This paper has three sections on computation revolutions in the information age. In the first section key examples of revolutions in the early history, i.e., how to overcome heat penalty arisen from high power consumption in silicon electronics will be presented to bring to light in the ever evolving nature of the computation. In the second section the optical interconnection in terms of silicon photonics will be discussed as the way in which computation have fundamentally altered the previous trend of power consumption. The final section will focus on a paradigm shift in computation, i.e., an architecture of logic implementation, binary decision diagram, to replace to current computation system without fighting transistor logic circuitry.

Recent progresses in silicon-based micro/nano photonic devices and optoelectronic integration are summarized, including Si-based light sources, waveguide devices, modulators, detectors and integration attempts. Surface plasmons, photonic crystals and other methods to realize these devices are described. The potential and possibility to realize all-silicon devices and monolithic integration or make the fabrication compatible with standard CMOS process are presented and discussed, and their applications in high density data communications are emphasized.

Ge nanostructures embedded in silica matrix are emerging as a promising material for new generation devices due to the unique electric and photonic properties. In this paper, Ge nanoclusters and nanocylinders with Ge shell were successfully formed by the high energy electron irradiation in the PECVD deposited glass. In addition, large area Ge nanoclusters were also created by heat-treatment of PECVD deposited glass film. These nanostructures were characterized in terms of size, composition, distribution and crystalline state by using TEM, HRTEM, EDS, SEM, Raman spectroscopy, and SIMS. Waveguides doped with Ge nanoclusters were fabricated and their absorption has been characterized in a wavelength range from 500nm to 1700nm.

A novel photonic buffer memory with a shift register function is proposed. The buffer memory consists of a two-dimensional array of polarization bistable vertical-cavity surface-emitting lasers (VCSELs), in which the bit state of the optical signal, "0" or "1" is stored as a lasing linear polarization state of 0 or 90°. Input data stored as the polarization states of the first VCSEL are transferred to the polarization states of the second VCSEL. In our experiment using 980 nm VCSELs, 10 Gbit/s optical buffering, 2-bit optical buffering, and a shift register function have been successfully demonstrated.

We proposed, experimentally and theoretically demonstrated all-optical two line-four line encoder and two bit-wise comparator of RZ data streams at 40Gb/s based on cross gain modulation (XGM) and four wave mixing (FWM) in three parallel SOAs. Five logic functions for digital encoder and comparator between two signals A and B: A¯ B¯, AB¯, AB¯, AB and A¯⊗¯B¯, were achieved simultaneously. The first three optical logics are realized based on XGM in SOAs, the fourth is realized with FWM, and the fifth is the mixing result of the first and the fourth. A detuning filter is employed to improve the output performance. The output extinction ratio (ER) for the XGM operation is above 10dB, and the ER for FWM operation is around 8 dB. Wide and clear eye patterns for the five logic outputs can be observed. The all optical digital encoder/comparator we demonstrated has advantages of simple structure, multifunctional optical logic functions and high speed.

We propose an ultrafast all-optical logic AND gate based on two cascaded SOA-OF configurations. Each SOA-OF consists a semiconductor optical amplifier (SOA) followed by an optical filter (OF) which reshapes the spectrum of the modulated probe light. A delayed interferometer (DI) and a tunable band pass filter (TNBPF) are chosen as OFs in the first and second SOA-OF. 40Gbit/s AND operation with 33% duty cycle return-to-zero (RZ) signals has been successfully demonstrated with SOAs whose 10%~90% recovery time are measured much larger than the time duration of one bit period. The quality factor (QF) and the extinction ratio (ER) of the eye diagram of the derived AND results were 6.3 and 8.8dB respectively. Numerical analysis and experimental demonstration with 40Gbit/s nonreturn-to-zero (NRZ) signals is also presented and shows that in order to achieve good AND result at the second stage, the differential time delay of the DI must be shorter than the single bit period of the input signals. The proposed AND gate takes the advantage of high speed WC realized by SOA-OF which displays flexibility to various data rates, pulse width as well as data formats.

Based on a three-band model, a new theoretical model is derived to describe the self-polarization modulation (SPolM) effect in semiconductor optical amplifier (SOA) on femtosecond time-scale. In the case of SPolM, the phase difference between the TE and the TM modes is created by the signal itself and results in changes of polarization states of the signal. Numerical experiments are carried out to analyze the gain and the signal induced phase shift when an ultrashort optical pulse injected into a tensile strained SOA. The higher injected power can induce a larger phase shift. The polarization-dependent effect has much more influences on strong optical pulses.

A tunable Mach-Zehnder wavelength duplexer has been realized based on P-i-n-N InGaAsP/InP. It has been made polarization insensitive by proper wafer layer stack and proper waveguide geometry. The layer stack for the duplexer was tested first with a waveguide phase shifter, which resulted in up to 36°/(V•mm) phase shifting efficiency for TE polarization, which is slightly more efficient than the most efficient phase shifter reported to date in bulk InP at 1.55 μm, and with much lower transmission loss[1]. The transmission loss was measured to be 4 dB/cm (5 dB/cm) for TE (TM) polarized light, for 2 μm wide shallowly etched waveguides, which is rather low compared to other reported high efficiency phase shifters for this material system. With this layerstack, we designed a Mach-Zehnder (MZ) duplexer with narrow, 1.5 μm wide, deeply etched phase shifters that meet the polarization insensitivity requirement. The measurement results showed that the phase shifting efficiency of this narrow and deeply etched duplexer is up to 34°/(V•mm) for both TE and TM polarization, and the transmission loss of this 1.5 μm wide waveguide is about 10 dB/cm for both TE and TM polarization. This is also the first reported deeply etched narrow phase shifter with high phase shifting efficiency and relatively low loss.

We propose and demonstrate 40Gb/s multifunctional all-optical logic gates based on single semiconductor optical amplifier (SOA) and a blue shifted optical bandpass filter (OBF), suitable for both return-to-zero (RZ) and nonreturn-to-zero (NRZ) format. The logic functions NOT, NOR and OR of RZ format are realized at the OBF detuning of -0.15nm, - 0.22nm, and -0.44nm, respectively. The logic functions NOT and NOR of NRZ format are realized at the OBF detuning of -0.24nm. The measured ER is around 7dB and Q factor is over 6. Our scheme has the potential advantages of multilogic functions, simple structure, and high tolerance to input pulsewidth.

Theoretical calculations of the optical properties of InGaAsP quantum well (QW) electroabsorption modulators (EAM's) operating at c-band (~1550 nm wavelength) is presented. Absorption coefficients of QW's are obtained from the linear optical susceptibility. Excitons are calculated in momentum space, which includes valence-band mixing, mixing of excitons originating in different subband pairs, and exciton spin-related optical selection rules. Various line-broadening mechanisms relevant to InGaAsP-QW's are also included. Investigations on asymmetric double QW's (ADQW's) show that the small-signal modulation efficiency, which is an important figure of merit for analog application, can be enhanced significantly at substantially reduced operating bias voltage. Simple optimization of ADQW band structure results in a maximum slope efficiency ~3.8 times larger than that of SQW EAM's at a reduced operating bias field of 34 kV/cm compared with ~70 kV/cm for comparable SQW's.

The semiconductor thin disk laser is a new type of semiconductor laser. This work gives the basic operation function of semiconductor disk laser, and analyses the heat effect by the experimentally measured photoluminescence spectrum of the laser chip at different pump power and different temperature. We can see that: with increasing pump power, the thermal effects of the gain material becomes seriously and causes the saturation of carrier lifetime, so the electron-hole pair created in the absorbtion layer have no enough time to rate to one of the wells, and the non-radiative recombination happens in the barrier. When the thermal effects becomes stronger, the chip will not lasing. This phenomenon is from the smaller energy offset between barrier and quantum well. We optimize the original structure design and experimental technology. A non-absorbing AlGaAs layer who is transparent to the pumping and laser wavelength is added to confine the carriers in the quantum wells. At the same time a DBR with double reflecting band is induced to improve the absorbing efficiency of the pumping light. The single QW is replaced by the three narrow QWs, This three QWs structure can add the quantum state of QW, increase the recombination probability of carriers in the QWs and reduce the heat effect. The chemical etch equipment is also improved to control the surface unevenness to be within 50 nm.

In this paper, we demonstrated for the first time variable 1.5μm wavelength conversion through cascaded second order nonlinear processes "SHG+DFG" by fan-out grating in lithium niobate waveguide. We fabricated the waveguide by annealed proton exchange in periodically poled LiNbO₃ (PPLN). The device used in this experiment is 4 cm long, has a QPM period from 14.8μm to 15.2μm, waveguide width of 12μm, proton exchange depth of 0.7μm, and was annealed for 32h at 350°C. After proton exchange in pure benzoic acid using a SiO₂ mask, the substrate was annealed in an oxygen atmosphere. The wavelength of signal light was set at 1551.3 nm. The wavelengths of tunable pump lights we used in experiment were 1543.2 and 1556.2 nm, and the corresponding grating periods were 14.87 μm and 15.03 μm, respectively. The temperature was set at 100.5°C to avoid photo refractive damage and to match the QPM peaks to the pump wavelengths. The conversion efficiency was about 10dB to be expected with the pump power 175mW in a similar device with a slightly different QPM period and operated at 125°C.

The design and fabrication of a Monothically integrated dual-wavelength tunable photodetector are reported. The dual-wavelength character is realized by introducing a taper substrate. The photodetector operating on long wavelength is Monothically integrated by using heteroepitaxy growth of InP-In_0.53Ga_0.47As-InP p-i-n structure on GaAs based GaAs/AlAs Fabry-Perot filter structure, which can be tuned by thermal-optic effect. High quality heteroepitaxy was realized by employing a thin low-temperature buffer layer. The integrated device with a dual-peak distance of 7nm (1530nm,1537nm) , a wavelength tuning range of 5.0 nm, and a 3-dB bandwidth of 5.9 GHz was demonstrated, according with the theoretical simulation.

In this report, the multiplication characteristics of InP/InGaAs avalanche photodiode (APD) with thick multiplication and charge layer have been studied theoretically and experimentally, considering the electric field distribution, carrier concentration, and different multiplication layer thickness. We find that ionization in the charge layer is very sensitive to avalanche multiplication (M) and breakdown voltage (V_br). Partial ionization in the charge layer has been suggested, which gives a good description of experimental results.

Abstract: We have explored the shared-layer integration fabrication of an resonant-cavity-enhanced p-i-n photodector (RCE- p-i-n-PD) and a single heterojunction bipolar transistor (SHBT) with the same epitaxy grown layer structure. MOCVD growth of the different layer structure for the GaAs based RCE- p-i-n-PD/SHBT require compromises to obtain the best performance of the integrated devices. The SHBT is proposed with super-lattice in the collector, and the structure of the base and the collector of the SHBT is used for the RCE. Up to now, the DC characteristics of the integrated device have been obtained.

The direct modulation characteristics and problems of a novel monolithically integrated transceiver for FTTH is analyzed in this paper along with the proposing of the novel structure and performance of this device. Through these analyses, its output beam spectrogram under 8GHz and 10GHz modulation are investigated and achieved.

All optical communication era is finally about to start. Penetration of optical communications into the residential area is under progress worldwide and the rate will be increased due to the bandwidth demanding services emerging consistently. To accommodate all the rapid increases of optical network expansion, vast of optoelectronic components are necessary. Costs and volume productions are key elements. One of the solutions is photonic integration which can reduce the cost of labors and have the potential of large volume productivity. In this paper, several photonic integrated devices which are currently under development in our lab are discussed.

InP-based vertical cavity surface emitting lasers (VCSELs) with AlGaInAs QWs and AlGaInAs/InP DBR have been demonstrated. Over 3 mW and ~1 mW powers at both 1.3 μm and 1.55 μm have been achieved at 20 °C and 85 °C, respectively. Tests for various applications have been performed with our 1.3 and 1.55 μm VCSELs. Error free transmission over 10 km under 10 Gbit/s, 85°C has been demonstrated with both 1.3 and 1.55 μm VCSELs. The effect of electrical dispersion compensation (EDC) with 1.55 μm VCSELs has been confirmed for transmission of medium range data transmission. Radio signal transmission with low error vector magnitude by 1.3 μm VCSELs has been achieved at 2.4 and 5 GHz-band radio frequency.

We demonstrate wafer-scale integration of a saturable absorber in a surface emitting semiconductor laser. Vertical external cavity surface-emitting lasers (VECSELs) have high quality circular output beams, 2D-array scalability, and high average power. To date, ultrafast VECSELs required a folded cavity with a separate saturable absorber device for passive modelocking. In the result presented here, we integrate the saturable absorber into the same semiconductor wafer, optimize its performance for integration with quantum dots and demonstrate stable passive modelocking in a simple straight external cavity which allows for a fully monolithically wafer-integrated structure to reduce cost and improve ease of mass production. We refer to this class of devices as the modelocked integrated external-cavity surface emitting laser (MIXSEL). Such devices would be ideally suited for many applications where the current ultrafast laser technology is considered to be too bulky and expensive.

High power vertical cavity surface emitting lasers with large aperture have been fabricated through improving passivation, lateral oxidation and heat dissipation techniques. Different from conventional three quantum well structure, a periodic gain active region with nine quantum wells was incorporated into the VCSEL structure, with which high efficiency and high power operation were expected. The nine quantum wells were divided into three groups with each of them located at the antinodes of the cavity to enhance the coupling between the optical field and the gain region. Large aperture and bottom-emitting configuration was used to improve the beam quality and the heat dissipation. A maximum output power of 1.4W was demonstrated at CW operation for a 400μm-diameter device. The lasing wavelength shifted to 995.5nm with a FWHM of 2nm at a current of 4.8A due to the internal heating and the absence of active water cooling. A ring-shape farfield pattern was induced by the non-homogeneous lateral current distribution in large diameter device. The light intensity at the center of the ring increased with increasing current. A symmetric round light spot at the center and single transverse mode operation with a divergence angle of 16° were observed with current beyond 4.8A.

InGaAsP/InP rectangular ring lasers based on active vertical coupler structure are demonstrated in terms of loss-reduction. Varied thresholds for different configuration devices with different coupling current have been measured. The smallest threshold current of 75mA has been achieved in the device with the coupler length of 300μm and coupling current of 30mA. Such variation has also been calculated assuming different fabrication loss. Their loss mechanisms have been investigated based on threshold analysis, which should benefit to further optimize loss-reduced semiconductor polygon ring lasers based on active vertical coupler structure.

In this contribution, a detailed analysis of optical gain in InGaN/GaN quantum structures with Indium content of 10% and 20% is presented. Experimental data are obtained from Hakki-Paoli characterization of edge-emitting Fabry-Perot lasers. A gain model that includes many-particle effects on a microscopic level, as well as combined quantum-well and quantum-dot density of states, is used to explain the experimental findings. Inhomogeneous broadening arising from local Indium clusters is included via a statistical fluctuation of the electronic density of states. Excellent agreement is obtained for the characteristic gain spectra from structures emitting at 405nm (10% In content) and 470nm (20% In content), and a systematic analysis of the microscopic physics shows signature of quantum-dot states.

We investigate the focal characteristics of several cylindrical microlenses made of anisotropically dielectric material illuminated by different wavelengths in close boundary based on rigorous boundary element methods. Several design schematics of the microlenses are performed and their focusing performances are analyzed. It is found that different focal lengths ans spot sizes are formed from different design schematics, and the separation of o-ray and e-ray can be well achieved by appropriately seleted geometrical parameters. It is believed that this analysis will provide useful information in the applications of micro-optis.

A swept light source is a type of light source whose wavelength can be quickly tuned . Typically, a swept light source is composed of a high optical gain-medium such as a semiconductor optical amplifier (SOA) or optical gain module and a wavelength selective component. Inphenix has developed swept light sources that cover from 800nm to 1550nm wavelengths and are cost effective, highly linear and ideal for medical imaging and industrial measurement applications. Applications of swept light sources have been widely investigated in real time optical imaging, optical fiber sensor interrogator and optical component testing.

Electroabsorption modulator has been widely used in modern optical fiber communication system and analog RF link system. In this paper, the design of a high-performance EAM with low coupling loss, high saturation power and high speed was demonstrated, which include the waveguide, active core and electrodes. A novel EAM with large optical cavity (LOC) waveguide structure, intrastep quantum well (IQW) active core and traveling wave electrodes was presented and fabricated successfully. Our results show that the LOC waveguide effectively improved the optical profile of EAM and reduced the coupling loss. The obtained traveling wave EAM achieved 21dBm saturation power and 23GHz 3-dB bandwidth.

A five-layer asymmetric coupled quantum well (FACQW) is a novel potential-tailored quantum well that is expected to exhibit giant electrorefractive (ER) index change in a transparency wavelength region. We studied the GaAs/AlGaAs and InGaAs/InAlAs FACQW theoretically and experimentally. A GaAs/AlGaAs FACQW was grown by molecular beam epitaxy (MBE). Giant ER sensitivity |dn/dF| as large as 1.7 × 10^-4 cm/kV was observed in the FACQW phase modulator. A Mach-Zehnder (MZ) FACQW modulator was fabricated and the operation voltage was successfully decreased. For 1.55 μm wavelength region, an InGaAs/InAlAs FACQWs was also proposed and studied. We found that the InGaAs/InAlAs FACQW is also expected to produce a giant ER sensitivity |dn/dF|. The InGaAs/InAlAs multiple FACQW was successfully grown by MBE and the results of its photoabsorption current measurements are consistent with the theory. We proposed an MZ modulator and a 2 × 2 switch with the multi-FACQWs in the core. Driving voltages of the FACQW modulator and the switch with 1mm-long phase shifters can be decreased as low as 0.1~0.2 V.

High performance InGaAsP/InGaAsP strained compensated multiple-quantum-well (MQW) electroabsorption modulators (EAM) monolithically integrated with a DFB laser diode have been designed and realized by ultra low metal-organic vapor phase epitaxy (MOVPE) based on a novel butt-joint scheme. The optimization thickness of upper SCH layer for DFB and EAM was obtained of the proposed MQW structure of the EAM through numerical simulation and experiment. The device containing 250^μm DFB and 170^μm EAM shows good material quality and exhibits a threshold current of 17mA, an extinction ratio of higher than 30 dB and a very high modulation efficiency (12dB/V) from 0V to 1V. By adopting a high-mesa ridge waveguide and buried polyimide, the capacitance of the modulator is reduced to about 0.30 pF corresponding to a 3dB bandwidth more than 20GHz.

We describe two Si based optical-electric modulators based on photonic crystals (PC), which are capable of monolithically integrated with Si photonic integrated circuits. One is a modulator based on Mach-Zehnder interferometer (MZI), the other is a modulator based on photonic band gap. These devices may enable the deployment of ultra-compact (-200 μm) devices with high extinction ration and low insertion loss.

A new kind of birefringence is found in a two-dimensional (2D) flat perfect photonic crystal (PhC). It is different from the one in the normal biaxial crystal, but qualitative, and comes from the positive and negative refraction in the 2D flat perfect PhC. The quantitative relationship between the refractive index and the incident angle are plotted, by the analysis of the equal-frequent surface (EFS) of the perfect PhC. The plot is consisted of three branches---the main across 0° to 45.53° of the incident angle, the upper across 33.3° to 38.53° and the lower across 38.53° to 45.53°. The upper reveals the positive refraction; the lower and the main reveal the negative ones. The finite-difference time-domain (FDTD) simulations are performed, and the relevantly quantitative measurement validates the quantitative relationship by the analysis of the EFS, but a 2.67° shift to the bigger incident angle. A novel beam guiding is observed, which is resulted not from the guiding in a defect photonic crystal (PhC) but from the negative refraction in a two-dimensional (2D) flat perfect PhC slab.

We investigate the local field enhancement properties of a 2D nano-cavity by using finite-difference time-domain (FDTD) method. The 2D nano-cavity is built by two pieces of silver slabs with periodical subwavelength corrugations on their inner boundaries. Numerical simulation result shows that the 2D nano-cavity can confine and enhance light field within the region smaller than the diffraction limit in a relatively broad range of wavelength. The result is useful for the research and development of new kinds of broadband optical sources used in WDM optical communication networks or micro-optical sensor system.

Transmission properties of slow light in one dimensional photonic crystal coupled resonator waveguide have been investigated. By inserting multiple half-wavelength cavities discretely to forming CROW, the slow light band in photonic crystal bandgap can be broadened effectively. Otherwise cavities distance and refractive difference can flatten the rough pass band. By modulating these parameters, an optimized broadened flat pass band with 20.99nm has been obtained. Within the band, the group velocity is in the range of 0.0142c to 0.02148c. The propagation mechanism has been investigated via analysis of the field distribution in CROW.

We have demonstrated a continuum-wave (CW) supercontinuum (SC) fiber light source with over 1000 nm bandwidth based on a low-cost erbium/ytterbium co-doped double-cladding fiber ring cavity laser. Based on the observation to the SC evolvement, we have experimentally analyzed the detailed contributions of several nonlinear effects within highly nonlinear dispersion-shifted fiber (HNLF). Our experimental results have clearly indicated that four-wave mixing (FWM) and stimulated Raman scattering (SRS) play key roles in CW-pumped SC generation. At the same time, self-phase modulation (SPM) mainly contributes to generate new frequency components near the peaks that appear in the form of the spectra broadening while cross-phase modulation (XPM) enhances the broadening of peaks.

A hybrid actively and passively mode-locking semiconductor optical amplifier fiber ring laser based on nonlinear polarization rotation was presented, where intensity modulator not only acted as modulator but also polarizer. Under the hybrid mode-locking mechanism, output pulse is with some new characters. So, a theoretical model that describes the SOA fiber ring laser was developed and system parameters effects on mode-locking pulse are discussed.

Liquid crystal photonic bandgap fibers represent a promising platform for the design of all-in-fiber optical devices, which show a high degree of tunability and exhibit novel optical properties for the manipulation of guided light. In this review paper we present tunable fiber devices for spectral filtering, such as Gaussian filters and notch filters, and devices for polarization control and analysis, such as birefringence control devices and switchable and rotatable polarizers.

The dynamic single-mode and modulation performance of λ/4 phase-shifted distributed feedback laser diode with chirped grating (QWS-CG-DFB) are analyzed theoretically. The numerical simulation shows that, In contrast to purely QWS-DFB laser, the enhanced dynamic single-mode suppression ratio (SMSR) can be reached by QWS-CG-DFB laser; Under the smaller biasing current, the modulation band-width in presence of chirped grating is narrower, this difference shrinks for larger biasing current; For large signal modulation, the chirped grating is helpful to increase the output extinction ratio, but worsens the frequency chirping.

We present the results of a transmission experiment, over 110 km of field installed fiber, for an all-optical 160 Gb/s packet switching system. The system uses in-band optical labels which are processed entirely in the optical domain using a narrow-band all-optical filter. The label decision information is stored by an optical flip-flop, which output controls a high-speed wavelength converter based on ultra-fast cross-phase modulation in a single semiconductor optical amplifier. The packet switched node is located in between two different fiber sections, each having a length of 54.3-km. The field installed fibers are located around the city of Eindhoven in the Netherlands. The results show how the all-optical switch can effectively route the packets based on the optical information and that such packets may be transmitted across the fiber with an acceptable penalty level.

Based on a three-band model, a polarization-dependent pulsed four-wave mixing (FWM) model which can be used to analyze the optical sampling process in semiconductor optical amplifier (SOA) is presented. The polarization-dependent characteristics and cross-polarization modulation (XPolM) of pump, probe and conjugate pulses are investigated in detail. The maximum sampling efficiency occurs, when pump and probe pulses are linearly co-polarized and parallel to the TE axis. When the pump, probe pulses with initial linear polarization states interacting in an SOA, their polarization states do not just rotate and the conjugate pulse is not just linearly polarized but with complicated polarization states during the propagation along the length of SOA.

We propose and demonstrate a novel approach generating ultrawideband (UWB) monocycle pulses using cross phase modulation (XPM) of the semiconductor optical amplifier (SOA). A pair of polarity-reversed UWB monocycle pulses is achieved by locating the probe carrier at the positive and negative linear slopes of the filter. We achieve different UWB spectrum width by 25ps- and 50ps-width Gauss pulse injection. The generated monocycle pulses can be controlled by either optical Gauss pulse or both injections, which function as logic OR-monocycle. It has potential applications in UWB-over-fiber communications.

A simple method of optical generation millimeter-wave signal employing optical phase modulator and band-elimination filter is proposed. This simple approach is capable generate millimeter wave signal of quadrupled or sextuple microwave source frequency with extremely high spectral purity. Millimeter-wave signal 26GHz (quadrupled fundamental frequency) or 39 GHz (sextuple fundamental frequency) is obtained respectively using different system chromatic dispersion when the microwave driver signal is at 6.5GHz.

Integration of mode-hop-free tunable laser array and a semiconductor optical amplifier is most reliable approach to realize widely tunable lasers. We have developed two types of tunable lasers, one is a thermally tunable DFB laser array for DWDM tunable transponders, which has shown high power and wide tunability covering Cband or L-band, housing in compact butterfly packages with robust wavelength locker. Another is a short-cavity DBR laser array for optical burst switching, whose lasing frequency can be monotonously tuned and locked on the ITU grid within 5 microseconds. Both lasers have demonstrated superior performances in system experiments.

Chip-on-carrier (CoC) sub-assemblies of Digital Ssupermode DBR (DSDBR) lasers are produced in high volume within Bookham manufacturing plants. These lasers can cover more than 100 ITU channels with a 50GHz channel range across the C or L band with a minimum 13dBm output power and 40dB side mode suppression ratio (SMSR). To guarantee a high quality and ensure a good yield, an automated screening process has been put in place at the CoC level to eliminate poor devices. Typical tuning maps and key performance features of the device are shown in this paper. We describe the general features of the tuning map, and indicate how a suitable operating point can be determined. The use of automated test kit is also described in this article. Finally, the performance of our device is presented in detail.

The Tuneable Transmitter Assembly is a high performance tuneable transmitter for use in the C or L wavelength bands of and suited to regional metro and long-haul applications. The key performance attributes of the TTA module is successfully demonstrated in this paper.

Widely tunable external cavity semiconductor lasers with sampled fiber gratings are investigated. Their static properties, such as threshold gain, tuning characteristic, emitting light spectrum, and side mode suppression ratio, have been simulated and discussed by a combined theoretical model, which is developed to match experimental results. Up to thirteen tunable channels can be obtained in this laser with high side-mode suppression ratios by tuning the injection current of the passive phase control section. With the decrease of AR-coating reflectivity, the narrow line-widths, wide tuning range, and high threshold gain in those tunable external cavity semiconductor lasers have been presented.

Based on the ray tracing method, the implicit expression of the output spectrum of the extremely short external cavity semiconductor Laser (ESECSL) is derived, and the output spectrum and P-I characteristic of the ESECSL are investigated. The results show that: when the length of external cavity is changed at the order of wavelength, the P-I characteristics of the ESECSL will undergo significant changes; with the variation of the external cavity length, the lasing wavelength of ESECSL will behave cyclical jump in the range of 10nm. Especially, for the external cavity length changed within the range of 40μm-70μm, the jump range of the lasing wavelength will reach the maximum. The simulations well agree with the experimental results reported.

Tunable semiconductor lasers have been under intense research interests for the past decades due to their vast applications in optical networks, optical characterization, and optical sensing. The required device characteristics can be very different for applying the tunable lasers to various areas. We classify the tunable lasers in terms of their tuning characteristics and switching speed. Four kinds of tunable lasers are described in this paper to manifest the application-dependent device structures and performance. The applications include the use of sampled grating based lasers to form multi-wavelength laser arrays, cascaded distributed-feedback lasers for multi-gas sensors, wavelength-selectable laser arrays for fast wavelength switching sources, and short cavity lasers for fault monitoring in passive optical networks.

In this paper we review our recent works on low cost lasers and remote modulators for Optical Network Unit in access network. Our work is carried out in the context of an FTTH PON migration scenario towards 10Gbit/s base rate as well as towards more capacity and flexibility using WDM technology. All components are based on the attractive thermal, gain and absorption properties of AlGaInAs/InP material system. As a first step to the speed increase we propose a new uncooled 10Gb/s laser based on a self thermal compensation principle. As a next step, new WDM PON architectures will require wavelength agnostic component for ONU. For this purpose, we demonstrate new colorless concepts on 10Gb/s remote amplified electro-absorption modulators. We show low-cost FTTH components may also be attractive for emerging access-metro WDM technology introducing colorless principle in reconfigurable add and drop multiplexers

The recent development of semiconductor laser technologies for cost-effective telecom/datacom applications is reviewed in details in this paper. This includes the laser design, laser chip technology, laser packaging technology and other low cost lasers (chip + packaging). Some design and simulation examples in Archcom laser production are described first. A latest trend in the wafer scale testing/characterization/screening technology for low cost semiconductor laser mass production is discussed then. An advanced long wavelength high power single mode surface emitting laser with wafer scale characterization using our unique mask free focused ion beam (FIB) etching technology is also demonstrated. Detailed descriptions on our wide temperature range (-50 °C to +105 °C) G-PON distributed feedback (DFB) semiconductor lasers with high performance and low cost wafer design are included. Cost reduction innovations in laser package with our beam profile improved laser and optical feedback insensitive (OFBI) laser are also addressed.

A new generation 980 nm pump laser module with a fiber output power more than 750 mW is presented. The module uses our generation-08 (G08) pump laser chip, which is designed for high output power and high reliability. The pump laser is stabilized by a fiber Bragg grating (FBG). A special thermo-electric cooler (TEC) is built into the package in order to enable operation of the device at high laser output powers.

A new method of coupling the light from a laser diode to a Single Mode Fiber (SMF) with large alignment tolerances and without using coupling lenses is presented. A pseudo vertical tapered coupler is designed for light coupling between laser diode and single mode fiber. It has a large input aperture which is about 100 times the size of the laser waveguide cross-section. The tapered coupler provides single mode output and matches the mode size with the single mode fiber. The tapered coupler is fabricated on a silicon optical bench and is located between the laser and the fiber through the silicon micrfabrication process. The misalignment between the fiber and taper coupler can be very small since this is controlled by high precision silicon optical bench patterning processes. The coupler relaxes the laser diode placement accuracies and eliminates the need for a coupling lens. Design Studies showed that the tolerance between the laser diode and taper coupler can be more than +/-5μm misalignment at x-y, and +/-0.5degree tilting angle tolerance and the fabricated assembly results are encouraging with good placement tolerances and coupling efficiency. The laser to single mode fiber coupling tolerances is greatly improved and passive alignment for laser and single mode fiber is realized. The technology can be useful for multi channel optical assembly where significant device and process cost saving can be achieved and is suitable for functional integration for silicon photonics.

By adding a little nitrogen in InGaAs / GaAs quantum well (QW), a strong bandgap-bowing in the QW is caused. However, the incorporation of nitrogen results in lower photonuminescence intensity and wider line width as a result of increased non-radiative centers. In order to increase the efficiency of radiative recombination and hence reduce the laser threshold, a post-growth heat treatment has to be applied. Such kind of heat treatment results in a big blue shift due to interdiffusion and other effects. During growth, in order to incorporate nitrogen into InGaAs, the growth temperature is much lower than normal InGaAs growth. Large number of point defects is induced under such low temperature. This is the main cause of the interdiffusion at the interfaces of InGaAsN / GaAs QW. There are some other facts to cause the blue shift during heat treatment, such as local neighbourhood redistribution called short range ordered. In our study, different blue shift behaviors were clearly observed due to different blue shift mechanism. Post-growth heat treatment also affects the laser performance dramatically. Lower temperature treatment mainly decreases the absorption loss and higher temperature treatment improves the conductivity of the cladding layers. Different heat treatment also results in very different burn-in behavior. An optimized heat treatment will be concluded after the annealing discussion on laser devices. In order to assure longer emission wavelength well as higher emission efficiency, many efforts have been tried and will be discussed in this paper.

The characteristics of intersubband transitions in III-nitride quantum wells are promising for detectors and all-optical switches through a high intrinsic speed (~1 THz), and can also provide a high optical saturation power and a desired small negative chirp parameter in electroabsorption modulators. The high LO-phonon energy allows to improve the operating temperature of THz emitters. Recent achievements and prospects for intersubband III-nitride photonic devices, mainly for λ=1.55 μm, are briefly reviewed. Further, means to enhance material quality by achieving crack-free growth of GaN/AlN multiple-quantum-well (MQW) structures, and by employing intersubband transitions in multiple-quantum-disk (MQD) structures incorporated into dislocation free GaN nanocolumns are discussed. We investigate the occurrence of cracks in MBE-grown GaN/AlN MQWs on GaN MOVPE templates with respect to the buffer layer, the number of QWs and the temperature reduction rate after growth. Intersubband absorption in GaN/AlN MQDs in the wavelength range 1.38-1.72 μm is demonstrated in three samples grown on Si(111).

A novel single-sideband (SSB) modulator using period phase reversal electrode and the theoretical analysis on the principle is proposed and presented. The results show that the SSB modulator can reach an optical sideband suppression ratio over 40dB in 60GHz radio-over-fiber (ROF) system, which successfully reduce the power penalty due to the chromatic dispersion.

A novel high-speed magneto-optic (MO) modulator which consists of an integrated wire grid polarizer (WGP), Bi-YIG waveguide with cladding layer and conducting micro-strip line is proposed. With the integrated WGP, this MO modulator is faster, more accurate and more stable because it is not only completely driven by electric signals but also has no mechanically moving parts. Moreover, it is compact-structured and low-cost. Large Faraday rotation is obtained with specific arrangement of the directions of the bias magnetic field and the modulation RF magnetic field. Optical route and optic-electrical detect circuit are also designed and analyzed.

Hollow core photonic crystal fiber (HC-PCF), fabricated according to a nominally non-birefringent design, shows a degree of un-controlled birefringence or polarization mode dispersion far in excess of conventional non polarization maintaining fibers. This can degrade the output pulse in many applications, and places emphasis on the development of polarization maintaining (PM) HC-PCF. The polarization cross-coupling characteristics of PM HC-PCF are very different from those of conventional PM fibers. The former fibers have the advantage of suffering far less from stressfield fluctuations, but the disadvantage of a higher loss figure and the presence of interface roughness induced modecoupling which increases in strength as birefringence reduces. Close to mode anti-crossing events of one polarization mode, the PM HC-PCF is characterized by high birefringence, a high polarization dependent loss and an increased overlap between the polarization modes at the glass interfaces. The interplay between these effects leads to a wavelength for optimum polarization maintenance, λ_PM, which is detuned from the wavelength of highest birefringence. By a suitable fiber design involving antiresonance of the core-surround geometry, λ_PM may coincide with a low-loss wavelength for the signal carrying polarization mode.

We report a novel double ring fiber laser incorporating a periodically poled LiNbO₃ (PPLN) waveguide. The double ring fiber laser is formed by placing two parallel-arranged tunable filters (TFs) followed by two variable optical attenuators (VOAs) inside the PPLN-based fiber ring cavity. Two continuous-wave (CW) lasing lights can be obtained from the double ring fiber laser with their wavelengths determined by two tunable filters. Using cascaded second-harmonic generation and difference-frequency generation (SHG+DFG), as one of the lasing waves is tuned at the quasi-phase matching (QPM) wavelength of PPLN, the third idler wave is generated. It is easy to perform tunable operation simply by changing the other lasing wavelength. Based on cascaded sum- and difference-frequency generation (SFG+DFG), it is also possible to realize tunable wavelength conversion. Both input signal and converted idler can be tuned by appropriately adjusting two lasing waves. With PPLN-based double ring fiber laser, we first demonstrate stable dual-wavelength generation with minimum wavelength spacing of 0.32 nm. Then we observe SHG+DFG-based tunable triple-wavelength generation. Finally, tunable wavelength conversion at 40 Gb/s based on SFG+DFG is successfully demonstrated in the experiment. No external CW optical waves are needed, which effectively reduces the complexity and cost of the wavelength converter.

The configuration of polymer light waveguide electro-optical printed circuit board(EOPCB) is proposed in this paper. An additional optical layer with light waveguide structure is used in conventional PCB to construct EOPCB. Light waveguide core layer mould is made with SU-8 photolithograph. Polymer light waveguide layer which is embedded between multiplayer PCB is made in experiment by Doctor-blading technology for large size application. Vertical cavity surface emitting laser (VCSEL) array is used as optical transmitter array. PIN photodiode array is used as optical receiver array. A MT-compatible direct coupling method is presented to couple light beam between optical transmitter/receiver with light waveguide layer. The optical signals from a processor element chip on the PCB can transmit to another processor element chip on the same PCB board through light waveguide interconnection in EOPCB. So optical interconnection between chip to chip for parallel multiprocessor system can be reailzed by EOPCB.

We recently proposed and demonstrated a saturable absorber (SA) incorporating carbon nanotube (CNT). CNT-based SA offers several key advantages such as: ultra-fast recovery time, polarization insensitivity, high optical damage threshold, mechanical and environmental robustness, chemical stability, and the ability to operate at wide range of wavelength bands. Using the CNT-based SA, we have realized femtosecond fiber pulsed lasers at various wavelengths, as well as the very short-cavity fiber laser having high repetition rate. Besides the saturable absorption, CNT has been shown to have high third-order nonlinearity, which is also attractive for realization of compact and integrated functional photonic devices, such as all-optical switches and wavelength converters. In this paper, we first present photonic properties of CNTs, and review our studies on CNT-based mode-locked fiber lasers. We also refer to fabrication methods of CNT-based photonic devices. We show our recent research progresses on novel photonic devices using evanescent coupling between optical field and CNT.

This paper summarizes recent advances on InAs/InP mode-locked quantum dashes (QD) lasers, and their applications for all-optical clock recovery, short pulse generation and millimeter wave generation. We demonstrate that QD FP lasers, owing to the small confinement factor and the 3D quantification of electronic energy levels, exhibit a beating linewidth as narrow as 15 kHz. Such an extremely narrow linewidth, compared to their QW or bulk counterparts, leads to the excellent phase noise and time jitter characteristics when QD lasers are actively mode-locked. We report also on an actively mode-locking tunnel injection quantum dash Fabry-Perot laser diode at 42.7GHz, generating nearly Fourier transform limited pulses with a pulse width of 2ps over 16nm.

Kinetic Monte Carlo simulations are applied on the investigation of the epitaxial growth of self-assembled InAs quantum dots on GaAs substrate with periodic strain-relief patterns. The study is focused on the initial stage when the first sub-monolayer is forming on top of the wetting layer. The flux is one of the most important growth parameters, which are studied in detail. It is demonstrated that uniformly sized and regularly ordered island arrays can be obtained by controlling flux, by means of analyzing the surface morphology, average island size, island size distribution and the standard deviation of island size distribution. If interruption is introduced, the influence of flux will significant different. The uniformity and order of islands will greatly affect the locating of quantum dots in sequent 3-D growth.

In this paper, we calculated the strain distribution of low dimension structure using the elastic continuum model. The strain energy density distribution on the different thickness of capping layer surface for the self-organized InAs/GaAs quantum dots system is investigated by the numerical finite element method. The influence of the different growth directions on the strain energy density distributions can be found from the calculated results. The results can explain some experiment results, such as the ordering array of the quantum dots supper-lattices. So the growth direction and spacing thickness can be regarded as another control parameters for strain engineering self-organized semiconductor quantum materials. As a comparison, the strain distributions of other low-dimension self-organized materials are also calculated.

We have designed, fabricated and characterized self-assembled InAs/InGaAsP QD-waveguide devices around 1.55 μm. In order to obtain optimal performance, we have investigated several QD-based semiconductor optical amplifiers (SOAs) / lasers with different core geometry and doped profiles. To make the fair comparison between QD-SOA and QW-SOA, InAs/InGaAsP QW-SOAs with the same structure and the doped profiles have been designed and characterized. The experimental results indicate the QD-SOA is much better than QW-SOA in term of optical spectral bandwidth, temperature sensitivity and output power stability. The 3-dB and 10-dB bandwidths of the amplified spontaneous emission (ASE) spectra of the QD-SOA are 150 nm and 300 nm around 1520 nm. By using CW pump and probe signals we have demonstrated a non-degenerated four-wave mixing (ND-FWM) process and the experimental results indicate that the asymmetry of the FWM conversion efficiencies is eliminated by using the QD-SOA. To make use of the inhomogeneous broadening which is one of the specific properties of QD waveguide devices, we have designed and investigated the QD-based multi-wavelength semiconductor laser. A stable multi-wavelength laser output with a 93-channel multi-wavelength laser with maximum channel intensity non-uniformity of 3-dB were demonstrated on the basis of a single InAs/InGaAsP QD F-P cavity chip. All channels were ultra-stable because of the inhomogeneous gain broadening due to statistically distributed sizes and geometries of self-assembled QDs.

Semiconductor quantum dots have been of major interest in recent years. This has largely been simulated by progress in quantum dot growth technology, whereby self-organized quantum dots array can be fabricated by MBE and MOCVD facilities using Stranski Krastanow growth mode. Quantum does material has achieved broad applications in optoelectronic devices and quantum information fields because of the unique 3D electron confinement. However, a good understanding about the electronic, excitonic and optoelectronics properties of the quantum materials are very important in fabrication nanostructure devices based on quantum dots. Based on the 1-band effective-mass theory, a finite element numerical technique is developed to calculate the electronic structure of truncated conical shaped InAs GaAs vertical aligned quantum dot molecular, including the wetting layer. Using the axis-symmetry model, the 3D effective-mass Schrödinger equation with step potential barrier can be reduced to a 2D problem by separating variable technique, which greatly reduced the calculation cost. Form the calculated results, we found that the coupling effects is obviously when the separation distance is in the range of the less than 10nm. The wave functions will exhibits large probability in the region between the quantum dots. In order to consider the effect of the distance between the two layers of quantum dots on the electronic state coupling, we calculated the results when the distance is 6nm, 11nm, 14nm and 17nm. The ground state, the second excited and the highest excited state will lower its energy with decreasing the distance between the quantum dots, but the second excited state will increase its energy. With increasing the distance between the two quantum dots, the coupling effect will become weaker, and for the ground state, the wave function distribution will tend to localized only in one of the quantum dot, the energy become something degenerate. The calculated results show that the ground state and the first excited state are degenerate. With decreasing of the distance, the degenerate states are broken, and the energy levels are separated. In our simulations, the strain effects are ignored. In the future woks, strain should be taken in to account as an easy way. The calculated results can help us to examine optoelectronic properties of the semiconductor nanostructure based on multi sheet of quantum dots with wetting layers.

With the interest in quantum structures, there is a need to have a flexible method that can help us to determine eigenfunctions for these structures. In this article, we present a method that accomplishes this by using the simulation of the Schrödinger equation based on finite-difference time-domain (FDTD). We choose one-and two-dimensional finite square well potential, and one-and two-dimensional harmonic oscillator potential as examples. Giving the initial condition, we determine the eigenfrequencies through a Fourier transform of the time domain data collected at the center point in the problem space. Another simulation implements a discrete Fourier transform at the eigenfrequencies at every point in the problem space, hence, the eigenfunctions can be constructed.

We propose a new scheme of resonant-cavity (RC) based monolithic white LED, it relaxes the hard requirement of high internal quantum efficiency of yellow multi-quantum (MQW) and offers an easy way to obtain high luminous efficacy white light emission. In the proposed white LED, the blue MQW and yellow MQW active layer are embedded in a resonant-cavity defined by the bottom distributed Bragg reflector(DBR) and top DBR. For a optimal design of RC-based white LED, the extraction efficiency for yellow light is enhanced, while that for blue light is suppressed, thus intensity ratio of yellow light in the emitting light is increased, which not only helps to obtain white emission in spite of the low internal quantum efficiency of yellow light, but also doubles luminous efficacy. The color coordinates and luminous flux of the emitting light from RC-based white LED are calculated and the performance dependence on directionality is investigated.

In this work, we analyzed and demonstrated a multi-quantum-well InGaNAs/GaAs RCE photodetector with a vertical taper absorption cavity operating at 1550nm. The GaAs/AlAs distributed Bragg reflectors and InGaNAs/GaAs quantum wells were grown on GaAs substrate by molecular beam epitaxy. The growth of InGaNAs/GaAs quantum wells maybe solves the problem that the GaAs-based materials can only response to short wavelength. The peak wavelength of the spectral response of our photodetector is at 1558.7nm, and the spectral linewidth is 3 nm.

High-power vertical-cavity surface-emitting lasers with InGaAs/GaAs quantum well active gain region are investigated. By using AlAs oxidation technology, the devices have been fabricated in experiment, and the characteristics of the device are carried out at room temperature. The 300μm-diameter VCSELs have the maximum room temperature continuous wave (CW) optical output power of about 1.1W, and the threshold current of the device is about 0.46A. The life test of the device is carried out in constant current mode. The life test of 300-μm diameter lasers shows that the average lifetime is about 1800h at 80°C. The device degradation mechanism is also discussed in detail.

White OLEDs (WOLEDs) have attracted much attention for several applications, such as backlights in LCD, full-color OLED display using on-chip color filters, and low cost illumination sources. OLEDs have typically very broad emissions, which makes them uniquely suitable for light source applications. In this paper, some fundamentals of the CIE colorimetry system including the color-rendering index are described. Given the spectral power distribution of WOLED, the parameters of a light source (chromaticity coordinate, CCT, CRI, and the luminous efficacy) can be calculated. A MATLAB program for this purpose is developed in this paper. WOLEDs utilizing two primary-color emitters are fabricated. NPB doped with 2% Rubrene is used as the red-emitting layer and anthracene derivative as the blue-emitting layer. With a structure of ITO/2TNATA(20nm)/NPB(20 nm)/ NPB:rubrene(2%)(10nm)/ anthracene (30 nm)/Alq₃(20nm)LiF(1nm)/Al(100nm), a white light with CIE coordinates of (0.34, 0.37) is generated. The color properties are presented in this paper. The results show that the white color can be created from numerous combinations of different spectra. Based upon this, the characterization of the WOLED is simulated and the design of WOLED for illumination is discussed. In contrary to light source applications where illumination quality white is the most important, all colors are equally important for display applications. The white spectrum of the two-emitter WOLED is transmitted through the typical red, blue and green color filters. The performance of this color display is simulated and the WOLED design for display application is discussed.

Introduced in this paper is an experimental system of packet-level self-synchronization using semiconductor optical amplifier based Mach-Zehnder interferometer (SOA-MZI). The function of SOA-MZI in the system is analyzed, as well as the relationship among key parameters of various components. A tunable optical delay line is used for the adjustment of phase difference. It reveals that SOA-MZI is an effective way to extract self-synchronization clock in both experiments and practical applications.

Transmission studies for quantum well (QW) structures composed by two-dimension photonic crystals are presented. The results indicate that, with the multiple QW structure, each resonant peak which appears in the single QW structure would split, and the splitting times increases in step with the increasing number of the wells. It is also found that the frequency spacing between adjacent splitting spectra can be finely tuned by adjusting the barrier widths properly, leading to great improvement of spectral efficiency. The physical explanation of the origin of the spectral splitting are provided. These results provide the prospects in super dense wavelength division multiplexing for optical communication and precise optical measurement for the purpose of maximization of channel density.

Pulsed anodic oxidation technique, a new way of forming current blocking layers, was successfully used in ridge-waveguide QW laser fabrication. We apply this method in 980nm VCSELs fabrication to form a high-quality native oxide current blocking layer, which simplify the device process. A significant reduction of threshold current and a distinguished device performance are achieved. The 500μm-diameter device has a current threshold as low as 0.48W. The maximum CW operation output power at room temperature is 1.48W. The lateral divergence angle θ_paralleland vertical divergence angle θ_{perpendicular} are as low as 15.3° and 13.8° without side-lobes at a current of 6A.

In this work, we reported the buffers optimum of high-quality InP epilayer grown on GaAs substrate for fabrication of InP-related devices. First, LT-GaAs (450°C, 15nm)/LT-InP((450°C, 15nm) double LT buffers were deposited on the substrate as the initial layers. The effects of double LT buffers were studied compared with the results of single LT-InP buffer scheme. It was demonstrated that: (i) with a proper LT-GaAs buffer thickness, the double LT-buffer became more "compliant" for strain accommodation than single LT-InP buffer; (ii) there existed an optimal thickness of LT-GaAs buffer for a given thickness of LT-InP layer at which the crystal quality reached the best, just like the conventional buffer optimum process. Second, in order to block the "escaped" dislocations from the buffer/substrate interface, In_xGa_1-xP/InP (x≈0.2) strained superlattices (SLS) were introduced as defect filtering layers before the growth of the final InP layer. We investigated the effects of the periods and inserting position of the SLS on the stress relaxation and the crystal quality of InP top layer. It was suggested that when the total thickness of the epilayer was fixed, both the thickness and the periods and the distance from the interface should be carefully designed to reduce the stress and improve the crystal quality of the epilayer simultaneously. Finally, a 2-μm-thick InP epilayer was grown on GaAs substrate using (450°C, 15nm)/LT-InP(450°C, 15nm) double LT buffers combined with inserting 15-period (4nm/6nm) In_0.8Ga_0.2P/InP SLS into epilayer. Then X-ray diffraction measurements showed the best result of the full width at half maximum (FWHM) was 203 arcsec with estimated dislocation density of 2.8×10⁷ cm^-2.

All-optical wavelength converters (AOWCs) are considered to be important components in future wavelength-division-multiplexed (WDM) networks. Cross gain modulation schemes in semiconductor optical amplifiers (SOA) are promising candidates for an all-optical wavelength conversion application due to the simple implementation and effective conversion. However, the slow gain recovery time of SOA limits the maximum operation speed and causes unwanted pattern effects. This paper provides a novel scheme for wavelength conversion enables ultra-fast conversion speed. On the one hand, we utilize a three-wavelength-device (TWD) to reduce the recovery time of the SOA. On the other hand, we use an optical band pass filter (OBF) which central wavelength is blue shifted with respect to the central wavelength of the probe beam to increase the frequency response. The combination of a reduction of the SOA recovery time and an increase of the frequency response enables conversion speed potentially to achieve 160 Gb/s or even faster.

We proposed a novel compact and integrated optical modulator, which consists of p-i-n silicon photonic crystals with triangular lattice and a line defect waveguide. The device operation is based on a dynamic shift of the photonic band gap (PBG), which is induced change in the silicon refractive index by the free carrier injection. We have numerically analyzed and investigated its light modulation performance by using the finite-difference time-domain method. Being a small size and high performance, the designed optical modulator can be used in photonic integrated circuits.

InGaN/GaN multiple quantum wells (MQWs) structure for ultraviolet emission has been grown on sapphire by metalorganic chemical vapor deposition (MOCVD). The High resolution x-ray diffraction (HRXRD), atomic-force microscopy (AFM) and photoluminescence (PL) are used to characterize the structural and optical prosperities of MQWs, respectively. HRXRD shows multiple satellite peaks to 3rd order indicates the high quality of InGaN/GaN layer interface. AFM measurement shows that there are some spiral growth hillocks and 3D nanostructures on the MQWs surface. They are related with the surface kinetics or thermodynamics of InGaN growth. Temperature-dependent PL results show that there exists a clear excition-localization effect in the InGaN/GaN MQWs. The fitted σvalue of InGaN/GaN MQWs is around 8meV. The emission peak was almost unchanged with the increase PL excitation power. Those results indicate there is almost none piezoelectric field-induced quantum-confined stark effect in the InGaN/GaN MQWs due to the low In content and thin quantum well thickness.

We investigate the optical gain properties of InGaN quantum well with different symmetry barriers and asymmetry barriers based on a self-consistent calculation which solves the Schrodinger equations and Poisson equations simultaneously. It is found that the Al_xIn_yGa_1-x-yN barriers which can eliminate the internal polarized field by adjusting the component x and y can improve the emission intensity in a large extent compared with other barriers. The internal polarized field is an important but not the only one factor to affect the emission power, the barrier confinement, the energy band are all have to be taken into considered. Otherwise, a quantum well which has proper asymmetry barriers also can obtain better emission efficiency than the well with symmetry barriers.

The stimulated Raman amplification of picosecond Stokes pulse is numerically investigated in ultra-small silicon-on-insulator optical waveguide. Numerical results show that we obtain the gain of up to 30-dB for weak Stokes pulse in the co-propagation configuration for 10-mm-longth waveguide using high intensity pump optical pulse. The peak gain, pulse width, rise time, and fall time of Stokes pulse will experience the variation course of decaying then increasing with increasing waveguide length. The time delay of output Stokes pulse is controlled by adjusting the initial time delay of both pump and Stokes pulses.

Silicon photodetector is easy to be integrated with all kinds of Silicon IC to get monolithically OEIC. And the photodetector array is also widely applied. A kind of CMOS-process-compatible N+/N-Well/P-Sub photodetector and its array are analyzed in this paper. Depended on the basic time-dependent equations of photodetctor and analyzed by Laplace transform method, the intrinsic frequency response characteristic is numerically calculated. The effect of reverse bias voltage on spectral responsivity is also discussed. The photodetector is fabricated in 0.5μm CMOS process. At 780nm wavelength incident light, the measured and calculated responsivity are 0.253A/W and 0.251A/W, respectively. The variety of measured responsivity with bias voltage is about 1.8mA/(W•V). At a reverse voltage of 5V, the maximum dark current is 0.148nA. And the junction capacitance and -3dB frequency are also measured. The crosstalk factor of photodetector with PN junction isolation and 5μm isolated space in CMOS technology is less than 5%.

Semiconductor optical amplifier (SOA)-based all-optical XNOR and AND gates using four-wave mixing (FWM) and cross-gain modulation (XGM) with improved dynamics are simultaneously realized. The effects of the input optical wave powers and injection current on the critical factors of the logic gate performances, such as the ON-OFF contrast ratio, the output power level of the logic "1", and the output power difference of the logic "1", are theoretically investigated in detail. In addition, the effect of the counter-propagating CW pump on the gain recovery is analyzed.

Free-standing ZnS:Cu and ZnS:Cu,Al quantum dots were prepared in the aqueous medium from readily available precursors. The construction, morphology and luminescence properties of the ZnS:Cu and ZnS:Cu,Al quantum dots were evaluated by XRD,TEM and photoluminescence spectra. The average particle size was calculated using the Scherrer formula to be 5nm, which is also observed from HRTEM image. In PL spectra, 0.7% Cu and 0.4% Cu -doped ZnS:Cu quantum dots have emission peak around 470 nm and 500 nm, which attributed to the transition from the shallow donor level (sulfur vacancy) to the e level and t² level of Cu²⁺ respectively. And PL spectra of ZnS:Cu,Al quantum dots is similar with the PL of ZnS:Cu quantum dots, but the luminescence intensity of quantum dots ZnS:Cu,Al increase, which arise from that Al ³⁺ ions as co-activated enhance donor level.

Free-standing ZnS:Er and ZnS:Er,Yb quantum dots were prepared in the aqueous medium from readily available precursors. The construction, morphology and luminescence properties of the ZnS:Er and ZnS:Er,Yb quantum dots were evaluated by XRD,TEM and photoluminescence spectra. The average particle size was calculated using the Scherrer formula to be 4nm, which is also observed from HRTEM image. The spectra of ZnS:Er and ZnS:Er,Yb quantum dots have broad emission between 1450 nm and 1650 nm centered at 1575 nm with the excitation wavelength 980 nm, which can be attributed to the ⁴I_13/2→⁴I_15/2 transition. But the intensity of ZnS:Er,Yb quantum dots significantly increases with the addition of Yb as a sensitizing ion into ZnS:Er quantum dots. Because that Yb³⁺ absorbed the energy and transfer energy from the ⁵F_5/2 level of the ⁴I_11/2 level (Er³⁺ ) and improve population accumulation on the ⁴I_11/2 level.

Photonic crystals(PCs)have many potential applications because of their ability to control light-wave propagation. In this paper, we theoretically investigate the tunability of light propagation in photonic crystal waveguides in two-dimensional photonic crystals with square lattices composed of heat-resistant silicon resin. Waveguides can be obtained by the infiltration of silicon resin into air regions in two-dimensional photonic crystals composed of air holes with square lattices of dielectric cylinders. The refractive index of silicon resin can be changed by manipulating the temperature of the sample. Numerical simulation by solving Maxwell's equations using the plane wave expansion(PWE) method shows that the band gaps can be continuously tuned by silicon resin, accordingly the light propagation in photonic crystal waveguides can be controlled. The band gap is analyzed in the temperature range of 20°C-120°C. In our work, the gap map for a square lattice of dielectric cylinders is also simulated. The method can separate TM- and TE-polarized modes in the waveguide. Such a mechanism of band gap adjustment should open up a new application for designing field-sensitive polarizer in photonic integrated circuits.

Gain-transparent semiconductor optical amplifier (GT-SOA) can be used as optical phase-modulators in Mach-zehnder interferometer configuration for 2R or 3R regeneration, wavelength conversion of differential phase modulated signals and all-optical format conversion from on-off keying (OOK) to binary phase shift keying (BPSK). Numerical simulation of the phase modulation effect of GT-SOA is performed using a wideband dynamic model and the performance is evaluated using the differential-phase-Q factor. Performance improvement by holding light injection is analyzed and non-return-to-zero (NRZ) and return-to-zero (RZ) modulation formats of the OOK signal are considered.

A novel fiber type traveling-wave modulator is developed. The finite element method (FEM) has been used to analyze the performance of the proposed modulator with coplanar waveguide (CPW) electrode structures. The optimized structures of the traveling-wave modulator are obtained. The results show that the novel modulator has a 3dB optical bandwidth of 112GHz, a half driving voltage of 2.7V, and characteristic impedance of 51.2Ω at 1.55μm wavelength.

A time series of experimental optical chaos signal with dynamic equation unknown and low SNR is obtained. The wavelet multi-resolution decomposition algorithm is applied here to reduce the noise mixed in the experimental optical chaos signal. The performance of the algorithm is verified by Lorenz chaos signal mixed with noise, which shows that the SNR is increased by 10dB or so. Some parameters of the optical chaos attributes are calculated before and after noise-reduction. It shows that the noise-reduction algorithm can improve the precision of the Lyapunov exponent calculated with small data method, and a completely opposite wrong result can be avoided by the noise-reduction process when computing the minimum embedding dimension with Cao method. The small data sets method is improved by Cao method (minimum embedding dimension) and mutual information method (delay time). As the result is shown, the error of the largest Lyapunov exponent is reduced by nearly 30%, and the largest Lyapunov exponent of the optical chaos signal is 0.3896 obtained with this method.

The static Transfer Matrix Method (TMM) and Multi-mode Rate Equations (MMRE) are combined together to simulate dynamic characteristics of three section DBR lasers. Dynamic Side Mode Suppression (SMSR) is introduced to study wavelength switching transients for the first time, along with the lasing spectrum characteristics.

An improved method of distinguishing chaotic signal based on Wigner distribution (WD) is presented. Except for the width of spectrum, the peak values speciality of time-frequency in 3-dimensional figure is employed to distinguish chaotic signal. The method is verified by Lorenz system. The Wigner distribution of the chaotic signal is compared with that of the gauss noise, and the result dedicated that the spectrums of them can be distinguished. The Wigner distribution is used to dispose the chaotic signal of the fiber ring laser base on Semiconductor Optical Amplifier (SOA) too, and the result is consistent with that of Lorenz system. In the conclusion, 3-dimensional WD distribution could more accurately show chaos spectrum characteristic and could be an effective method for distinguishing the chaos and other signals.

Two gradient-doping GaAs photocathodes were designed and activated, the achieved highest integral sensitivity for the gradient-doping cathode is 2178μA/lm, which is much higher than that of uniform-doping cathode. The increase in the integral sensitivity is attributed to the electric field induced in the active layer of gradient-doping cathode. We analyze the transported mechanism of gradient-doping cathodes and solve the quantum efficiency equations of exponential-doping cathode, which is a special gradient-doping cathode with a constant induced electric field, from the one-dimensional continuity equations. According to these equations, we calculate the theoretical quantum yield of the exponential-doping cathodes, and compare the performance of exponential-doping cathodes with that of uniform-doping cathodes. The theoretical results show that the exponential-doping structure can increase the quantum yield of photocathodes evidently, for the transmission-mode cathodes the increase is even more pronounced.

Four-section sampled-grating DBR (SGDBR) lasers, including two sampled grating DBRs, gain and phase tuning sections, were designed by a InGaAlAs / InP strained multiple quantum well microstructure. The lengths of gain and phase sections are 280 μm and 150 μm, respectively. One SGDBR, accommodating 10 sample periods with a total length of 270 μm, is situated adjacent to the gain, and another SGDBR, containing 11 sample periods with a total length of 330 μm, is to the right of phase section. The estimated threshold current and characteristic temperature, T₀ , were 10 mA and 166 °K respectively. High T₀ reflects excellent electron confinement of the MQW structure. The optical coupling strength and the cavity-mode spacing of DBR lasers were carefully chosen. With appropriate tuning mechanisms, the tunable SGDBR lasers produced 172 transmission channels, each with side-mode suppression ratios higher than 40 dB. Under 100 mA current driving, the lasers could be operated with a bandwidth of 20 GHz and the channel switching could be completed within 1 nS. Moreover, from the observations on the relative intensity noises as well as on the eye diagram, the lasers do exhibit excellent dynamic characteristics.

For device analysis and optimization, a model based on Transfer Matrices Method was built to simulation the performance of a new integrated device. This novel DFB laser was composed of two serial sections to provide selectable wavelengths. The periods of two Bragg gratings were different and were chosen to achieve a spacing of 20nm between the two corresponding Bragg wavelengths. The model is more simple and convenient to simulate optical integrated device than other direct simulation methods used before. The simulation results prove that this novel waveguide structure of the serial DFB lasers is feasible. The integrated optical device was fabricated and two wavelengths of 1.51um and 1.53um were realized under different work conditions.

Optical receiver of high speed and high sensitivity is indispensable for long distance fiber-optic communication systems of transmitting rate up to Gb/s. In recent years, OEIC (Optoelectronic Integrated Circuits) optical receiver has attracted more and more attention for its advantages over conventional optical receiver such as minimized parasitics, lower cost, higher reliability and compact size. In this paper, design of preamplifier for InP-based PIN/HBT OEIC optical receiver with share layer structure was presented. As a basis for design of preamplifier, HBT large signal model (GP model) was investigated and model parameters were extracted and optimized. The extracted GP model shows a good agreement with measured characteristics of HBT fabricated. Based on this GP model, the preamplifier was designed and fabricated which exhibits good high-frequency characteristics of −3dB bandwidth is up to 2.0 GHz.

A tunable ultraviolet laser source in the spectrum range of 0.36-0.388 μm was obtained as second harmonics from a frequency doubled Alexandrite laser whose output covers the wave range over 0.72-0.78 μm. A LBO crystal was used as frequency doubling crystal. The phase mateching angle in the wide spectrum range of the crystal was calculated, and the crystal was cut in the way that the normal incidence at the center wavelength of the fundamental wave at the crystal. The output spectrum line was measured and the highest second harmonics conversion efficiency reached 1.2% from long pulse fundamental wave at the center wavelength.

Compared to the non-magnetic ordinary dielectrics, the negative-index metamaterials have not only a dispersive electric permittivity but also a dispersive magnetic permeability. The purpose of this paper is to identify the role of dispersive magnetic permeability in nonlinear propagation of ultrashort electromagnetic pulses in metamaterials. Firstly, we derived a generalized system of coupled three-dimensional nonlinear Schroedinger equations suitable for few-cycle pulse propagation in the metamaterial with both nonlinear electric polarization and nonlinear magnetization, which clearly demonstrates the role of dispersive permeability in nonlinear pulse propagation: In the linear propagation aspect, its contribution is buried in the ordinary dispersive terms; while in the nonlinear propagation aspect, the dispersive permeability manifests itself as a nonlinear polarization dispersion, although it is a linear parameter. Secondly, by exemplificatively using the coupled nonlinear Schroedinger equations in the Drude dispersive model, we quantitatively discussed the influence of dispersive permeability on pulse propagation in metamaterials.

Ultrawideband (UWB) is an attractive technology for short-range high-capacity wireless communication systems. A novel all-optical method for generating UWB pulses is proposed and theoretically analyzed using cascaded periodically poled LiNbO₃ (PPLN) waveguides. The operation principle relies on the sum-frequency generation (SFG) in the first PPLN and the cascaded second-harmonic generation and difference-frequency generation (SHG+DFG) in the second PPLN. We simulate the proposed PPLN-based UWB pulses generation using the well-known coupled-mode equations describing the SFG and SHG+DFG processes. A pair of polarity-reversed UWB monocycle pulses is generated with a central frequency of 5 GHz and a 10 dB bandwidth of 8.75 GHz. Thus the fractional bandwidth is 175%. Moreover, a pair of polarity-reversed UWB doublet pulses is also obtained. One has a central frequency of 5 GHz, a 10 dB bandwidth of 7.5 GHz, and a fractional bandwidth of 150%, and the other has a central frequency of 5 GHz, a 10 dB bandwidth of 3.75 GHz, and a fractional bandwidth of 75%. It is found that all generated pairs of polarity-reversed UWB monocycle and doublet pulses match well with the UWB definition of Federal Communications Commission (FCC).

Microbolometer detector is very competent as uncooled infrared detector for a wide range of thermal imaging applications, since it has been found to be more sensitive and has the advantage of using standard Si micro-fabrication process compared with pyroelectric or ferroelectric technology. The heart of microbolometer detector is a two dimensional array of thermal sensitive thin-film layers, which can change their temperatures and resistivities depending on the radiation absorbed. During the entire thermal imaging process, the microbolometer detector's substrate temperature, calibration temperature and ambient temperature are the key parameters which determine the thermal-electrical performance and the ultimate imaging quality of the microbolometer detector. In this work, based on the analysis of the characteristics of these parameters, the experiment has been conducted with the uncooled infrared thermal imaging system based on 320×240 amorphous silicon microbolometer detector working at different substrate temperatures, adopting different calibration temperatures for different ambient temperatures. The corresponding measurement results of the system's NETD, residual nonuniformity and power consumption, as well as the system's imaging results are presented, which all have a great agreement of the theory analysis above.

Ultrahigh-speed all-optical AND and NOR gates for three-input polarization-shift-keying (PolSK) signals, based on orthogonal dual-pump four-wave mixing (FWM) in semiconductor optical amplifier (SOA), is proposed and theoretically investigated. This scheme is simple, compact, and feasible for ultrahigh-speed operation.

A numerical model for the investigation of ultrafast gain and phase dynamics in InGaAsP bulk semiconductor optical amplifiers for transient cross phase modulation is presented in considering non-equilibrium carrier distribution and carrier temperature dynamics for different probe wavelengths. Results indicate that the nonlinear patterning of transient cross phase modulation can be effectively suppressed by an assist light, especially for the up-conversion and the assist light in counter-propagating scheme.

Thin Sn films in the thickness range 0.3-2nm are deposited by r.f.-sputtering on porous silicon (PS) anodized on p-type silicon. Microstructural features of the samples before and after r.f.-tin-sputtered are investigated with scanning electron microscopy (SEM). The photoluminescence (PL) studies showed that a broad luminescence peak of PS near the 621nm region gets a reduction in intensity, and a new peak at 441nm was produced at first and then disappeared. The FTIR spectra on the PS/Sn structure revealed no major change of the native PS peaks.

All-optical wavelength conversion based on four-wave mixing (FWM) is one of the key techniques for building dynamic optical networks. In this paper, the cavity enhancing effect of the residual F-P cavity mode on the non-degenerated four-wave mixing (NDFWM) in a distributed-feedback semiconductor laser diode (DFB-LD) have been investigated both experimentally and theoretically. The conversion efficiency of NDFWM is obtained at small or large frequency detuning range. The results show that the NDFWM can be enhanced obviously when the probe wavelength matches one of the F-P cavity modes, and the high conversion efficiency can be achieved even if the frequency detuning between the injection probe frequency and free-running frequency of the DFB-LD is up to THz.

Based on rate equations for carrier density in the active region of the QDSOA, the performance of a XOR gate using a quantum-dot semiconductor optical amplifier- based Mach-Zehnder interferometer (QDSOA-MZI) is analyzed in terms of Q factor through numerical simulations. The control pulse energy, the pulse width and the carrier capture time from the wetting layer into the dots are examined, which prove to be relevant to the Q factor. Our numerical results show that at 160Gb/s a high quality output signal with a Q factor over 5.9dB can be achieved.

Proposed in this paper is a high efficient 160Gb/s all-optical wavelength converter based on terahertz optical asymmetric demultiplexer with quantum dot Semiconductor optical amplifier (QDSOA -TOAD). The performance of the wavelength converter under various operating conditions, such as different injected current densities, input pulse widths and input control pulse energies, is analyzed in terms of contrast ratio (CR) through numerical simulations. With the properly chosen parameters, a wavelength-converted signal with CR over 19.48 can be obtained.

The spatial resolution of a space communication system is constrained by the diffraction limit of the telescope aperture. In a frequency-modulated continuous wave (FMCW), the frequency of the laser is ramped, and the frequency difference between the reflected wave and a local-oscillator wave is monitored. For maximum performance the frequency ramping should be linear. Linear frequency modulation (LFM) of the laser emission is a commonly used method for improving the detection sensitivity. Because of the available technology, techniques that use relatively low modulation frequencies were implemented first. In the early 1980's, an elegant measurement method based on frequency modulation opened up new applications for spectroscopy with spectrally modulated laser light. In this paper we analyzed systematically the principles of saw tooth-wave optical FMCW. For optical FMCW, because all the practical optical waves are from single-mode lasers, and because the laser beam from a single-mode laser is coherent in space due to the nature of stimulated emission, the spatial coherence is always satisfied, and therefore only temporal coherence need be considered. The chip signal, experimental system, and results are analyzed and discussed.

Laser Diodes (LDs) are becoming increasingly attractive as small and reliable laser-beam sources, with applications that include pumping of solid-state lasers, materials processing, and medicine. However, because of the poor beam quality of its output beam, which affects its direct applications, thus people pay much attention on how to shape the beam of the high power laser diode bar effectively. In this paper some typical beam shaping methods, shaping principles, key techniques and shaping effects are discussed. Based on geometric optical analysis, the collimation properties of the off-axis light wave through a cylindrical lens and elliptical cylindrical lens are separately analyzed in detail by rays tracing formulas, the collimation effects of them are also made a comparison.

We investigate the focused light from a metallic nanostructure by the coupling of a metallic nanoparticle and a metallic nanoslit flanked with periodic sinusoidal grating on one surface. We employ the boundary element method (BEM) to simulate the optical field distribution and the transmission resonant spectrum. The numerical results show that focused light are formed from the metallic nanoslit system. Such device can be used for miniature optical antennas in the optical regime, which can transmit or receive light along a specific direction for a given wavelength. Potential applications include that the coupling light in or out of fibers and the achieving the miniature optical source.

A thorough simulation and evaluation of phase noise for optical amplification using semiconductor optical amplifier (SOA) is very important for predicting its performance in differential phase shift keyed (DPSK) applications. In this paper, standard deviation and probability distribution of differential phase noise are obtained from the statistics of simulated differential phase noise. By using a full-wave model of SOA, the noise performance in the entire operation range can be investigated. It is shown that nonlinear phase noise substantially contributes to the total phase noise in case of a noisy signal amplified by a saturated SOA and the nonlinear contribution is larger with shorter SOA carrier lifetime. Power penalty due to differential phase noise is evaluated using a semi-analytical probability density function (PDF) of receiver noise. Obvious increase of power penalty at high signal input powers can be found for low input OSNR, which is due to both the large nonlinear differential phase noise and the dependence of BER vs. receiving power curvature on differential phase noise standard deviation.

The preparation process of GaAs photocathodes is very complicated, in order to prepare the high performance cathodes, it is crucial to obtain information enough to evaluate the preparation process in real time. Based on a particular transfer light setup and a flexible communication network, we develop an on-line measurement system for GaAs cathode preparation, which is used to measure the pressure of activation chamber, sample temperature, photocurrent, spectral response curves, and currents heating Cs and oxygen dispensers during the heat-cleaning or activation processes of cathodes. According to these signals, we present some simple and real-time evaluation techniques for cathode preparation. Several peaks of pressure are observed in the pressure variations measured during heat cleaning. These peaks corresponding to the desorption of AsO, As₂O₃, Ga₂O and Ga₂O₃ from the sample surface at different temperatures, respectively, are used to evaluate the effect of heat cleaning very well, while the signals measured during activation can be used to analyze and optimize the activation technique. Based on a revised quantum efficiency equation, many performance parameters of cathodes are obtained from the fitting of spectral response curves. According to these parameters, the performance of cathode material and the effect of activation can be evaluated.