may enable new promising functionalities for potential applications.

Using in situ electric-field-modulated anisotropic magnetoresistance measurement, a large reversible and non-volatile in-plane rotation of magnetic easy axis of $\sim$35$^{\circ}$ between the positive and negative electrical poling states is demonstrated in Co$_{40}$Fe$_{40}$B$_{20}$/(001)-cut Pb(Mg$_{1/3}$Nb$_{2/3}$)O$_{3}$-0.25PbTiO$_{3}$ (PMN-PT). The specific magnetoelectric coupling mechanism therein is experimentally verified to be related to the synchronous in-plane strain rotation induced by 109$^{\circ}$ ferroelastic domain switching in the (001)-cut PMN-PT substrate.

GaN nanorods are successfully fabricated by adjusting the flow rate ratio of hydrogen (H$_{2}$)/nitrogen (N$_{2}$) and growth temperature of the selective area growth (SAG) method with metal organic chemical vapor deposition (MOCVD). The SAG template is obtained by nanospherical-lens photolithography. It is found that increasing the flow rate of H$_{2}$ will change the GaN crystal shape from pyramid to vertical rod, while increasing the growth temperature will reduce the diameters of GaN rods to nanometer scale. Finally the GaN nanorods with smooth lateral surface and relatively good quality are obtained under the condition that the H$_{2}$:N$_{2}$ ratio is 1:1 and the growth temperature is 1030$^{\circ}\!$C. The good crystal quality and orientation of GaN nanorods are confirmed by high resolution transmission electron microscopy. The cathodoluminescence spectrum suggests that the crystal and optical quality is also improved with increasing the temperature.

The effect of a self-organized SiN$_{x}$ interlayer on the defect density of (11$\bar{2}$2) semipolar GaN grown on $m$-plane sapphire is studied by transmission electron microscopy, atomic force microscopy and high resolution x-ray diffraction. The SiN$_{x}$ interlayer reduces the $c$-type dislocation density from $2.5\times10^{10}$ cm$^{-2}$ to $5\times10^{8}$ cm$^{-2}$. The SiN$_{x}$ interlayer produces regions that are free from basal plane stacking faults (BSFs) and dislocations. The overall BSF density is reduced from $2.1\times10^{5}$ cm$^{-1}$ to $1.3\times10^{4}$ cm$^{-1}$. The large dislocations and BSF reduction in semipolar (11$\bar{2}$2) GaN with the SiN$_{x}$ interlayer result from two primary mechanisms. The first mechanism is the direct dislocation blocking by the SiN$_{x}$ interlayer, and the second mechanism is associated with the unique structure character of (11$\bar{2}$2) semipolar GaN.

We propose a self-organized optimization mechanism to improve the transport capacity of complex gradient networks. We find that, regardless of network topology, the congestion pressure can be strongly reduced by the self-organized optimization mechanism. Furthermore, the random scale-free topology is more efficient to reduce congestion compared with the random Poisson topology under the optimization mechanism. The reason is that the optimization mechanism introduces the correlations between the gradient field and the local topology of the substrate network. Due to the correlations, the cutoff degree of the gradient network is strongly reduced and the number of the nodes exerting their maximal transport capacity consumedly increases. Our work presents evidence supporting the idea that scale-free networks can efficiently improve their transport capacity by self-organized mechanism under gradient-driven transport mode.