K two systems. fermionic species mixture of two spin states. K 6 Li mass imbalance! cold atoms: superfluidity in Fermi gases

Transcription

1 Bad Honnef, 07 July 2015 Impurities in a Fermi sea: Decoherence and fast dynamics impurity physics: paradigms of condensed matter-physics Fermi sea fixed scalar impurity orthogonality catastrophe P.W. Anderson, Phys. Rev. Lett. 18, 1049 (197) moving scalar impurity heavy-impurity problem A. Rosch, Adv. Phys. 48, 295 (1999) Rudolf Grimm Center for Quantum Physics in Innsbruck fixed impurity with spin ondo problem J. ondo, Progress of Theoretical Physics 32, 37 (194) University of Innsbruck Austrian Academy of Sciences cold atoms: strongly interacting Fermi gas fermionic species mixture of two s cold atoms: superfluidity in Fermi gases collective modes (2004) Duke Innsbruck Fermi condensates (JILA, 2004) Feshbach resonance vortices (MIT, 2005) quenching of moment of inertia (Ibk, 2011) Li two systems and many other fantastic experiments critical velocity (MIT, 2007) second sound (Innsbruck, 2013) much more than superfluidity Fermi polarons: key experiments population-imbalanced Fermi gas ground state at T=0? MIT 2009 Schirotzek et al., PRL 102, 2302 (2009) Observation of Fermi polarons in a tunable Fermi liquid of ultracold atoms Li Innsbruck 2012 ohstall et al., Nature 485, 15 (2012) Metastability and coherence of repulsive polarons in a strongly interacting Fermi mixture Li mass imbalance! Landau s Fermi-liquid picture: quasiparticles as building blocks for the many-body state Cambridge 2012 oschorrek et al., Nature 485, 19 (2012) Attractive and repulsive Fermi polarons in two dimensions 1

2 typ. expt. conditions (impurity regime) spin channels Li: N = 3.5 x 10 5 E F 2 µ T 300 n heavy in Fermi sea of Li : N = 2 x 10 4 E F = 500 n optical trap is 2.5x stronger for Li 155G lowest third-to-lowest nearly homog.!!! Li Fermi energy our leading energy scale! E F /h khz 1/ F 3000 a 0 Feshbach resonance in Li - impurity in a Fermi sea resonance parameters: Naik et al. EPJD 5, 55 (2011) position B 0 = (1) G width = G background scatt. length a bg = 3.0 a 0 differential magnetic moment = 2.35(2) MHz/G resonance length (Petrov) R* = 250(25) a 0 resonance is closed-channel dominated (two-body physics) and has intermediate character in the many-body sense F R* 1 weak (k F a <<1): simple mean-field description? stronger: quasiparticle (polaron) à la Landau Fermi liquid theory? strong (k F a > 1): polaron or molecule? Innsbruck FeLix team (2011) Metastability and coherence of repulsive polarons in a strongly interacting Fermi mixture C. ohstall et al., Nature 485, 15 (2012) theory collaboration energy diagram (T=0) theory: P. Massignan and G. Bruun repulsive polaron Florian Schreck Andreas Trenkwalder Michael Jag Christoph ohstall Pietro Georg Bruun Massignan U Aarhus, Denmark ICFO, Spain Matteo Zaccanti Rudi Grimm E + E m molecule-hole continuum (MHC) E m - F attractive polaron E - 2

3 radio-frequency spectroscopy spectral response in non-interacting in (strongly) interacting 1 ms p-pulse (w/o interaction) radiofrequency NB: different from standard rf probing! repulsive polaron observed up to -1/ F a = -0.3 polaron-molecule transition -1/ F a = 0. spectral response (strong rf) spectral response: what do we understand???? the big mess??? idea for time-domain impurity spectroscopy FeLix group + theory collaborators Harvard USA R. Schmidt M. nap E. Demler Aarhus, Denmark Ramsey spin echo atom interferometer (with non-int. state as phase reference) R. Sørensen G. M. Bruun Monash, Australia Marko Cetina Isabella Fritsche RG Rianne Lous Michael Jag ARHUUS UNIVERSITY Jook Walraven M. Parish J. Levinsen HARVARD UNIVERSITY 3

4 interferometer fringes loss of coherence non-interact. state moderate interaction strength = 1.9(1) interacting state fits to a simple exponential decay how to measure very fast dynamics on resonance? problem: decoherence too fast for rf pulses 10µs 20µs 10µs optical resonance shifting differential light shift induced by trap light Feshbach resonance atom pair trap light up-shifts Feshbach resonance optical resonance shifting! experimental implementation switching between two traps apply rf-pulses somewhat away from resonance (1/ F a 2) 200ns switching time switch equiv. to mg (up to 100mG) loss of coherence on resonance = 0.1(1) decoherence rate T/T F 0.1 Fermi time τ F = ħ/ε F = 4.5 µs initial quantum evolution fits to exponential decay after ~1.5 F how to understand decoherence? 4

5 elastic scattering decoherence rate T/T F 0.1 We have measured the quasiparticle scattering rate in a 3D fermionic system! elastic scattering provides which way information loss of fringe visibility Greenberger, Yasin (1988); Englert (199) theory on quasiparticle scattering without / with medium corrections Christensen and Bruun, PRA 91, (2015) decoherence rate temperature dependence T/T F but what causes the fast decoherence on resonance? Pauli blocking in a degenerate Fermi gas decoherence thermally activated no decoherence in the T=0 limit moderate interaction strength -1/ F a = 2.3 theory on quasiparticle scattering not applicable 0.04 temperature dependence Fermi time: relevant time scale decoherence at T=0: mechanism different from el. scattering on resonance ultracold Fermi gas (E F = 1.7 µ) 4.5 µs τ F = ħ/ε F traditional condensed-matter system (E F = 5 ev) 130 as we can observe the time dynamics of excitations deep in the Fermi sea! 5

6 experiments on fast quantum dynamics what happens on a very fast time scale? τ F = ħ/ε F = 4.5 µs dynamics of quasiparticle formation ( birth of a polaron )? Ramsey interferometer 21 July 200 optical control of interaction birth of a polaron (what we would expect) Ramsey interferometer not on resonance, but not too far away 1 optical resonance shifting! Z T=0 finite T interaction parameter during rf pulses: X 0 5 between rf pulses: X < t / F polaronic regime theory Chevy ansatz, experimental data, functional det. theory on-resonance regime Chevy ansatz X = (X 0 = -3.9) X = +0.8 (X 0 = 5.8) X = (X 0 = 4.8) functional det. expt. data remarkable match

7 interferometer contrast Chevy ansatz vs. functional determinants interacting impurities: what we have learned Chevy ansatz single particle-hole excitations only (no decay into molecular excitations) zero-temperature theory describes finite mass ratio functional determinants multiple particle-hole excitations (no decay into molecular excitations) finite-temperature theory restricted to infinitely heavy impurity (under our conditions problem fixed by R* correction) conventional rf spectroscopy (frequency domain) energies of attractive and repulsive quasiparticle branch lifetime of (metastable) repulsive polaron determination of quasiparticle residue via Rabi oscillations time domain spectroscopy quasiparticle scattering rate ultrafast decoherence on resonance birth of a polaron : dynamics of quasiparticle formation ultrafast dynamics on resonance: observation of beating observation of Anderson s orthogonality catastrophe? initial quantum dynamics future experiments ultrafast impurity dynamics cntn d power-law decay signals OC interactions between impurities high concentration fermions vs. bosons thermal decoher. impurities pinned in a species-specific lattice immobile and infinite mass long-range vs. short-range interactions nap et al., PRX 2, 020 (2012) Thank your for your attention! 7