We show that an optical pulse inherently computes three-dimensional classical fluid dynamics. Taking optical diffraction, dispersion and nonlinearity into account, one can define the metaphoric fluid density, velocity and vorticity in the optical pulse. We propose the use of the group-velocity-delayed time to represent the third dimension of the fluid, and the "splitstep" method to combine optical devices as a configurable system that simulates fluid flow. Optical systems, with the inherent speed, parallelism and configurability, may one day be utilized to assist the study of fluid dynamics.

We present two potentially interesting new venues in all-optical signal processing. First, we demonstrate experimentally that collisions between vector (Manakov-like) solitons involve energy exchange; this feature could be explored for all-optical signal processing. Second, our detailed theoretical studies show how inserting materials that support electro-magnetically induced transparency into microcavities enables design of microcavities with extraordinarily long lifetimes, and enables all-optical signal processing at single photon power levels.

We present design and simulation results on two types of inplane, resonant 2 x 2 optical switches in the strip-waveguided silicon-on-insulator (SOI) system. Both switches are intended for cascading into N x N matrices, and the intersecting strips in the second floating ring switch facilitate construction of an N x N crossbar cross-connect. We examine complex index perturbations of both rings due to electrical injection of free carriers into the rings or to the generation of electron-hole pairs induced by above-bandgap control light focused upon the rings. Switching behavior produced by free-carrier-plasma dispersion, Franz-Keldysh, quantum-confined Stark, Kerr, and thermooptic effects are investigated and compared at the 1.33, 1.55, and 10.6 mm wavelengths. A study of design tolerances and their consequences shows that the 10.6 mm wavelength is preferred from the standpoint of photolithographic feasibility and the stronger carrier-optic effect. The SOI high index contrast requires small coupling gaps for the 1.33 and 1.55 mm wavelengths, which in turn necessitates a high photolithographic precision.

Electromagnetically induced transparency is accompanied by extraordinary dispersion that introduces slow light. In an optical cavity, this slow light dispersion can cause line center cavity pulling as well as pulse delay and resonance detuning in response to the associated modest excursion of the refractive index, magnified by the finesse of the cavity. The use of this nonlinear optical effect in a modulator device to transfer small amplitude electrical signals onto optical carriers is examined. Although many aspects of the concept fall short of useful targets, aspects of waveguide coupling and ring resonator design are revealed that could find utility in other device designs.

We predict the propagation of slow and fast light in two co-resonant coupled optical resonators. In coupled resonators, slow light can propagate without attenuation by a cancellation of absorption as a result of mode splitting and destructive interference, whereas transparent fast light propagation can be achieved by the assistance of gain and splitting of the intracavity resonances, which consequently change the dispersion from normal to anomalous. The effective steady-state response of coupled-resonators is derived using the temporal coupled-mode formalism, and the absorptive and dispersive responses are described. Specifically, the occurrence of slow light via coupled-resonator-induced transparency and gain-assisted fast light are discussed.

We investigate theoretically and experimentally the characteristics, performance and possible applications of coupled resonator optical waveguides (CROWs). The ability to engineer the dispersion properties of a CROW and especially the ability to realize ultra-slow group velocities paves the way for various applications such as delay lines, optical memories and all-optical switching. Simple analytic expressions for the time delay, usable bandwidth and overall losses in CROW delay lines are derived and compared to exact numerical simulation. Good quantitative agreement is found between the theoretical transmission function obtained by transfer matrix formalism and the measurement of a CROW interferometer realized in polymer material.

Photonic circuits require elements that can control optical signals with other optical signals. Ultra-low-light-level operation of all-optical switches opens the possibility of photonic devices that operate in the single-quantum regime, a prerequisite for quantum-photonic devices. We describe a new type of all-optical switch that exploits the extreme sensitivity to small perturbations displayed by instability-generated dissipative optical patterns. Such patterns, when controlled by applied perturbations, enable control of microwatt-power-level output beams by an input beam that is over 600 times weaker. In comparison, essentially all experimental realizations of light-by-light switching have been limited to controlling weak beams with beams of either comparable or higher power, thus limiting their implementation in cascaded switching networks or computation machines. Furthermore, current research suggests that the energy density required to actuate an all-optical switch is of the order of one photon per optical cross section. Our measured switching energy density of ~4.4 × 10^-2 photons per cross section suggests that our device can operate at the single-photon level with modest system improvement.

We show that, by simple modifications of the usual three-level Λ-type scheme used for obtaining electromagnetically induced transparency (EIT), phase dependence in the response of the atomic medium to a weak probe field can be introduced. This gives rise to phase dependent susceptibility. By properly controlling phase and amplitudes of the drive fields we obtain variety of interesting effects. On one hand we obtain phase control of the group velocity of a probe field passing through medium to the extent that continuous tuning of the group velocity from subluminal to superluminal and back is possible. While on the other hand, by choosing one of the drive fields to be a standing wave field inside a cavity, we obtain sub-wavelength localization of moving atoms passing through the cavity field.

This paper outlines the scientific basis on which the DARPA Slow Light program is formed. Two fundamental application classes are mentioned, and the scientific motivation for each developed to indicate overall goals and prospects for insertion into specific devices.

We present a preliminary experimental study of slow and stored light in Rb vapor cells under the conditions of electromagnetically induced transparency (EIT). We study the efficiency of light storage as a function of pulse duration, storage time, retrieval field intensity, etc. We demonstrate that atomic diffusion in-and-out of the laser beam plays an important role not well described by previous analysis.

For many applications of slow or stopped light, the delay-time-bandwidth product is a fundamental issue. However, existing slow light demonstrations do not give a satisfactory delay-time bandwidth product, especially in room temperature solids. Here we demonstrate that the use of artificial inhomogeneous broadening has the potential to solve this problem by simultaneously slow down all the frequency components of the input pulse. The proof of principle experiment was done using three-wave mixing in a photorefractive crystal Ce:BaTiO₃ where Bragg selection is used to provide the inhomogeneity.

A reduced-density-matrix description is developed for linear and non-linear electromagnetic interactions of quantized electronic systems in the presence of environmental decoherence and relaxation phenomena. Applications of interest include many-electron atomic systems (in electron-ion beam interactions, gases, and high-temperature plasmas) and semiconductor materials (bulk crystals and nanostructures). Time-domain (equation-of-motion) and frequency-domain (resolvent-operator) formulations are developed in a unified manner. The standard Born (lowest-order perturbation-theory) and Markov (short-memory-time) approximations are systematically introduced within the framework of the general non-perturbative and non-Markovian formulations. A preliminary semiclassical treatment of the electromagnetic interaction is introduced. Compact Liouville-space operator expressions are derived for the linear and the general (n’th order) non-linear electromagnetic-response tensors, allowing for coherent initial electronic excitations and for the full tetradic-matrix form of the Liouville-space self-energy operator representing the environmental interactions. It is emphasized that quantum-coherent many-body interactions cannot be adequately treated as environmentally induced phenomena.

We propose a "channelization" architecture to achieve wide-band electromagnetically induced transparency (EIT) and ultra-slow light propagation in atomic Rb-87 vapors. EIT and slow light are achieved by shining a strong, resonant "pump" laser on the atomic medium, which allows slow and unattenuated propagation of a weaker "signal" beam, but only when a two-photon resonance condition is satisfied. Our wideband architecture is accomplished by dispersing a wideband signal spatially, transverse to the propagation direction, prior to entering the atomic cell. When particular Zeeman sub-levels are used in the EIT system, then one can introduce a magnetic field with a linear gradient such that the two-photon resonance condition is satisfied for each individual frequency component. Because slow light is a group velocity effect, utilizing differential phase shifts across the spectrum of a light pulse, one must then introduce a slight mismatch from perfect resonance to induce a delay. We present a model which accounts for diffusion of the atoms in the varying magnetic field as well as interaction with levels outside the ideal three-level system on which EIT is based. We find the maximum delay-bandwidth product decreases with bandwidth, and that delay-bandwidth product 1 should be achievable with bandwidth 50 MHz (5 ns delay). This is a large improvement over the 1 MHz bandwidths in conventional slow light systems and could be of use in signal processing applications.

Using coherence enhanced nonlinear optics we observe absorptive switching in hot Rubidium atoms. Electromagnetically induced transparency helps create a larger absorptive Kerr nonlinearity enabling strong absorptive switching with laser intensities of 10 microwatts per square centimeter. Switching is interpreted in terms of optical pumping into and out of the "dark" state.

We have experimentally demonstrated anomalous stimulated scattering of sound waves with optical waves of the same direction in a cell of Rb vapor. The optical wave is propagating under the conditions for ultra-slow group velocity. The phase-matching condition for the interaction of the waves is achieved when the group velocity of the probe field to approximately the speed of the sound wave.

We review our work on electromagnetically induced transparency (EIT) as a potentially key enabling science for few-qubit Quantum Information Technology (QIT). EIT systems capable of providing two-qubit phase shifts as large as pi are possible in a condensed matter system such as NV-diamond, but the potentially large residual absorption necessarily arising under this condition significantly reduces the fidelity of a nonlinear optical gate based on EIT. Instead, we emphasize that a universal set of quantum gates can be constructed using EIT systems that provide cumulative phase shifts (and residual absorptions) that are much smaller than unity. We describe a single-photon quantum nondemolition detector and a two-photon parity gate as basic elements of a nonlinear optical quantum information processing system.