Séminaires à venir / Upcoming

    • Inverse methods for the characterisation of anisotropic porous media and coupled systems, par Jacques CUENCA (LMS International - Siemens Industry Software)

      2019
      12
      12

      SEMINAIRE A 14H

      This seminar aims at providing an overview of activities performed as a cooperation between UEF, KTH and Siemens on the estimation of material and geometrical parameters of complex structures. Particular attention is given to anisotropic poroelastic media and coupled resonant structures, whose behaviour exhibits material orientation symmetries and frequency periodicity, respectively. The estimation of physical properties of these systems is thus hindered by the similarity of their behaviour at different observation orientations or frequencies. From a model inversion point of view, these similarities result in multiple possible solutions or local minima, thus rendering the search for a unique set of model parameters difficult. The present work investigates the use of an incremental model inversion approach, consisting in solving successive sub-problems, starting with a convex asymptotic case, and terminating at the full problem of interest. In the case of resonant coupled systems, a low-frequency asymptotic sub-problem is used as a starting point and the solution is guided towards the true solution by progressively extending the frequency range of the observation. In the present work, two complementary model inversion approaches are used, namely gradient-based optimisation and a Bayesian inversion framework. Examples of impedance tube characterisation of resonant duct elements including porous media will be given.

    • Electron-phonon interaction in a terahertz acoustic phonon cavity, par Sandeep SATHYAN (Post-Doc, LAUM)

      2019
      12
      17

      Interaction of a two-level system (TLS) and acoustic phonons in a semiconductor material could lead to a new quantum entity called ‘Phoniton, ‘[1] The theoretical model proposed to realize the Phoniton is based on a strained Si cavity formed between two distributed Bragg Si/Ge reflectors. To create a suitable TLS in the cavity, they had proposed to split the ground state of “donor” impurity, such as phosphorus in Si using strain. It was proposed that when the TLS is in resonance and is strongly coupled to the cavity, it will be possible to observe Rabi-splitting of the cavity phonon mode. The major challenge to implementing this model in the laboratory is bringing TLS in resonance with the acoustic phonon cavity mode. However, P:Si energy level used in the proposed model as a TLS can’t be tuned once fabricated and this puts some practical difficulties while trying to make the cavity mode in resonance with the TLS.


      As the first step towards the realization of the Phoniton, we have designed an acoustic cavity structure formed using GaAs/AlAs superlattices. In this work we seek the possibility of driving the cavity mode using an external THz acoustic phonon source. We have performed an ultrafast pump-probe experiment on the acoustic cavity structure to study the phonon dynamics [2]. An interesting resonant behavior of ~ 0.5 THz longitudinal acoustic (LA) phonons in the acoustic nanocavity is observed [3]. Acoustic phonons were generated in the top superlattice using ultrafast laser excitation. These modes generated outside the cavity were used to force the cavity into oscillation. We observed a buildup of the driving mode in the cavity. Since the quality factor of the cavity used in this experiment is low, we could not observe the presence of Phoniton experimentally. Even though, the Q of the cavity is low, this experiment paves the way for the observation of a more interesting physical phenomena like phonon analogue of Cavity-quantum electrodynamics.


      Based on the above consideration, we have proposed a ‘Phoniton’ structure using a very high-Q phonon cavity (Q ~ 10^4) with a tunable TLS [4]. We have used moderately doped GaAs/AlAs double quantum well system as the TLS. Energy levels of this TLS can be externally tuned by changing the electrical bias across the quantum wells [5]. The TLS is enclosed in a phonon cavity formed using GaAs-AlAs superlattice (SL) phonon mirrors. When the stark splitting between the ground states of the quantum wells exactly matches the cavity phonon energy, the electrons should continuously shuttle between the ground state of the two quantum wells with the absorption and emission of longitudinal acoustic phonons, leading to the observation of the Rabi splitting of the cavity phonon mode.


      [1]. Soykal Ö O, Ruskov R and Tahan C “Sound-Based Analogue of Cavity Quantum Electrodynamics in Silicon” Phys. Rev. Lett. 107 235502 (2011)
      [2]. Beardsley R, Akimov A V., Greener J D G, Mudd G W, S Sandeep, Kudrynskyi Z R, Kovalyuk Z D, Patanè A and Kent A J “Nanomechanical probing of the layer/substrate interface of an exfoliated InSe sheet on sapphire” Sci. Rep. 6 26970 (2016)
      [3]. S Sandeep, Sarah L Heywood, Richard Campion, Anthony J Kent, Rajeev N Kini, “Resonance of Terahertz phonons in an Acoustic nanocavity” Phys Rev B. 98 235303 (2018)
      [4]. S Sandeep and R N Kini, “Phoniton in a GaAs-AlAs acoustic phonon cavity with a tunable two-level system” Semicond. Sci. Technol. 34 075010 (2019)
      [5]. Kent A J, Kini R N, Stanton N M, Henini M, Glavin B A, Kochelap V A and Linnik T L “Acoustic Phonon Emission from a Weakly Coupled Superlattice under Vertical Electron Transport: Observation of Phonon Resonance” Phys. Rev. Lett. 96 215504 (2006)

    • Modèles acoustiques pour la reconnaissance du locuteur, par Anthony Larcher (LIUM)

      2020
      01
      07

      Le traitement automatique de la parole et la reconnaissance du locuteur en particulier tendent à extraire du signal de parole des informations pertinentes pour la caractérisation des personnes. Nous verrons les hypothèses utilisées dans ce domaine pour modéliser l’information acoustique et l’évolution en cours dans le contexte du changement de paradigme actuel de l’apprentissage profond.

    • Journée RAMDAM, = pas de séminaire

      2020
      01
      14

      + d'infos bientôt

    • Optoacoustic phenotyping of cells and tissues: from cancer to metamaterials, par Thomas Dehoux (Institut Lumière Matière, Lyon)

      2020
      01
      28

      Despite deeper understanding of cancer metabolism, 90% of experimental drugs fail in clinical studies, mostly due to lack of efficacy. This stems from the lack of predictability of in vitro and in vivo models that are used to design generic drugs at preclinical stages, and from the limited histophysiological clues that can guide clinicians in adapting the generic therapy to each patient. At the same time, it is now well established that the mechanical properties of tumors control their physiology and phenotype. In this talk I will first present a novel quantitative, label-free microscopy technique based on Brillouin light scattering (BLS) to decipher the link between mechanical phenotype and drug efficacy.


      In the second part of this talk, I will discuss how the mechanical phenotype of tissues can be harnessed to engineer acoustic metamaterials. Biological composites have been known to possess remarkable mechanical performance that far exceed those of their constituents, due, in part, to their complex, hierarchal microstructure, composed of fibrous polymers, and other supra-molecular arrangements. While more is known about their mechanical properties in quasi-static settings, their phononic properties remain unexplored. In this study, we investigate the phononic behavior of micron-thick onion cells walls using laser generated, sub-GHz surface acoustic waves (SAWs). Our measurements reveal the presence of an acoustic band gap for SAWs, and suggest our biological material behaves as an organic, locally resonant metamaterial. Moreover, we show that this band gap can be phenotypically tuned, and anticipate these findings will yield new biologically-derived, "green", phononic materials.