Selective Optical Generation of Coherent Acoustic Nanocavity Modes M. F. Pascual Winter, 1,2 G. Rozas, 1 A. Fainstein, 1, * B. Jusserand, 2 B. Perrin, 2 A. Huynh, 2 P. O. Vaccaro, 3,† and S. Saravanan 3 1 Centro Ato ´mico Bariloche and Instituto Balseiro, C.N.E.A., 8400 S. C. de Bariloche, R.N., Argentina 2 Institut des Nanosciences de Paris, CNRS, Universite ´s Paris 6 et 7, Campus Boucicaut, 140 Rue de Lourmel, 75015 Paris, France 3 ATR Wave Engineering Laboratories, 2-2-2 Hikaridai, Keihanna Science City, Kyoto 619-0288, Japan (Received 30 January 2007; published 28 June 2007) Femtosecond pump-probe experiments on a Ga 0:85 In 0:15 As nanocavity enclosed by two Ga 0:85 In 0:15 As=AlAs phonon Bragg mirrors reveal selective generation of terahertz confined acoustic modes and regular folded phonons. Selective generation of the confined modes alone is achievable for laser excitation at certain energies below the mirror absorption edges, corresponding to electronic transitions within the cavity layer only. Calculations based on the photoelastic effect explain the experimental results. Decay times of cavity and regular modes evidence longer decay times and anharmonic effects for the cavity mode. DOI: 10.1103/PhysRevLett.98.265501 PACS numbers: 63.22.+m, 43.38.+n, 78.47.+p, 78.67.Pt Acoustic nanocavities opened the possibility of confine- ment and amplification of acoustic phonons in the megahertz-terahertz range [1]. They are interesting in the development of coherent monochromatic acoustic sources and as feedback systems for sound amplification by stimu- lated emission of radiation. Experiments on the coherent control of acoustic phonons have shown attractive poten- tialities [2], particularly if performed on monochromatic oscillations. Important progress has been recently achieved in the search for monochromatic sources. Emission of terahertz phonons in superlattice nanostructures under ver- tical electron transport has been reported [3]. Furthermore, phonon propagation through a nanocavity has been studied by means of transmission experiments by picosecond acoustics where broadband phonon generation is attained by strong light absorption in an aluminum cap layer [4]. In the present work, we provide a method for selective optical excitation of coherent nanocavity modes by direct genera- tion in the heterostructure. Direct generation presents two main advantages. First, resonance to electronic transitions provides selectivity in the generation of the cavity modes, allowing all other modes to be silenced by appropriate selection of laser energies. The generation process is thus selective with respect to both the laser and the phonon energies. Second, it avoids terahertz energy limitations intrinsic to metallic layer light-to-sound transducers. The reported experiments give access to the dynamics of the vibrations, enabling studies of the decay times and anhar- monicity affecting the cavity modes. A nanocavity consists of an intermediate layer enclosed by two acoustic Bragg mirrors or superlattices (SLs). Direct generation of sound by femtosecond light pulses in a nanocavity is to be understood in terms of the kind of transducer it provides. Three different mechanisms can lead to the transduction of light pulses into sound: the deformation potential [5], thermal stress [5,6], and photo- elastic processes [7]. In the point that follows, we will only consider the third (equivalent conclusions can be surmised if the other mechanisms are regarded). The quantity that determines the transducing properties of the heterostruc- ture is the photoelastic constant p of the different layers. It is strongly dependent on the light energy, being negligible far below the electronic band gap and varying strongly near the gap. If the pump pulse central energy E L is around the fundamental electronic transition E SL of the SL quantum wells, p will thus result in the profile depicted in Fig. 1(a). Such a quasiperiodic transducer will generate phonon modes with displacement fields u extended throughout the whole heterostructure, as the one shown in the figure. This type of acoustic mode will be referred to as a SL mode, since it does not vary much from the kind of mode into which a perfect SL would transduce a light pulse. On the contrary, a cavity mode presents a displacement field that is maximum in the intermediate cavity layer [1], as shown in Fig. 1(b). To selectively produce such a mode, the transducing properties of the heterostructure need to be enhanced in the cavity layer, as in the photoelastic profile proposed in the lower curve of Fig. 1(b). Such a transducer (a) E E L SL ~ (b) E E L CAV ~ E SL E CAV (c) || u 2 p Transition Energy z || u 2 p z z FIG. 1 (color online). (a) Photoelastic constant (p) profile of the sample along an axis z perpendicular to the layers, and typical displacement field u generated when E L  E SL . (b) Same as (a) for E L  E CAV . (c) Scheme of the fundamental transition energies as a function of the position. PRL 98, 265501 (2007) PHYSICAL REVIEW LETTERS week ending 29 JUNE 2007 0031-9007= 07=98(26)=265501(4) 265501-1  2007 The American Physical Society