Search

Optically generated 20-GHz microwave carriers with phase noise lower than −75 dBc/Hz at 10 kHz offset and lower than —90 dBc/Hz at 100 kHz offset are obtained using single- and double-sideband injection locking. Within the locking range, the effect of sideband injection locking can be regarded as narrow-band amplification of the modulation sidebands. Increasing the current of slave laser will increase the power of beat signal and reduce the phase noise to a certain extent. Double-sideband injection locking can increase the power of the generated microwave carrier while keeping the phase noise at a low level. It is also revealed that partially destruction of coherence between the two beating lights in the course of sideband injection locking would impair the phase noise performance.

The blue band (BB) in low temperature photoluminescence of Mg-doped GaN films with different Mg concentrations is investigated. The BB peak of as-grown samples with higher Mg concentration centres at lower energy. A shift of the BB peak energy is observed after annealing in N2 at different temperatures. Meanwhile, the difference between the BB peak energies diminishes for raised annealing temperature, and the BB peaks for different samples converge to 2.92 eV after annealing at 850°C. These experimental results can be accounted for by a model based on compensation effect. The shift of BB lines provides a useful criterion for the optimum annealing temperature of the Mg-doped GaN material, and the value is taken to be 850°C in our case.

A novel ridge-waveguide multisection (MS) distributed feedback (DFB) laser, which consists of two identical DFB sections but different ridge widths, is proposed to generate beating-type self-pulsations (SPs). The spatiotemporal dynamic response of such a multisection DFB laser is calculated based on a large-signal travelling-wave model. Self-pulsating output at about 150 GHz is predicted, and evidences for the beating mechanism of the SPs are provided. To the best of our knowledge, this is the first report on SP generated by MS-DFB lasers with varied ridge width. Compared to other alternatives, such devices are much easier to implement and also enjoy the advantages of lower cost and higher design freedom.

In recent decades, infrared (IR) detection technology has been widely used in many fields such as weather monitoring, environmental protection, medical diagnostics, security protection, etc. With the progress and mature of the technologies, more attention has been paid to the imaging detections of weak IR signals. So the higher efficiency of the device is required. Moreover the next-generation IR photodetection technology focuses on large-scale, high-speed and low-dark-current imaging. The mechanical bonding between infrared detector chip and silicon readout circuit inevitably causes a thermal mismatch problem. Up-conversion IR photodetectors can solve the problem about the performance deterioration of photodetector and the thermal mismatch with silicon-based readout circuit, hence they have great advantages in realizing large-format focal plane array detection.However, the poor light extraction efficiency due to total reflection severely restricts the overall efficiency of the up-conversion device, which has become one of the bottlenecks to improve the device efficiency. In this paper, surface microstructures with micro-pillar morphology are designed and fabricated on quantum-cascade up-conversion IR photodetectors. The effect on the up-conversion efficiency is investigated by enhancing the light extraction efficiency.Firstly, by the optical ray retracing method, the influence of surface microstructure on light extraction efficiency is studied when considering different morphology parameters, and optimized surface microstructure is designed to possess a pillar base length of 150 nm, height of 105 nm and side wall angle of 75.Then based on the results of simulation, up-conversion IR photodetectors with surface microstructures are fabricated using polystyrene nanospheres as mask. The self-assembled monolayer nanospheres are first etched to a proper size by using O2 plasma, then the patterns are transferred to SiNx film, which acts as an ICP dry etching mask of the micro-pillars. Finally, the up-conversion device and a silicon detector are together loaded on a cold finger of a cryogenic dewar. The characteristics of the up-converter and up-conversion system are evaluated using a blackbody source.The experimental results show that the devices with and without surface microstructure exhibit similar IR responses and dark currents, while the emission of device with microstructure is obviously increased. Taking into consideration other factors related to external quantum efficiency, the light extraction efficiency of the device with micro-pillar structure on surface can be increased by up to 130%. Therefore it can be concluded that this method is an efficient way to improve the efficiency of up-conversion IR photodetector. The finding in this paper can also be applied to other semiconductor device with light extraction efficiency.

We study the room-temperature dry, etching of InP by inductively coupled plasma (ICP) using Cl2/CH4/Ar mixtures. Etches were characterized in terms of anisotropy and surface roughness by scanning electron microscopy and atomic force microscopy, respectively. It is found that the flow ratio between Cl2 and CH4, ICP power, rf chuck power, and table temperature can greatly influence the, etching results. By adjusting, etching parameters, vertical sidewall and smooth surface can be obtained simultaneously, together with a moderate, etch rate and a good select ratio. The root-mean-square surface roughness is measured to be as low as 0.27 nm. To the best of our knowledge, this is the best result for InP to date. The, etch rate is 855 nm/min, and the selectivity ratio over SiO2 is estimated to be higher than 15:1. The stoichiometry of the, etched surface has also been investigated by Auger electron spectroscopy. The, etched surface is found to manifest a slight P deficiency, and the ratio between P and In reaches the stoichiometric value within about 0.75 nm depth into the surface.

We study the external optical feedback resistance of both index-coupled (IC) and gain-coupled distributed feedback (DFB) lasers using a very simple experimental set-up. Though IC-DFB lasers with large coupling coefficients are found to be less sensitive to external reflection, they suffer from mode instability at high injection level. On the other hand, the introduction of gain-coupling mechanism can significantly improve both the immunity to external optical feedback and the single mode yield of the device.

A backside illuminated mesa-structure InGaAs/InP modified uni-traveling-carrier photodiode (MUTC-PD) with wide bandwidth and high saturation power is fabricated and investigated. The device structure is optimized to reduce the capacitance and resistance. For the 22-μm-diameter device, the maximum responsivity at 1.55 μm is 0.5 A/W, and the 3-dB cutoff frequency reaches up to 28 GHz. The output photocurrent at the 1-dB compression point is measured to be 54 mA at 25 GHz, with a corresponding output radio frequency (RF) power of up to 15.5 dBm. The saturation characteristics of the MUTC-PD are also verified by the electric field simulation, and electric field collapse is found to be the cause of the saturation phenomenon.

Infrared (IR) photodetectors have been widely used in the fields of both civil and military applications such as environmental monitoring, medical diagnostics, satellite remote sensing and missile guidance, etc. In conventional large scale focal plane array (FPA) IR imaging, the thermal mismatch between IR photodetectors and silicon readout circuits will inevitably lead to the degradation of the device performance. An up-conversion IR photodetector, which converts IR photons to short-wavelength photons for Si-CCD-based imaging, can avoid thermal mismatch caused by hybridization with silicon readout circuits, resulting in a low-cost way for large array IR imaging. The operation principle of the semiconductor up-conversion IR photodetector is based on electron transitions and carrier transportation in different functional sections including absorption section, transportation section and emission section, hence the carrier distribution in the device structure has a crucial influence on the device performance. In order to achieve low dark current, carriers are expected to be non-uniformly distributed in the up-conversion device structure. Designing and optimizing the carrier-blocking structure are usually the key issues to acquire inhomogeneous carrier distribution. In this paper, up-conversion infrared photodetectors with various hole-blocking structures are investigated both theoretically and experimentally. Firstly the carrier distributions are calculated by self-consistently solving the Schr?dinger equation, Poisson equation, current continuity equation and carrier rate equation. Then the influence of the carrierblocking structure on the device performance is analyzed by electroluminescence measurements on the corresponding epitaxial structures. According to the theoretical and experimental results, it is found that a 2-nm-thick AlAs barrier layer can block holes effectively without hampering the electron transportation, which is necessary for the up-conversion infrared photodetectors. However, other attempts to block holes, such as light n-doping in the transportation section or lowering the injection barrier, do not work well. In addition, the influences of the thickness and height of the blocking barrier and the operation temperature on the carrier distributions are also studied. When the thickness of the blocking barrier is less than 2 nm, the thicker or the higher is the barrier, the better is the blocking effect. However, when the thickness of the blocking barrier is larger than 2 nm, the blocking effect is not persistently enhanced with increasing thickness because the tunneling process is almost fully suppressed. Furthermore, with the same blocking barrier parameters, lowering the operation temperature can lead to better blocking effect. This work demonstrates the utilization and effect of carrier-blocking structures in semiconductor devices which deamnd an inhomogeneous carrier distribution.

The design and fabrication of multilayer antireflection (AR) coating based on plasma enhanced chemical vapor deposition (PECVD) is studied for its applications in optoelectronic devices. Deposition conditions for obtaining SiO2/SiNx thin films with large refractive index difference is determined through systematic study of factors influencing the refractive index of deposited SiNx. Four-layer SiO2/SiNx AR coating is designed to exhibit a reflectivity of less than 10-4 over 70 nm bandwidth. Reflectivity of the thin film structure at the center wavelength of 1550 nm remains less than 5×10-4 when the thickness deviation of any single layer is within ±5 nm from the designed value. Based on the simulation results, SiO2/SiNx multilayer AR coating is deposited on the end facet of a Fabry-Perot laser. By analyzing the output spectra of the laser, the residual reflectivity of the AR coating is determined to be on the order of 10-4 over the wavelength range of 1535—1565 nm.

In this paper, GaN films are grown by metal organic vapor phase epitaxy. Different N2 carrier gas percentages were used in the high-temperature growth process of the bulk GaN, and the dependence of sheet resistance on N2 carrier gas percentage is studied. It is found that the sheet resistance of the GaN film increases dramatically with N2 carrier gas percentage. When N2 carrier gas percentage is 50%, the sheet resistance of the GaN film is 1.1×108Ω/sq, and the surface root-mean-square roughness is as small as 0.233 nm. Secondary ion mass spectroscopy measurements reveal that the concentration of carbon and oxygen impurities is almost the same in the samples with different N2 carrier gas percentage. The density of edge-dislocation-related defects increases with N2 carrier gas percentage, while the density of screw-dislocation-related defects shows no obvious difference for all the samples. Our results indicate that the increase of the sheet resistance of GaN film is mainly due to the increase of edge threading dislocations, which act as acceptor centers in the GaN material.

Precise measurement and control of GaN-film thickness is very important for GaN-based material epitaxy and device fabrication. However, GaN-films heteroepitaxially grown on large-mismatch substrates, such as sapphire, SiC and Si, etc., usually show a mosaic structure, which causes great difficulty to the GaN-film thickness measurement. Combining the advantages of Williamson-Hall plot method and diffraction profile shape analysis method, a new strategy was presented to effectively distinguish the X-ray diffraction broadening factors of finite crystallite size and inhomogeneous strain, which can be used to precisely and reliably determine the thickness of epitaxial films. The thickness of a series of GaN films grown on sapphire substrates in the range of 0.7—4.2 μm were measured by this method. Comparing with the thickness obtained from spectroscopic ellipsometry measurements, the difference was found to be within 4%, which shows the excellent performance of this method.

In this work, the abnormal double-peak in the electroluminescence spectra of In xGa1-xN/GaN multiple quantum wells (MQWs) light emitting d iode s tructure materials with high indium contents was studied under different injecti on currents. The results show that the screening of internal electric field by i njection current plays an important role in the radiation recombination process of InxGa1-xN/GaN-MQWs materials.