We report on a hybrid organic/inorganic platform that allows the integration of passive and active optical functions to form planar lightwave subsystems. The integration approaches include chip-to-chip attach, flip-chip mounting, and insertion of films in slots formed in optical waveguiding circuitry. The materials integrated include polymer, silica, silicon, silicon oxynitride, lithium niobate, indium phosphide, gallium arsenide, yttrium iron garnet, and neodymium iron boron. The functions enabled by the hybrid integration approaches span the range of building blocks needed in optical circuitry, while using the highest-performance material system for each function. We demonstrate a number of subsystems on a chip, including fully reconfigurable optical add/drop multiplexers and tunable optical transmitters.

An InGaAs photodetector array interconnected with a silicon readout IC is the industry standard for 1.2-1.6 μm imaging applications. However, the indium-bump technique it employs for interconnection makes it expensive. An alternative approach is to combine a CCD with a device that upconverts 1.2-1.6 μm radiation to a wavelength below 1 μm. Here we report the realization of a 1.5 μm to 0.87 μm optical upconversion device using wafer fusion technology. The device consists of an InGaAs/InP PIN photodetector and an AlGaAs/GaAs light emitting diode (LED). Incoming 1.5 μm light is absorbed by the InGaAs photodetector. The resulting photocurrent drives the GaAs LED, which emits at 0.87 μm. The PIN and LED structures are epitaxially grown on an InP and a GaAs substrate, respectively. The two wafers are wafer fused together, the GaAs substrate is removed, and the sample is processed using conventional microfabrication technology. In this paper, we first present the design and fabrication process of the device. We then discuss the approaches to increase device efficiency. We show, both experimentally and theoretically, that the active layer doping affects the LED internal quantum efficiency, especially under low current injection. An optimum doping value is obtained. The LED extraction efficiency is increased using several approaches, including micro-lens and surface scattering. Overall device efficiency is further improved by introducing a gain mechanism into the photodetector. Our results show the potentials of this integrated photodetector-LED device for 1.2-1.6 μm imaging applications.

The advantages of monolithic integration of passive and active optoelectronic components into multifunctional PICs (photonic integrated circuits) are numerous and well known. In WDM transmission systems, where up to hundreds of wavelength channels may be present and must be manipulated on a per wavelength basis, the advantages of PICs containing wavelength (de)multiplexers and arrays of active components rapidly multiply. This has been exploited in our earlier reported SurePath family of InP-based PICs for optical channel monitoring / equalization, which contain an echelle grating demultiplexer and single-mode vertically integrated waveguide photodetectors / electroabsorption attenuators inserted into each of its output channels. Now, the same design principles and fabrication techniques have been utilized for the development of a dense WDM data receiver. This paper concentrates on the key building block of such a component, the single-mode vertically integrated waveguide photodetector. A detailed characterization and analysis of the device performance are presented.

Compact and low-cost semiconductor laser sources have significant potential for use in applications which are currently dominated by expensive solid state lasers. In such high power lasers, the optical mode size of the laser source has to be expanded in order to reduce the junction temperature, and the nonlinear effects and the optical power density at the laser facets. Broad-area tapered devices have been proposed to achieve high output power, together with achieving better quantum efficiency and higher beam quality [1]. To study the evolution of the optical beam in an axially non-uniform structure a beam propagation method (BPM) is essential as part of the design process. Thus a full-vectorial FEM-based BPM [2] is used to obtain the transformation of optical beam profile in the tapered region of the device. In this work a deeply-etched semiconductor tapered structure is considered and an expansion of the beam shape has been observed. This characteristic, together with the optical power loss for various taper designs has been studied will be presented. Mode beating also generates beam filamentation in the transverse direction and can deteriorate the overall beam quality. Our subsequent analysis by the use of overlap integral approach has showed the generation of higher order odd modes and their interference with the fundamental mode. The axial variation of the spot-size has also been correlated with the beat lengths between these modes. REFERENCES [1] E S Kintzer et al., IEEE Photonics Technology Letter, 5, pp.605, 1993. [2] S S A Obayya et al., J Lightwave Technol., 18, pp.409-415, 2000.

Resonant cavity enhanced photodetectors (RCE-PDs) are promising candidates for applications in high-speed optical communications and interconnections. The parasitics effects on these high-speed photodetectors must be carefully considered since they can significantly degrade the performance of the photodetector. In this paper, we will present a complete accurate model for the time response of the RCE-PDs. We will also study the effects of the parasitics of RCE-PDs on their time response and how we can compensate for the performance degradation from these parasitics. This study has been done for both RCE-PIN-PDs and RCE-avalanche photodetectors (RCE-APDs). RCE-separated absorption graded charge multiplication-APD was taken as an example of RCE-APDs. The time response of these RCE-PDs has better performance when compared to those of non RCE-PDs. The parasitics effects include the effects of both of the load resistance and the capacitance of the photodetector. The effects of the inductor that may be added in series with the load are also studied. It is shown that adding an external inductor results in higher performance of the photodetectors and this inductor can compensate some of the degradations resulting from other parasitics. The effects of the parasitics have been investigated for different dimensions of the photodetectors, different values of both the load resistance and the added inductor and also for different multiplication gains for the case of RCE-APDs.

We present simulation results on how power output-input characteristic bistability in Distributed FeedBack -DFB semiconductor laser diode SLA can be employed to implemented Boolean logic device. Two configurations of DFB Laser diode under external optical injection, either in the transmission or in the reflective mode of operation, is used to implement different Optical Logic Cells (OLCs), called the Q- and the P-Device OLCs. The external optical injection correspond to two inputs data plus a cw control signal that allows to choose the Boolean logic function to be implement. DFB laser diode parameters are choosing to obtain an output-input characteristic with the values desired. The desired values are mainly the on-off contrast and switching power, conforming shape of hysteretic cycle. Two DFB lasers in cascade, one working in transmission operation and the other one in reflective operation, allows designing an input-output characteristic based on the same respond of a self-electrooptic effect device is obtained. Input power for a bit "1" is 35 μW(70μW) and a bit "0" is zero for all the Boolean function to be execute. Device control signal range to choose the logic function is 0-140 μW (280 μW). Q-device (P-device)

We have developed a new scanning probe microscopy - based technique, scanning differential spreading resistance microscopy (SDSRM) and applied it to profile the free carrier distribution inside operating optoelectronic devices. The results of our SDSRM study of multi-quantum-well (MQW) buried heterostructure (BH) lasers under zero and forward biases are presented. SDSRM scans with high spatial resolution over the MQW active region of a BH laser yielded quantum-well-resolving differential spreading resistance measurements. The SDSRM results show the changes of internal carrier distribution within the MQW active region in BH lasers upon the application of forward biases. In combination with scanning voltage microscopy results, the SDSRM measurements provide direct experimental evidence of electron overbarrier leakage due to heterobarrier lowering, which has been speculated in theoretical modelings. Our results demonstrate the strength of SDSRM in probing the inner workings of operating quantum optoelectronic devices on nanometer scale, in which conventional analytical techniques such as secondary ion mass spectroscopy (SIMS), scanning spreading resistance microscopy (SSRM), and electron beam induced current microscopy can either apply only to devices under zero bias or provide only qualitative pictures.

A subtle roughening of the surface of a buried 60 nm InGaAs epitaxial layer was detected using a combination of sample cleaving, selective chemical etching and Field Emission Scanning Electron Microscopy (FESEM). In our technology, InGaAs is the photo-absorbing layer of Metal Organic Chemical Vapor Deposition (MOCVD) grown layers used in the monolithic integration of active photo detectors and a passive mux/demux. Conventional Photo-Luminescence (PL) and X-Ray Diffraction (XRD) techniques used to monitor and optimize the growth of epitaxial layers did not show this microscopic surface roughness. The appearance of roughness in the InGaAs layer was linked to very large changes in the dislocation density of the layers grown over the rough surface. Increases of up to three orders of magnitude in the Etch Pit Density (EPD from 10⁴ to 10⁷ cm^-2) were revealed using a standard Huber Etch. The Huber Etch also showed the preferred formation of "pairs" of dislocations threading out from a common point on the rough InGaAs surface. Changes in growth conditions resulted in the complete elimination of roughness and of excessive dislocation densities

Transparency current density (J_tr) was studied in GaInNAs ridge waveguide lasers. The devices employ Ga_1-xIn_xN_yAs_1-y multiple quantum wells and were grown on GaAs substrates using solid-source molecular beam epitaxy (MBE) with an RF plasma cell. The transparency current density is sensitive to material quality: defects, traps and other sources of non-radiative recombination. It is also dependent on the rate of thermionic emission from quantum wells. Wavelength, polarization and temperature dependence of transparency carrier density of annealed material was studied. Record low transparency carrier densities of 20 and 90 A/cm²/well were observed (for TM and TE polarizations) in devices based on GaInNAs material designed for emission at 1340 nm after optimized rapid thermal annealing. This low value of J_tr confirms the excellent quality of the GaInNAs material and demonstrates that GaInNAs lasers with excellent material properties can be grown for long wavelength applications provided appropriate annealing is applied. It is believed that the low transparency current density is a unique feature of GaInNAs and is due to the band structure and band alignment of the material system.

Subcarrier multiplexed fiber-optic systems using direct modulation of semiconductor lasers have attracted much attention for analog and digital broadband services. Since analog modulation is adopted, the system performance can be degraded seriously due to inherent nonlinearities of semiconductor lasers. In this study, the injection-locking technique is applied to reduce the nonlinearities in directly modulated semiconductor lasers. In particular, the characteristics of the second harmonic distortion (SHD) and the third harmonic distortion (THD) are experimentally investigated. First of all, at a modulation frequency of 5 GHz, both SHD and THD in the injection-locked semiconductor laser are observed to reduce significantly and equally for a large range of modulation power, which are 15dB- and 23 dB-decreased, respectively. Second, at a fixed modulation power of 6 dBm, the reduction of SHD and THD is, however, found to vary with modulation frequency. The reduction of SHD is more substantial at around one-half of the relaxation resonance frequency of the free-running laser. The decrease in THD is more significant at around one-half and one-third of the relaxation resonance frequency of the free-running laser. Finally, how the reduction of harmonic distortions varies with the operational parameters of the laser system is also investigated. The results demonstrate the feasibility of the injection-locking method in reducing harmonic distortions for high-speed and high-power analog modulation applications.

The importance of semiconductor optical amplifiers (SOAs) as key components in optical communications and integrated optics, covering a wide large of applications for the 1550- and 1300-nm optical windows, has grown in recent years. Polarization sensitivity of the optical of gain in SOAs is an issue that needs to be addressed to improve their performance and enhance their suitability for monolithic integration. We report on optical gain spectra measurements of polarization-insensitive SOAs that have been designed in our laboratory. Polarization insensitivity has been achieved employing a combination of both tensile and compressively strained quantum wells in the active layer of an InGaAsP/InP - based device. The SOA chips were characterized with a continuous wave input signal of a tunable laser at wavelengths between 1430 and 1600 nm. The gain saturation properties were experimentally investigated in order to determine how well the amplifier maintains its polarization insensitivity in the saturation regime. The experimental results were compared with theoretical values. The optical gain dependence on the current density and the length of the amplifier has been studied. The calculated device gain based on amplified spontaneous emission (ASE) spectra measurements was compared with the amplified signal measurements. Broad area lasers with lengths ranging from 500 to 1500 μm were also fabricated and tested to check the material quality and obtain information about the optical gain uniformity. Our amplifiers have an unsaturated gain of 22 dB and in the saturation regime the maximum observed gain difference between the TE and the TM mode gains was 0.5 dB within a spectral width of 60 nm. The measured 3 dB saturation output power was 6 mW.

Directly modulated lasers (DMLs) have two high performance applications: 1310 nm 10 Gb/s uncooled and 1550 2.5 Gb/s extended reach. Two key elements are gain coupled gratings and buried heterostructures. Gain coupled gratings simultaneously increase the DML's intrinsic relaxation oscillation frequency and damping, while the buried heterostructure reduces thermal chirp and parasitic capacitance. Large relaxation oscillation frequencies and reduced parasitic capacitance allow 85 °C operation; large damping and reduced thermal chirp enable extended reach.

Characterization of DWDM photonic integrated circuits presents many challenges and trade-offs. High channel numbers significantly increase the required time to screen and qualify devices at different stages within their production. The SurePath Monitor product requires accurate calibration of the absolute responsivity of each of its 43 100GHz-spaced channels, across the C-band (1525 to 1565nm), to meet its power monitoring accuracy specification of ±0.5dB. Further specifications such as input optical power range (0 to -40dBm) require photocurrent measurement capabilities ranging from mA down to sub-nA levels, whilst the high efficiency of the integrated demultiplexer filter requires large electrical dynamic range (>65dB). Characterization stations developed to ensure high throughput and low cost without sacrificing measurement accuracy and repeatability were based upon matrix-based measurement methodologies to ensure this balance is achieved.

A novel ultrahigh-speed electro-optic polarization modulator is introduced. The modulator uses a mode converter and a static polarization controller to change the output polarization state in a circular path, following a great circle, around the Poincaré sphere. Any two states on the Poincaré sphere can be connected. The mode converter is constructed using an AlGaAs ridge waveguide combined with slow-wave travelling wave electrodes. The travelling wave electrodes are designed to match the velocity of the electrical modulating signal, the data signal, to the optical carrier signal over a broad frequency range. This modulator demonstrates a 3 dB bandwidth in excess of 40 GHz. The polarization modulator exhibits extremely low differential group delay, on the order of a few 10s of femto-seconds, and low drive voltage, on the order of 5 V.

We use a quasi-three-dimensional numerical model combining finite element calculations in the x - y plane and a longitudinal optical model for the design and the simulation of wide band superluminescent InGaAlAs/InP light emitting diodes (SLEDs). It is shown that by using an active region with a continuously varying composition, bulk devices can provide singlelobe spectra of more than 100 nm full-width-at-half-maximum (FWHM) and output powers of a few tens of mW. This is broader than multiple quantum-well (MQW) device singlelobe spectra which do not exceed ~70 nm FWHM in the same power range.

We have measured a net gain of 19.5 dB in a 4 mm long piece of Cd₃P₂ Semiconductor Cylinder Fiber (SCF) at a wavelength of 1550 nm. The fiber was pumped from the side with a 100 mW, 832 nm laser. Side pumping is very inefficient since only a small portion of the pump light is absorbed by the very thin, approximately 6.694 nm thick, semiconductor film. However, this pumping arrangement is very convenient and does not require wavelength sensitive input and output couplers. We also measured the absorption spectrum. The absorption spectrum is in good agreement with a theoretical model. The absorption spectrum exhibits a step due to the two direct energy gap conduction bands of the Cd₃P₂ semiconductor film.

We report a unique low-cost technique for broadband gain and noise figure characterization of erbium-doped fiber amplifiers using an amplified-spontaneous-emission (ASE) source and a tunable filter with multiple steep slopes. The filter edge is tuned in steps of 10 pm and a series of output versus input power spectral density data points are taken at a fixed wavelength. Gain and noise figure of the amplifier are obtained by extracting the slope and intercept of output versus input power spectral density. The results obtained over a 20 dB total input power range are in good agreement (within ± 0.2 dB) with those obtained using conventional spectral-interpolation technique employing multiple DFB lasers at 100 GHz spacing over the C-band. The required filter depth is about 35 dB. Our method has several major advantages: (1) Low cost, as there is no need for multiple DFB laser sources or high-speed AOM modulators and RF drivers; (2) Immune to steady-state noise in the source; (3) Can be used to characterize amplifiers with fast dynamics as its accuracy is in principle not affected by the response time of the amplifying medium.

Homogeneous gain broadening suppression and super mode noise reduction in a multiwavelength active mode-locked erbium-doped fiber ring laser are investigated. By incorporating a semiconductor optical amplifier that is biased to operate just above the transparent point, the gain spectral hole burning of the erbium-doped fiber ring laser is effectively suppressed and the super mode noise is significantly reduced. Active mode locking of 8 wavelengths at room temperature with improved noise figure is demonstrated.

We have described a high-power hybrid fiber amplifier, which comprising an erbium / ytterbium co-doped double-cladding fiber amplifier (EYCDFA) as a post-amplifier and a conventional erbium-doped fiber amplifier (EDFA) as a pre-amplifier. At the signal wavelength of 1550 nm, the signal gains of up to 71 dB and the maximum output powers of 36.4 dBm or 4.37 W have been demonstrated when the total pump laser power was 12.3 W.

Q-switched rare-earth-doped single-mode fiber lasers attract much attention in many applications. It is well know that in actively Q-switched solid-stage lasers, single output pulses observed in experiments are in good agreement with analytical solutions and numerical simulations. However, in actively Q-switched fiber lasers, the output pulses break into several sub-pulses in Q-switched envelopes. These phenomena have been reported in various actively Q-switched fiber lasers under different cavity schemes, pump and fiber conditions. Since dynamic distributions and evolutions of photon density and inversions in the laser cavities have not been effectively measured at the initial stage of Q-switching, the mechanism of these phenomena has remained uncertain in past two decades. Since the understanding of the related mechanism is critical to the design and development of laser systems, some conjectures have been proposed in the literature, based on experimental observations and analyses. Unfortunately, without careful examinations on the pulse initiation and evolution processes, the mechanism has not been exactly explained. In this paper, by investigating the initiation and formation of split pulses in typical actively Q-switched fiber lasers, the mechanism is exactly illustrated for the first time to the best of our knowledge. It is related to the evolution processes of the injected perturbation, and determined by the rise time of switching and conditions of photon density and inversions in Q-switching processes. Furthermore, based on this principle, some solutions to realize single-pulse output are proposed in this paper, and the simulated results are compared with the experimental ones.

The Raman gain spectrum of Phosphorous-doped optical fibers (PDF) is known to exhibit, in addition to the pure silica response, a strong 40 THz frequency-shifted Stokes peak relative to the pump. This large shift reduces the number of cascades required in Raman lasers using nested cavities. We report a measurement of the Raman gain coefficient for a PDF. Our measurement scheme has the advantage of requiring only a pump source and no signal. We developed a model taking into account both contributions from spontaneous and stimulated Raman scattering. Raman spectra gathered at several pump powers are fitted against this enhanced theoretical model. The experimental values of the Raman gain coefficient obtained are in agreement with published results.

Coupling light into and out of small high index contrast waveguides is a fundamental challenge to implementing practical microphotonic waveguide and photonic crystal devices. Previous approaches include three-dimensional tapers, inverse taper waveguides, and grating based couplers. We propose and describe a much simpler coupler based on a short length of graded index (GRIN) material deposited on top of a silicon-on-insulator (SOI) microphotonic waveguide. The GRIN coupler has a refractive index that decreases from the index of silicon at the waveguide-coupler interface, to an optimized value at the coupler surface. Beam propagation method calculations are used to evaluate the coupling efficiency from a 4 μm thick coupler section to the fundamental mode of a 0.5 μm thick SOI waveguide. Coupling efficiencies are compared for couplers with smoothly varying quadratic index profiles and with one, two and three index steps. Coupling efficiencies of 75% (1.3 dB) or better are predicted using a three step GRIN structure with indices ranging from n=3.30 to 3.41 (Si). This index range is easily accessed using a-Si layers deposited by PECVD at varying deposition conditions, or by using composite digital alloys of high and low index films. With this method, microphotonic waveguide couplers can be designed and fabricated using only PECVD deposition and one patterning etch step with very modest tolerances. Efficiency increases to 87% (0.6 dB) when the index range of the 3-step coupler is extended to 3.0.

Spot-size converter (SSC) is an important building block of InP-based photonic integrated circuits since it allows a standard single-mode fiber with a large and symmetric mode spot to be efficiently coupled with high displacement tolerance to a semiconductor waveguide with a small and asymmetric mode spot. Having an on-chip SSC integrated with a semiconductor waveguide is practically advantageous since such an element greatly simplifies the packaging process while increasing its reliability. In this paper, a SSC utilizing two-step lateral tapering is proposed for converting the semiconductor waveguide device mode into that suitable for fiber coupling without compromising the designs of the device and coupling waveguides. This is achieved by inserting a transient taper between the device and coupling waveguide as an impedance matcher. This paper describes the design principles and characterization results for such a two-step SSC, compatible with earlier reported InP-based photonic integrated circuits for WDM. Transmission, Fabry-Perot fringes and photoresponsivity measurements (the last - by using an on-chip waveguide photodetector monolithically integrated with the SSC) show excellent performance of the two-step SSC. In good agreement with simulations, it was experimentally demonstrated that using this integrated component for fiber coupling can reduce the coupling loss to below 1dB, which includes the taper radiation loss of ~0.2 dB.

To miniaturize optical passive components or to have optical interconnects replace the current copper/low k interconnects for clock distribution, super high index contrast optics are needed because they allow optical waveguides with small bending radius, ie. < 50um. Silicon nitride core on oxide cladding has loss of <0.1dB/180° for 20um bending radius. However, coupling loss from the fiber to SiN waveguides, with 0.7umx0.7um cross section for single mode, is very large, > 20dB. To reduce the coupling loss, our approach is to have a double-core architecture, where fiber is first coupled to fiber matched waveguide, and then coupling from fiber match waveguide to SiN waveguide through a spot size mode converter. We have found the mode converter loss is reduced by 8dB by reducing the tip of the taper from 0.35um to 0.15um. In this paper, we are reported results of tips with less than 0.1um. We also describe the fabrication technology that enables us to make such fine tip with smooth surfaces.

Waveguide Photodetector (WGPDs) are considered leading candidates to overcome the bandwidth-efficiency trade-off presented in conventional photodetectors. In this paper, we present a physical model of the waveguide-separated absorption charge multiplication-avalanche photodetector (WG-SACM-APD). Both time and frequency modeling for this photodetector are presented. The frequency response has been simulated for different thicknesses of the absorption and multiplication layers and for different areas of the photodetector. The gain-bandwidth characteristic of WG-SACM-APD is studied for different areas and different thicknesses of both the absorption and the multiplication layers showing the dependence of the performance of the photodetector on the dimensions, the material parameters and the multiplication gain. In addition, the characteristics of WG-SACM-APD are studied for the case of an inductor added in series to the load resistor and better performance is achieved in comparison to the case with no inductor. The results obtained from the model that is presented in this work are compared with published experimental results and good agreement has been obtained.

We study the use of multi-section distributed feedback (DFB) lasers with integrated external cavities for the optical generation of millimeter-wave signals. Using a longitudinally-dependent time-domain model, we obtain the spectral characteristics of three basic designs through FFT analysis of the steady-state response. We then compare the stability and tunability of these devices and show the superiority of gain-coupled lasers with integrated active feedback.

Long-period fiber Bragg gratings (LPG) where the grating period is much longer than the wavelength of light have many unique characteristics and find uses in gain-flattening filters and mode converters. This paper describes the characteristics of the initial LPGs fabricated at the University of Adelaide using an infrared CO₂ laser. The optical system implemented promotes uniform irradiation of the full circumference of the fiber, avoiding many of the non-uniformities, associated with a single sided system. Some initial gratings have been made using this method, which typically show an attenuation of 10dB within a wavelength range (FWHM) of 8 nm. Work is now focused on improving these devices through an understanding of the writing process and its effect on the transmitted spectrum.

Traditionally in the process of writing fiber Bragg gratings with a phase mask the fiber is placed near or in close contact with the mask. With low coherence excimer sources this is necessary because the fringe visibility is greatly reduced beyond 500 μm. As a result of these limitations there has been increased interest in understanding the interference phenomena associated with a phase mask. During the past year we studied the beam interference phenomena associated with ultrafast gratings. We observed that with these coherent sources it was possible to write gratings remotely (Phase Mask-Fiber distance of ≈1 cm). In addition to this we observed evidence of walk-off between mask orders that significantly affected the interference patterns. In this paper we demonstrate that a frequency doubled Argon-ion laser, being a coherent source, can be used to inscribe fiber Bragg gratings at large distances from the phase mask (> 15 mm). We will demonstrate that walk-off between mask orders will change the interference profile along the grating length. We show that the spectral profile correlates with the calculated interference pattern. Beam walk-off effects play a role in the inscription of any photonic device with a phase mask. This remote writing technique can be used to tailor the index modulation pattern in the fiber and could potentially be used to produce two beam interference gratings even in the presence of a significant zero order amplitude.

The long-span and high-speed optical transmission networks require high-quality optical cables and components to reduce the detrimental effects of polarization-mode dispersion (PMD) and polarization-dependent loss (PDL) on system performance. In addition, many high-accuracy fiber-optic sensing networks also require low polarization-dependent components. As a result, polarization-insensitive transmission fibers and components have been pursued recently. A well-established technique to produce low-PMD and PDL fibers is to spin the fiber during the fiber drawing process, known as spun fibers. As key optical components, fiber Bragg gratings (FBGs) play important roles in optical fiber communications and fiber sensors. However, FBGs written into traditional single-mode fibers exhibit relatively high polarization dependence due to the birefringence introduced by side UV exposure in FBG fabrication process. In this paper, we report for the first time the characteristics of FBGs written into spun fibers that were formed by spinning fibers with certain intrinsic birefringence. Distinct properties of FBGs in spun fibers are found in the experiments, compared to those in traditional single-mode fibers. Based on Jones matrix method, the FBG spectra are simulated and compared with the measured ones. Preliminary experimental results indicate that the spun fiber gratings suffer lower PMD under the same fabrication conditions. Moreover, our theoretical analysis shows that the PMD and PDL characteristics of spun fiber gratings can be further improved by optimizing the structures of spun fibers.

We theoretically and experimentally study the polarization properties and the spatial and wavelength dependence of side scattered radiation from a fibre Bragg grating (FBG). Antenna theory is used to model the radiation pattern from the FBG. Recently, it was reported by other authors that side scattered light from a fibre Bragg grating exhibits a remarkable anisotropy in its azimuthal distribution, proposed to be due to an inhomogeneous transverse grating profile. We show that the angular distribution of the scattered light depends on the actual phase-matching angle - thus the observation angle - and if any, on the grating tilt. Hence, inhomogeneities in the cross-section of the fibre grating are not necessarily responsible for the anisotropic light distribution as was reported. It is shown that the scattered light pattern from any grating (even a uniform un-tilted one) possesses a non-uniform intensity distribution. The spatial distribution of the scattered light is then treated as the Fraunhofer diffraction pattern. In an analogy to dipole radiation, the presented model explicitly shows that the scattered light is located in specifically defined regions of a cone depending on the principal parameters of the FBG. The technique has interesting implications for several optical fibre Bragg grating devices.

It has been known for sometime that the tap angles associated with slanted, tilted and blazed Bragg grating structures can be affected by the guided mode's state of polarization (SOP). Recently this polarization dependent out-coupling has been employed in order to develop a number of useful devices including in-line polarimeters and PDL equalizers. Although a variety of tools are available to model blazed fibre Bragg grating (FBG) characteristics, a simplified explanation of the fundamental dependencies and potential behaviour has never been fully presented in the literature, making the optimization of these devices difficult and elusive at times. In this submission we present a thorough, intuitive discussion of these trends and possibilities as observed through an extensive theoretical analysis rooted in the Volume Current Method (VCM). In addition to discussing the potential limitations and shortcomings of this formulation, some rough guidelines for the manufacture of various devices are also disclosed.

The photonic crystal fiber (PCF) is a micro-structured fiber, where arrays of holes running along the waveguide length, has more controllable fabrication parameters than a standard single mode fiber. Increasing interest is being shown in such PCFs for a range of applications in optical communications, sensing and signal processing. This includes the control and guidance of optical beams, taking advantage of their unique transmission characteristics, including being continuously single-moded, with controllable spot-sizes and with tailored group velocity dispersion characteristics. To date, most of the research into these fibers has a strong experimental basis [1], which has recently been complemented by various modal solution approaches to their characterization, but mostly using scalar formulations or being limited to specific types of structures. The modal solution approach based on the powerful finite-element method (FEM) [2] is more flexible, can represent any arbitrary cross-section more accurately and has been widely used to find the modal solutions of a wide range of optical waveguides [2]. The optical modes in a high-index contrast PCF with two-dimensional optical confinement are also hybrid in nature. To accurately characterize such fibers a full-vectorial approach is necessary and a H-field based full vectorial approach [2] has recently been extended to study the polarization issues in such PCFs. Polarization dependent single mode operation, variation in the spot-size, modal field profiles, modal hybridism, birefringence, and the beat length have been calculated for these fibers. REFERENCES [1] J C Knight et al., Opt. Lett., 21, pp.1547-1549, 1997. [2] B M A Rahman and J B Davies, J. Lightwave Tech., 2, pp.682-688, 1984.

We present the optical performances of a compact variable optical attenuator (VOA) developed at Photintech. The presented VOA’s operation principle is based on the guided wave evanescent field manipulation. This approach allows the cost effective fabrication of VOAs with extremely low insertion loss, below 0.1 dB. The access to the evanescent portion of the guided radiation is achieved by removing a portion of the original waveguide’s cladding and replacing it by a thermo-optic composite polymer material. By changing the temperature of the thermo-optic material we create a partial leakage of the guided radiation into the replaced cladding area, attenuating thus the guided radiation. It is well known however, that the application of polymer materials for optical component fabrication creates significant birefringence due to the shrinkage and thermal stresses. As a result, the polarisation dependence of such devices is relatively high. We apply a specific cladding geometry and heating electrode (pending patent of Photintech inc), which ensures axial compensation of the birefringence, providing thus very small polarisation dependence. The design provides also a high dynamic range operation (above 50dB). An in-house designed electronic board allow the VOA operation in three different regimes: direct driving, constant output power or constant attenuation coefficient with a precision as better as 0.1dB. The developed VOA device can be used in agile optical networks, for such applications as dynamic gain equalisation, dynamic channel equalisation, optical transmitter power control and receivers protection in the telecommunication systems.

Integrated planar concave gratings are promising candidates for Wavelength Division Multiplexing (WDM) devices. However, device insertion losses reported to date are higher than other options such as Arrayed Waveguide Gratings (AWGs) mainly due to the strict requirement for a deep vertical etch of grating facets. We present a novel planar concave grating demultiplexer design based on total internal reflection and the focusing properties of ellipses which simplifies the fabrication of the device by avoiding grating facet metallization and improves grating efficiency by reducing its sensitivity to non-vertical etching of the facets. As a proof of concept of the grating design, a 4-channel coarse WDM demultiplexer working near 850nm with a channel wavelength spacing of 20nm has been fabricated in PMMA polymer using the LIGA process, a deep x-ray lithography technique using synchrotron radiation. Smooth and extremely vertical sidewalls 220 microns deep have been obtained. Test results for this preliminary device show a -7dB on-chip loss for the best channel. Based on the test data, a –3dB loss can be reasonably expected for a device with optimized material processing. In the future, the LIGA process can be used to fabricate a metal stamp that can massively produce devices in polymers with plastic embossing or molding techniques.

Polymeric materials have been widely used for the fabrication of photonic devices, in particular for applications in short haul optical networks employing coarse wavelength division multiplexing (CWDM). However, the molecular design and processing of polymeric materials to have all the properties required for the fabrication of high performance photonic devices continue to present challenges. This paper presents data on the design, fabrication and characterization of waveguide devices using novel fluorinated poly(arylene ether ketone) materials. These materials exhibit low optical loss (slab loss ~ 0.5 dB/cm at 1550 nm), high thermal stability (1 wt% loss at temperatures up to 430 °C), and are easily processed at temperatures lower than those previously reported for poly(arylene ether)s (< 200 °C). High quality waveguides have been fabricated using standard photolithographic processes. Issues affecting polymer layer and device birefringence and optical loss have been investigated, including molecular structure, processing conditions and substrate selection. Coupling devices sensitive to waveguide dimensions have been designed and fabricated, and their output compared to numerical simulations. Characterization of these devices allows further optimization of the materials and the waveguide process and assists with the design of more complex polymer photonic components.

With the development of telecommunications and high-speed computations, polymeric materials for optical applications are attracting much attention in highly integrated optical waveguides and circuits. In comparison with current inorganic waveguiding materials (e.g., silica and other III-V semiconductor materials), organic polymers offer several advantages including cheap fabrication, tunable properties, good processability and the ability for integration into large scale semiconductor circuits. In this presentation, we describe the design and synthesis of novel crosslinkable optical polymers for use in optical waveguides based upon a bisphenol monomer containing crosslinkable tetrafluorostyrol units. The introduction of this crosslinkable bisphenol into perfluorinated poly(arylene ethers) allows the synthesis of crosslinkable fluorinated polymers with adjustable refractive index and controllable high crosslinking densities. These polymers have been shown to exhibited low optical loss at 1550 nm, low birefringence, high glass transition temperatures, good mechanical properties, and excellent processability. In the presence of a suitable initiator, these polymers can undergo rapid crosslinking either thermally or optically allowing for multilayer device fabrication.

We have developed a widely tunable, narrow-linewidth, simultaneous triple-wavelength oscillation erbium-doped fiber ring laser (EDFRL), which can produce double-wavelength oscillations with the same polarization output, as well as another widely tunable wavelength oscillation with orthogonal polarization from 1522.2 nm to 1595.9 nm. By using this EDFRL along with a method of measuring polarization-mode dispersion (PMD) in optical fibers based on a broad-band orthogonal-pump four-wave mixing in a semiconductor optical amplifier (SOA), we have measured the PMD values of optical fibers, which are in good agreement with values measured by means of commercial PMD testing equipment. We have also proposed several novel devices for in-field PMD measurement and monitoring on dense wavelength-division multiplexed (DWDM) traffic-carrying links, which will significantly reduce the cost and time of the PMD testing in the running DWMD networking systems.

Erbium-doped fiber lasers can find many applications such as WDM systems, fiber-optic sensors and microwave photonics, thanks to their high power, narrow linewidth and low noise. The key factor that limits the number of stable lasing wavelengths achieved at room temperature is the strong gain homogenous broadening of erbium-doped fiber. Different approaches have been approached. These approaches include the use of a phase shifter in the fiber ring cavity to reduce the homogenous broadening, and the use of Nitrogen to cool the erbium-doped fiber with a reduced homogenous broadening. Clearly, the approach using a phase shifter increases the system cost and the approach using Nitrogen is not suitable for practical applications. In this paper, we propose a novel approach to achieving stable multiwavelength lasing at room temperature using a semiconductor optical amplifier in the fiber ring laser cavity. In the fiber laser, the erbium-doped fiber is used as a gain medium. The semiconductor optical amplifier is used as a phase shifter. The semiconductor optical amplifier is driven by a sinusoidal wave. The refractive index change of the semiconductor optical amplifier leads to a phase shift, which effectively suppresses the homogenous broadening and cross-gain saturation. In addition, since the semiconductor optical amplifier is biased at the transparent point, no insertion loss is introduced. In the experiment, the modulation voltage to the semiconductor optical amplifier can be as low as 10 mV, which is much lower than that used in a phase modulator (10 V). Experiment is being carried out and more experimental results will be reported.

Recently, multimode fiber (MMF) and components based on MMF have attracted much attention due to their potential applications in future optical access networks. Fiber Bragg gratings (FBGs) are considered to be key components in both telecommunication and sensing applications. Although single-mode fiber based FBGs (SMFBG) have been studied thoroughly, few studies have been reported on MMFBGs, mainly because of the complexity and multiple mode nature of the MMF. In this paper, transmission and reflection spectra of MMFBGs are studied systematically. Relationships between transmission/reflection spectra and the excitation conditions are clearly demonstrated by observing the far-field pattern. Different launching methods including the lateral offset launching and angular offset launching and light sources with different spectrum width are used in experiments. Furthermore, the transmission/reflection spectra dependence on the polarization state of excitation light and asymmetric refractive index perturbation profile are studied in detail. Theoretical simulations are used to compare these experimental results.

1x4 optical waveguide power splitters were fabricated in a fused silica glass sample using a focused 800nm-amplified femtosecond laser, and characterized in terms of their optical properties at the wavelength of 1550nm. A shifted multi-scan writing technique was developed to substantially reduce the excess loss of the involved Y junctions and the polarization dependent loss of such integrated splitters.

We report on the possibility of strong relief grating fabrication in amorphous chalcogenide As₂S₃ glass using polarization modulated near bandgap light illumination. Such gratings are created using low intensity and spatially uniform light illumination. The obtained relief structures are strong, very smooth and do not require post-exposure development procedures. They are stable to the heat treatment and the uniform photo-exposure. The possible mechanism of such relief structure formation is also briefly discussed.

Actuators with large strokes and fine resolution have many applications, such as precise positioning, high resolution lithographing, sub-micro machining and fine manipulation of cells in biomedical fields. In this paper, a compact design of an elliptic-shaped actuator using 16 flexural hinges is presented. The device used two piezo stacks in long axis to generate amplified displacement in short axis. Overall displacement of more than 1 mm was obtained from a piezo-stack displacement of 40 μm. A feedback loop was designed to control the motion using a linear variable differential transformer (LVDT). The design achieved a fine resolution of ~20 nm. One unique feature of the design is its thermal stability. Its invar frame minimizes the change of the short axis length when temperature varies. The actuator was applied to drive an optical tunable filter. Wide tuning range and excellent wavelength tuning repeatability of ± 15 pm were achieved with the LVDT feedback loop.

Long-period fiber gratings (LPFG’s) find applications in optical fiber communication systems and fiber sensor systems. Among others, it can be used as gain flattening filters (GFF’s) in the communication systems. Depend on the amplifier design, the GFF’s need to be either athermal, or have specific temperature sensitivities. The temperature sensitivity requiement sometimes can be very stringent. It has been known that the temperature and strain sensitivity are dependent on the fiber parameters and the order of the cladding modes it is used. In this paper we will describe the general method for finding suitable cladding mode in a specific fiber for specific requirements. We found that the polarization dependent losses (PDL) in high sensitivity modes are remarkably higher than the ones in common LPFG’s. In those high sensitivity filters achieved by the UV-beam side illuminating, the birefringence-related resonant wavelength separation (RWS), which is the central wavelength separation corresponding the slow and fast axis state of polarizations (SOP), can be in the range of 5 nm, which is remarkably larger than the reported values in other LPFG’s. There are two sources of birefringence which lead to PDL: fiber core ovality induced birefringence, which is intrinsic, and the anisotropic UV-beam exposure induced birefringence. We proposed methods to deal with those birefringence sources. The leads to almost complete remove of the RWS in the high sensitivity LPFG’s.

This work presents the performance analysis of ring resonator-based tuneable optical filters, namely the influence of the design parameters such as coupling coefficient (k), coupler loss coefficient (a) and operating temperature (t) on the optical performance of the filter. Computed results have shown that a and k dramatically influences the performance of the add-drop filter, the operating temperature can effectively be used as a mechanism of selecting the desired frequencies within the dynamic range of the filter, and special attention should be paid to the operating temperature range of the filter in order to avoid a decrease of its free spectral range below the value of the desired pass bandwidth. Preliminary sensitivity analysis has also revealed that the single ring add-drop filter design is almost insensitive to the relative position of the two optical couplers onto the ring, and therefore very versatile for integration into a photonic chip.

Fibonacci sequences constructed from high-index-contrast GaAs and Al₂O₃ quarter-wavelength layers are used as unit cells in a novel multilayer system. Quasi-periodic heterostructures, formed by concatenating repeated Fibonacci sequences of different order, have properties markedly different from classic Fabry-Perot bilayer systems. We employ the transfer matrix method, including imaginary components of the refractive index, to extract transmission and reflection spectra, and consider their sensitivity to material and geometrical variation. We find that these quasi-periodic heterostructures may have a very high quality factor and deep extinction in reflection. By contrast, bilayer structures of similar dimension are so strongly evanescently damped that the coupling to the cavity is negligible. Due to the coupled geometric resonances in the Fibonacci-based heterostructure, the spectral properties are easily tuned by altering the imaginary component of the refractive index in a single layer. We discuss the viability of possible technological applications.

Wireless access systems operating at higher microwave/mm-wave frequencies (10-60 GHz) are attractive because of ample spectrum availability and the potential for large bandwidth allocations. The difficulty with such a system is in the high cost associated with the microwave sources and signal distribution. Optical fibers are an obvious means of alleviating distribution problems, but optical modulation techniques are not currently capable of economically generating mm-wave frequencies. One potentially cost-effective method to fabricate such a system is via optical heterodyning. In such a system, it is relatively simple to generate microwave and mm-wave carriers. However, the difficulties in generating a high-quality signal are two-fold: The first is in maintaining a specific frequency difference (i.e. microwave signal) between the lasers for a prolonged period of time. The second is in narrowing the inherent linewidth of the laser from the MHz values typically produced by conventional semiconductor lasers, down to values more practical for a communication system. The applications in which this microwave signal can be used are determined by the degree to which the linewidth can be reduced. This presentation will discuss the design and implementation of an 11 GHz discriminator-aided, optical phase-lock loop constructed from commercially available external-cavity lasers incorporating fiber Bragg gratings.

There is a strong need for a simple and reliable characterization technique of nonlinear optical effects applicable to thin films. The second-order susceptibility is the parameter that plays a key role not only in SHG but also in cascading phenomena and optical bistabilities. The measurement methods of second-order susceptibility are divided into absolute and relative techniques. Relative measurements were performed to obtain the relative values of second-order susceptibility by comparison with a reference material, which is often crystalline quartz. There is a different approach for nonlinear optical characterization that depends on the type of sample: crystal or thin film. There are two methods used for second-order susceptibility measurement of thin films: a Maker fringe and reflective SHG. The Maker fringe is limited to the investigation of thin films on transparent substrates. Here we discuss an experimental protocol based on reflective SHG for nonlinear optical characterization of thin films. Since most measurements are performed relative to a reference material, the establishment of a well-accepted value for a standard material is important. The SHG in reflection of z-cut quartz is discussed in detail. The method is simple and reliable but limited to thin films. Measurements at different wavelengths and mapping will be reported.

We report general hot-carrier mechanisms for electroluminescence (EL) in metal-insulator-silicon tunnel diodes. We demonstrate these effects using various combinations of Si-oxide and Al-oxide tunnel-barrier insulators. In addition to an EL peak near the 1.1-eV Si band gap, we observe broad-spectrum EL that can span the detector-limited range from 0.7 eV to 2.6 eV (1780 nm to 480 nm). The maximum above-band-gap photon energy increases with the forward bias, consistent with hot-carrier recombination in Si. Below-band-gap EL is likely due to (i) hot-electron inter-conduction-band radiative transitions in Si and/or (ii) radiative recombination via localized interface states. Light emanates from specific sites with apparent size < 1 μm that appear during high-forward-current electrical stress. The number of sites can be in the hundreds, and is in direct proportion to the stress current, as anticipated for tunnel barrier dielectric breakdown. Current-voltage characteristics can be fit using a model appropriate to localized breakdown sites. Virtually all current is thought to cross the barrier at such sites, with local current densities as high as 10⁸ A/cm². We also describe novel devices where tunnelling occurs at predetermined sub-micron sites formed in 18-nm-thick SiO₂ using electron-beam lithography and wet-chemical etching.

Interdigitated metal-semiconductor-metal (MSM) photodetectors operating at 1550 nm were fabricated by depositing metal contacts on top of polycrystalline silicon (polysilicon), which was deposited by a variety of techniques. The highest responsivity was 0.66 mA/W, corresponding to an external quantum efficiency of 0.16%, obtained from a 2 μm thick polysilicon sample. Using Raman spectroscopy, it was found that polysilicon with grain sizes between 6 nm and 13 nm provides the best photoresponse at 1550nm.

A novel periodic leaky-wave microstrip antenna is presented that is capable of dynamic beam scanning using optical control techniques. Optical control is achieved by illuminating silicon elements placed periodically along the length of the antenna with variable optical intensity. Incident photons absorbed into the surface of the silicon create an excess carrier concentration whose behavior within electromagnetic fields is characterized by a complex permittivity that is derived from first principles using Maxwell’s equations. Preliminary simulations validate the antenna concept.

Silicon rich silicon oxide thin films have been fabricated by electron cyclotron resonance plasma enhanced chemical vapor deposition. Following their deposition, these films were subjected to thermal anneals at temperatures up to 1100°C for times of up to 120 minutes. Annealing of the films causes a phase separation of the material to form Si precipitates, which nucleate to form Si nanocrystals, within an amorphous SiO₂ matrix. The nucleation of the nanocrystals was analyzed as a function of the composition of the films, as determined by Rutherford backscattering and elastic recoil detection analysis experiments, and the annealing conditions. The bonding structure of the films was analyzed using Fourier transform infrared spectroscopy. Surface morphology, including analysis of the size and distribution of the nanocrystals, was determined through the use of atomic force microscopy. Spectroscopic ellipsometry, in the range from 600 to 1100 nm, was used to examine the effects of the formation of nanocrystals on the optical properties, i.e., index of refraction and extinction coefficient, of the films. Photoluminescence spectra were used to show that due to quantum confinement effects the nanocrystals exhibit luminescence, making them a potential candidate for integrated photonic emitters.

In this paper we present a novel 1V current reference for a transimpedance amplifier (TIA) which provides a small dependence of the output response on temperature and process variation. The circuit is designed to drive the current tail of a photoreceiver TIA. The proposed architecture is based on a bandgap voltage reference and use an nMOS transistor at its output to produce a temperature compensation of mobility and threshold voltage (V_TH) over a wide range of temperature (-40 to 100°C) and process variation affecting V_TH. This technique has been confirmed by theoretical analysis and SPICE simulations using the TSMC CMOS 0.18μm process.

The progress in the development of Si-based Surface Plasmon Resonance sensing technology is reported. This technology uses multi-layer structures with a gold film and a silicon prism in the Kretschmann-Raether geometry and makes potentially possible the miniaturization and integration of the sensor device on a silicon-based microplatform. We show conditions of the simultaneous excitation for two plasmon polariton modes over both sides of the gold film using different intermediate layers, between the high-refractive index silicon prism and the gold, and examine their response in configurations of the conventional and nanoparticle-enhanced sensing. The system has been calibrated in real-time measurements of protein (Concanavalin A) adsorption.

The optical properties of devices derive from the flow of energy within their structure, a flow that is governed by the interaction between geometry and material properties. Through consideration of optical bound modes in simple buried-core waveguiding structures embedded in strained multilayers, we investigate how the symmetries of the system interact through the stress-optical coupling. We do so using finite element analysis, first extracting the stress distribution in the system, and then finding the eigenmodes of the fully-vectorial anisotropic Maxwell's equations. In particular, we explore how, and to what degree, various symmetry breakings compensate for one another. We illustrate this by the distinction between global and local metrics that delimit device operability. Modal birefringence, which arises from the difference in transport velocities associated with the fast and slow axes of the waveguide, is an example of the former, whilst the optical power density illustrates the latter. We demonstrate that these, though coupled, are independent metrics.

The current trend towards integrating CMOS circuitry and photonic devices on silicon-on-insulator (SOI) wafers into monolithic optoelectronic circuits requires modification of the traditional rib waveguide to provide a planar surface. One possible modification involves creating a trench on either side of the rib and then filling this trench with a planarizing oxide. The resulting planar surface is much more compatible with the photolithographic systems required to print the small dimensions found in state-of-the-art CMOS. We report the fabrication of prototype trench-isolated waveguides, measurement of optical performance and comparison with simulation. 2.5 μm thick Si film was utilized with 0.5 μm deep trenches ranging from 1 to several microns in width. Rib widths ranged from 2 μm to 3 μm (the maximum value providing single mode propagation). Allowed modes were determined with FEMLAB, while beam propagation was studied using Optiwave BPM. Simulation indicated mode confinement would be lost for trench widths less than 1.5 μm. The narrowest trench width which could be fabricated was 2.0 μm, and qualitative optical testing shows good mode confinement to the central rib for a trench geometry greater than 3.0 μm.

A novel type of waveguides utilizing simultaneously Bragg reflection and TIR (total internal reflection) for light confinement is proposed and studied. The BRAGGATIR waveguides combine the advantages of Bragg waveguides and lateral ARROWs (antiresonant reflecting optical waveguides) -- namely, large core sizes allowing easy and low loss fiber coupling and their reduced sensitivity to fabrication errors, with the properties of high index contrast ridge waveguides which tolerate small bending radius. Since the proposed waveguides have only one Fabry-Perot layer, they are more compact compared to the Bragg waveguides. Moreover, the large dispersion of the BRAGGATIR waveguides near the resonances may find applications in optical signal processing. Beam propagation method (BPM) and mode solver numerical simulations were performed and the bending losses of different waveguides, namely BRAGGATIR, ridge and ARROW, were compared for different bending radii down to 10 microns. The results demonstrate advantage of the proposed waveguides allowing very small radius ring resonators with large free spectral range and more compact integrated optics devices.

Early development work in the design of optical power splitters, likely influenced by similar construction in the microwave regime, placed heavy emphasis on Y-branch designs with the output waveguides immediately branching from the input waveguide at non-zero angle. This design approach, which is still prevalent, is fundamentally flawed from the perspective of both optical power flow and fabrication, as it leads to significant excess loss and/or a large statistical variance. If inherent broadband response is not a critical requirement, directional-coupler or multimode-interference splitters are usually chosen instead. We demonstrate, choosing a minimal function perspective where the optical design is sensitive to the smallest possible set of critical fabrication parameters, that robust and low-loss Y-branch designs are indeed possible. The minimum gap width between waveguides being the critical parameter, we reveal the dependence of the irreducibly simplest design on all elements of the parameter space, as they relate to the critical one. In so doing, we show that the concept of bending angle is irrelevant.

In the present paper, we take a closer look at the design of long period fiber gratings (LPFG) which present several advantages compared to similar components. In fact, the LPFG has stronger wavelength stability when opposed to temperature variations and a better resistance against mechanical strength. These qualities offer them more importance in the industry. We use an electric arc to change the fiber characteristics and to create a fringe which presents a very thin section of the fiber that has a different refractive index. We use a computer based setup which allow us the control of the arc intensity and the discharge time and the gap between two neighbor fringes for the LPFG varies in the range of 400 to 700μm. A various optical fiber components designs and implementations are possible using electrical arc. In this paper we will expose the use of electrical arc and micro deformations for writing the Long Period Fiber Gratings (LPFG) on a single mode fiber. We take a closer look at the design of this component which present several advantages compared to similar components.

We present experimental results demonstrating the possibility of obtaining low-loss splices of microstructured optical fibers (MOFs) by using conventional electric-arc splicers. We show evidence of the effectiveness of the method by splicing two MOFs together as well as a MOF with a standard single mode fiber (SMF). The results are presented in terms of fusion losses and tensile strength. Theoretical calculations of the losses attributable to mode mismatch between the MOF and the SMF suggest that the splicing losses could be further reduced by optimizing the MOF design parameters. For the case of a MOF-MOF splicing, the loss that could be due to a possible rotational misalignment that comes with the non-cylindrical symmetry of the modal distribution is also evaluated.

The refractive index is assumed to be a linear function of the dopant concentration. We investigate the core dopant evolution during the fused-fiber coupler fabrication by solving the convective diffusion equation. The tapering and the fusion of the coupler are considered as coupled phenomena. The slenderness of the geometry makes it possible to simplify the equations and to obtain a three-dimensional model for the velocity field. We present numerical solutions of the resulting equations and compare them with experimental measurements of the refractive index profile.

The dispersion curves of tapered fibers at various geometries are characterized by using white-light interferometry. In this presentation, the white-light interference patterns at frequencies near the taper’s second zero-dispersion frequency are measured and discussed in detail. A more convenient formula is proposed to fit the experimental data at these frequencies. As a result, small dispersion values can be calculated more efficiently.

2-dimensional (2-D) near hexagonal close packed optical fiber bundle (NHCP-OFB) is proposed for parallel optical interconnection (POI). Conventionally in 2-D POI, fibers must be accurately aligned to 2-D light sources and detectors using precisely fabricated hole-plate or stacked V-grooves. It is time consuming to thread each fiber into hole-plate, as well as precisely stack the V-grooves. For polymer optical fibers with large diameter, the whole dimension of the bundle may be too huge to connect ICs if they are positioned using V-grooves or hole-plate. The purpose of NHCP-OFB is to pack the fibers closely into a 2-D lattice, and precisely positioned them at the nodes of the lattice. Due to the fluctuation of the diameter of the commercial optical fibers, it can not be packed into perfect 2-D hexagonal close packed bundle. The proposed NHCP-OFB slightly stretch the perfect 2-D hexagonal close packed bundle in one dimension and still have all fibers packed tightly with high packing fraction. The position of fibers in NHCP-OFB is very important when aligned to 2-D light sources and detectors. This paper analytically analysis the positioning offset of fibers in NHCP-OFB. The possibility distribution of positioning offset is related to variance of fiber diameter and stretched length of NHCP-OFB. Experiment samples of NHCP-OFB have been made using polymer optical fiber showing that fibers can be precisely positioned.

We have developed a deep ultraviolet (DUV) lithography technique for fabricating super dense silicon based photonic crystals. Binary mask is used to create nano scale patterns of very high density. Based on the simulation, photonic crystals with both square and triangular lattice of air cylinders are designed and fabricated to work in communication frequency range (λ within 1.3 to 1.55μm) on amorphous silicon. In order to pattern circular hole we designed different kind of polygons on the mask and layout pattern was under sized at constant pitch. Bottom anti reflection coating (BARC) recipe was developed to improve circularity of the pattern and reduce interhole spacing.

We characterize coupling between two identical collinear hollow core Bragg fibers, assuming T₀₁ launching condition. Using multipole method and finite element method we investigate dependence of the beat length between supermodes of the coupled fibers and supermode radiation losses as a function of the inter-fiber separation, fiber core radius and index of the cladding. We established that coupling is maximal when fibers are touching each other decreasing dramatically during the first tens of nanometers of separation. However residual coupling with the strength proportional to the fiber radiation loss is very long range decreasing as an inverse square root of the inter-fiber separation, and exhibiting periodic variation with inter-fiber separation. Finally, coupling between the T₀₁ modes is considered in a view of designing a directional coupler. We find that for fibers with large enough core radii one can identify broad frequency ranges where inter-modal coupling strength exceeds super-mode radiation losses by an order of magnitude, thus opening a possibility of building a directional coupler. We attribute such unusually strong inter-mode coupling both to the resonant effects in the inter-mirror cavity as well as a proximity interaction between the leaky modes localized in the mirror.

The optical propagation of a pulse through one dimensional finite gratings and photonic crystals is discussed. In the case of shallow gratings the light transport properties are derived within the frame of the Coupled Mode Theory, while the Transfer Matrix Method is used for investigating photonic crystal (PC) structures. The so-called superluminal tunneling of wave-packets through the band-gap region is investigated, and the dwell time is computed. Because of the analyticity of the wave equation, the Einstein causality principle is not violated, although the group velocity can exceed the speed of the light in vacuum. We show that the dwell time is a propagation phenomenon and not a quasi-static process in which the incident pulse envelope modulates the amplitude of an exponentially decaying standing wave. Nevertheless, the group velocity cannot be always used to compute the transmission group delay, because the latter does not represent the time spent by the energy to propagate through the band-gap.

We present a dual beam multiple exposure technique that can generate complex 2-D and 3-D bandgap template structures in a photosensitive material. The system parameters related to the planar interference pattern produced by the two laser beams and reorientation effect of the sample relative to these planes is discussed. Photonic crystal structures such as the 2-D square and hexagonal arrays of dielectric "rods" and "holes" and the 3-D Yablonovite and other profiles are given. We show band gap calculations for these structures and discuss a technique for increasing the band gap through laser sculpting the dielectric template profile. In addition the paper discusses how the dual beam multiple exposure technique can be used to design 2-D quasicrystal structures of low to high rotational symmetry.

A study of the modal reflection, transmission and radiation properties of deeply etched second-order Bragg waveguide gratings in silicon-on-insulator (SOI) slab waveguide is presented. Both the eigenmode expansion method and the finite-difference time-domain method have been used to model this structure. High reflection can be obtained with a very short grating structure. The out-of-plane coupling efficiency and directionality can be controlled by the grating depth and groove shape. This structure has potential for use in micro-cavity laser, compact outcouplers and incouplers, and surface-emitting lasers. We propose an application example which involves a double-grating coupler to realize efficient coupling between vertically integrated SOI waveguides over short distances. Experimental results will be presented.

An aspheric collimating slab waveguide lens is designed using a diverging planar 1-D photonic crystal. An approximation method for analysis of such structures has been developed. A lens design procedure (which minimizes area) is also introduced. For illustration purposes, we use Silicon on insulator technology with the minimum feature size of 100 nm. We show that fast lens with 130 μm focal length, f/# = 1.3 is achievable with an etching area of only 658 μm².

A scalar diffraction model was developed that simulates throughput and phase information vs. wavelength in MEMS based multi-channel gain equalization applications. Group delay and chromatic dispersion are determined and compared to that measured on a commercial device. Basic features and magnitudes of the model and device data are found to correlate extremely well enabling device performance to be better predicted during the MEMS modeling phase of new product development.

This paper explores a novel way of sensing the angular velocity rate change of MEMS based gyroscopes with optical methods. In the capacitive sensing mechanism, displacement due to angular rate induces differential capacitance, requiring many supporting filtering and amplification sub-systems and circuits in the device to eliminate the noise. One of the most important aspects of using optical sensing elements with MEMS technology is the elimination of capacitive sensing used in conventional MEMS based gyroscopes. Conventional bulky optical gyroscopes based on Sagnac effect, with very small drift-rate parameters, is difficult to be implemented with MEMS technology due to the requirement of large area enclosed by the laser ring and fiber optic paths. Light interference effect of optical properties can permit the phase shift measurement of the light wavelength. In this paper, the combination of optical and mechanical systems is studied for estimating angular acceleration rate by Coriolis Effect.

Micro-electro-mechanical-systems (MEMS) offer many advantages for sensing a variety of physical parameters such as accceleration, pressure and temperature. Their small size allows them to operate in close proximity where conventional sensors cannot be introduced especially for thermal measurements. Temperature measurement and control is of fundamental importance to the optimal operating conditions of materials and machinery such as gas turbine engines, space exploration, etc. The temperature characterization will allow proper diagnosis of operating conditions and hence the optimization of controls and environment in order to augment performance and useful lifetime. MEMS based thermal measurements will be very useful as they are sensitive to small fluctuations in the operating conditions. Here, this paper proposes a novel MEMS based bimorph optical device as a thermal sensor. The paper includes the theoretical and experimental analysis on the thermal behavior of optical MEMS devices under different geometrical and parametric conditions. The paper also presents the static and dynamic behavior of optical MEMS based devices under different thermal environments. The results obtained verify the validity of the proposed designs for thermal sensing.

High bit-rate laser communications have been increasingly studied for applications ranging from short-distance transmissions to inter-spacecraft links. Optical communications involving moving parties require precise beam pointing and mutual tracking of communicating transceivers. Current approaches based on electro-mechanical beam steering are limited by the need for large volumes of beam-addressing computing and difficulties in providing automatic tracking/pointing capabilities to compensate for rapid changes in directivity patterns, transmitters’ relative misplacement and jitter [1]. An all-optical adaptive beam-tracking approach, proposed by some of the authors earlier, is based on the double phase conjugation effect (DPC) [2]. No mechanical steering, positioning or addressing computing are needed for fine tracking in such a bi-directional optical link. The approach efficiency strongly depends on non-linear properties of the used optical materials, which have been thoroughly studied [3]. This paper presents the results of theoretical analysis and further experimental studies of the DPC all-optical tracking technology. In the experiment, two optical terminals were linked with a modulated laser signal at a telecommunication wavelength. A DPC-mirror was a multi-layer liquid-crystal stack with a giant optical nonlinearity. The tracking and communication capabilities were simultaneously demonstrated in a range of angles, transmission rates and laser power levels. The experiment was in good agreement with the theoretical model. REFERENCES 1. E Lerner, Laser Focus World 36 11 2000 2. A Dudelzak, A Kuzhelev, A Novikov, G Pasmanik, Patent Application 12346-US-Prov 2002 3. A Kuzhelev, A Dudelzak, J Opt A: Pure and App Opt 5 L5 2003

Continues-wave photomixing phenomenon in ultra-fast photoconductors and high-temperature superconductors (HTS) is studied and photomixing efficiencies of these materials are investigated. Photocurrent distributions in both photoconductor and superconductor based photomixers are calculated and their common characteristics are compared in detail.

Terahertz emission from n-type (100), (110) and (111) InAs crystals have been measured as a function of the sample orientation. Emission was excited using 120 fs Ti:Sapphire laser pulses at an incident angle of 45° with fluences of approximately 1-2mJ/cm². The data is shown to match the behavior expected for optical rectification at the surface, with small contributions from bulk optical rectification and photo-carrier diffusion. Thus, at fluences employed in the present study, it appears that the dominant mechanism for generating THz radiation is optical rectification at the surface.

The dynamical behavior of the nonlinear optical loop mirror (NOLM) with feedback and low birefringence twisted fiber in the loop is examined. It is found that the output of the NOLM with feedback depends on many parameters, including the fiber beat length, the polarization state of the counter-propagating beams in the loop, as well as the length, twist rate, and nonlinearities of the loop fiber. The placement of a quarter-wave plate (QWP) asymmetrically in the loop allows for the tuning of the bistable and chaotic output from the optical resonator. As well, the output polarization state of the NOLM with feedback is shown to rely on the QWP angle as well as the input power, which is of importance when using the NOLM with feedback in optical systems that have polarization sensitivity. As all fibers exhibit some degree of twist and birefringence, the addition of a QWP in the NOLM with feedback allows for an easy and practical measure of control of the bistable and chaotic regions of the nonlinear optical resonator, which is important when implementing the device in an optical system.

This paper reports, on the use of a tapered single mode fiber as a sensing element for the detection of acoustic emission (AE) and ultrasound. When an acoustic wave impinges on the mode-coupling region of a fiber, the coupling coefficient is modulated via the photo-elastic effect. Therefore, the transfer function of the fiber is modulated by an acoustic wave. The sensitivity of the sensor at 156 kHz was approximately 1.2 V/mbar. Because of the resonant condition of the coupled-mode theory, this sensor is almost immunity of the environment perturbations, and the output is a DC signal instead of AC, so it is very suitable for those situations such as faulty protection which need very fast responsibilities.

This paper describes work investigating the impact of lattice defects on the attenuation of optical signals at wavelengths around 1550nm in silicon rib waveguides. Using Fourier transform infrared spectroscopy it is shown that high energy proton irradiation of silicon induces excess optical absorption peaked at a wavelength of 1800nm, but extending below 1600nm. This absorption is related to the introduction of silicon divacancy defects. It is further demonstrated that silicon divacancy concentration is accurately determined for a range of proton doses using positron annihilation spectroscopy and successfully predicted using an analytical expression proposed previously. Low loss rib waveguides were fabricated in silicon-on-insulator substrates. These waveguides were subsequently implanted with silicon ions at an energy of 2.8MeV through photolithographically defined mask windows of various lengths. The additional optical loss as a result of the defects introduced by the implantation process was accurately determined. For a dose of 2.5x10¹⁴cm^-2, the loss is greater than 500dBcm^-1. Finally, it is shown that excess absorption can be predicted using the same analytical expression for the determination of vacancy concentration, thus providing a straightforward method for the design of integrated, on-chip optical absorbers in silicon photonic circuits.

Optical bandpass filters with ripple-free spectral response are highly desirable for dense wavelength-division-multiplexed (DWDM) systems. We study and analyze the transmission characteristics of an optical bandpass filter, based on a three-mirror Gires-Tournois resonator (GTR) in a Michelson interferometer (MI). The three-mirror GTR (R₁, R₂, R₃) is actually an all-pass filter with the reflectivity of the end mirror R₃ being unity. In this paper, we present an analytical expression for the optimum design which has a ripple-free spectral response. The results show that flattop spectra can be obtained by suitably choosing the reflectivities R₁, R₂ of the two mirrors. Whereas a two-mirror GTR-based bandpass filter has only one unique optimum reflectivity, a three-mirror GTR-based filter has many sets of optimum values, making it easier to be designed to give much better performance. The effects of the reflective coefficients of the mirrors on the optical performance were also discussed.

In wavelength division multiplexing (WDM) systems, such as the fiber-to-the-home (FTTH) service (e.g., wavelength 1.31/1.55/1.49 mm in the EPON and wavelength 1.31/1.545/1.5x mm in the APON), coarse and dense WDM are two key technologies. In general, it takes two steps to accomplish this mixing (coarse/dense) WDM functions. In order to make the WDM device compact and low loss, it is necessary to integrate these two WDM functions into one step. In this paper, we propose two typical designs of the coarse/dense multiplexers, which multiplex the coarse/dense-WDM (C/D-WDM) signals simultaneously. We believe that, although their operation principle is simple, the proposed structures simplify the design of integrated photonic devices and will be needed in future WDM systems, specially for the WDM filter in the FTTH service. The proposed devices consist of a grating-assisted symmetrical or asymmetrical Mach-Zehnder interferometer with the wideband coupler (WBC) and/or the 3dB coupler. Depending on different operation principles, the WBC coupler operates as a 3dB coupler in two wavelength bands or operates as a 3dB coupler in one wavelength band and a cross-state coupler in another one. In this paper, firstly, the operation principles of the proposed C/D-WDM Multiplexers are demonstrated though an example of multiplexing three signals (l1/l2/l3) in two wavelength bands, where l1 (e. g., 1.31mm) belongs to the first band and l2/l3 (e. g., 1.55/1.5x mm) belong to the second band. Then, the design procedure and simulation results of those C/D-WDM multiplexers and related components such as the WBC coupler and planar gratings, which are based on the silica material, are given. Finally, some related issues such as polarizatin dependence and material dispersion are discussed.

Porous thin films of TiO₂ exhibit interesting and useful optical properties when the glancing angle deposition (GLAD) technique is used to impart controlled structural variations on the nanometer scale. Specifically, helically structured thin films possess optical properties sensitive to the polarization state of incoming light, including selective reflection of circular polarizations and optical rotation of the vibration ellipse of light as it passes through the film. By adjusting the deposition parameters, the helical structures can be transformed into vertically aligned columns with nanometer diameter variations. These films possess a continuously varying refractive index along the substrate normal. This index profile can be tailored so that it varies sinusoidally along the substrate normal to form a rugate interference filter. With the addition of a constant index layer of thickness equal to the sine period located in the center of the film, a narrow bandpass appears within the filter’s larger reflectance band.

Resonant reflection filters -- also known as grating waveguide structures -- are characterised by a multilayer configuration including a substrate, waveguide layer and grating(s) at the top of and, in this investigation, also under the waveguide layer. For a specific wavelength at a specific angular and polarisation orientation an incident beam is partly diffracted, guided and rediffracted, leading to vanishing transmission due to destructive interference with the directly transmitted beam, while most of the light is reflected. Since this resonance is a guided mode phenomenon these devices can be used as tunable filters or dichroic elements (reflected wavelength as a function of incident angle) as long as the guided mode condition holds. In this experimental study the behaviour of ultrashort pulses of ~100 fs within structures with various grating depths and, therefore, different spectral resonance bandwidths was investigated under resonance conditions. Spectral and time-resolved measurements in transmission as well as reflection geometry revealed that the ultrashort pulses leaving the structures are time-bandwidth limited, i.e. the spectral bandwidth of the resonant filter determines the pulse length. Group velocity dispersion (GVD) has no important influence since the light is immediately rediffracted after having been coupled into the waveguide layer of the sample.

The increase of DWDM wavelength channel count has resulted in significant increases in power applied in telecommunication optical fibers. As higher powers and longer distances are required, there is a need for stronger fiber amplifiers. Here we present passive components for controlling the optical power in high power networks. High power amplifiers are used to increase the signal power injected into the input port. The output signal is a function of the input signal (including spectral and temporal shape), as well as the operating conditions of the amplifying unit. When the input signal is composed of many wavelength channels, each having its own power level, power management (or equalization) is required at the input port. When the channels are equalized, amplifiers are exploited most efficiently. Power unbalance due to non-equalized amplification of amplifiers causes degradations of signal to noise ratio. Output ports of fiber amplifiers are sometimes susceptible to high power spikes. This can occur, for example, when the input to an operating amplifier is suddenly turned on. These spikes can be destructive to components and equipment on the network. Moreover, phenomena known as the ‘fiber fuse’ can destroy the fibers in a network due to over-power. We introduce novel passive optical components, which can control and regulate the optical powers in the input to and at the output from a fiber amplifier. These components include power limiters that protect against excessive optical power transmission, and power equalizers. They can either replace or complement existing active power feedback control loops. Power control components for high power networks will be described with emphasis on both their properties and their application in the network.

Diffraction gratings are often used in fiber-optic telecommunication, especially as wavelength (de)multiplexers and three-dimensional micro-optical coupling devices. The use of diffraction gratings made out of anisotropic materials is particularly attractive when trying to design high efficiency, polarization insensitive components. We show that by incorporating perpendicular grating structures of subwavelength dimensions, it is possible to enhance or reduce the polarization dependence of a traditional grating structure. This three-dimensional, form birefringent structure is analyzed with the use of rigorous coupled wave analysis (RCWA) and effective medium theory (EMT). The application of other analytical techniques, such as the coordinate transformation method and the classical differential method, are discussed, in the context of this particular type of grating structure. We present grating structures in which it is possible to enhance or reduce the polarization dependence of the diffraction efficiency, in a given wavelength range.

The traditional output waveform of a pulsed laser is a Gassian beam both in temporal domain and in radial distribution. However, in application, many users need more uniform laser pulse in energy distribution. We propose the capability of square wave output of solid-state laser in temporal domain using the smooth cavity dumping, the reduction of output loss compensating the photon decrease in cavity. In experiment, we used two sets of avalanche transistor chains to control the functional voltage applied on the Pockels cell for getting the fundamental wave pulse-width variable range from 15ns to 140ns, obtained the laser energy output 1000mJ/pulse, and used the square wave pulse-width to pump the nonlinear optical crystal LBO that is obtained the energy output 300mJ/p of second harmonic generation (SHG) and 160mJ/p of third harmonic generation (THG) for 1064nm and 532nm difference frequency in BBO. Comparing with the traditional Gaussian pule, the pulsewidth variable square wave laser has some advantages in waveform, polarization, energy uniformity, so it can be used in high power, high conversion efficiency lasers without damaging the rods, mirrors and crystals.

The most commonly-used technique for measuring the refractive index profile of an optical fiber is the refracted near-field method. This standard method cannot be directly used for integrated optical waveguides such as silica-on-silicon or LiNbO₃ because of the geometrical constraints imposed by the slab waveguide. A modified method was described in previous work and subsequently implemented with some improvements (e.g. use of a calibrated solid refractive-index reference element; a simplified waveguide identification) in a commercial apparatus. However, the non-availability of suitable index-matching liquids having an index of refraction greater than about 1.8 prevent this apparatus being used with high-index DUTs. In this paper, we propose and experimentally verify a modified instrument that permits the characterization of the index profile of high refractive-index waveguides such as LiNbO₃. Provided that the waveguide is written in a homogeneous bulk substrate with a known index, this modified approach allows for spatial and refractive-index resolutions that are practically as good as those obtained with the standard technique applied to optical fibers.

Future sensing technologies are needed to provide higher accuracy, lower power consumption and occupy small real estate within munitions. The novel ideas being supported at the Army Research Development Engineering Center (ARDEC) at Dover, New Jersey, uses principles of electromagnetic propagation and the properties of waveguide cavities with various geometries to develop a new class of sensors for onboard direct measurement of the angular orientation and position of objects in flight and applications such as mobile robotic platforms. Currently available sensors for munitions are based on inertia, optics or heat. Inertia based sensing generally suffers from drift, noise and the currently available sensors cannot survive high firing accelerations while maintaining the required measurement sensitivity. Optical technologies generally have short range and require line-of-site. The sensing technologies presented in this paper employ radio frequency, make direct measurement of position and orientation, and do not require added information for their operation. The presented sensors employ waveguide cavities that are embedded into the structure of munitions. It is shown that the geometry of the waveguide cavity can be designed to achieve high angular orientation sensitivity with respect to a reference, polarized electromagnetic field. In this paper, the theoretical fundamentals describing the operation of the developed sensors are described. Studies of the interaction of the polarized signals with various waveguides and cavity geometries are presented. Simulations results as well as experimental results validating the theoretical and the simulation results are provided. The simulation and experimental results clearly demonstrate the potentials of the developed position and angular orientation sensors in general, and to munitions in particular.

A simple numerical method is developed to analyze changes in intrinsic birefringence of ridge waveguides with waveguide dimensions and UV irradiation. Experimentally, Bragg gratings were written on different core size ridge waveguides using the phase mask technique and ArF laser irradiation. By monitoring the shifts in Bragg wavelength with UV irradiation, the variation of the waveguide birefringence with waveguide dimension and UV processing is observed and quantified. The mechanism of the waveguide birefringence controlled with UV irradiation is verified both in the theoretical analysis and experiment.

1x2 and 2x2 optical waveguide directional couplers were fabricated in a silica glass sample by a femtosecond laser and characterized at the wavelength of 1.5 μm. The coupling coefficient was determined by systematically writing series of 1x2 directional couplers. After the process and design optimization, the two types of waveguide directional couplers successfully achieved the coupling ratio of about 3.0 dB with the average IL and PDL 6.55 ± 0.25 dB and 0.28 ± 0.12 dB respectively. The multi-scan writing technique with inter-scan shifts was used to reshape the magnitude and profile of the refractive index change so that excess loss of S bends and the PDL for such couplers were substantially reduced.

This paper presents a simulation study of a true-time-delay system for wideband phased array antenna employing a waveguide Bragg grating prism. The Bragg grating prism is constructed of chirped Bragg gratings strategically positioned in order to achieve beam steering capabilities. The effects of double- and single-sideband modulation on the performance of the system are studied. Finally, the effects of apodization of the gratings are also discussed.

A model for a multiple quantum well electroabsorption modulator is developed based on measurements of the escape time of photogenerated carriers, the dependence of the fiber-to-fiber loss on the applied voltage, the dependence of the α-parameter of the modulated signal on the applied voltage, and the intensity modulation frequency response. The accuracy and computational efficiency of the model in describing the intensity and phase modulation properties of optical signal make it suitable for computer simulations aimed at system design and performance evaluation.

We are reporting the investigation on the degradation of heterojunction laser diodes as they were subjected to electron beam irradiation. The research was done under the European Union’s Fusion programme, and targets the possible use of semiconductor lasers for remote sensing and robotics, under irradiation conditions. A total irradiation dose of 90 Mrad was achieved at room temperature, and the irradiation geometry was an axial one. The measurements were performed off-line. An automatic measuring set-up was developed including various instruments: a laser diode driver, a laser power meter, a wavelength meter, a fiber optic spectrometer and a laser beam analyzer. After each irradiation step, the following characteristics were monitored, as function of the driving current and laser case temperature: the emitted optical power, the wavelength of the emitted radiation, the embedded photodiode current, the longitudinal and transversal mode structure, as well as the temporal behavior of all these parameters. For each irradiation dose, the laser diode serial resistance, threshold current, and quantum efficiency, and the photodiode responsivity were plotted for different operating conditions (extended graphical information is provided in the paper). The following changes were noticed: an increase by 6 % for the threshold current, a drop by 7 % for the quantum efficiency, a decrease by 9 % for the photodiode responsivity, and a slight modification of the wavelength of the emitted radiation.

A tunable multiwavelength laser source can find several applications in spectroscopy, wavelength-division multiplexing and optical metrology. We developed a fabrication technique for holographic gratings designed to allow simultaneous lasing at two wavelengths in Littrow configuration. The design of these gratings is based on recording a constant groove spacing grating and a variable groove spacing grating on the same photoresist. Using the gratings as coupler in the external cavity of a semiconductor laser, we obtained simultaneous laser emission at two wavelengths. The gain medium we used was a commercial laser diode emitting at 675 nm. By translating the grating, we demonstrated tuning of the spectral separation of the dual-wavelength output from 0.8 nm to 6.2 nm. The fabrication technique of holographic gratings and the results obtained with the laser diode will be presented.

In this research a novel modulation technique for free-space laser communication system called Intensity Position Modulation (IPM) is carried out. According to TEM₀₀ mode of a laser beam and by linear fitting on the Gaussian function as an approximation, the variation of linear part on the reverse biased pn photodiode produced alternating currents which contain the information. Here, no characteristic property of the beam as intensity or frequency is changed and only the beam position moves laterally. We demonstrated that in this method no bandwidth is required, so it is possible to reduce the background radiation noise by narrowband filtering of the carrier. The fidelity of the analog voice communication system which is made upon the IPM is satisfactory and we are able to transmit the audio signals up to 1Km.

In this work we investigate the use of low-frequency noise (LFN) spectroscopy as a sensitive tool for investigation of semiconductor avalanche photodiode (APD) quality and reliability for high-speed communication applications. Samples from several manufacturing runs pre-screened through the standard production batch validation showed very low start of life low-frequency noise levels. The LFN test indicates the high quality of the devices with respect to noise characteristics. We also investigated the noise characteristics of production reject devices, some of which exhibit a 1/f type noise peak at the guard ring punch-through voltages, and an increase in noise intensity at punch-through voltages after long term accelerated lifetesting (>1000 h at 200°C and 100 μA). Useful information on device quality can thus be obtained from the noise measurement results performed in the deep breakdown bias range (27-30 V), where a sharp peak of the Lorentzian type noise may be observed. This feature is believed to be due to intensive recombination processes at defects in interlayer regions. It is also shown that avalanche photodiodes containing some defects exhibit not only increased dark current and low-frequency noise level, but also increased multiplication excess noise factor which decreases with increasing input light intensity. This work shows that LFN spectroscopy is very useful for the localization of noise sources, and provides important information for further product quality improvement.

A method of Bragg gratings written in silica-on-silicon planar waveguides to be used to monitor the overall uniformity of the waveguides and grating processing is presented. By measuring the shift of Bragg wavelength with UV exposure time, the initial effective index n_Oeff and birefringence B₀ of the planar waveguides are measured accurately. With one phase mask, Bragg gratings induced on different waveguides with widths that vary from 4.6 to 8.8 μm, result in variations of N_Oeff and β_O of 1.5 x 10^-3/μm and 1 x 10^-4/μm, respectively. The result is used as a way of improving control over the waveguide dimensions obtained from the photolithographic and RIE processes, and optimizing the design of ridge waveguide structures to compensate the waveguide birefringence. This will improve the quality of the PLCs that include symmetric Bragg grating structures: MZI-OADM etc. By writing Bragg gratings on the linear taper planar waveguide, a chirped grating response is realized.

We investigate finite element modeling of MEMS micromirrors actuated electrostatically by means of tools available in ANSYS finite element modeling software, and compare numerical results with analytical solutions for the static analysis of MEMS micromirrors. MEMS micromirrors must be accurately modeled in order to achieve precise optical positioning. Analysis of MEMS micromirrors leads to the study of structural and electrostatic fields. Finite Element (FE) method is an effective technique to model structural and electrostatic fields. The FE analysis of these coupled fields is accomplished by several tools in ANSYS. This paper models torsional and flexural-torsional micromirrors by different methods in ANSYS. These methods include: (a) a sequential coupled electrostatic and structural field tool; (b) a directly coupled electrostatic and structural field tool employing one-dimensional (1D) transducer element; and (c) a directly coupled electrostatic and structural field tool utilizing a 2-D or 3-D reduced order model. The torsional micromirror is of 1000 by 250 microns square, and the flexural-torsional micromirror is of 100 by 100 microns square. The numerical results are compared with analytical solutions. Comparisons show advantages and disadvantages of these tools for MEMS micromirror modeling. These comparisons allow a selection to be made of the most suitable tool for a given modeling task and assess the accuracy of analytical solutions.

Recently, we have demonstrated a new approach to fabricate a variable focal lens using polymer-stabilized liquid crystals [1, 2]. The approach is based on curing of a polymer/liquid crystals mixture with a circularly symmetric (e.g. Gaussian) shaped laser beam to induce spatially inhomogeneous polymer network. Applying a uniform voltage to the non-pixilated cell leads to circular-symmetric (lens-like) distribution of refractive index in the cell with plane parallel substrates. In this paper we study and optimize the electro-optical characteristics of such lens by varying the wavelength of the polymerizing laser, temperature regime of the process of polymerization as well as frequency of the lens driving voltage. Obtained results are applied to develop lenses that have no moving components and allow the electro-optical zooming. REFERNCES [1] V.V.Presnyakov, T.V.Galstian, K.E.Asatryan, A.Tork. "Polymer-Stabilized Liquid Crystal for Tunable Microlens Applications", Optics Express, 10, 17, 865-870, 2002. [2] V.V.Presnyakov, T.V.Galstian. "Variable Focal Length Lens Based on Polymer-Stabilized Nematic Liquid Crystals", 19th International Liquid Crystal Conference, Edinburgh, UK, June 30 – July 5, 2002, Book of Abstracts, P754.

Our developed thermal curing adhesives were reported having excellent performance in coupling optical fibers to waveguides. The fiber-to-waveguide coupling process based on these adhesives is reported in this paper. The process consisted of three major steps in the process, including loading waveguide dies and fiber arrays onto the sample holders, aligning the fibers to waveguides at a constant temperature to reach minimum loss, bonding the fiber arrays to waveguide dies. The sample holders, which used ceramic spaces to isolate heat and springs to damp stress, were specially designed to keep fiber arrays and waveguide dies at a constant temperature up to 120°C with minimum shift. When the fiber arrays and waveguide dies were equilibrated with the set temperature, a rough alignment was conducted manually, followed by an automatic alignment controlled by a Melles Griot system. Then, the adhesives with proper viscosity and curing rate were applied to the gaps between fiber array and waveguide dies to bond them together. The curing temperature was optimized so that the adhesives could be distributed rapidly and cured at a speed that still allowed a small alignment adjustment during the curing. Such a temperature optimization was achieved by studying the adhesives’ curing kinetics with a DSC.

The bulking or stressing of fiber ribbons in the packaged waveguide-based components is associated with the performance deterioration of the components. The current industrial practice of avoiding the problem is to keep fiber ribbons movable regarding the packaging house. This approach, however, makes the component vulnerable to external load during component handling and does not complied with Telcordia test standard. A special technology based on incorporating a soft gasket was developed in our laboratory to solve the problem. The gasket is made of low-modulus elastomer foam with certain thickness and is positioned between the packaging house and strain relief boots, on which fiber ribbons are bonded with an in-house developed epoxy adhesive that has passed Telcordia test. In the packaged components, any effect caused by the mismatched coefficient of thermal expansion between the packaging house and fiber ribbons are compensated by the gasket, and no bulking or stressing occurs in the fiber ribbons. Meanwhile, since the fiber ribbons are firmly bonded to the strain relief boots, any external force applied on the fiber ribbons is transferred to the packaging house, instead of the fiber arrays and waveguide dies. The packaged component with this technology meets the Telcordia test standard and is cost-effective.

In this work we discuss a method of smooth waveguide fabrication by wet etching, using a novel mixture of organic and inorganic acids. We fabricated 4.5μm-deep large single mode waveguides with walls and edges roughness comparable with the roughness of non-etched regions. Surface and wall roughness as well as etching rates and propagation losses are presented and compared to characteristics achieved by traditional wet etching recipes. The method was developed as part of the effort to implement fast and widely tunable AWG-based wavelength filter with channel spacing suitable for dense WDM networks.