InP-based vertical cavity surface emitting lasers (VCSELs) with AlGaInAs QWs and AlGaInAs/InP DBR have been demonstrated. Over 2 mW and 0.8 mW single-mode powers at 1.3 μm have been achieved at 25 °C and 85 °C, respectively. A high eye-diagram margin at 10 Gbit/s modulation was demonstrated using 13 GHz maximum relaxation oscillation frequency VCSELs. This high relaxation oscillation frequency enables the achievement, without optical isolator, of error-free transmission under 10 Gbit/s modulation with high optical reflection. To achieve stable polarization operation, 1.3 μm VCSELs on InP(311)A substrate have been realized and showed very stable polarization behavior.

We report on bottom-emitting vertical-cavity surface-emitting lasers (VCSEL's) and laser arrays providing high output powers in the 980-nm wavelength regime. Single devices with active diameters of 500 μm show high output powers of 1.95 W at room temperature. Its threshold current is 510 mA, and the maximum spatially averaged optical power density is 0.93 kW/cm². A 16 elements array with 200μm aperture size (250μm center spacing) of individual elements shows a CW output power of 1.21W at room temperature, resulting in a average optical power density of 1KW/cm². The threshold current of the array is 1.32A and the lasing peak wavelength is 981.9 nm. The distinction of emission spectrums between the single device and the array is discussed.

We describe the design, fabrication, and calculation characteristics of the 980nm high-power diode-pumped vertical external-cavity surface-emitting laser(VECSEL).From our calculation, the VECSEL with active region of InGaAs/GaAsP/AlGaAs system can operate near 1w in a single transverse mode.

We describe design, numerical simulation and characteristics of high-power optical pumped VECSELs at different wavelength (980nm, and 1300nm). The device design realizes the integrating diode-pumped lasers with vertical-cavity surface-emitting laser structure, drawing on the advantages of both. With periodical gain element structure, optical pumped VECSEL is scalable to watt level output. The characteristics such as threshold condition and output power are calculated theoretically. An optimum number of quantum wells and external mirror reflectivity are obtained from the calculation results, and the thermal characteristic is also considered. Finally the calculation results also predict high output power in this kind of device structure.

Polymeric electro-optic materials have many potential advantages over their inorganic counterparts in both industrial and military applications for the reason of bandwidth, driving voltage, and cost. To fulfill the device processing and performance requirements, electro-optic materials need to possess high electro-optic activity, low optical loss, adequate thermal and photochemical stability as well as long term alignment stability. In this work, first we have designed and synthesized a series of chromophores with various acceptors. Their thermal stability and electro-optic activity were compared. Detailed chromophore loading studies were carried out. An electro-optic coefficient of 160 pm/V was achieved. Secondly we modified our chromophore bridge by adapting 3,4-ethylenedioythiophene moiety. Both thermal stability and electro-optic activity are improved. Finally, we synthesized two polycarbonates with higher glass transition temperatures and lower optical loss. Their guest-host systems show improved EO activity, optical loss and long term alignment stability.

Polymeric electro-optic materials have made significant progress in increasing electro-optic co-efficient, enhancing temporal stability, and lower optical loss at operating wavelength. To fully realize the electro-optic activity in a waveguide device, appropriate cladding materials need to be designed carefully and a compatible fabrication process should be developed accordingly, which has been challenging and tedious. It would be desirable if glass-based optical fiber or silica substrate can be utilized as parts of device materials. The challenge is to match the relatively lower refractive index of fiber and silica substrates. Thus, a low refractive (~1.44) of electro-optic material was designed. The efforts include the synthesis and characterization of the low index host materials via sol-gel process and silanization of a chromophore with large hyperpolarizability. In this study, an electro-optic sol-gel material with a refractive index ~1.44 at the wavelength of 1.55 μm was developed. It was highly processible and compatible with glass-based fiber or silica-based materials. An interdigitated electrode configuration is designed and a substrate with such a configuration was fabricated. Combined with the in-situ SHG monitoring setup, poling process can be readily monitored and optimized. The r₃₃ obtained at the wavelength of 1.3 μm is 19 pm/V. Thus, it becomes possible to design optical devices using glass-based fiber or silica substrate as part of the device structure, which could simplify the material and process development.

In EO polymer materials, the second-order nonlinear optic chromophores must be oriented in one direction in order to be electro-optically active. Interchromophore electrostatic interactions, which encourages the formation of non-active and light-scattering crystalline domains through the antiparallel stacking of dipoles, have long been an obstacle to the translation of large molecular optical nonlinearity into corresponding bulk nonlinearity. Great progresses have been made in the design and synthesis of chormophores with reduced electrostatic interactions. New effort toward the complete elimination of the destructive effect of the electrostatic forces is on the way. On the contrary, strong intermolecular electrostatic interaction (e.g. the force responsible for π-π stacking of conjugated systems) is often desired in LED, transistor and photovoltaic (PV) devices since high mobilities of the charge carriers are the key to their high performances and often come from ordered stacking of pi-conjugated systems. For PV applications, bi-continuous "bulk heterojunction" of electron donor (D) and acceptor (A) is ideal for efficient charge carrier generation, transport and collection. Differential electrostatic interactions between D and A is the key to the formation of such morphology. We have synthesized a novel type of block copolymer having a basic unit of D-B-A (B is a non-conjugate bridge) for solar cell application. D and A can be designed in such a way that D-D and A-A interactions are stronger than the D-A interaction, and therefore, have a strong tendency to phase separate.

A method for measuring the electro-optic coefficient of poled polymer films on the basis of an asymmetry Fabry-Perot cavity is described. Two aluminum films were deposited on glass substrate by thermal evaporation in high vacuum to form high reflection mirror and low reflection mirror, respectively. Nonlinear polymer thin films are spin coated on low reflection mirror and poled by corona poling in order to break the internal centrosymmetry. Then the sample layers are placed upside down on the high reflection mirror. The sandwich structure is objected to a laser beam, and a variable voltage is applied the aluminium films resulting in a modulation of the transmitted laser power. The electro-optic coefficient γ₁₃ of the poled polymer film can be calculated by evaluating the Fabry-Perot equation. The spatial resolution is tested with a polymer film that was poled by a needle corona discharge in air through a metal grating with a period of 120 microns. By scanning the sample plate in the direction perpendicular to the grating lines, the spatial resolution is also demonstrated according to the spacing of the poled structure.

Wuhan National Laboratory for Optoelectronics (WNLO) is one of the five national laboratories initiated and sponsored by the Ministry of Science and Technology, China in 2003. WNLO is an important part of the Wuhan • Optics Valley of China (WOVC). It is also an integrated part of Chinese national scientific innovation system. The mission of the WNLO is to become a top innovation base for the field of optoelectronics in China, to promote and lead the complete commercialization system (R&D&P) for "Wuhan • Optics Valley of China", and to contribute to the growth of optoelectronic industries through technology transfer. So far, eight research thrusts (RTs) are designed for the WNLO, which are related to Optoelectronics, Photonics, Microelectronics, Microsystems, or Nanotechnology. One of the major research themes in the Laboratory is miniaturization. This paper will introduce the newly established Wuhan National Laboratory for Optoelectronics and the status of the current research and development activities.

An ultra-wide-band frequency response measurement system for optoelectronic devices has been established using the optical heterodyne method utilizing a tunable laser and a wavelength-fixed distributed feedback laser. By controlling the laser diode cavity length, the beat frequency is swept from DC to hundreds GHz. An outstanding advantage is that this measurement system does not need any high-speed light modulation source and additional calibration. In this measurement, two types of different O/E receivers have been tested, and 3 dB bandwidths measured by this system were 14.4GHz and 40GHz, respectively. The comparisons between experimental data and that from manufacturer show that this method is accurate and easy to carry out.

According to the wavelength tuning characteristic of widely tunable sampled grating distributed Bragg reflector (SG-DBR) lasers, a kind of wavelength control system is developed. A look-up table, the file including data of picked operation points, is established to store the relationship between lasing wavelengths and tuning currents for each SG-DBR laser in the system. By rapidly looking up the look-up table and providing the corresponding driving currents, the particular lasing wavelength of SG-DBR laser is obtained. Wavelength tuning experiments have been done with eight SG-DBR lasers, and the results show that this system has high precision of wavelength control.

This paper describes in detail the design and performance of the DS-DBR monolithic full band tunable laser and its application in tunable laser and transmitter modules

We investigate response flatting of cascaded sum- and difference- frequency generation (SFG/DFG)-based wavelength conversion in quasi-phase-matched (QPM) periodically poled lithium niobate waveguides for wavelength division multiplexing optical communication systems. The cascaded SFG/DFG-based configuration shows more robustly combined properties than conventional DFG-based and cascaded second-harmonic generation and DFG (SHG/DFG)-based configurations. The response fluctuation, generated while enhancing the conversion bandwidth, can be efficiently reduced by the method of pump detuning or QPM period detuning.

Apart from gain dependence on polarization state, phase modulation in semiconductor optical amplifier (SOA) is usually polarization dependent as a result of birefringence. This paper analyzes the factors that determine the polarization dependence of phase modulation and proposes the necessity of fabricating SOA with polarization-insensitive gain and polarization-insensitive phase modulation. Based on energy band theory, the strain effects on polarization dependences of gain and differential refractive index in multiple quantum-well (MQW) SOA are studied, it is demonstrated that SOA with polarization insensitive gain and polarization insensitive phase modulation can be achieved by optimal design of active region and waveguide structure.

High power InGaAsN triple-quantum-well strain-compensated lasers grown by metal organic chemical vapor deposition (MOCVD) were fabricated with pulsed anodic oxidation. A maximum light power output of 304 mW was obtained from a 10-μm stripe width uncoated laser diode in continuous wave (CW) mode at room temperature. The characteristic temperature of the lasers was 138 K.

Beryllium was incorporated in InGaAsN single quantum well (SQW). Comparing with the conventional InGaAsN SQW structures, photoluminescence (PL) investigations show a significant improvement. After 3000 sec of annealing at 700 °C, the PL peak area is about 20 times higher while the wavelength keeps 25 nm longer. After 800 sec of this annealing, the PL quenched slowly for the conventional structures because of the strain relaxation, while the PL of the new structures increased rapidly and show no saturation after 3000 sec of annealing. Laser processing based on the new InGaAsN structures resulted in one half of the threshold current density compare to conventional InGaAsN.

The low temperature bonding of GaAs/InP wafers is successfully realized by a new surface treatment at 380°. The properties of the bonded structures are studied in terms of the interface shape, electrical and optical characteristic. The low temperature bonding of GaAs/InP wafers is successfully realized by a new surface treatment at 380°. In this method, the surfaces of two wafers are etched by sulfuric solution. Then following a thermal annealing process. The properties of the bonded structures are studied in terms of the interface shape, electrical and optical characteristic through transmission electron microscope (SEM), interface I-V curve, X-ray diffraction (XRD), photoluminescence (PL) spectra, and so on. The optical characteristic of In_0.53Ga_0.47As/InP multi-quantum wells (MQWs) grown on an InP substrate, which was bonded to GaAs substrate was investigated by measuring photoluminescence (PL) spectra at room temperature. After bonding, neither wavelength shift nor degradation of full-width at half maximum (FWHM) was observed. And the PL intensities of the bonding sample increased about 50%. The bonding strength was found to be sufficiently high and could withstand the subsequent etching and polishing procedures of the bonded wafers. This low temperature wafer bonding method is very attractive to realize optical devices such as lasers, photodetectors and optical waveguides on lattice-mismatched substrates.

Improved transmission line laser model has been developed to study the polarization characteristics by incorporating into spin-flip model (SFM). Theoretical studies on the influence of the Bragg reflectivity on the polarization switching of VCSEL have been discussed in this paper. We have demonstrated that reflectivity of Bragg reflectors can be optimized to provide suitable hysteresis loops width which indicates the tolerance of the input power.

A Very-Small-Aperture Laser with a 250X500 nm² aperture has been created on a 650nm edge emitting LD. The highest far-field output power is 1.9mW and the power per unit emission area is about 15mW/μm2. The special fabrication process and high output power mechanism are demonstrated respectively. The near-field distribution properties are also analyzed theoretically and experimentally.

A 1.55-μm ridge DFB laser and electroabsorption modulator monolithically integrated with a buried-ridge-stripe dual-waveguide spot-size converter at the output port for low-loss coupling to a cleaved single-mode optical fiber was fabricated by means of selective area growth, quantum well intermixing and dual-core technologies. These devices exhibit threshold current of 28 mA, side mode suppression ratio of 38.0 dB, 3-dB modulation bandwidth of 12.0 GHz, modulator extinction ratios of 25.0 dB dc. The output beam divergence angles of the spot-size converter in the horizontal and vertical directions are as small as 8.0°×12.6°, respectively, resulting in 3.2 dB coupling loss with a cleaved single-mode optical fiber.

A novel device of tandem multiple quantum wells (MQWs) electroabsorption modulators (EAMs) monolithically integrated with DFB laser is fabricated by ultra-low-pressure (22 mbar) selective area growth (SAG) MOCVD technique. Experimental results exhibit superior device characteristics with low threshold of 19 mA, output light power of 4.5 mW, and over 20 dB extinction ratio when coupled into a single mode fiber. Moreover, over 10 GHz modulation bandwidth is developed with a driving voltage of 2 V. Using this sinusoidal voltage driven integrated device, 10GHz repetition rate pulse with a width of 13.7 ps without any compression elements is obtained.

Combined the characteristics of Mach-Zehnder interferometer (MZI) with the advantages of microring resonator, a new and simple method for improving the performance of MZI electro-optic modulators (EOMs) is proposed. Using the coupled-mode theory and transfer matrix method, the transfer function of the device is analyzed and parameters are optimized. Numerical results demonstrate that Q-factor exceeding 1x10⁴ and modulator bandwidth above 40Gbit/s is achieved by choosing the proper microring radius. And the 3rd-order nonlinear harmonic distortion of the modulator curve is cancelled and linear range higher than 90% is obtained by setting biasing point and choosing coupling coefficients. Compare with the single ring structure, performances of the designed device can be improved further by integrating multiple microring resonators.

In this talk, identical epitaxial layer high speed electroabsorption (EA) modulator integrated light sources are studied systematically. Firstly, a 1.55 μm 10 Gb/s distributed feedback (DFB) laser integrated EA modulated laser is presented, which adopts a partially gain-coupled DFB laser to improve single-mode yield. The typical measured small signal response indicates a 3 dB electrical bandwidth over 10 GHz. Secondly, a 1.55 μm laterally coupled DFB laser is fabricated to eliminate the necessity of additional regrowth, in order to reduce the process complexity and fabrication cost. A stable single-mode operation has been demonstrated with a side mode suppression ratio over 45 dB at an injection level of 60 mA. Finally, an SOA integrated EA modulator with a planar ridge waveguide is fabricated for 40 Gb/s optical fiber communication systems. The device is chip-level packaged and tested on a coplanar waveguide based submount. The small signal frequency modulation responses of the integrated EA modulator are measured and simulated based on an equivalent circuit. The measured 3 dBe modulation bandwidth exceeds 40 GHz, and it is estimated to be over 45 GHz.

We study the transmission of a one-dimensional ternary left-handed photonic crystal which is consisting of three alternating slabs in the form of {ABC}, where A and C represent two kinds of positive- refractive-index materials, and B represents the negative-refractive-index material. Firstly, we obtain the dispersion equation based on the boundary conditions and Bloch theorem. By construing the obtained dispersion relation theoretically we demonstrated that a zero averaged refractive index (ZARI) gap which is around the frequency where the average refractive index vanishes appears. Secondly, we investigate the dependence of the transmission on the thickness of slabs by changing the thickness of slab A, B and C respectively. We find that the transmission of such a structure can be made tunable. This novel property may be very useful in designing tunable filters because the position of band gaps in such a structure can be chosen expediently. Furthermore, we study the property of defect mode of such a structure, and find that the defect mode doesn't always appear simultaneously in both Bragg and ZARI gaps of the transmission spectrum. It can be selected to appear either in the Bragg gap or in the ZARI gap.

We proposed a novel quasi-three dimensional (3D) photonic crystal (PC), which is composed of 2D PC and 1D distributed Bragg reflector (DBR) structures with a matching layer set between them. The band structures and reflectance of 1D DBR with changed thickness of the matching layer were calculated by 2D plane wave method and transfer matrix method respectively, while the dispersion diagram of the quasi-3D PCWs both in GaAs/Al_xO_y (or Si/SiO₂) and GaAs/AlAs PCWs were investigated using 2D plane wave method with supercells. Through adjusting the thickness of the matching layer, out-of-plane confinement by the effects of Bragg mirroring and total internal reflection can be achieved simultaneously. Using this quasi-3D PC waveguide structure, a wide transmission band was demonstrated both in GaAs/Al_xO_y (or Si/SiO₂) and GaAs/AlAs material group.

We investigated the group-velocity dispersion of a one dimensional uniform photonic crystal by the optical transmission method. For application in optical communications, the wavelength should be near one of the two edges of a photonic bandgap. Four kinds of dispersion-compensation may be obtained with a photonic crystal. Huge negative and positive group-velocity-dispersion (GVD) about a zero-dispersion-point as large as 5.1 Tera- ps/nm/km by a photonic crystal of 100 periods can be realized. Such a value is about 50 Giga times the GVD of conventional dispersion-compensation fibers. The GVD reaches a maximum when the optical length ratio of the high refractive index material to the low refractive index material is 1.2 for given operating parameters. When we keep the optical length of each layer being constant, the GVD is found to increase rapidly with the refractive index ration of the high refractive index material to the low one and even more rapidly with the number of periods of a photonic crystal. Under quite common operating parameters, a thin piece of photonic crystal of 100 periods may play the role of an ordinary dispersion-compensation fiber with a length over 158 kilo-meters.

An important class of optical waveguides are those refractive index profile is not continuous, such as Bragg fiber, photonic crystal fiber and 2-D photonic crystal waveguides. These microstructure fibers and waveguides become more and more important in the future optical devices for their novel and spurious optical characteristics. To give a full understand of these devices, the exact mode fields solver is very critical. In this paper, we use the full-vector finite difference approach free of spurious modes to solve mode field characteristics of the photonic crystal and 2-D plane photonic crystal waveguides. The photonic band structures within an irreducible Brillouin zone are investigated for both in plane and out plane propagation. The out of plane propagation can be used for the photonic crystal fiber. The coupled difference equations are in terms of the transverse magnetic filed components. Based on the appropriate transparent boundary conditions, a unique set of couple five-point difference equation are developed, and an efficient numerical technique to solve the deterministic equations by the Eispark in Matlab. For the in plane propagation, the guides mode are either TE or TM modes, the difference equations are uncoupled. Using the appropriate period boundary condition and combined with the transparent boundary condition, we derived the five point difference equations that can be used for the 2-D plane photonic crystal waveguides. Based on these finite difference equations, we analyzed the optical mode field characteristics of the crystal fiber and the plane optical crystal waveguides. The filling materials either dielectric or air are also analyzed. Good agreements are obtained compared the numerical results with the experiment data and the published literature. The mode fields' solver can also be used for the other waveguides such as Bragg fibers.

Width varied quantum wells show a more flat and wide gain spectrum (about 115nm) than that of identical miltiple quantum well. A new fabricating method was demonstrated in this paper to realize two different Bragg grating in an identical chip using traditional holographic exposure. A wavelength selectable DFB laser based on this material grown by MOVPE was presented. Two stable distinct single longitudinal mode of 1510nm and 1530nm with SMSR of 45 dB were realized.

Ultra-short optical pulse sources are key components in optical time-division multiplexing(OTDM) system. In this paper, a wavelength and repetition frequency tunable actively mode-locked fiber ring laser based on cross-gain modulation in a Reflective Semiconductor Optical Amplifier(RSOA) is demonstrated. Theoretical model for this scheme is also established and output characteristics are calculated. Stable picosecond pulse trains at 10-80 GHz repetition frequency are obtained. 30nm wavelength tuning range is achieved in our proposed actively mode-locked laser.

A novel wavelength converter based on cross-polarization modulation is optimized for all-optical network application. Stable error-free application result is achieved for 10 Gbit/s NRZ signal after transmission through 80 km single-mode fibers between network nodes without dispersion compensation.

In this proceedings the Finite Difference Time Domain (FDTD) and frequency domain Finite Element (FE) methods are used to model both linear chirped pulse and arbitrary chirped pulse propagation in 2D Photonic Crystal (PhC) waveguides. An in-house FDTD code has been implemented which allows the study of pulse propagation in a very direct way. The carrier wavelength of the pulse is swept across the bandwidth of a mini-stopband feature and pulse compression behaviour is observed. In the case of linear chirped pulse, both round hole and square hole PhC waveguides are studied with the latter giving increased pulse compression. An input pulse is then derived from a SOA model which has arbitrary chirp. This is passed through a mini-stop band in a narrowed W3 PhC waveguide and pulse compression is observed.

A two-dimensional spatially independent rate equation model of vertical-cavity surface-emitting lasers (VCSELs) is derived and then used to analysis the multi-mode behaviour of VCSELs. The transverse mode characteristics of VCSELs, the carrier distribution in both the radial and the azimuthal directions, and the effects of the azimuthal non-uniformity of the injection current on the transverse mode behaviours are investigated in detail. By using both Bessel and Fourier expansion of carrier density, the 2D spatially independent rate equations for transverse mode are formulated, which take into account carrier diffusion both in the radial and in the azimuthal direction as well as gain non-uniformity in the lateral direction. The equations are numerical solved self-consistently using the Runge-Kutta method for different spatial periodic injection current. Results show that a proper current injection profile can separate the sine mode and cosine mode of the same order transverse modes observably. It is found that an injection current with periodic change in the azimuthal direction is favourable for the excitation of the modes whose mode profile match the current profile best. The results are useful to the design and control of transverse mode characteristics of a VCSEL.

We systematically investigated the strain field distribution of conical-shaped InAs/GaAs self-organized quantum dot using the two-dimension axis-symmetry model. The normal strain, the hydrostatic and biaxial components along the center axis path of the quantum dots was analyzed. The dependence of these strain components on volume, height-over ratio and cap layer (covered by cap layer or uncovered quantum dot) are investigated for the quantum grown on the (001) substrate. The dependence of the carriers' confining potentials and electronic effective mass on the three circumstances discussed above is also calculated in the framework of eight-band k • p theory. The numerical results are in good agreements with the experiment data in published literature.

Performance of quantum dot based on cross-gain in SOA NOR optical logic gate has been simulated. The saturation power and cross gain modulation of a QD SOA are numerically analysed with the rate equation model which takes into account the effect of the excited state and the wetting layer (WL). The calculated result will be used to study the NOR performance.

Zero-n photonic gap in a periodic structure of positive and negative refractive index materials is studied with a transfer matrix method. A zero-n gap is found in the photonic spectra of this periodic structure. It is demonstrated that the zero-n gap is less sensitive to the incident angle and the scale in contrast to the Bragg gap.

For a photonic crystal, the larger the photonic band gap (PBG), the greater the bandwidth for manipulating the optical wave propagation. Therefore, to enlarge the photonic band gaps would be an important research topic. In this paper, we adopt the simulated annealing (SA) algorithm to synthesize two-dimensional photonic crystals for large band gaps. The SA is an iterative procedure and its probabilistic nature provides an opportunity to escape from the local minima and to reach the global minimum in a nonlinear optimization problem. To speed up the numerical calculations, we also use the fast plane wave expansion method for calculating the band diagram of a photonic crystal at each iteration step of the SA procedure. Numerical simulations are carried out to demonstrate the feasibility and effectiveness of our synthesis algorithms.

We show theoretically that the upper and lower cutoff frequencies in a two-dimensional photonic crystal coupled-cavity waveguide (CCW) can be moved synchronously by changing two configuration parameters of the waveguide simultaneously. The CCW is formed in a two-dimensional photonic crystal (2-D PhC) consisting of a square array of dielectric rods in the air. Efficient methods have been demonstrated for moving the two cutoff frequencies in the same direction over equal distances up to 25.7% of the photonic band gap. The results can be applied to the design of integrated optical devides based on photonic crystal waveguides.

Optical microcavities based on zero-group-velocity surface modes in photonic crystals are studied. It is shown that high quality factors can be easily obtained for such microcavities in photonic crystals. With increasing of the cavity length, the quality factor is gradually enhanced and the resonant frequency converges to that of the zero-group-velocity surface mode in the photonic crystals. Different with other microcavities mentioned in the literature, microcavities proposed in this paper can be considered as open cavities in the sense that one of the in-plane boundaries is exposed to air.

The antiphase states were observed experimentally in a laser-diode pumped sub-nanosecond microchip Cr,Yb:YAG self-Q-switched multimode laser which is resulted from the spatial hole burning effect. The stable two-mode, and three-mode oscillation are obtained with the increase of the pump power ratio. The modified multimode rate equations including the spatial hole-burning effect in the active medium and the non-linear absorption of the saturable absorber is proposed. The numerical simulations of the antiphase dynamics of such laser are in good agreement with the experimental data and the antiphase dynamics was explained by the evolution of the inversion population and the bleaching and recovery of the inversion population of the saturable absorber.

A high-power-output and high polarization rate Nd:YVO⁴/KTP green laser has been presented. By analyzing the factors which affect the conversion efficiency and polarization rate, we used a short folded-cavity resonator, the maximum output power of 320 mW is obtained at 532 nm with the LD pumping power of 1 W, the optical-to-optical conversion efficiency is 32% and the polarization rate is 550:1.

Fiber optic amplifier is a field of new growing interest owing to its application to the optical metro and access networks for amplifying a weak optical signal. In order to design short-fiber-length and broadband optical amplifier, a new amplified fiber with semiconductor inner cladding has been presented. The semiconductor optical amplifier (SOA) is usually pumped by electric currents, but an optical pump is used to the semiconductor inner cladding fiber amplifier in our work. The rate equation is derived in the optical pumped amplifier, and the properties of the amplifier are analyzed. Comparison with conventional optical pumped amplifier, the semiconductor inner cladding fiber amplifier can be pumped by a wide-band optical pumped power, and high pump efficiency can be designed.

Dynamic model and experimental scheme are presented for a wavelength converter (WC) based on cross-polarization-modulation (CPM) in a semiconductor optical amplifier (SOA). The frequency chirping and transmission characteristics of this WC are investigated.

We analyze the all-optical wavelength converter (AOWC) based on a single-port-coupled (SPC) semiconductor optical amplifier (SOA). A comprehensive dynamic model is developed by considering longitudinal variations of carrier density, the residual rear-facet reflectivity of SOA and wide-band spontaneous noise emission (ASE). The numerical simulations for the novel wavelength conversion at 10Gbit/s are presented based on the model. The impact of the rear facet reflectivity on extinction ratio and conversion efficiency is analyzed. We find that when the reflectivity is in the range of 0.2 and 0.3, the extinction ratio over 10dB and the conversion efficiency over 0dB can be achieved. The pattern effect is also discussed to optimize the reflectivity. The wavelength conversion experiments employing two single-port-coupled SOAs with different reflectivity are demonstrated. The extinction ratio and conversion efficiency are measured. The experimental results are agreement with the simulation results.

Optical signal regeneration will be key technology for extending the total transmission distance and improvement of optical signal quality for next-generation high-speed optical transmission and optical packet networks. In addition, wavelength conversion will be advantageous for contention resolution to process each data packets with low packet loss. We have proposed and investigated the performance of a cascaded wavelength converters called a cross-gain-modulation type wavelength-converter cascaded optical regenerator (CGCOR), consisting of two-stage XGM wavelength converters based on semiconductor optical amplifiers (SOA's). By this configuration, power fluctuation in both low and high level for input light power can be suppressed, and so the improvement of the signal to noise ratio and input dynamic range can be achieved. On the other hand, a cascaded gain-saturated SOA with a XGM wavelength converter is also proposed for suppressing chirping and improving transmission characteristics. Both offset of input timing between XGM and input signal, and offset of the center wavelength of the optical bandpass filter from the input signal wavelength can result in the reduction of chirping and the error free performance in the 50km SMF transmission. The operation performance of two types of cascaded wavelength converter-type optical regenerator will be presented.

We have developed ultra small form factor plastic package fiber optic transceivers which are designed for large core fiber system (PCS-fiber, PF-POF). The VCSEL based fiber optic transmitter exhibits an output power stability within 0.5dB over an operating temperature from -40°C to +105°C. The fiber optic receiver consists of a GaAs MSM-photodetector with a large active area and high bandwidth. Giga bit transmission over a 10m PCS-fiber is demonstrated with the transceivers. The modules are applicable to automotive, consumer electronics, home networking, and datacom applications.

A novel microwave packaging technique for 10Gb/s electro-absorption modulator integrated with distributed feedback laser (EML) is presented. The packaging parasitics and intrinsic parasitics are both well considered, and the packaging circuit was synthetically designed to compensate for the intrinsic parasitic of the chip. A butterfly-packaged EML module has been successfully developed to prove that. The small-signal modulation bandwidth of the butterfly-packaged module is about 10 GHz. Optical fiber transmission experiments have shown that the module can be used for 10Gb/s optical transmission system. After transmission through 40km, the power penalty is less than 1 dBm at a bit-error-rate of 10^-12.

Recently there have been market announcements on communication equipment mounted 40-Gbit ports. The requirements for 40-Gbit/s transceivers (transponders) have been increasing to include (a) low cost, (b) compact size, and (c) low power solution. The latest technology trends of transceivers, optical devices and SERDES IC's are described, focusing on the 300-pin MSA SERDES transceivers. These technologies are not mature but are making steady progresses. Finally, future trends and requirements will be discussed.

Arrayed waveguide gratings and MMI couplers are key components in Photonic Integrated Circuits. However, mass application of PICs still has to come. The road to such a broad application is reviewed.

SOI (Silicon on Insulator) based photonic devices, including stimulated emission from Si diode, RCE (Resonant Cavity Enhanced) photodiode with quantum structure, MOS (Metal Oxide Semiconductor) optical modulator with high frequency, SOI optical matrix switch and wavelength tunable filter are reviewed in the paper. The emphasis will be played on our recent results of SOI-based thermo-optic waveguide matrix switch with low insertion loss and fast response. A folding re-arrangeable non-blocking 4×4 matrix switch with total internal reflection (TIR) mirrors and a first blocking 16×16 matrix were fabricated on SOI wafer. The extinction ratio and the crosstalk are better. The insertion loss and the polarization dependent loss (PDL) at 1.55 μm increase slightly with longer device length and more bend and intersecting waveguides. The insertion losses are expected to decrease 2~3 dB when anti-reflection films are added in the ends of the devices. The rise and fall times of the devices are 2.1 μs and 2.3 μs, respectively.

Optical communication systems operating at 10Gbit/s require transceivers of low cost, size and power consumption, driving a "hot" source solution. This paper describes the current status of these "hot" devices for different applications. 10 Gb uncooled FP (Fabry Perot) to be used in conjunction with an EDC (Electronic Dispersion Compensation) receiver for LRM transceivers and CWDM (Course Wavelength Division Multiplexing) 3.125 Gb DFB for LX4 transceivers are the devices chosen for enterprise network (link up to 300m), which can address the need to transmit data at 10 Gb on the legacy multimode fibers. 10 Gb uncooled 1300 nm DFB (Distributed FeedBack) and 10 Gb 1300nm uncooled EML (Electro-absorption Modulator Laser) for LR transceivers are the devices chosen for LAN (Local Area Network) applications (link up to 10 km). Finally 10 Gb 1550 nm EML are the devices chosen for Metro applications (links up to 40 Km and potentially 80 km), for single wavelength and DWDM (Dense Wavelength Division Multiplexing) applications, so for those applications in which a directly modulated laser can not be used, due to the chromatic dispersion. All these laser sources can be used for different transceiver form factors: XENPAK, X2 and XFP.

In this paper, we report the high-temperature uncooled and high-speed directly modulated 1.3μm wavelength AlGaInAs/InP MQW-DFB ridge waveguide laser diodes. By optimizing the structure of active region based on AlGaInAs strained MQW, and the design of DFB grating, such as the position, coupling coefficient and the detuning with respect to the material peak gain, we have successfully fabricated high-speed and uncooled 1.3μm DFB laser diodes. Large bandwidth of 15GHz was achieved at room temperature. Large characteristic temperature of 80K and small degradation of slope efficiency of 1.2dB from 25°C to 85°C have been realized.

An extended subtraction method of scattering parameters for characterizing laser diode is introduced in this paper. The intrinsic small-signal response can be directly extracted from the measured transmission coefficients of laser diode by the method. However the chip temperature may change with the injection bias current due to thermal effects, which causes inaccurate intrinsic response by our method. Therefore, how to determine the chip temperature and keep the laser chip adiabatic is very critical when extracting the intrinsic response. To tackle these problems, the dependence of the lasing wavelength of the laser diode on the chip temperature is investigated, and an applicable measurement setup which keeps the chip temperature stable is presented. The scattering parameters of laser diode are measured on diabatic and adiabatic conditions, and the extracted intrinsic responses for both conditions are compared. It is found that the adiabatic intrinsic responses are evidently superior to those without thermal consideration. The analysis indicates that inclusion of thermal effects is necessary to acquire accurate intrinsic response.

After taking into account the multiple reflections of light in external cavity, the influence of the optical feedback on the large signal modulation characteristics of the external cavity semiconductor laser (ECSL) has been theoretically investigated. The numerical simulations show that, with the increase of the modulation index, the peak photon number of the ECSL tends to higer level on the whole. When the optical feedback is strong or weak, the peak photon number of the ECSL shows single period. However, the peak photon number exhibits bifurcation or chaos for different modulation index when the optical feedback is intermediate.

The theory of Light Emitting Diodes(LEDs) life tests and mathematic model of life tests were introduced. The performance of LEDs was affected by the drive current and by the ambient temperature. Life tests of tunnel junction regenerated AlGaInP LEDs were performed at different currents and ambient temperatures. On axis output intensity of tunnel junction regenerated LED had decreased 35.53% after 5203 hours at 30mA and 25°C. At the ambient temperature of 80°C, on axis output intensity of tunnel junction regenerated LED had degraded 19.26% after 3888 hours at 20mA. According to the results mentioned above, the normal working lifetime of tunnel junction regenerated LEDs were concluded. Moreover, the main Failure Mechanisms of it were described. Our work reviews the failure analysis that was performed on the degraded LEDs and the degradation mechanisms that were identified. The results show a thermal degradation mechanism that dominates degradation at high ambient temperatures.

In this paper, we calculated the strain distribution of low dimension structure using the elastic continuum model in the first part we focused on the lens-shaped quantum dot, and discussed the strain distribution of the quantum dot with and without capping layer. In another section, we gave a detail analysis about the strain distribution for quantum wire and quantum dot intergrowth structure, and there are two situations: the first, a little bigger quantum dot grown within a single quantum wire, the second, Y shaped quantum wires where quantum dot grown in the cross section of the wires, and the angle of the Y shaped wires lessen than 12 degree. All the low-dimension structures discussed have been observed in the laboratory. For the lens-shaped quantum dot with and without capping layer, the results showed that in both circumstances, the strain distribution would become more uniform in the quantum dot if the transverse size becomes larger. Compared with the open quantum dot (without capping layer), the capping layered quantum dot has a more sufficient strain relaxation, and even excessive strain relaxation is observed in the simulation. This phenomenon can be used to qualitative explain the quantum dot cave in and cavern out in the sequential capping layer growth interruption observed in laboratory. It is very promising for the quantum wire and quantum dot intergrowth structure. The phenomena of photoluminescence spectrum not changing with the variation of temperature have been observed recently in the intergrowth structure. For convenience, we used a simplified model to calculate such a structure, which we adopted rectangle quantum wire and rectangle quantum dot. The growth is along the [001] direction, and the wire is along the [110] direction. The results showed that the strain in quantum dot is greatly enhanced in the intergrowth structure compared with the single quantum dot system. There are strain-focusing region around the interface of quantum dot and quantum wire.

This paper introduces ZnO semiconductor optical amplifiers. In this paper, stimulated emission from zinc oxide nanoparticles are discussed in low and high powder density samples. 4 nm linewidth emission peak are observed in ZnO nanoparticles pump by Nd:YAG laser with 8 ns pulse width in high powder density sample. Scattering is strong when the nanoparticle density is high. And strong scattering is the reason of narrow peak in the nanoparticle. We also investigate the influence of the excitation spot area and the scatter concentration on the laser threshold in the TiO₂/Rhodamine B gain media, with frequency-doubled output of a mode-locked Nd:YAG laser (20 Hz repetition rate, 6 ns pulse width). Monte Carlo methods with ring laser were employed to calculate the threshold gain required for modeling the input-output characteristics in scattering gain media. We performed a Monte Carlo simulation of random walks in the geometry. With the increased pump area, the laser threshold is decreased. If we add more scatterers in the Rhodamine B, the laser threshold will reduce.

The localized synthesis of 4.2-5.6 nm-Si nanocrystals (nc-Si) in Si-rich SiO₂ (SRSO) by CO₂ laser annealing at laser intensity of below ablation-threshold (6 kW/cm²) is demonstrated. Since the SRSO exhibits a high absorption coefficient of up to 0.102 cm^-1 at wavelength of 10.6 μm, a direct-writing CO₂ laser annealing system with focusing spot size of 0.2 mm² is used to locally anneal the SRSO and precipitate the nc-Si. A thermophysical model reveals that the surface temperature of SRSO ranging from 130^oC to 3350^oC is achieved by varying the laser power densities from 1.5 to 13.5 kW/cm². The CO₂ laser-ablation-threshold power density is about 6 kW/cm², corresponding to the optimized annealing temperature 1285^oC at the ablation threshold. The CO₂laser annealing is capable of the precise control on power density and spot size, which benefits from the in-situ and localized annealing temperature control of SRSO film, and also prevents from the eternal damage of the other electronic devices nearby the annealing site. The nc-Si dependent photoluminescence (PL) were observed at 806 nm or longer, whereas the laser-ablation damaged SRSO film exhibits significant blue PL at 410 nm due to the oxygen-related structural defects. The refractive index of the lasertreated SRSO film is increasing from 1.57 to 2.31 as the laser intensity increases from 1.5 to 6.0 kW/cm² which is mainly attributed to the increasing density of nc-Si embedded in SRSO. High resolution transmission electron microscopy (HRTEM) analysis reveals that the average size of nc-Si embedded in SRSO film is about 5.3 nm, which correlates well with the theoretical prediction of a corresponding PL at 806 nm. The HRTEM estimated square density of the nc-Si in SRSO film under the laser intensity of 6 kW/cm² is about 10¹⁸ cm^-3.

A novel InP-based micromechanied tunable photodetector with the structure of OMITMC (One-Mirror-Inclined Three-Mirror-Cavity)is presented. The tuning characteristic of the device is analysed in the way of electrical actuation. Through simulation of the filter transmission spectra and the quantum efficiency, the characteristics of high tunability, narrow bandwidth and high quantum efficiency are analysed. As a result, the photodetector is tuned with 30nm,with a quantum efficiency of 59% and a linewidth of 1.2nm, when actuated by 10 volts.

A novel top-illuminated In_0.53Ga_0.47As p-i-n photodiodes (MM-PINPD) grown on GaAs substrate by using a linearly graded metamorphic In_xGa_1-xP (x graded from 0.49 to 1) buffer layer has been demonstrated on the SONET OC-192 receiving performance. With a cost-efficient TO-46 package, the MM-PINPD at data rate of 10 Gbit/s can be obtained at minimum optical power of -19.5 dBm. At wavelength of 1550nm, the dark current, optical responsivities, noise equivalent power, and operational bandwidth of the MM-PINPD with aperture diameter of 60 μm are 13 pA, 0.6 A/W, 3.4×10^-15 W/Hz^1/2, and 8 GHz, respectively. All the parameters are comparable to those of similar devices made on InP substrate or other InGaAs products epitaxially grown on an InGaAlAs buffered GaAs substrate. The performances of the MM-PINPD on GaAs are analyzed by impulse injecting of 1.2-ps pulse-train, eye pattern at 10Gbps, and frequency response from VNA.

Two types of waveguide photodiodes (WG-PD) - an evanescently coupled photodiode (EC-PD) and a separated-absorption-and-multiplication avalanche photodiode (SAM-APD) - have been developed for use in 40-Gbps receivers. The EC-PD simultaneously exhibited high external responsivity of 0.96 A/W and wide bandwidth of >40 GHz. The SAM-APD showed wide bandwidth of 30-35 GHz and gain-bandwidth (GB) product of 140-180 GHz. High receiver sensitivities of -11.3 dBm at 43 Gbps and -19 dBm at 40 Gbps have been successfully achieved for the EC-PD and the SAM-APD, respectively.

Adding resonant cavity to increase quantum efficiency of the monolithically integrated PIN/HBT-Receiver is described. Between the InP buffer and device epitaxial structure, InP/InGaAsP quarter wavelength stack (QWS) are used to form DBR (Distributed Bragg Reflectors) mirror. The PIN-PD is integrated within a Fabry-Perot cavity and the incident light is reflected many times by the Fabry-Perot cavity and consequently absorbed many times by the absorption layer. Therefore, the quantum efficiency of this detector is enlarged, meanwhile other performances such as frequency response are not influenced. We discuss the method to fabricate the resonance cavity, make theory simulation, optimize design on it, and analyze the advantage of this device.

The effect of bias voltage on the characteristic of GaAs photoconductive semiconductor switches (PCSS) was investigated theoretically and experimentally. The outputs of the switches for different bias voltages were obtained by solving the basic equations of transient model of PCSS. With a bias voltage of 2400V and triggered by a laser diode, the high gain PCSS switched a electric pulse with voltage up to 1700V. The simulated results agree with the experiment observations well. A new phenomenon of carriers accumulation effect was found.

In this paper the performance characteristics of compact optical 40 GHz pulse laser modules consisting of a monolithic mode-locked MQW DBR laser on GaInAsP/InP are reported. The monolithic devices were fabricated as tunable multi-section buried heterostructure lasers. A DBR grating is integrated at the output port of an extended cavity in order to meet the standardized ITU wavelength channels allocated in the spectral window around 1.55 μm in optical high speed communication networks. The fabricated 40 GHz lasers modules not only emit short optical pulses (< 1.5 ps) with very low amplitude noise (<1.5 %) and phase noise levels (timing jitter: 50 fs) but also enable good pulse-to-pulse phase and long-term stability. A wavelength tuning range of 6 nm is possible and large locking bandwidths between 100 ... 260 MHz are observed. All data have been achieved by operating the lasers in a hybrid mode-locking scheme with a required minimum micro-wave power of only 12 dBm for pulse synchronization. Details on laser chip architecture and module performance are summarized and the results of a stable and error free module performance in first 160 Gb/s (4 x 40 Gb/s OTDM) RZ-DPSK transmission experiments are presented.

A configuration of semiconductor fiber ring laser (SFRL) is used to achieve a tunable all-optical logic NOR gate at 20 Gb/s. Both gain and gain modulation is provided by a single semiconductor optical amplifier (SOA) in the ring cavity. In the theoretical simulations, sectional model is used for SOA and iterative condition is used for ring cavity to simulate the output from SFRL. All-optical NOR operation is successfully demonstrated experimentally at 20 Gb/s. There is a good agreement between theoretical simulations and experimental results. The extinction ratio higher than 14.9 dB is obtained as the output wavelength is tuned from 1530nm to 1560nm.

A novel scheme for all-optical frequency multiplication/recovery based on the semiconductor optical amplifier (SOA) ring cavity is proposed and investigated numerically. The results show, for a 2.5GHz driving pulse train, it can be generated 5-25GHz repetition rate pulse trains with low clock amplitude jitter (CAJ), polarization independence and high peak power. Furthermore, the extraction of the clock signal from a pseudorandom bit sequence (PRBS) signal can be realized based on the proposed scheme.

We investigate the optical properties of a one-dimensional semiconductor quantum wire in the presence of a static electric field applied along the axis. Optical absorption spectra near the band gap are nonperturbatively calculated by solving the low-density semiconductor Bloch equations in real space. The influence of the Franz-Keldysh effect and the exciton ionization on the continuum of the absorption spectrum in semiconductor quantum wire is investigated. We found substantial and tunable absorption oscillations appear above the band gap for reasonable electric strengths. This shows semiconductor quantum wire has great promise of the potential applications in electric-optical devices.

From the theoretical analysis for symmetric coupled quantum well, a novel coupled quantum well structure with low driving electric-field, low absorption loss and large field-induced refractive index change at 1.55 μm is put forward.

The characteristics of evanescent waves in metamaterial photonic crystal composed by PIM and NIM layers are discussed. Firstly, a single period of photonic crystal is considered as the barrier in FTIR structure, and the transmittance, the post-tunneling position shift and the tunneling time of various frequency incident waves through the barrier are calculated. In the case of the barrier of more periods of photonic crystal, the transmittance and the tunneling time also vary with layer number, which are analyzed in detail.

The static and dynamic characteristics of two-mode vertical-cavity surface-emitting lasers (VCSELs) under mode-selective feedback are investigated numerically. The model includes the spatial dependences of the optical modes, the carriers and the injection current. The research is respectively done for two different cases: weak coupling and strong coupling cases. For weak coupling case, the disc injection current is introduced and the LP₀₁ and LP₁₁ modes are included; while for strong coupling case, LP₁₁ and LP₂₁ modes are excited simultaneously by using a ring injection current. Besides, LP₀₁ and LP₁₁ modes are selectively reflected back by the external mirror for the weak and strong coupling cases, respectively. For short external cavities, the laser keeps its continuous wave (CW) operation. The modal power varies periodically with respect to the external cavity length and increases (or decreases) with the increase of the external mirror reflectivity. The variation trend of modal power is determined by the external cavity length, which decides the relative phase of the reflected light and the in-cavity one. Moreover, we find that the two modes under consideration exhibit anti-phase behaviors with the variation of the feedback conditions. Especially, for the strong coupling case, the enhancement of one mode can result in the thorough suppression of the other mode. For long external cavities, however, the relaxation oscillation process of laser becomes undamped, which induces rich nonlinear dynamics for both the reflected mode and the one without feedback. Through computer simulation, many typical dynamics, such as the CW, period-one, period-two, period-four, quasi-period, and chaotic states, are observed for both modes. It is also shown that fixing the feedback level, the dynamics shown by the laser exhibit periodic evolution with respect to the external cavity length; however, for a given external cavity length, a distinct period-doubling route to chaos is found with increasing external mirror reflectivity. Therefore, although just one mode is directly affected by the external feedback, the other mode still exhibits the similar nonlinear behaviors due to the carriers dynamics induced by the mode with feedback.

In AlGaInP/GaInP multi-quantum well (MQW) lasers, the electron leakage current is a much more serious problem than that in laser diodes with longer wavelength. To further improve the output performance, the leakage current should be analyzed. In this letter, the temperature dependence of electrical derivative characteristics in AlGaInP/GaInP multi-quantum well lasers was measured, and the potential barrier for electron leakage was obtained. With the help of secondary ion mass spectroscopy (SIMS) measurement, theoretical analysis of the potential barrier was presented and compared with the measurement result. The influence of p-cladding doping level and doping profile on the potential barrier was discussed, and this can be helpful in metalorganic chemical vapor deposition (MOCVD) growth.

Absolute frequency reference in 1.5-μm wavelength region is very useful for DWDM system. We have studied on producing artificial 1.5-μm wavelength region absolute frequency reference by calibrating the F-P etalon. Double acetylene absorption line calibrating technique was adopted. An experiment of calibrating the Fabry-Perot etalon using double acetylene absorption lines has been demonstrated. In the experiment, Mode 1996 of the F-P etalon was locked to acetylene absorption line 1533.4621 nm. The frequency stability of the calibrated F-P etalon was about 7×10^-8. After calibrating, a series of artificial absolute frequency references in 1.5-μm wavelength region was obtained. To reduce the complexity of calibrating, a novel frequency difference measurement method was introduced.

Beam quality is very important in laser technology and its application. In practice a good beam quality evaluating method can indicate which beam is good and which laser is suitable. Therefore evaluating beam quality is of huge significance. Because the beam of laser diode has many characteristics differ from other lasers, such as the angle of the beam is too big, and the angle in the direction parallel to the junction plane disagrees with it in the perpendicular direction, a new quality parameter Q(θ,ω,d) for evaluating beam quality of laser diode is given. The difficulty of collimation is well expressed by use of Q. The smaller the value of Q is, the more easily the beam is collimated.

980nm InGaAs/GaAs separate confinement heterostructure (SCH) strained quantum well (QW) laser with non-absorbing facets is fabricated. The microchannel coolers is designed and fabricated with a five-layer thin oxygen-free copper plate structure. We report the operating characteristics of 980 nm high power semiconductor laser stacked arrays packaged by microchannel coolers. A highest CW output power of 200 W for 5-bar arrays is obtained.

The laser drive is one of the key components needed to generate the high-current, high-speed signal required for driving a transmitter, This paper propose a novel parallel laser drive circuit model with adaptive bandwidth bus architecture based on hybrid current/voltage mode. By using the highly parallel characteristics of the VCSEL and the bus structure, the system becomes very compactable. Several improvements can be achieved in signalling speed by using current mode in comparison to voltage mode. Therefore we use the current mode control module. By using the on hybrid current/voltage mode in the data path, the static power dissipation is reduced greatly when the data path is idle. Overall power dissipation improvement is attained over voltage-mode signalling schemes at high data rates while minimizing the power dissipation.

The temperature greatly influences the characteristics of external cavity laser diode (ECLD). The high accurate temperature control of ECLD is one of the pivotal technologies of stable output wavelength. In this paper, the principle and method for high accurate temperature control of ECLD are described. The scheme of bistable temperature control is given, which is combined by the analog circuit with PID control theory and the digital circuit design based on microcontroller, and it is studied and analyzed.

A reference model is setup for a high stability temperature controlling in LD by testing the response characteristics of LD refrigerator, and an SoC based digital temperature controller is designed and realized, which is simple in hardware, and flexible in software. Experiment results demonstrate that the temperature drift of LD is within ±0.05 °C.

A wavemeter based on Michelson interferometer accurately measure static wavelength of a tunable laser. Its operation principle is formulated in details. Double longitudinal-mode He-Ne laser with frequency stabilization is used as the reference optical source of the wavemeter. Voice-coil motor using PID means can realize to move in uniform motion. Phase-locked loop circuit including NE564 and 74LS193 is used to enhance resolution of the wavemeter. Data processing is carried out by the counter unit including two 8254 programmable timer, a MCU, a LCD. The test shows that its measurement accuracy is 1×10^-6 and is higher than those of other wavemeters such as Fizeau interference and Fabry-Perot wavemeter.

With the improved theoretical model for wavelength conversion, we demonstrate that the characteristics of wavelength conversion such as extinction ratio, signal-noise ratio and bit error rate is relative to the mutual coupling parameter introduced into the rate equations.

Wavelength converter is key equipment in dynamic optical networks and the injection locking technology in semiconductor lasers is a promising low cost candidate. High speed injection-locked wavelength conversion (10 Gbit/s, NRZ, PRBS) has been achieved in our simulations and experiments, both logically inverting and non-inverting. In this paper, injection-locked model in FP lasers is analyzed and the results are consistent with those in both static and dynamic experiments. The situation when input and output lights are at different facets is also theoretically compared with that at the same facet and it's derived that the latter may have better performance.

A deep self modulation output in the diode pumped Bi-directional CW Nd:YVO4 ring laser is observed. The self-modulation has multiple styles and in certain style it shows the distinctness sensitive of rotation. The phenomenon in our experiment is reported in detail. There will be a potential application to the rotation sensor or gyroscope.

In this paper, after considering the facet residual reflectivity of the semiconductor optical amplifier (SOA), a theoretical model of the SOA based harmonic mode-locked fiber ring laser has been established. Using this model, the influence of the facet residual reflectivity on the shape, peak power, and pulse width of the pulse output from the harmonic mode-locked fiber ring laser, has been investigated.

A novel all-optical header and payload separation technique that can be utilized in un-slotted optical packet switched networks is presented. The technique uses a modified TOAD for packet header extraction with differential modulation scheme and two SOAs that perform a simple XOR operation between the packet and its self-derived header to get the separated payload. The main virtue of this system is simple structure and need not any additional continuous pulses. Through numerical simulations, the operating characteristics of the scheme are illustrated. In addition, the parameters of the system are discussed and designed to optimize the operation performance.

Cross modulations such as cross polarization modulation, cross gain modulation, cross phase modulation are import nonlinear effects in semiconductor optical amplifier (SOA). In this paper, the relationship among them is explored systematically; theoretical simulation is made based on a multi-section SOA model which takes into birefringence effect account, and the results agree well with experimental results. The linear relationship among cross gain modulation (in dB unit), cross phase modulation and cross polarization modulation is indicated. A more significant cross modulation effect is observed when pump wavelength is nearer to peak-gain wavelength region, while it becomes smaller when pump wavelength is more away from peak-gain region. No perceptible polarization state change of output probe beam is observed when the input polarization of pump beam is varied due to that the gain of the SOA is polarization insensitive.

AlGaAs/GaAs material diode lasers grown by MOCVD using TBA as the group-V source and N₂ as the carrier gas, was reported. Lasing has been successfully achieved with a low threshold current density of 506 A/cm².

This paper presents the structure design and fabrication technology of 850nm wavelength high power wide spectrum Superluminescent Diodes (SLDs) as non-coherent light source, for the application of fiber gyroscope and other areas. Quantum Well epitaxial structure, unpumped absorbing region structure and facet coating methods have been adopted for enhancing the gain coefficient, output power and the reduction or elimination of lasing oscillation. As typical device performance results, SLDs have been demonstrated with central wavelength of 848-851nm, spectrum FWHM no less than 20nm, and no less than 7mW output under 120mA injection current. The devices operated up to 100°C.

This paper presents the structure design and fabrication technology of 850nm superluminescent Diodes (SLDs).Various ways have been tried for the suppression of F-P lasing oscillation to realize superluminescence: Tilted-stripe structure, tandem-type structure and non-injection section near the rear facet are introduced. Three structures are also compared and combined with each other. The device not lasing at maximum injection current 200mA is realized. At injection current of 150mA, the maximum output power can be 7.8mW and the device can still work at 100°C.

In this letter, we propose a novel optical modulator based on GaAs photonic crystals and investigate its optically properties numerically by using the finite-difference time-domain method. The position of the cutoff frequency can be varied by free carriers injection, and the band gap shift can be observed. Band gap shift is used to modulate light. Bing several micrometers length, low insertion loss, and large extinction ratios, the modulator can be used in ultra-small and ultra-dense photonic integrated circuits.

Narrow frequency sharp angular filters with integral times of quarter wavelength layer structure, which can be fabricated with commonly used coating machines and work in visible and near infrared, are designed. These filters possess not only a narrow frequency pass band but also a sharp angular pass breadth in visible and near infrared. The experimental result agrees with the theoretical calculations well. These dual functional filters have many potential applications.

In this paper, the development and filter characteristic of photonic band-gap (PBG) structure and defected ground structure (DGS) are analyzed. The non-periodic structure has simpler structure, smaller size and smaller ripple compared to the periodic structure. Though the stop band of non-periodic structure is narrower, it can meet the application. What's more, C-shaped structure with two stop bands can realize selection in special frequency band. So it can meet the need of two stop bands.

We propose a kind of planar photonic crystal micro-cavity which is truly two-dimensional and is composed of a line of periodic air-holes and a set of periodic air grooves both with defect. Unlike the photonic crystal micro-cavity formed by distributed Bragg reflection (DBR) layers and two dimensional holes in a semiconductor substrate, the micro-cavity proposed is a true two-dimensional planar structure that is easy for mass-production through die-press copying. Numerical simulations by the FDTD method show that there exist resonance modes inside the photonic bandgap. The resonance wavelength is mainly decided by the period and radius of the holes. The quality factor of the cavity is mainly decided by the number of periods of the holes and the grooves.

Silicon photodetector can be integrated with all kinds of Silicon circuits to get monolithic OEIC. A CMOS-process-compatible silicon double-photodetector with structures of P⁺/N-well and N-well/P-substrate, called PD1 and PD2 respectively, is designed in this paper. The theoretical absolute spectral response and response speed of this double-photodetector are calculated and analyzed. Simulation results in 0.5um standard CMOS process show that the responsivity of the double-photodetector is above 0.2A/W from 400 to 900nm wavelength range without ARC (Anti-Reflection-Coating). Both the effects of the insulated medium layers (SiO₂ and Si₃N₄) in CMOS process and reverse voltage on spectral responsivity are also discussed. When the optical window area is 16.54μmx16.54μm², the capacitance of PD1 is about 100fF at a reverse voltage of 2.5V. Yet the capacitance of PD2 is almost 1/10 of PD1. With a load-resistor of 50Ω, the response speeds of PD1, PD2 and double-photodetector are 0.628, 2.04 and 2.05ns at 650nm wavelength (corresponding bandwidth about 276MHz, 85MHz and 84.7MHz), respectively. Finally, the co-design of Monolithic OEIC is also discussed. A full CMOS monolithic OEIC for optical-disc signal pickup is designed with this double-detector.

Thin films of vanadium dioxide (VO₂) were selected for microbolometers. The thin films were fabricated with a novel method mainly including ion-sputtering and annealing. It is found that the electrical properties of these thin films can be controlled by adjusting the time of ion-sputtering and annealing. A standard microbolometer pixel structure of micro-bridge has been applied. Two-dimensional arrays of microbolometers have been fabricated on silicon integrated circuit wafers using a surface micromachining technique. A new type of on-chip readout integrated circuit (ROIC) for 32×32 pixel bolometric detector arrays has been designed and fabricated using a 1.5μm double metal poly complementary metal oxide semiconductor (CMOS) processing. The readout circuit consists of three stages, which provides low noise, a highly stable detector bias, high photon current injection efficiency, high gain, and high speed. Several prototypes of 32×32 pixel bolometric detector arrays have been designed and fabricated. These arrays consist of detectors with lateral dimensions of 50μm 50μm, and each bolometric detector is on a 100μm pitch. The results of measurement show that the fabricated uncooled infrared focal plane arrays (UIRFPAs) have excellent performance. The frame rate is 50Hz, the pixel operability is above 96%, the responsivity (R) @ f/1 value is up to 15000V/W, the noise equivalent temperature difference (NETD) @ f/1 and 30Hz is about 50mK, and the average power dissipation is only 24.7mW. The results indicate that the technology of fabricating these 32×32 UIRFPAs has potential to be utilized for fabricating low cost and large-scale UIRFPAs.

Analysis of monolithic integrated MQW waveguide and passive waveguide with ATG structure is presented by using normal modes theory. It is based on the analysis of normal modes of asymmetric five-layer waveguide, combined with theoretic model of MQW waveguide. The electromagnetic field distribution, dispersion equation, calculation and simulation results are reported.

Quasi-phase-matched parametric downconversion via cascaded optical nonlinearities in an aperiodically poled MgO:LiNbO₃ superlattice was studied in theory and experiment. Coupling equations reveal that multiple-wavelength parametric downconversion in a wide infrared spectrum range can occur in a single crystal. Enhancement of the conversion efficiency and output stability through coupling of two nonlinear processes is demonstrated. The result also reveals that cascaded parametric downconversion process can be used to efficiently downconvert the fundamental wavelength to longer wavelength of infrared region.

In this paper, we present two new photonic crystal structures, which are composed of fractal Cantor multilayer with defects embedded in its middle. Optical transmission matrix method is used to calculating the transmittance and reflectance. Compared with general Cantor multilayer, we find these new structures have wider stopbands and show super narrow bands in the middle of wider stopbands. They can be served as super narrow bandpass filters. The pass band obtained can be less than 0.6nm near the infrared 1530 nm when there is a defect embedded in the cantor multilayer. The optical transmission in the center wavelength is higher than 99%. This means a very low insert loss. If there are three detected layers, three super narrow peaks can be found in the middle of the stopband. The center wavelengths are 1232.4 nm, 1372.8nm and 1538.3 nm, respectively. It is more superior to other kind narrow band filters. These kinds of photonic crystal super narrow band optical filters may find applications in super dense wavelength division multiplexing for optical communications and precise optical measurement.

In this paper an Optical Circular-Polarization Modulator (CPM) based Liquid Crystal (LC) is presented. Using the electric birefringence of LC material, the device can output polarized light with two circular-polarization states. Detailed illumination of the device's structure and its testing investigation are also presented.

The GaSb characteristics grown by molecular beam epitaxy (MBE) on GaAs substrates was reported. The abruptness of the interfaces, the degree of intermixing and the anion incorporation greatly affect the material quality. The RHEED patterns provide information on the surface structure and morphology of the sample and dictate surface reconstruction, accumulation and segregation. The structure parameters of samples are obtained from the rocking curve. The first and second satellite peaks appear around the main 0th-order peak. The experimental and simulated results of samples A and B with x-ray rocking curves show there is a GaAsSb layer because of As-for-Sb exchange at the GaSb/GaAs interface.

Two-step growth method was used to grow InP epilayers directly on GaAs (001) substrates. By employing double-crystal x-ray diffraction (XRD) to characterize the epilayers and analyzing the value of full width at half maximum (FWHM) of ω scan rocking curve, we found the initial buffer layer act a key role on the quality of epilayers. Depending on optimizing the thickness and growth temperature of the initial buffer layers, we have succeeded in improving the crystallinity of InP epilayers. When the low temperature buffer layer was 10 nm thickness and grown at 450°C, the quality of InP epilayers for 1μm thickness were the best, its FWHM of XRD ω scan rocking curve was only 512 arcsec and 201arcsec for ω-2θ scans, the room temperature photoluminescence spectrum shows the band edge transition of InP, its central wavelength is 921nm and the FWHM is only 38 meV.These results indicate high quality of InP epilayers on GaAs substrates.

In this thesis, the Kinetic Monte Carlo method is applied for the study of epitaxial growth. In order to study the effect of a homogeneous substrate, the lattice is divided into square-shaped domains. From the island morphologies, we can see that the confinement effect are found to have strong dependence on the growth conditions, such as decrease deposition flux F, increase lateral nearest neighbor atom term E_N or increase temperature T which can make a more formal substrate. The island size distributions are also be calculated for a detail illumination.

Hydrogenated nanocrystalline silicon thin films have been deposited by helicon wave plasma chemical vapor deposition technique over the range of magnetic field strength (0-200Guass) using a mixture of silane and hydrogen gas. The effects of magnetic field on the structural and optical properties of the deposited films are characterized by Raman spectroscopy, x-ray diffraction and ultraviolet-visible transmittance spectroscopy. It has been found that the crystalline fraction and optical band gap E_g of the films change with magnetic field strength monotonically, while the deposition rate, the grain size and the B factor relating to the overall structural disorder reach a maximal value at 150 Guass. The variation of Raman scattering intensity ratio between silicon TA and TO mode indicates that the inner microstructure of the films becomes more orderly with increasing magnetic field strength until 150 Guass. The decrease trend of the optical band gap E_g and the changes of the B value with magnetic field strength from the ultraviolet-visible transmittance spectroscopy analysis are related to the increase of the grain size, the crystalline fraction and the hydrogen. The observed results have been discussed by the increase of hydrogen weak bond etching and dangling bond passivating effect with increasing magnetic field strength.

Nanocrystalline cubic silicon carbide thin films have been fabricated by helicon wave plasma enhanced chemical vapor deposition (HWP-CVD) on Si and Corning 7059 glass substrates using the mix plasma of SiH4, CH4, and H2. The effect of negative radio-frequency (rf) bias voltage on the optical and structural of the deposited hydrogenated nanocrystalline SiC (NC-SiC:H) films has been investigated by Fourier transform infrared (FTIR) spectroscopy, ultraviolet-visible (UVVIS) transmittance spectroscopy, and photoluminescence (PL) spectroscopy. It is found that with increasing the negative rf substrate bias, the NC-SiC:H thin films become denser and have fewer defects. The PL measurement indicates that all the deposited film present a strong light emission at the room temperature under an excitation of the 370 nm line of a Xe lamp. The blue-green PL peak can be ascribed to quantum confine effect of small size SiC nanocrystal in the film.

Hydrogenated amorphous silicon nitride (a-SiN_x:H) thin films are deposited by helicon wave plasma chemical vapor deposition technique. The structural and photoluminescence properties of these films have been characterized by X-ray photoelectron spectroscopy (XPS), Photoluminescence (PL) and ultraviolet-visible (UV-VIS) spectroscopy. It is shown that the silicon atom bonds exist in the Si-Si and Si-N configurations and the amorphous silicon regions appear separately in the Si-rich a-SiN_x films. All the PL spectra of the deposited films manifest itself as several interference peaks superposed on an energy-dependent Gaussian distributed band. The PL and absorption results of the deposited films with different nitrogen content support that the luminescence of the Si-rich a-SiN_x:H films is related to the photo-excited carriers radiation process in the separated amorphous silicon potential well region, while the blue shift of PL main peaks and the enlargement of PL intensity with increase nitrogen content are ascribed to the size reduction of amorphous silicon separated regions and the enhancement of confinement effect.

A Surface Plasmon Resonance(SPR) sensor based on Kretschmann configuration has been setup. In this setup, Ag was applied as supporting metal, and incident angle was fixed. There are two main method, one is Surface Plasmon Microscopy(SPM) based on optical intensity, the other the Surface Plasmon Interference Microscopy (SPIM) based on the theory of Surface Plasmon Resonance and optical interference. SPM and SPIM were analyzed both theoretically and experimentally. Comparing and analyzing the result of theory and experiment, the result shows that SPIM has higher spatial resolution than SPM, and more powerful and immune to noise due to ambient light. So the SPIM is more fitful for sensor applications than SPM.