A new compound Bragg grating structure with periodic nonlinearity and gain is presented. It works as a DFB laser structure that could be switched between two resonance frequencies. A mathematical model has been developed and used to investigate the structure. The laser can be switched between two resonance frequencies by a triggering optical pulse. A numerical simulation of the switching dynamics in the time domain is described. At one of the frequencies, the structure is found to have a self-pulsing behavior.

We report what is, to our knowledge, the first demonstration of strong Bragg reflection gratings operating at ~1550 nm written using standard phase mask techniques in a novel multimode fiber. A reflectivity of >98% and a bandwidth of <0.5 nm in very good agreement with modeling predictions have been observed.

We present a new architecture for a programmable disperions matrix for optical beamforming. A 2-channel prototype is constructed to demonstrate beam steering for angles of ±70° in the receive/transmit mode in RF0.1-1 GHz frequency range.

We present a tunable second-order Polarization Mode Dispersion (SOPMD) module with constant mean first-order differential group delay (DGD). The module consists of a variable DGD line (variable from τ^min_variable to τ^max_variable), a polarization controller, and a high birefringent (HiBi) fiber segment with a DGD value of τ_fixed. A Muller transitional matrix has been employed to describe the system. Theoretical calculations predict that after careful calibrations, by tuning the variable DGD line, the module can provide SOPMD values variable from Φ₀ to 1/2τ²_fixed+Φ₀(Φ₀ is a constant), while its mean first-order DGD is kept at a constant value of about τ_fixed. We implemented two of such modules, e.g. module No. 1 and No.2, which have different tuning range of SOPMD and different mean DGD values. For module No. 1, a variable DGD line (variable from 0.68 to 15.00 ps) and a HiBi fiber with a DGD value of about 10.5 ps were used. Module No. 1 demonstrates variable SOPMD values from about 37.0 to 522.0 ps² with an accuracy of about ±1.0 ps. Module No. 2 demonstrates variable SOPMD values from about 10.0 to 63.0 ps² with an accuray of about ± 5.0 ps², while its mean first-order DGD is kept at a ocnstant value of about 10.5 ± 1.0ps. The implications of this tunable Second-order PMD Module on high-speed (≥ 10 Gbit/s) optical communications are discussed.

A new robust approach is presented for achieving very large fiber port count all-optical crossconnect switches. This three dimensional optics-based switch has built-in alignment capabilities with fault-tolerance, allowing graceful port count scaling.

We present an analytical study of the output intensity fluctuations caused by the polarization mode dispersion (PMD) and the polarization dependent loss (PDL). An analytical expression for the Jones matrix autocorrelation function (ACF) in presence of the PMD and PDL is derived. The analytical ACF can be applied to different modulation formats leading to the analytical evaluation of the eye diagram.

In this paper, we review first, the different noise terms involved in amplified and filtered spontaneous emission detection using the standard beating theory (or the semi-classical field expansion theory) of the noise in optical amplifiers. We derive BER calculations using the following approaches: 1) non-ideal electrical and optical filters combined with a Gaussian approximation for the resultant noise process and an ideal integrate and dump electrical filter and an ASE source combined with non-Gaussian Erlang noise statistics.

We report an extremely flat (better than 3-dB), ultra-wide band (over 160-nm) supercontinuum source for use ni the C and L wavelength bands. The supercontinuum is generated using approximately 2.2 Km of commercially available single mode non-zero dispersion shifted fiber by exploiting soliton decomposition and Raman self-frequency shift nonlinear effects. A supercontinuum spectral flatness of better than 1-dB over a 115-nm wavelength range is achieved.

No abstract available.

The resolution of chromatic dispersion (CD) measurement using the phase shift technique is limited due to the light modulation. It is the most severe if CD varies significantly within the bandwidth of the modulation. A simple experimental relationship between the transmission coefficient and its CD (measured via phase shift technique) is shown for different optical fitlers within their reflection band.

We propose a novel architecture for an optical programmable dispersion matrix (PDM) able to process simultaneous independent RF beams. The proposal is demonstrated by processing the transmission of two independent RF signals in two-channels with a resolution of 3-bit. The optical beamformer processes two independent RF-beams, for eight different angular directions, and it is based on a binary array of three delay lines. Each delay line is composed of four fiber Bragg gratings whose center wavelengths are channels 30 to 33 of the ITU grid. Beampatterns are characterized in transmit mode for two simultaneous RF beams in 0.6-1.5 GHz frequency range and azimuth angles from 0° to 70°.

An all-digital laser beam profiler instrument is proposed and demonstrated that for the first time can provide real-time and full-repeatability multi-technique profiling with high power (e.g.,>10kW/cm²) laser beams.

A high gain, high speed single-avalanche-stage APD based on heterojunctions is proposed.

In this work, we study a high efficient four-junction solar cells using AlAsSb(1.91eV), AlInAs(1.46eV), ALGaInAs(1.0eV) and InGaAs(0.75eV) materials. The >35% efficiency is calculated under AM0, and the thickness of each layer is adjusted to equalize the short circuit current. We also experimentally fabricated a 1.0eV cell and studied its characteristics.

A new optical detection system for microfluidic lab-on-a-chip applications is described. The photomultiplier tube (PMT) system is controlled by LabVIEW and has output/primary current gains that are programmable from 10⁴ to 10⁷. Light is delivered to and from microfluidic systems by a custom launch-and-detect fiber probe fabricated from one-millimeter plastic optical fibers. The noise characteristics are descrbied and the detection of fluorescent 15μm polystyrene spheres is demonstrated. Using the measured static spatial response of the tip, the velocities of moving microparticles can be calculated from dynamic measurements of their fluorescence.

A novel and low-cost time-of-flight system is presented. The system, which is based on a vertical cavity surface emitting laser (VCSEL) array as the light source, images the near field of all the VCSEL beams into the measurement volume. Each VCSEL provides a light sheet or a light spot in the measurement volume, and thus adds to the total optical power present in the measurement volume. The multiple-beam time-of-flight system provides operation at a lower signal-to-noise ratio and determinant of target-motion direction with a single photodetector. The concept provides a simple and compact sensor for surface velocity and fluid flow measurments, where the calibration of the system is almost temperature independent. This paper will demonstrate the concept and quantify the system tolerance with respect to temperature changes.

Quantum mechanics can revolutionize the art of code-making by breaking standard encryption schemes and by making unbreakable codes. In this document, I survey the motivation, power and limitation of quantum code-making.

We propose an interferometric fiber-optic sensor for absolute measurements of temperature, hydrostatic pressure, and elognation. The sensor is based on single-mode holey fiber with hexagonal symmetry and uses a spectral decoding method to directly measure the changes in optical group imbalance induced by the measurand. We also determined the susceptibility of the group refractive index to temperature, elogation and pressure, which equals respectively to dN/dT=+1.28×10^-5 1/K, dN/dε=-0.40×10^-7 1/strain, and dN/dp=-1.0×10^-5 1/MPa.

The principles of the method of single expression (MSE) for boundary problems solution in classical electrodynamics are presented. In the MSE the solution of the Helmholtz's equation is presented in the special form of a single expression describing resultant amplitude and phase distributions in the medium. This form of solution presenation permits to pass over the restrictions of the superposition principle and to solve both linear and nonlinear problems with ths same ease. In the MSE the Helmholtz's equation is reformulated to the set of first order differential equations and the boundary problem is solved numerically. No approximations are implied either in Helmholtz's equation or in boundary conditions. Using the MSE steady-state boundary problems are modeled for wavelength scale multilayer and modulated 1D photonic structures including amplification and nonuniformity evoked by intense electromagnetic field.

Quantum cryptography, or more specifically, quantum key distribution (QKD), has attracted a lot of attention in the recent years with the discovery that it can provide absolute secrecy for communications. We propose a new architecture for implementing a fiber-based network of quantum key distribution using optical wavelength division multiplexing in the fiber. We discuss the advantages over previous proposals and we report experimental work demonstrating the feasibility of the proposed architecture.

A simple and handy technique to produce optical fiber components such as Bragg gratings has been used. A CO₂ laser beam, focused on a single mode fiber, can change its optical properties locally, allowing therefore the creation of a Bragg grating inside the fiber itself. In the present study we investigate the effect of this laser beam on the fiber and show how optical fiber components can be produced. When exposing a single-mode optical fiber to a high-power CO₂ laser beam a small device is developed insdie this fiber. Inside this micro-structure, the optical power is exchanged between the core and the cladding modes. This device may exhibit either a very selective rejection behavior similar to Bragg grating or an oscillating behavior like tapered fiber. The coupled mode theory combined with a mathmatical algorithm can be used to study the propagation of the modes involved in the fiber.

A high power UV LED array operating at 396nm with an output greater than 1 Watt has been developed. Performance characteristics of the device are presented. It is also shown that the device is as effective as traditional arc lamps in curing acrylic adhesives as demonstrated through microhardness and DSC testing.

When we say that a process is "controlled" we mean that we know as much as it is possible to know about that process. When we say that a process has been "characterized", we mean that we know exactly how many critical steps there are, and that we know exactly what happens in each of those steps. Every step has a metric and a tolerance. When that metric is monitored and measured, preferably on a real-time basis, manufacturing management has the necessary data to control that process, and to reap the benefits that come with such control. Developing and installing the instrumentation needed to measure processes required the better part of the 20th Century. Today's process instrumentation can measure every conceivable physical parameter, record it, and communicate it anywhere in the world. This method does have its drawbacks: such instrumentation can be expensive, and often the parameters being measured are only indirectly related to the process. The benefits, however, of a stable, reliable process are worth the effort. The technology that supports instrumentation continually improves, as indeed, it has to if it is not to become obsolete in short order. For some time, however, the greatest improvement in the field has been in improving measurement of physical parameters based on physics, engineering and computer science. However better that focus becomes, process instrumentation development now has an alternative path. Although the end hardware, of course, is still physical, the path is based on a different model: biology, rather than physics. This new path emulates the neural network--that is, the brain. Using neural networks for process control is as old as processes themselves. Long before today's sophisticated instruments, a person would examine a process for some time, learn how it worked, and make a judgement about how well it was actually doing what it was supposed to do. Decisions about process performance were based on what someone observed, and recommendations for improvements were based on that person's experience. Instrumentation is now available that makes judgements in exactly the sme way a person would, but with the precision and tireless reliability of modern electronic equipment. The first of these instruments is called ZiCAM.

Low coherence interferometry (LCI) can be used to measure the profile of industrial products, the measured sample being scanned under the LCI probe. Axial distance measurements are made using the light reflected by the surface and collected on the same optical axis used for its illumination. Therefore, when the transverse resolution is not an issue, only a narrow laser beam is required and hard-to-reach surfaces can be probed with axial accuracy in the range of 1μm. With other techniques such as triangulation the surface requires to be visible from another point-of-view and complex shapes often become inaccessible. When the LCI instrument is assembled with optical fibers, delicate instrumentation may be kept away from harsh environments and only a single optical fiber needs to get close to the measurement location. However, optical fibers are particularly sensitive to temperature and a reference is required to compensate for path length drifts. Furthermore, industrial mechanical displacement systems typically induce positioning errors much larger than the LCI instrument accuracy. One approach to circumvent these problems consists in measuring the location of another surface close to the region of interest. Such a reference surface is not always available and typically requires a second probe. We found a more practical approach by using an optical quality window located over the sample surface and moving with the sample. The laser beam from the LCI instruments travels across the window just before it reaches the samples surface. The window surface induces a first reflection (4% of the incident power) and its distance is measured by the LCI as well as the sample surface distance. Since the location of the window is fixed relative to the sample while the entire surface is scanned, out-of-plane movement of the motorized slide is compensated. High-resolution measurements are obtained by simply subtracting the window plane from the sample surface.

No abstract available.

High-speed transmission optical devices and equipment often use a single-mode fiber as receiver or emitter. The fiber must be accurately aligned over five degrees of freedom with respect to an optical field. In this paper, we investigate a model-based approach which takes advantage of the properties of the optical coupling physical phenomenon. Using a Gaussian field approximation and an approximated solution of the overlap integral, two fundamental properties are first derived. Both are then validated by numerical simulations and experiments. A very simple alignment procedure is the proposed and tested experimentally using two single-mode fibers.

Laser diode intensity noise as a function of laser current and temperature is studied. It is shown that the intensity noise in current-temperature coordinates forms "islands" amid larger area of relatively low-noise operation. Although these "islands" of noise are showing some movement in time, generally it is possible to find an operational point characterized by long-term low noise. We show also that laser noise deteriorates signal-to-noise ratio (SNR) in photodiode circuit causing random fluctuations of SNR. The comparison of noise distributions for isolated laser module and the laser incorporated into Doppler velocimeter gauge allowed us to reveal design faults contributing to back reflection and enhancing the noise.

Mode-locked and relative intensity noise (RIN) of hybrid soliton pulse source (HSPS) utilizing Gaussian apodized fiber Bragg grating (with zero dc-index change) is described. The HSPS is modeled by a time-domain solution of the coupled-mode equations including spontaneous emission noise. RIN is calculated using numerical solutions of these equations. It is found that for these gratings soliton type pulse are not generated at the mode-locking frequency. However, these pulses can be generated only at limited tuning range with and without noise. RIN peak does not locate at the fundamental frequency that shows optical resonance due to cavity roundtrip time. Also, obtained soliton pulses are not directly related to magnitude of noise.

Gradient index (GRIN) lenses are used both in imaging and beam shaping optical systems. Usually, to design a GRIN lens system, we would find the equivalent lens effective focal length and the principal planes position. Then if we design an imaging system the Newton formula gives the image position or, if we design a beam shaping system the Kogelnik's ABCD law gives the Gaussian beam waist size and position. The Hamilton's phase plane approach gives the same result as these sets of formulas, but is more spectacular and clearer. Hamilton's phase plane approach gives a vivid and clear illustration of GRIN lens imaging and beam shaping properties and does not require substitution by an equivalent lens. Hamilton's phase plane approach allows easy derivation of transparent and simple direct formulas for a GRIN lens created image and beam transformation, including Gaussian beams and multimode fiber beams. These formulas and graphical illustrations facilitate solving, and give a clearer understanding of practical problems.

The effect of ruthenium doping on optical properites of Bi₁₂TiO₂₀ (BTO) single crystals is studied. A low concentration of ruthernium does not change absorption behavior of BTO, however higher ruthenium addition significantly shifts the optical absorption to the near infrared spectral region with an absorption shoulder appearing around 1.8 eV. The absorption coefficient increases with increasing the Ru concentration. Light-induced absorption changes are observed and the dynamics of build-up and dark decay processes are investigated. Results show that concentration of shallow levels increase with Ru content. Evolution of the build-up and dark decay of light induced absorption consists of fast and slow components: the first process occurs within few seconds, while the second one needs several hours (or days). It is established that light-induced absorption can be modified by preliminary illumination and thermal treatments thus multi-level lead to the transient and persistent parts of light-induced absorption. A model, assuming one deep and two shalow levels is proposed to explain the bi-exponential behavior of the build-up and the dark decay of light-induced absorption when the crystals are preliminary annealed in oxygen atmosphere.

This paper presents the realization and characerization of a matrix of DFB lasers realized by ion-exchange on Er/Yb codoped glass substrate. Output power characteristics such as power efficiency and Relative Intensity Noise are presented, wavelength characteristics including linewidth, wavelength stability versus temperature and emission wavelength calibration are also investigated. Concerning output power, a 4.5 mW fiber coupled emission was achieved and a low RIN falling below -150bB/Hz after 4 MHz was measured. The matrix contains 15 lasers emitting at 100 or 25 GHz spaced frequencies with a very narrow linewidth measured to be 3 kHz. Finally, the temperature dependency of the comb of wavelength is measured to be 1.3 GHz/°C.

We demonstrate a novel technique for controllable coupling to the second-order guided mode of an index guided photonic crystal fiber (PCF). Efficient coupling of light to the PCF depends on synthesizing the appropriate input field distribution at the focus of a coupling lens, corresopnding to the spatial Fourier transform of the phase-encoded incident light. A π-modulated binary grating encoded on a computer-programmable spatial light modulator generates the appropriate field distribution for contorllable coupling to the second-order guided mode of a commercially available index-guided PCF wiht a triangular lattice air-hole structure.

Cladding mode coupling loss below 0.1 dB for a 30 dB fiber Bragg grating are reported for a wide range of H₂ treatment pressure up to 1500 psi. Extension of the photosensitive region into the clading was used for reducing the cladding mode coupling. It is shown that to take proper advantage of this, one must use a wider laser scan beam. Otherwise, the cladding mode coupling loss may increase as the H₂ pressure treatment is raised. The benefit of a properly matched photosensitivity obtained by matching the Ge concentration in the core and cladding regions is also highlighted. The fiber was also designed to match a standard single-mode fiber in order to lower the average splicing loss below 0.03 dB and the attenuation to about 0.2 dB/km at 1550 nm.

We report the results of a numerical analysis of the pump absorption in a double-clad fiber amplifier based on a beam-propagation method. Various inner-cladding shapes are compared in terms of absorption efficiency. This incldues the circular shape iwth an offset core, the hexagonal and D-shaped fibers, a circular fiber with two truncations and a polarization-maintaining circular fiber.

We present a new idea for minimizing temperature sensitivity of a fiber-optic interferometric sensor based on highly birefringent side-hole fibers. The sensing part is composed of a side-hole fiber characterized by a very high pressure-to-temperature sensitivity ratio equal to 25°C/bar, which is about two orders of magnitude higher than in the case of other highly birefringent fibers. We took advantage of this unique property to assure good insensitivity to temperature effects, which are always associated with fast pressure changes. The presented sensor requires no additional temperature-compensating fiber, which has been a normal procedure until now for all fiber-optic pressure sensors. We used a digital demodulation system based on the coherence-addressing principle to decode a differential phase shift introduced by the sensing fiber and the quartz plates, which compensate an optical path delay of the fiber. The demodulation system allows unambiguous measurement of a phase shift with a resolution of 1/8 interference fringe in a range of 40 fringes. In this paper we also show that the side-hole fibers can be applied for measurements of high dynamic pressures at an operating range of 110 bar. We compared the dynamic characteristics of the fiber-optic sensor to the responses of a calibrated and temperature-compensated piezoelectric sensor with a sampling rate up to 200 kHz. The dynamic characteristics of the fiber-optic sensor are in good agreement with the reference piezoelectric sensor, which shows the great utility of the side-hole fibers in accurate measurements of high dynamic pressures.

Brillouin scattering spectra in a photonic crystal fiber with partially Ge-doped core is demonstrated, for the first time to our knowledge, using pump-probe technique. We have observed and characterized Brillouin spectra in the photonic crystal fiber at 1320 nm. One main peak and several sub-peaks are observed. Brillouin loss spectrum of main peak has a Lorentzian shape. The bandwidth Δν_B is 66 MHz and the Brillouin frequency shift ν_B is 12.054 GHz at room temperature. The peak originating from a higher order guided longitudinal acoustic mode in the graded-Ge-doped core of the photonic crystal fiber is observed for the first time. The temperature related behavior of the photonic crystal fiber is investigated. It is shown a linear dependence for the photonic crystal fiber.

Our optical waveguides are silica capillaries filled with xenon at atmospheric pressure. Outside the capillary helium gas "outer cladding" insures a positive refractive index difference Δn between core and cladding. The propagation losses have been calculated and compared to some preliminary measurements. In such fibers, the attenuation is high because once infrared light is injected, several hgih order modes are excited. We are studying the low loss properties of tapered xenon-core capillary.

This paper investigates the application of a tapered single-mode optical fiber as an optical filter for wavelength division multiplexing applications. The transmission characteristics of the tapered single-mode fibers are studied and observed to be oscillatory due to the non-adiabatically of the tapered region. The experimental results are examined in the 1300 nm to 1600 nm wavelength region with special attention around 1550 nm region. In order to validate the experiment a theoretical analysis is carried out for the same device.

An all-fiber dynamic slope compensator is demonstrated, based on Mach-Zehnder interferometer design. Low voltage and electrical power are used to tune the attenuation slope via thermo-optic effect. A constant average insertion loss mode is demonstated.

This paper reports planar waveguide fabrication using a novel fluorinated aromatic polyether bearing cross-linkable pentafluorostyene groups. This polymer exhibits high thermal stability and low optical loss, and the crosslinker used allows thermal processing at low temperatures (<200°C). The influences of the fabrication process and the substrate choices on the refractive index and birefringence of polymer layers have been investigated. Polymer ridge waveguide structures have been fabricated and characeterized, and device optimization through molecular design, fabrication process and substrate choice is discussed.

We introduce a new class of siloxane-based epoxy polymers for thin film optical waveguide applications. The thickness of spun-on films can be controlled by varying either spin speed or viscosity using solvents. Cured films exhibit excellent adhesion to silicon oxide and Al and excellent thermal and chemical stability. Waveguides of ~2 micron thickness on silicon oxide exhibit <0.2dB/cm loss at 830 nm. We demonstrate the formation of a 45° vertical mirror using reactive ion etching slope transfer from photoresist to epoxy polymer film.

In order to find the optimal traveling-wave electrode formation for directional-coupler-based polymeric modulators, four electrode structures are evaluated by simulating the modulator's driving voltages and bandwidths, and by experimentally measuring their driving voltages. First, the characteristic parameters of these four structures are analyzed. Then, the driving votlages for modulators using these four structures are computed using beam propagation method, and their bandwidths are obtained using small-signal analysis. Finally modulators are fabricated using DR1/PMMA and CLD1/APC respectively, and their driving voltages for different electode structures are measured.

The concept of metronomic photodynamic therapy (mPDT) is presented, in which both the photosensitizer and light are deliverd continuously at low rates over extended periods in order to increase selective tumor cell kill through apoptosis. The focus of the present work is on mPDT treatment of malignant brain tumors, in which selectivity between damage to tumor cells versus normal brain tissue is critical. Previous studies have shown taht low-dose PDT using aminolevulinic acid (ALA)-induced protoporphyrin IX (PpIX) can induce apoptosis in tumor cells within causing nectrosis in either tumor or normal brain tissue or apoptosis in teh latter. In order to produce enough tumor cell kill to be an effective therapy, multiple PDT treatments, such as hyperfractionation or metronomic delivery, are likely required, based on the levels of apoptosis achieved and model calculations of tumor growth rates. mPDT poses two substantial technical challenges: extended delivery of ALA and implantation of interstitial devices for extended light delivery while allowing free movement. In rat models ALAL administration via the drinking water has been accomplished at significant doesse (up to 10 times therapeutic dose) for up to 10 days, and ex vivo spectrofluorimetry of tumor, normal brain and other tissues post mortem demonstrates a 3-4 increase in the tumor-to-brain concentration of PpIX, without toxicity. Prototype light sources and delivery devices are also shown to be practical, either using laser diode or light emitting diode (LED) coupled to an implanted optical fiber in the case of the rat model or a directly-implanted LED in rabbits. The combined delivery of both drug and light over an extended period, with survival of the animls, is demonstrated. Preliminary evidence of selective aopotosis of tumor under these conditions is presented.

This study describes the development of novel, fluorescent-based actinometer encapsulates as a means of discerning volumetric Photodynamic therapy (PDT) dosimetry relative to the incident light and reactive oxygen species (ROS) production. PDT relies on three main ingredients; oxygen, light and photo-activatable commpounds, although, the PDT response is definately contingent on the site and level of ROS generation. Providing a localized, in situ measurement of luminance and ROS generation is therefore critical when deciphering targetd photodynamci therapy (PDT) protocols in vivo. Toward this end, alginate-poly-L-lysine-alginate encapsulates were made using ionotropic gelation of sodium alginate droplets ranging from 75 to 200 μm in diameter. Two candidate dyes, ADS680WS (ADS) and R-phycoerythrin (RPE) were chosen based on photochemistry, chemical stabilty and sensitivity to changing pH and oxygen environments. Alginate beads were constructed with ADS conjugated to the inside and RPE attached to the outside layer. The production of ROS was initiated either chemically using increasing concentrations of potassium perchromate or photochemically using tetra-sulphonated aluminium phosphorescence (AlPcS₄). The generation of singlet oxygen was confirmed by the presence of a phosphorescence peak at 1270 nm. The resulting photodegradation and subsequent decrease in fluorescence of RPE was found to correlate very closely (p<0.001) with increasing perchromate or fluence respectively. This effect was independent of pH (6.5-8) and could be inhibited using sodium azide. RPA was not susceptible to photobleaching with light alone (675 nm; 150 J/cm²). Meanwhile, ADS680WS, which absorbs light at 670-690 nm, showed a direct correlation between diminished fluorescence (photobleaching) and incident fluence (675 nm; 0-100 J/cm²). This effect was independent of fluence rate (10-40 mW/cm²). We propose that actinometer encapsulates may prove useful for implanting into potential target areas such as the brain in order to determine the delivered dose of PDT at specific sites within that target area.

Drugs used in disease treatment can cause damage to both malignant and normal tissue. This toxicity limits the maximum therapeutic dose. Drug targeting is of high interest to increase the therapeutic efficacy of the drug without increasing systemic toxicity. Certain tissue abnormalities, disease processes, cancers, and infections are characterized by high levels of activity of specific extracellular and/or intracellular proteases. Abnormally high activity levels of specific proteases are present at sites of physical or chemical trauma, blood clots, malignant tumors, rheumatoid arthritis, inflammatory bowel disease, gingival disease, glomerulonerphritis, and acute pancreatitis. Abnormal protease activity is suspected in development of liver thrombosis, pulmonary emphysema, atherosclerosis, and muscular dystrophy. Inactiviating disease-associated proteases by the administration of appropriate protease inhibitors has had limited success. Instead, one could use such proteases to target drugs to treat the condition. Protease mediated drug delivery offers such a possibility. Solubilizing groups are attached to insoluble drugs via a polypeptide chain which is specifically cleavable by certian proteases. When the solubilized drug enounters the protease, the solubilizing moieties are cleaved, and the drug precipitates at the disease location. Thus, a smaller systemic dosage could result in a therapeutic drug concentration at the treatment site with less systemic toxicity.

In this work, the deposit technique of amorphous films for production of optical waveguides is presented, as well as the characterization of these waveguides. The basic theory is also presented for propagation of light in waveguides.

The autocorrelation function (ACF) for the direction of the principal state of polarization (PSP) vector is reported. The analytical results are compared with the simulation results. It is shown that the magnitude and the directional ACF of the PSP vector are not independent. The directional correlation bandwidth for the PSP vector is verified to be narrower than that of the PSP magnitude.

We report the autocorrelation function (ACF) of polarization dependent loss (PDL) in optical systems having polarization mode dispersion (PMD). A characteristic PDL ACF is derived. It is shown that the PDL autocorrelation bandwidth is proportional to the inverse of the mean PMD and the proportionality constant is a function of mean PDL. Monte Carlo simulations of the autocorrelation functions confirm the analytical results.

A combination of polarization mode dispersion (PMD) and polarization dependent loss (PDL) in a fiber optic system will lead to a complex principal state vector. We report the autocorrelation function (ACF) for such a complex principal state vector. It is shown that the correlation bandwidth decreases with the increasing PDL.

The holographic properties of mixed dichromated [polyvinylalcohol (PVA)/polyacrylic acid (PAA)] were investigated and compared to the pure photosensitive materials DCPVA and DCPAA. It appeared that the addition of 25% in weight of PVA in PAA strongly increase the diffraction efficiency (DE) of the resulting hologram. On the contrary, the addition of PAA in PVA only slightly increases DE values. In the last case, the addition of a complexing agent such as dimethylformamide DMF improves the hologram quality.

Nuclear pore membranes are polymer films with nano- or micro-metric size diameter pores, which are created through the film by irradiation with heavy ions and then chemical etching of the corresponding nuclear traces. The potential of creation of color filters with spectral characteristics controllable by an electric signal was recently demonstrated. To achieve this control it is necessary to obtain nuclear pore membranes with pores thar are oblique, rather than perpendicular to the polymer film. In this paper we present the methods of fabrication of such structures as well as the investigation of their morphological and spectral properties.

Benzocyclobutene (BCB) polymers have demonstrated excellent planarization capabilities. They are procesed using photo patterning or dry etching. A combination process involving both these methods has been developed and optimized for photonic integrated circuits (PICs). Inductively Coupled Plasma (ICP) and Reactive Ion Etching (RIE) are compared as well as soft and hard mask processes. An orthogonal design experiment is demonstrated as an excellent method for evaluating the parameter space of an etch process with variables of table power, ICP power, pressure, and O₂/(O₂+CF₄) ratio. The sensitivities of these variables on etch rate, selectivity and etch profile are discussed. Etch rates of over 1.0 μm/min were achieved with the ICP source, which was in order of magnitude improvement to that of RIE and vertical to positive sloped BCB profiles with excellent mask selelctivity were established.

The IR-140 dye has a maximum of absorption in the near infrared spectral range. It can be used successfully for the initiation of photopolymerization in various organic complexes based on polymers and liquid crystals. In this paper we report the experimental observations of the strong influence of the nematic liquid crystal matrix on the spectral properties of this dye. The character of interaction between the dye and liquid crystalline host and its influence on the wavelength dependence of photopolymerization cross section are analyzed.

The surface layer of InP-based quantum well (QW) laser structures were implanted with As or P ions at energies ranging from 200 to 360 keV. The structures were then annealed at temperatures ranging from 650 to 750°C during 120s, allowing the defects created by implantation to diffuse, resulting in intermixing at the barrier-QW interfaces. The consequence of the intermixing is the blue shift of the QW photoluminescence wavelength. The blue shfit was studied as a function of the implantation temperature (25 or 200°C). Implantation-induced damage in the samples was characterized by Rutherford Backscattering in channeling mode (RBS-c) and correlated with the observed blue shift. It is found that blue shift is more efficient at higher implantation temperature, even if the resulting defect concentration is much lower. This is attributed to the diffusion of defects during high-temperature implantation, leading to a larger region containing defects contributing to intermixing. Also, when the implanted dose is too high, no blue shift is observed. This could be due to the formation of defect clusters that inhibit the subsequent diffusion of defects. Finally, the defect creation mechanisms within InP and InGaAs layers are found to have a significant impact on the resulting wavelength blue shift.

We report a non-destructive in-line monitoring method developed for Cd diffusion into InP on SACM-APD structure. Photocurrent vs voltage measurement are taken directly via proving diffused diodes on a wafer. We demonstrate that there is linear correlation between punch-through voltages Vpt on the photo I-V curves and diffusion depth measured by SIMS and Polaron profiles. It has been established that Vpt can be extracted easily from I-V curves and used for re-diffusion to approach target depth.

By using the selective area growth (SAG) technique, the grating over growth technique, and the buried hetero-structure (BH) regrowth technique, we have successful fabricated integrated distributed feedback laser and electro-absorption modulator. A suitable tensile strain is introduced during the multiple quantum well (MQW) growth to compensate the compressive strain caused by the SAG effect. To obtain better control on the growth condition and grating coupling coefficient, a quaternary grating laser is used. The best DFB laser shows a threshold current of around 12 mA and slope efficiency up to 0.13 mW/mA from one facet.

With the development of photonic devices technology, optical communication networks require highly integrated photonic devices. The key problems in integrated devices are how to integrate these active devices and passive devices together and achieve the low loss and high performance. In this paper, we are talking about a new processing technology to the integrated device which shows a great potential in the application of integrated devices. The basic idea is to use semi-insulating material as the buried heterostructure and cladding material at the same time, then open a hole on the active region to define and confine the current flow. We did lots of experiments to prove this method, and the results showed the feasibility. This new method can have wide applications not only on the issues of the integrated devices. It can be used in very narrow waveguide devices and other applications.

The method of using Si delta doping to obtain high n-type doping for InP and In_0.53Ga_0.47As lattice matched to InP was studied. With the multiple delta doping, we can obtain ~10¹⁹ cm^-3 N-type doping for both the InP and InGaAs materials by using TMI, PH3, Si2H6 and TEGa precursors in Metalorganic Chemical Vapor Depositions growths. Current results show that the delta doping technique is critical for getting extremely high N-type doping in InP and InGaAs.

The incorporation of carbon into In_0.53Ga_0.47As, In_0.52Al_0.48As and In_0.52Al_0.2Ga_0.28As lattice matched to InP was investigated using carbon tetrabromide (CBr₄) as the carbon source in Metalorganic Chemical Vapor Depositions growth. The parameters and growth conditions are optimized to get high p-type doping for photonic device applications. This is among the first few studies on C-doping in InAlAs and InAlGaAs, and the results show that the presence of Al also efficiently helps to obtain high p-type carbon doping.

Bulk InGaAs and InGaAsP and quantum well structure have been grown on different oxide pattern by MOCVD. We found that because of material composition variation caused by the oxide pattern compressive strain s built, so that the PL intensity is reduced. By introducing tensile strain to compensate for the compressive strain caused by the SAG effect, the PL intensity has been improved. Active and passive integrated devices can be fabricated without significantly suffering the material degradation by using this method.

Bulk InGaAs and InGaAsP and quantum well structure were grown on different oxide pattern. The material composition varies with the width of oxide pattern. The PL peak wavelength of the grown SAG quantum well is determined by the variation of both the material composition and the well width. We have experimentally identified the ratio of the contribution from the two sources, which agree well with the theoretical calculation from the measured thickness changes.

Based on the sp³s tight-binding method, the electronic band strcutures of both common-anion and common-cation II-VI superlattices (SLs) are investigated. As models, we took for the former the case of CdTe/ZnTe(001) SLs where the common anion is confirmed to yield a vanishing or a small valence-band offset (VBO) and the biaxial strain, of course, contributes in the valence-band splittings and yield type-I SLs in most of the studied cases. Whereas, we tok as a second model two different SLs: the ZnS/ZnSe(001) and ZnSe/ZnTe(001) SLs. We have confirmed that the common-cation SLs cannot have a vanishing conduction-band offsets (CBO), as speculated, but rather the CBO could be as large as the VBO. The biaxial strain, again, can participate here in the formation of the band offsets and yield either type-I SLs, as in the case of the ZnS/ZnSe, or type-II SLs, as in the case of ZnSe/ZnTe. Moreover, the reason why some type-II SLs, such as ZnSe/ZnTe(001), could be useful as photonic devices is explained by the tendency of the carriers to confine near the interface as a result of a strong photoluminenscence data and conclusions have been drawn about the strain morpholgy and the structural and optical qualities of the experimental samples.

A thin layer of INGaAs is the critical photo-absorbing layer in epitaxially grown layers used in the monolithic integration of active photodetectors and a passive mux/demux. This thin InGaAs active layer requires tight thickness control that is beyond the accuracy and reproducibility of conventional thickness measurement techniques such as a step-height profilometer. In order to address this issue, a technique using Field Emission Scanning Electron Microscopy (FESEM) combined with a chemical delineation etch (use to reveal thin layers) was developed. The etch used was dilute H₂SO₄:H₂O₂:H₂0 which will etch all ternaries and quaternaries but does not etch InP. The selectivity of this etch, between the ternary InGaAs material and the InP, was measured to be approximately 500:1. We found that using the FESEM technique provided a minimum measurable thickness of approximately 50 nm, with a precision of approximately 5 nm. The accuracy of our measurement was verified by TEM analysis on selected samples. We also found that SIMS data, obtained from an independent analytical laboratory and originally performed to monitor doping levels in the layers, could be used to monitor the thickness and/or the variations in composition of the thin InGaAs photo-absorbing layer. The key is to make use of the integrated SIMS peaks rather than the full-width-half-maximum (FWHM) of the peaks. For an InGaAs layer, the integrated arsenic SIMS signal varies with the layer thickness and the arsenic concentration. The variations in either of these, in different growth batches, can be detected by SIMS. We found that using integrated SIMS profiles gives a fully independent thickness and/or compositional control monitor to compare with our FESEM measurements and can be used as a trusted control fo the layer quality. This comes at no additional cost since in our case, SIMS profiles are required to monitor doping levels. In this paper, a correlation is shown between the thickness of a thin layer of InGaAs measured by FESEM and the integrated arsenic SIMS peaks from several independent epitaxial growths. Deviations from the norm are explained from small thickness or compositional changes as supported by FESEM and photo-luminescence measurements.

This paper shows the use of the Rice representation to obtain the correlation functions of the phase and power fluctuations of the optical noise of a laser oscillator when the Henry's factor is taken into account.

We will report on a new type of tunable semiconductor laser, which is based on the electronic selection of one Bragg grating among an array of such gratings in silica. The device that we have built operates at 120 Mb/s but extension to 1 Gb/s for Gigabit-Ethernet applications would be straightforward. In comparison with tunable semiconductor lasers using gratings in the III-V materials, silica gratings offer two significant advantages: 1-wavelength stability and predictability, 2-the ability to phusically overlap many gratings in a compact space in order to enable the selection of a large number of wavelengths for wavelength division multiplexed communications systems. The time required to chagne the wavelength in our laser has not been measured for lack of the necessary electronics but it is expected to be in the microsecond range on the basis of a straightforward calculation. The robust all solid-state nature of our device and its expected microsecond random-access tuning capability make it a strong candidate for agile wavelength routed largely optical networks. With microsecond tuning of precisely and robustly defined wavelengths, optical routing would be possible on a packet basis in a network using state-of-the-art add/drop filters or other types of wavelength selective elements, such as arrayed waveguide gratings.

The effects of using multiple electrodes to reduce longitudinal spatial hole burning (LSHB) in a complex-coupled distributed feedback (DFB) laser are investigated. Using a time-domain model, we show that LSHB can be compensated by injecting more current into the center of the device. We also find that differntial current injection has an effect on the threshold current and quantum efficiency of the DFB laser. Spectral characteristics are obtained by fast Fourier transform of the time-domain data and we show that inhomogeneous current injection can also have an effect on the laser linewidth.

Short optical pulse generation at 40GHz and 1540nm wavelength is achieved using fully packaged bulk quaternary electro-absorption modulator modules. Experimental results obtained with broadband and narrowband optimized packaged modules are presented and compared against empirical model predictions. Pulse duty cycle, extinction ratio and chirp are studied as a function of sinusoidal drive voltage and detuning between operating wavelength and modulator absorption band edge. Design rules and performance trade-offs are discussed. Low-chirp pulses with a FWHM of ~12ps and sub-4ps at a rate of 40GHz are demonstrated. Optical time-domain demultiplexing of a 40GHz to a 10GHz pulse train is also demonstrated with better than 20dB extinction ratio.

The degradation of the side-mode suppression ratio (SMSR) in an integrated DFB laser and semiconductor optical amplifier (SOA) cavity is investigated. A numerical model combining finite element calculations in the x-y plane and a longitudinal model based on the Green's function approach is used for that purpose. An expression is derived that expresses the degradation of the SMSR in the case of a perfectly AR-coated SOA facet. It is shown that the ASE backcoupling can have dramatic effects and degrade and SMSR of single-mode devices to unacceptable levels.

We propose a double linewidth pseudo-Voigt profile to handle the spectral shape variances caused by short pulse interactions during the Brillouin scattering process in optical fibers. When short pulses are used, their broad frequency spectra contribute to the overall Brillouin spectrum. Hence, the convoluted spectrum consists of CW light scattering plus the broad frequency spectrum from the short pulse.

Multi-parameter measurement is important for building health monitoring. In this paper, we use fiber Bragg grating (FBG) and extrinsic Fabry-Perot interferometer (EFBI) sensor to realize the measurement of strain, temperature and vibration. A universal demodulate method, which combines the low coherence interference demodulation of EFBI and fiber Fourier transform demodulation of FBG, is proposed. Thus it is possible to extract the signals from FBG and EFBI parallely with only a Michelson interferometer. A data fusion model is established to process the origin data to get more precise results and distinguish the different parameters. The system is tested on a equi-intensity cantilever, and the results of experiment show that strain and temperature can be obtained efficiently.

The extinction ratio of an Electro Optic Modulator (EOM) limits the accuracy of BOTDA measurement of Brillouin-based fiber optic sensors. An EOM with a finite extinction ratio yeilds Brillouin frequency peaks from outside the spatial resolution. This result in confusion in locating the true strain and temperature and error in detecting the Brillouin frequency. The effects of different parameters such as extinction ratio, pulse power and pump power wil be discussed.

Gain crosstalk in semiconductor optical amplifiers (SOAs) developed for WDM networks has in the past been a limiting factor in SOA deployment. One recently-proposed solution is the gain-clamped SOA, where the amplifier gain is stabilized in a complex cross-cavity configuration. In this paper, we describe the design of conventional SOAs for low multi-channel gain crosstalk (GXT). Specifically, we compare alternatives for reducing crosstalk, including reductions in SOA length, current density, and the number of quantum wells. Numerical modeling shows the reduction in well number to have the largest impact; experimental result for a 4-well device show 2-channel GXT of -20 dB at a modulation frequency of 10 GHz, a fiber-coupled output power of +5 dBm (+2 dBm per channel), and a module gain of 15 dB.

In_1-xGa_xAs_yP_1-y nipi Structure has been grown by MOCVD and been characterized by photoluminescence. The two PL profiles from the direct and the indirect recombination channels were clearly observed. The excitation intensity and temperature dependence of the PL profiles are studied. A calculated carrier lifetime, as long as 71μs is possible to be realized. With such a long carrier lifetime, we can push the XT noise out of the signal band down to frequencies < f=f=1/(2πτ)=2.2kHz. Equivalently, the XT is eliminated.

Ultra-broadband gain materials can be made by using selected area growth (SAG) techniques. Material gain peak along the waveguide can vary as much as 300nm (from 1300nm to 1600 nm) in InGaAsP/InP quantum well waveguides. This gain peak variation is obtained by both material composition changes and quantum well width changes. This kind of broad band materials can be used to make ultra broad tunable lasers, amplifiers, and wavelength converters. Due to large band gap variation, the current injection will not be constant along the waveguide. The average serial resistance perpendicular to the waveguide direction will play an important role in the current distribution. In our simulation, a large serial resistance device will have a uniform current injection, therefore a uniform carrier distribution. But for a small serial resistance device, the carrier distribution along the waveguide is mainly decided by the local waveguide material bandgaps.

We have demonstrated a very simple scheme to excite new lines in the near IR range via the DFM of 2nd harmonic of Nd:YLF laser and Stokes seed, in particular obtaining 943 nm emission. Energy efficiency for wave interaction in RTP crystal was equal to 27%. Being measured, nonlinear coefficient of Double Photon Absorption in this crystal showed 6×10^-4 cm/MW.

This paper reports data of green (549.8 nm) fluorescence and lasing of Erbium (Er³⁺) doped in Yttrium Orthoaluminate (YAlO₃) crystal pumped by near IR light at 8 and 40 degrees Kelvin. The fluorescence amplitude of Er³⁺ is measured at IR light pump wavelengths between 780 and 820 nm at 8 and 38 degrees Kelvin. Fluorescence peaks are related to the energy diagram of Er³⁺ and shown to originate from the ground energy state and a meta-stable energy state at 8 degrees Kelvin. Fluorescence peaks resulting from the first excited energy state are also present at 34 degrees Kelvin. The lasing of Er³⁺ pumped by these peak fluorescence IR wavelengths is measured as a function of pump power at 8 and 40 degrees Kelvin. 600 mW of green laser light are produced from 3 W oif IR pump pouwer indicating 20% lasing conversion efficiencies.

An effective method to implement room temperature operation of multiwavlength erbium-doped fiber ring laser (MW-EDFL) employing equivalent low frequency shift feedback (ELFSF) is presented. Simultaneous multiwavelength lasing with 0.5nm wavelength spacing is experimentally demonstrated by applying a sawtooth signal of the order of 10kHz to a phase modulator inserted in the ring cavity to prevent single-wavelength oscillation. A LiNbO₃ multifunction chip of fiber gyro has been used as the phase modulator, the output coupler, and the polarizer of Lyot periodic filter. The MW-EDFL output is linearly polarization light that meets the requirement of external modulation for WDM applications.

A novel lightsource to provide the excitation radiation for fluorescence microscopy is presented and its performance is compared to the current de factor standard in the field: mercury short arc lamps. This novel light source is remote to the microscope, and the radiation is coupled to the microscope via a liquid lightguide or fiber optic cable using special coupling optics. We present measurements made on some common fluorescent microscopes that show the new light source provides for higher overall optical power delivered to the sample and provides more uniform illumination of the microscopes' field of view in comparison to the standard short arc lamps. Using the definition of the Koehler illumination rules it is shown that the inherent design of the remote source makes it resistant to many non-uniformities and misalignments commonly enountered with the short arc lamp sources; thereby providing for a consistent, uniform irradiance and intensity distribution of the entrance pupil to the microscope. The experimental method used to quantitatively measure the uniformity of the excitation radiation at the microscope's objective plane is also discussed and shown to be far more reliable than other techniques which rely upon fluorescent radiation from synthetic samples placed at the objective plane.

Preventive oncology is in need of a risk assessment technique that can identify individuals at high risk for breast cancer and that has the ability to monitor the efficacy of a risk reducing intervention. Optical transillumination spectroscopy (OTS) was shown to give information about breast tissue composition and tissue density. OTS is non-invasive, and in contrast to mammography, uses non-ionizing radiation. It is safe and can be used frequently on younger women, potentially permitting early risk detection and thus increasing the time available for risk reduction interventions to assert their influence. Before OTS can be used as a risk assessment and/or monitoring technique, its predictive ability needs to be demonstrated and maximized through the construction of mathematical models relating OTS and risk. To establish a correlation between OTS and mammographic density, Principle Components Analysis (PCA), using risk classification, is employed. PCA scores are presented in 3D cluster plots and a plane of differentiation that separates high and low tissue density is used to calculate the predictive value. Stratificaiton of PCA for measurement position on the breast in cranial-caudal projection is introduced. Analysis of PCA scores as a function of the volunteer's age and body mass index (BMI) is examined. A small but significant correlation between the component scores and age or BMI is noted but the correlation is dependnet on the tissue density category examined. Correction of the component scores for age and BMI is not recommended, since a priori knowledge of a women's tissue density is required. Stratification for the center and distal measurement positions provides a predictive value for OTS above 96%.

We present a tunable wideband bismuth-substituted yttrium iron garnet (Bi-YIG) waveguide magnieto-optic (MO) modulator. An 800 nm optical beam is modulated through active control of the magnetization of the Bi-YIG film. Large bandwidth optical modulation is achieved by driving the device in a non-resonant mode that is well below the ferromagnetic resonance frequency of the film. The MO modulator is capable of operating at bandwidths higher than 1 GHz by tuning the applied static magnetic field.

We demonstrate a high-speed magneto-optic (MO) beam deflector baseo n iron garnet materials. In this design, the two films are dynamically activated by external magnetic fields to achieve the necessary phase shift to enable beam deflection. Using standard material parameters, a deflection efficiency of 20% is predicted at an operating frequency of 1.0 GHz. Diffraction effects and temporal response are discussed and illustated using a quasi-time-dependent beam propagation method. This theoretical framework is not only useful for the demonstration of the MO beam deflector, but it is also applicable in the design and optimization of other magneto-photonic devices.

The feasibility of ultrafast (10 THz) communication by time multiplexing using femtosecond pulses is discussed. Such a system calls for stable mode-locked sources in frequency and repetition rate, that can be synchronized with sub fs jitter. Preliminary experiments and theory show this goal to be achievable.

Model calculation of second order susceptibilities for FA color centers in wide band gap materials are reported. The second order optical nonlinearity in crystals of alkali halide due to FA color centers evaluated theoretically. The density matrix formalism is employed and the equation of motion is solved by second order perturbation to evaluate the nonlinear optical susceptibilty for second harmonic generation as well as frequency mixing. It is found that the system shows large resonance-enhanced second order susceptibility (~10^-16 mV^-1) for color center concentration of ~10²³ m^-3. A scheme of phase matching in terms of anomalous dispersion of the centers and coherent length are discussed.

A classical phase-amplitude description of the optical field is proposed to discuss the optical amplifier noise generation. This description allows pointing out the contributions to output noise of the vacuum fluctuation input noise and of the intrinsic amplification and attenuation noises mechanisms. Comparison is done with the standard Amplified Spontaneous Emission (ASE) and associated beat noise approach. The mode is applied to the theoretical noise figure discussion for distributed amplifiers, as a function of built in internal loss and values of the net gain achieved, and to the derivation of the equivalent lumped amplifier noise figure.

We obtained amorphous carbon (a-C) thin films, deposited on glass and silicon substrates, by sputtering radio frequency (RF) from a pure graphite target. The properties of these layers are studied by x-ray reflectometry. We present here a precise and simple method to measure layer thicknesses and densities.

A new generalized Poincare sphere method is proposed to measure the complex principal states of polarization vector for a system with polarization dependent loss or gain. The spectral resolved measurements with our propsoed test-set agree with the well known Jones eigenvalue analysis method.

GaAs/AlAs superlattices (SLs) were grown by MOCVD. The temperature dependence of photoluminescence was measured. The type II-transition dominated PL spectrum was achieved by controlling the layers thickness of GaAs and AlAs at low temperature. Such SLs with long carrier lifetime is very attractive for low crosstalk semiconductor optical amplifier applications.

Measurement of two aerial fibers fundamental frequency is performed using the state of polarization. Comparisons are made with basic wave theory and reasonable agreement is obtained. The affect of sag on the fundamental frequency is investigated.

The angle between the principal state of polarization fast and slow modes is investigated. In the presence of polarization dependent loss the angle is shown to become non-orthogonal and vary statistically. It is found that the statistics of the angle only depends on polarization dependent loss and is independent of polarization mode dispersion. This phenomena is investigated through experiments and simulations.

PMD characteristics of four commercial photonic crystal fibers are measured using Jones Matrix eigenanalysis. It is found that PMD value could vary up to two orders of magnitude due the asymmetry of the photonic cyrstal fiber fabrication. The PMD temperature dependent characteristics is presented for a 1.55μm crystal fiber.

BER degradation, caused by a dynamic PMD emulator, is investigated in a 10Gb/s optical system. It is shown that dynamic PMD emulation more closely emulates rare field fiber BER events whereas standard emulators average away these events.

Spectrally resolved polarization mode dispersion is measured in emulators and spool fibers for different temperatures. The PMD emulators, made of randomly spliced PM fiber sections, are found to be correlated. The field fibers are found to be random.