Carrier-diffusion-limited remote optical addressing of single quantum

Abstract. We present a scheme for remotely addressing single quantum dots (QDs) by means of near-field optical microscopy that ... ity of confining photons to extremely small dimen- ... pled to a laser source and is positioned in the near field.
231KB taille 2 téléchargements 240 vues
Available online at www.sciencedirect.com

Physica E 21 (2004) 219 – 222 www.elsevier.com/locate/physe

Carrier-diusion-limited remote optical addressing of single quantum dots M. Bruna , N. Chevaliera , A. Drezeta , S. Huanta;∗ , J.-F. Mottea , H. Marietteb , M. Starka , F. Tinjodc , J.C. Woehla a Laboratoire

de Spectrometrie Physique, Universite Joseph Fourier Grenoble and CNRS, BP 87, St Martin d’H&eres 38402, France b Nanophysics and Semiconductors, CEA-CNRS-UJF group, LSP, St Martin d’H& eres 38402, France c Nanophysics and Semiconductors, CEA-CNRS-UJF group, CEA Grenoble, 38054 Grenoble Cedex 9, France

Abstract We present a scheme for remotely addressing single quantum dots (QDs) by means of near-6eld optical microscopy that simply makes use of the polarization of light. A structure containing self-assembled CdTe QDs is covered with a thin metal 6lm presenting sub-wavelength holes. When the optical tip is positioned some distance away from a hole, surface plasmons in the metal coating are generated which, by turning the polarization plane of the excitation light, transfer the excitation towards a chosen hole and induce emission from the underlying dots. In addition, our procedure gives valuable insight into the diusion of photo-excited carriers in the QD plane that can put limits to the addressing scheme. ? 2003 Elsevier B.V. All rights reserved. PACS: 78.67.−n; 07.79.Fc; 73.20.Mf Keywords: Quantum dots; Near-6eld optics; Optical addressing

1. Introduction Ultimate control of light requires the combined ability of con6ning photons to extremely small dimensions, i.e., much smaller than the wavelength, and manipulating them in a well-controlled state to address at will optically active single nanometre-scaled objects such as semiconductor QDs. In this paper, we present a method aimed at remote addressing of single QDs by means of near-6eld scanning optical microscopy (NSOM). The principle of our method is as follows [1]. A thin opaque



Corresponding author. Fax: +33-476-63-5495. E-mail address: [email protected] (S. Huant).

metal 6lm hollowed by sub-wavelength apertures is deposited onto the QD sample. An NSOM tip is coupled to a laser source and is positioned in the near 6eld of a well-chosen nanohole as shown schematically in Fig. 1, or at a well-controlled position on the metal 6lm (a typical “feedback” tip-to-surface distance at low temperature is 15 nm). The 6rst possibility allows us to gain valuable insight into photo-carrier diusion in the QD plane as described in Section 3. The second possibility allows for selective launching of surface plasmons towards a chosen nanohole by orienting the light polarization at the apex tip accordingly. Subsequent plasmon scattering at the hole boundaries allows to recover the plasmon energy under the form of light with the initial wavelength [2]. We demonstrate in Section 4 that this selective

1386-9477/$ - see front matter ? 2003 Elsevier B.V. All rights reserved. doi:10.1016/j.physe.2003.11.003

220

M. Brun et al. / Physica E 21 (2004) 219 – 222

microscope used in this study has been described previously [4]. In this microscope, the spatial resolution is ensured by using the optical tip for local excitation, while the luminescence collection is towards the far 6eld without particular spatial resolution.

3. Photo-carrier diusion

Fig. 1. Principle of the experiment. An NSOM tip is brought in the near 6eld of the hollowed metal 6lm and positioned at some selected point, e.g. in front of a particular nanohole to induce luminescence emission from the underlying semiconductor QDs, or at any point far away from any hole.

launching and scattering of 2D plasmons allows for remotely addressing the underlying QDs.

Information on the diusion of photo-excited carriers in the QD plane is gained by launching the optical excitation into one nanohole, collecting the luminescence from the QDs located underneath this hole, and repeating this procedure for various holes. An example of such a study is shown in Fig. 2 where the NSOM optical tip has been successively positioned in the near 6eld at four dierent points on the surface. When the tip is positioned directly above one nanohole, e.g. point B or C, the spectra consist of a set of very sharp and intense peaks. In contrast, there is almost no luminescence when the tip is positioned far from any hole (distance in excess of 1 m), e.g. point A (the very weak residual luminescence can possibly

2. Sample and experimental details We used self-assembled CdTe semiconductor QDs grown by atomic-layer epitaxy [3]. The growth sequence includes a ZnTe buer layer, a 2.1-nm-thick CdTe layer, and a 30-nm ZnTe cap layer. In the CdTe/ZnTe system, carrier con6nement to zero dimension is ensured by Cd-rich islands (typical density of 5 × 1010 cm−2 , average height and width of 7 and 15 nm, respectively) that form due to elastic relaxation of the strained CdTe layer. A hollowed metal 6lm is deposited onto the semiconductor surface according to the following procedure. We 6rst disperse commercially available latex beads with a diameter of 240 nm on the surface, then evaporate a 70-nm-thick aluminum layer by vacuum evaporation, and 6nally dissolve the latex nanospheres in an organic solvent. Optical information is subsequently gained from this QD medium by a far-6eld collection of the low-temperature (all measurements presented here have been performed at 4:2 K) luminescence emitted by single CdTe QDs located underneath a selected nanohole. The low-temperature NSOM

Fig. 2. A 2 m × 2 m reLection image (left) of the aluminum mask with two nanoholes separated by 800 nm. They are seen as dark spots and marked with two circles. The optical tip is successively positioned above four points labelled A–D. Point A is far away from any hole, point B is on the upper nanohole, point C is on the lower hole and point D is located half-way between the two holes. The corresponding luminescence spectra are shown on the right. The intensity scales are given by double arrows on the right. All spectra use the same excitation ( = 515 nm) intensity estimated at 30 nW at the tip apex. The optical aperture at the tip is 100 nm.

M. Brun et al. / Physica E 21 (2004) 219 – 222

221

be induced by light diusion through the metal grains in the rough Al mask), because QDs can hardly be excited through the 70-nm-thick aluminum mask. The emission spectra for points B and C are typical for radiative recombination of excitons or exciton complexes taking place in a few (typically 5 –10) QDs located under the metal mask [4]. Close inspection of spectra B and C in Fig. 2 shows that there is no spectral overlap of individual sharp peaks which turn out to be unique for the set of QDs that are directly excited by the optical tip through the nanohole. As such, the luminescence spectrum of a nanohole, or more precisely of the QDs located under this nanohole, forms a 6ngerprint for this nanohole. This allows us to infer an upper limit of 800 nm for the diffusion of photo-excited carriers in the semiconductor structure.

4. Remote optical addressing of QDs The very fact that a substantial luminescence intensity is collected when the optical tip is located half-way between nanoholes B and C, i.e. point D in Fig. 2, is intriguing. A close inspection of spectrum D shows that it is a spectral combination of spectra B and C or, in other words, that QDs under nanoholes B and C are simultaneously excited. Therefore, the optical excitation is mediated to the nanoholes by some process. As shown in Fig. 3, this process is light-polarization dependent: an electric-6eld polarization directed towards the holes magni6es the luminescence intensity. In addition, the comparison of the radial (tip-to-hole distance) and angular (polarization) dependence of the luminescence intensity measured for a single nanohole geometry with the calculated surface plasmon intensity [1] con6rms the surface plasmon interpretation in which the optical excitation is mediated by surface plasmons from the tip to the nanohole.

5. Conclusion Our approach for remotely addressing single nano-objects has been demonstrated on CdTe QDs.

Fig. 3. A second 2 m × 2 m reLection image of the aluminum 6lm with two nanoholes on the upper and left borders. It is found that rotating the light polarization when the NSOM optical tip is positioned above the upper hole, point 1, has no eect on the luminescence intensity collected from the underlying QDs. When the tip is positioned away from the hole, points 2–5, there is a preferential polarization (marked with an arrow), pointing towards the hole, that enhances dramatically the luminescence intensity.

However, there is no restriction to apply it to other optically active nano-objects (single molecules, nanocrystals) in dierent environments. In addition, most of the parameters of the hollowed mask (e.g., the nanohole diameter or the hole density) can be easily tuned over large ranges. It would also be of interest to use better plasmon metals other than aluminum in which the propagation length of plasmons is limited to approximately 4 m in the green region of the optical spectrum. In this respect, silver would be a good candidate by allowing for addressing the objects from more remote positions. Finally, it is worth noting that our method, although versatile, can be limited by the photo-carrier diusion described in Section 3. For all of these reasons, we expect the new nano-optical phenomenon reported in this paper

222

M. Brun et al. / Physica E 21 (2004) 219 – 222

to 6nd important optical applications in the nanoworld [5]. Acknowledgements We acknowledge the support from IPMC Grenoble. MS acknowledges support from the Alexander-vonHumbolt foundation and from CNRS.

References [1] M. Brun, A. Drezet, H. Mariette, J.C. Woehl, S. Huant, Europhys. Lett. 64 (2003). [2] C. SQonnichsen, A.C. Duch, G. Steininger, M. Koch, G. von Plessen, J. Feldmann, Appl. Phys. Lett. 76 (2000) 140. [3] L. Marsal, et al., J. Appl. Phys. 91 (2002) 4936. [4] M. Brun, S. Huant, J.C. Woehl, J.-F. Motte, L. Marsal, H. Mariette, Solid State Commun. 121 (2002) 407. [5] J.R. Guest, et al., Science 293 (2001) 2224.