Transcript Defense
Soliton Solutions for HighBandwidth Optical Pulse Storage and Retrieval Elizabeth Groves University of Rochester Thesis Defense February 11th, 2013 Soliton Solutions for HighBandwidth Optical Pulse Storage and Retrieval Elizabeth Groves University of Rochester Thesis Defense February 11th, 2013 Soliton Solutions for HighBandwidth Optical Pulse Storage and Retrieval Elizabeth Groves University of Rochester Thesis Defense February 11th, 2013 Normal vs. Short Optical Pulse Propagation Beer’s Law of Absorption • • • • Atoms absorb laser pulse energy Pulse may be too weak to promote atoms to excited state Atoms dephase, return little or no energy to the field Laser pulse depleted Stopped light? Maybe, but not useful. We want storage. Normal vs. Short Optical Pulse Propagation Long weak pulses Short strong pulses • • • • • Atoms absorb laser pulse energy Pulse may be too weak to promote atoms to excited state Atoms dephase, return little or no energy to the field Laser pulse depleted • • • Atoms initially absorb laser pulse energy Laser pulse drives atoms to excited state Atoms don’t have time to dephase; return energy to the field coherently Laser pulse undepleted Normal vs. Short Optical Pulse Storage Long weak pulses Short strong pulses • • • Storaged achieved using Electromagnetically-Induced Transparency (EIT) and related effects Linear equations, adiabatic, steady-state conditions • • • Storage of high-bandwidth pulses is desirable Enable higher clock-rates, fast pulse switching Nonlinear Nonlinear equations equations hard! Support soliton solutions We derived an exact, second-order soliton solution that is a reliable guide for short, high-bandwidth pulse storage and retrieval. Solving Nonlinear Evolution Equations (PDEs) Analytical Methods Approaches Separation of variables, symmetry arguments, clues from related linear system Problems Hard!!, idealized conditions, cannot linearly superimpose solutions to find a general solution. Each equation seems to require special treatment. Numerical Methods Approaches Finite difference method, method of lines, spectral method (uses Fourier transforms) Problems What’s a numerical artifact? Have you really sampled the solution space? How important are the initial conditions you’re using? Solving Nonlinear Evolution Equations (PDEs) Analytical Methods Approaches Certain nonlinear evolution equations can be solved exactly by soliton solutions. Problems Hard!!, idealized conditions, cannot linearly superimpose solutions to find a general solution. Each equation seems to require special treatment. Numerical Methods Approaches Finite difference method, method of lines, spectral method (uses Fourier transforms) Problems What’s a numerical artifact? Have you really sampled the solution space? How important are the initial conditions you’re using? What are Solitons? In 1834 John Scott Russell, an engineer, was riding along a canal and observed a horse-drawn boat that suddenly stopped, Stable solitary wave causing a violent agitation, giving rise to a lump of water that rolled forward with great velocity without change of form or diminution of speed. Such, in the month of August 1834, was my first chance interview with that singular and beautiful phenomenon which I have called the Wave of Translation. Russell’s Wave of Translation • • Experiments showed that the solitary wave speed was proportional to height. Data conflicted with contemporary fluid dynamics (by big deals like Newton) http://www.bbc.co.uk/devon/content/images/2007/09/19/horse_465x350.jpg Solitons What are Solitons? Russell’s Wave of Translationwas largely ignored until the 1960s. 1965 Numerical integration by Zabusky & Kruskal Korteweg-de Vries (KdV) Equation QuickTime™ and a PNG decompressor are needed to see this picture. Speed is proportional to height QuickTime™ and a PNG decompressor are needed to see this picture. QuickTime™ and a PNG decompressor are needed to see this picture. Balanced solitary wave solutions to nonlinear evolution equations Solitons What are Solitons? Russell’s Wave of Translationwas largely ignored until the 1960s. 1965 Numerical integration by Zabusky & Kruskal Korteweg-de Vries (KdV) Equation Speed is proportional to height Collision between two solutions Minimal energy loss QuickTime™ and a PNG decompressor are needed to see this picture. Both solitary waves recovered Balanced solitary wave solutions to nonlinear evolution equations that survive collisions Summarizing Solitons Special solutions to nonlinear evolution equations (PDEs) that: • • Are stable solitary waves (pulses/localized excitations) Maintain their shape under interaction/collision/nonlinear superposition Solitary waves Solit Collide like particles electrons, muons, ping pongs ons Summarizing Solitons Special solutions to nonlinear evolution equations (PDEs) that: • • Are stable solitary waves (pulses/localized excitations) Maintain their shape under interaction/collision/nonlinear superposition Solitary waves Collide like particles electrons, muons, ping pongs Solitons in Nature Alphabet Waves • • Not as unusual as once thought May play a role in tsunami and rogue wave formation What I Did On My Summer Vacation http://www.douglasbaldwin.com/nl-waves.html (Speculated) Solitons in Nature Morning Glory Clouds Jupiter’s Red Spot Strait of Gibraltar Deep and shallow water waves, plasmas, particle interactions, optical systems, neuroscience, Earth’s magnetosphere... I will use solitons to describe solutions to integrable nonlinear equations generated by the Darboux Tranformation method. http://en.wikipedia.org/wiki/File:MorningGloryCloudBurketownFromPlane.jpg http://www.universetoday.com/15163/jupiters-great-red-spot/ http://www.lpi.usra.edu/publications/slidesets/oceans/oceanviews/slide_13.html Solving Integrable Equations Integrable nonlinear systems can be characterized by the Lax formalism Lax Form Analytic Methods • • • • Inverse Scattering Transform (AKNS Method) Zakharov-Shabat Method Bäcklund Transformation Darboux Transformation Solving Integrable Equations Integrable nonlinear systems can be characterized by the Lax formalism Lax Form Seed solution Darboux Transformation New solution Generates soliton solutions! Solving Integrable Equations Integrable nonlinear systems can be characterized by the Lax formalism Lax Form Seed solution 1. Solve linear Lax equations 2. Construct Darboux matrix New solution Solving KdV Equation First-Order Soliton Solution Seed solution 1. Solve linear Lax equations 2. Construct Darboux matrix New solution QuickTime™ and a PNG decompressor are needed to see this picture. Darboux parameter determines velocity/height Solving KdV Equation Second-Order Soliton Solution QuickTime™ and a PNG decompressor are needed to see this picture. Seed solution 1. Solve linear Lax equations but potentially hard!! 2. Construct Darboux matrix New solution QuickTime™ and a PNG decompressor are needed to see this picture. Darboux parameters determine velocities/heights Solving KdV Equation Nonlinear Superposition faster first-order soliton slower first-order soliton Useful for colliding/combining solutions with desirable properties for more complicated systems like short optical pulses QuickTime™ and a PNG decompressor are needed to see this picture. Algebraic Nonlinear Superposition Rule QuickTime™ and a PNG decompressor are needed to see this picture. Darboux parameters determine velocities/heights Short Optical Pulse Propagation Dipole moment operator d Wavefunction ψ (pure states) Density matrix ρ (mixed states) 2 1 Long Collection of Atoms Short Optical Pulse Propagation Optical frequency ω Slowly-varying envelope E Laser Pulse Short Optical Pulse Propagation Laser Pulse 2 Optical frequency ω Dipole moment operator Slowly-varying envelope 1 E Rabi frequency Resonant Atoms Ω dE d Short Optical Pulse Propagation Laser Pulse Resonant Atoms 2 Optical frequency ω Dipole moment operator Slowly-varying envelope d 1 E Rabi frequency Ω Pulse area time dE Short Optical Pulse Propagation Short pulses allow us to neglect atomic decay mechanisms and focus on coherent effects 2 Dipole moment operator Slowly-varying envelope Atom-field coupling d E μ 1 Ω dE Rabi frequency von Neumann’s equation Maxwell’s slowly-varying envelope equation Integrable Nonlinear Evolution Equations First-Order Soliton Solution Darboux Transformation Method Zero-Order Soliton Solution 1. Solve linear Lax equations 2. Construct Darboux matrix First-Order Soliton Solution First-Order Soliton Solution McCall-Hahn Self-Induced Transparency (SIT) Pulse 2 2 2 2 1 1 1 1 Zero-Order Soliton Solution Darboux Transformation First-Order Soliton Solution 1 0 First-Order Soliton Solution McCall-Hahn Self-Induced Transparency (SIT) Pulse Pulse travels at a reduced group velocity in the medium 2 QuickTime™ and a PNG decompressor are needed to see this picture. Temporal pulse width Absorption coefficient The 2 -area hyperbolic secant pulse shape induces a single Rabi oscillation in each atom 1 0 1 Two-Frequency Pulse Propagation in Three-Level Media 3 1 2 Opportunities for interesting dynamics and pulse-pulse control Two-Frequency Pulse Propagation in Three-Level Media Nonlinear Evolution Equations First-Order Soliton Solution Q-Han Park , H. J. Shin (PRA 1998) B. D. Clader , J. H. Eberly (PRA 2007, 2008) Equal atom-field coupling parameters 3 Control Signal 1 2 Two-Frequency Pulse Propagation in Three-Level Media Nonlinear Evolution Equations First-Order Soliton Solution Q-Han Park , H. J. Shin (PRA 1998) B. D. Clader , J. H. Eberly (PRA 2007, 2008) Equal atom-field coupling parameters Same form as two-level equations Temporally matched pulses 3 Control Signal 1 2 First-Order Soliton Solution Darboux Transformation Method Warning! Soliton solutions are labelled by the number of applications of the Darboux transformation. Order corresponds to maximum number of solitary waves of a particular frequency. Zero-Order Soliton Solution Darboux Transformation First-Order Soliton Solution First-Order Soliton Solution Darboux Transformation Method 3 1 3 3 2 1 2 1 3 2 1 2 Zero-Order Soliton Solution Darboux Transformation First-Order Soliton Solution First-Order Soliton Solution Optical Pulse Storage 13 QuickTime™ and a Photo - JPEG decompressor are needed to see this picture. 23 Ratio of pulses at any x is given by a simple relationship 11 Absorption Depths 13 33 23 QuickTime™ and a Photo - JPEG decompressor are needed to see this picture. 22 12 Important for finite-length media Absorption Depths 3 3 Signal 2 1 Slow SIT pulse Signal Control 2 1 Both pulses active 3 Control 2 1 Decoupled pulse First-Order Soliton Solution Optical Pulse Storage 13 23 11 Absorption Depths 13 33 23 Long-lived atomic ground states store pulse information 22 12 Absorption Depths Interesting, but what else can we do with it? Second-Order Soliton Solution 3 Control Signal 1 2 Seed solution 1. Solve linear Lax equations but potentially hard!! 2. Construct Darboux matrix New solution Second-Order Soliton Solution Two first-order soliton solutions Algebraic Nonlinear Superposition Rule Second-order soliton solution Second-Order Soliton Solution Optical Pulse Storage and Memory Manipulation 3 3 Control Signal 1 Control 2 Signal and control pulse durations 1 Control pulse duration Two first-order soliton solutions Algebraic Nonlinear Superposition Rule Second-order soliton solution 2 Second-Order Soliton Solution Optical Pulse Storage and Memory Manipulation 3 3 Control Signal 1 2 Signal and control pulse durations Control 1 2 Control pulse duration Warning! The concept of collision is much more complicated than it was for the KdV equation. We should think carefully about when we want the fastermoving control pulse to catch up with the slower storage solution If we are clever, we can arrange for the signal pulse to be stored before the faster-moving control pulse catches up. Second-Order Soliton Solution Anticipated Behavior If we are clever, we can arrange for the signal pulse to be stored before the faster-moving control pulse catches up. Before Collision After Collision Faster-moving control pulse catching up to the storage solution Faster-moving control pulse moving ahead of the pulse storage solution How will the imprint change? Second-Order Soliton Solution Analytic Results • We can choose integration constants cleverly so the signal pulse is stored before the new control pulse arrives/collides • Faster-moving control pulse hits the stored signal pulse imprint and recovers the stored signal pulse • Recovered signal pulse soon re-imprinted at a new location • Distance the imprint is moved is given by the phase lag Second-Order Soliton Solution Analytic Results Relation to finite-length media Ratio Location of original imprint fixed by injected pulse ratios Distance the imprint is moved is given by the phase lag New imprint location is Warning! If these guides are unreliable, we may push the imprint too close to the edge of the medium – recovering part of the signal pulse before we are ready for it! Numerical Solution High-Bandwidth Optical Pulse Control Step 1: Pulse Storage Signal pulse area 13 23 Control pulse area Pulse ratio Imprint location Theoretical Numerical Percent Error 11 22 12 QuickTime™ and a GIF decompressor are needed to see this picture. 33 13 23 Absorption Depths Numerical Solution High-Bandwidth Optical Pulse Control Step 1: Pulse Storage Signal pulse area 13 Control pulse area 23 Pulse ratio 11 22 12 33 13 23 Imprint location Theoretical Numerical Percent Error Original Imprint Absorption Depths Numerical Solution High-Bandwidth Optical Pulse Control Step 2: Memory Manipulation Control pulse area 13 23 Control pulse duration Distance Moved Theoretical Numerical Percent Error 11 22 12 QuickTime™ and a PNG decompressor are needed to see this picture. 33 13 23 Absorption Depths Numerical Solution High-Bandwidth Optical Pulse Control Step 2: Memory Manipulation Control pulse area 11 22 12 Control pulse duration Distance Moved Theoretical Numerical 11 22 12 Original Imprint Manipulated Memory Absorption Depths Percent Error New location Our second-order soliton solution gives us remarkably tight control of the imprint! Numerical Solution High-Bandwidth Optical Pulse Control Step 3: Pulse Retrieval Control pulse area Signal pulse is recovered! 13 Control pulse duration 23 Choose control pulse width so that the new storage location is outside the boundary of the medium 11 22 12 Distance Moved Theoretical New location QuickTime™ and a PNG decompressor are needed to see this picture. 33 13 23 Absorption Depths well outside medium. Conclusions • Demonstrated control possibilities to convert optical information into atomic excitation and back again, on demand, without adiabatic or quasi-steady state conditions • Focused on broadband pulses, enabling faster pulse-switching and higher clock-rates • Combined numerical and analytical methods to develop a novel three-step procedure to store, move, and retrieve a signal field with high-fidelity • Our new, second-order soliton solution indicates how to control the imprint location by adjusting injected pulse ratios and temporal durations • Numerical studies indicate the general procedure works even for nonidealized input conditions, including pulse areas and shape Experimental Realizations Lifetime ~ 26 ns F=3 2 5 P3/2 Lifetime ~ 26 ns 2 5 P3/2 266.650 MHz F=3 F=3 266.650 MHz F=2 156.947 MHz 72.2180 MHz 2 5 P3/2 266.650 MHz F=2 Lifetime ~ 26 ns F=2 156.947 MHz 156.947 MHz F=1 F=0 72.2180 MHz F=1 F=0 72.2180 MHz F=1 F=0 384.230 THz - 2.56005 GHz 384.230 THz 384.230 THz 384.230THz + 4.27168 GHz F=2 F=2 2.56301 GHz 2.56301 GHz 2 2.56301 GHz 2 2 5 S1/2 5 S1/2 5 S1/2 6.83468 GHz 4.27168 GHz 6.83468 GHz Rb D2 Line Transition Focus on coherent effects by using laser pulses shorter than excited-state lifetime 𝜏 < 26 ns 6.83468 GHz 4.27168 GHz F=1 87 F=2 4.27168 GHz F=1 Two-Level Model Large bandwidth pulse cannot resolve ground or excited hyperfine states 𝜏 < 2 ns F=1 Three-Level Model Pulse bandwidth chosen to resolve ground but not excited hyperfine states 0.15 ns < 𝜏 < 2 ns Integrable MaxwellBloch Equations Lax Form Maxwell-Bloch Equations Lax Operators Traveling Wave Coordinates Second-Order Soliton Solution Storage and Retrieval 13 23 QuickTime™ and a Photo - JPEG decompressor are needed to see this picture. Absorption Depths κx 13 23 11 22 12 Absorption Depths κx Absorption Depths κx 13 23 11 22 12 Second-Order Soliton Solution Two First-Order Soliton Solutions Linear (but potentially hard!!) Lax equations Darboux Transformation Nonlinear Superposition Rule Hermitian unitary matrix combines two first-order soliton solutions no integration required!! Second-Order Soliton Solution Second-Order Soliton Solution second-order soliton Rabi frequency QuickTime™ and a PNG decompressor are needed to see this picture. Asymptotic behavior of the solution The phase lag is the only remnant of the collision Second-Order Soliton Solution second-order soliton Rabi frequency first-order soliton Rabi frequency first-order soliton Rabi frequency QuickTime™ and a PNG decompressor are needed to see this picture. Asymptotic behavior of the solution The phase lag is the only remnant of the collision EIT