First observation of transverse Anderson localization in an optical fiberWe utilize transverse Anderson localization as the waveguiding mechanism in optical fibers with random transverse refractive index profiles. Using experiments and numerical simulations, we show that the transverse localization results in an effective propagating beam diameter which is comparable to that of a typical index-guiding optical fiber. Despite more than 50 years of research on Anderson localization, its curious and interesting properties are yet to be harnessed in practical devices. Here, we have taken a step towards utilizing this phenomenon in a device application. In addition to the practical importance of the device-level implementation of the localization of light, the availability of well-established fiber-optic characterization techniques and tools enables a more detailed assessment of Anderson localization effects in future studies. Therefore, the possibility of an unparalleled wealth of data in a fiber-optic platform provides a unique opportunity to unwrap the complexity of wave transport and localization. We anticipate that the ideas and methods presented here can be implemented in other platforms where disorder can play a functional role to induce wave localization.
Gain guided fiber lasersSiegman has recently shown that in the presence of sufficient gain, an index-antiguiding fiber with an arbitrarily large core diameter, can operate in a robust single transverse mode. While the idea has been confirmed experimentally, the measured values of the beam quality factor (M2 or M-squared) differ substantially from theoretical predictions, and fiber lasers have resulted in very low power efficiency. Recently, we attributed the differences between theory and experiment in M2 to the truncation of the long cladding tail of the mode at the cladding-jacket interface. Presently, as part of her Ph.D. thesis, Parisa Yarandi (Ph.D. student) is actively pursuing a solution to the pump confinement problem in this index-antiguiding fiber to improve the power efficiency; her recent results have been very promising.
Nonlinear fiber opticsNonlinear phase noise:
We recently developed a fast, accurate, and easy-to-implement scheme to estimate the impact of nonlinear phase noise in WDM systems; brute-force Monte Carlo simulations are computationally expensive and cannot be used to optimize an optical fiber communication (OFC) link in the presence of nonlinear phase noise. This work successfully led to a generalization of the Eigenfunction Expansion Method to model the performance of WDM OFC systems in the presence of nonlinear phase noise.
Multimode optical fibersLow-loss coupling between two single-mode optical fibers with different mode-field diameters using a graded-index multimode optical fiber:
We present a method for ultra-low-loss coupling between two single-mode optical fibers with different mode-field diameters using multimode interference in a graded-index multimode optical fiber. We perform a detailed analysis of the interference effects and show that the graded-index fiber can also be used as a beam expander or condenser. The results are important for devices in which optical fibers with different mode-field diameters are coupled in series, such as in ultrashort-pulse fiber ring lasers, or in optical fiber communication links. Bandwidth improvement in multimode fibers via scattering from core inclusions:
The graded-index profile in GIMFs is often defective near the center of the core, giving rise to an unbalanced distribution of group velocities and lowering its measured bandwidth. This defect can place the ultimate limit on the bandwidth of the multimode OFC systems. I showed that a controlled intentional mode coupling induced via scattering from core inclusions can substantially improve its bandwidth, e.g., from 693-MHz.km to 2.5-GHz.km with less than 1-dB additional power loss. The desired level of coupling can be obtained by exposing the photosensitive core of the fiber to a UV laser, therefore creating the micro-inclusions, even after the draw.
Photonic crystal fibersPhotonic crystal fibers with a general Bravais lattice
Recently, with K. Koch of Corning, we presented a comprehensive analysis of PCFs with a general Bravais lattice. Our work resolved a common misconception in the literature that claimed that the PCF GVD depends on the shape of its lattice (square or triangular). We disputed this claim and showed that the difference stems only from an unnatural choice in the unit lat- tice area, which is rooted in the scale invariance of the Maxwell equations, and has little to do with the lattice shape. As such, properly scaled square and triangular lattices with equal unit lattice areas result in almost identical normalized GVD. The study also revealed that when presented in the proper length scale, the endlessly-single-mode critical radius remains almost invariant under lattice shape deformations.
Photonic bandgap structuresImpact of lattice-shape moduli on band structure of photonic crystals
We conduct a comprehensive study of the effect of the Bravais lattice-shape moduli on the band structure and in particular the band gap of photonic crystals. Unlike the conventional comparisons between triangular and rectangular photonic crystals, where the effect of the volume modulus is not separated, we rigorously decouple the volume modulus and determine the differences that can be attributed only to the shape of the lattice. We observe that the triangular lattice enjoys the largest band gap owing to its unique symmetry properties. We also show that the band gap decreases when the ratio of the lattice constants differs from unity. The use of an appropriate parametrization of the lattice-shape moduli combined with the inherent scaling invariance in the Maxwell equations allows us to cover Bravais lattices of all shape and volume moduli completely and without redundancy.
Plasmonic sensorsOptimizing plasmonic grating sensors for limit of detection based on a Cramer-Rao bound
We explore the impact of the geometrical features such as the period, duty cycle, and thickness on the performance of a metal-dielectric plasmonic grating sensor. The sensor is designed to operate at the wavelength of 850 nm for water-monitoring applications. Limit of detection (LOD) is chosen as the performance metric for the optimization of the sensor design. The LOD metric is based on a Cramer-Rao bound (CRB) and offers a theoretical lower limit on the estimation uncertainty of the spectral shift in the proposed sensor. We show that the lowest values of LOD correspond to grating designs, which result in step-like spectral features. An optimum grating design corresponds to gold stripes that are 160 nm thick and 456 nm wide on glass where the stripes are separated by 74 nm. Although our study is focused on a particular design and application, our methods and observations are applicable to a wide range of plasmonic grating sensor designs.