The investigation of optical microcavities has gained considerable interest due to their ability to confine and influence light on the length scale of its wavelength. This project deals with the investigation of a novel type of microcavity, a rolled-up microtube resonator as shown as scanning electron micrograph in the left part of Fig. 1. In our former work we investigated microtubes with emitters that were embedded inside the microtube wall during the growth of its underlying layer system. Our special interest here at the Institute of Physical chemistry is to study the hybrid system consisting of the epitaxially grown rolled-up semiconductor microtube resonators and chemically synthesized nanocrystals (see right Fig. 1) as active medium.
Our semiconductor microtubes are fabricated from strained layer systems grown by molecular beam epitaxy. The strain is built up by growing materials with a larger lattice constant (yellow layer in Fig. 2a) on top of a material with a smaller lattice constant (blue and red layer in Fig. 2 a). By selective wet-chemically etching a sacrificial layer (red layer in Fig. 2 a) below the strained layers the incorporated strain bends up the layer system and leads to the self-rolling of a microtube.
Due to the large refractive index in the order of 3.3 and the transparency of the material below their bandgap multiple total internal reflections (see Fig. 3a) lead to waveguiding of the light inside the microtube wall (see video Fig. 3b) and to the formation of ring modes by constructive interference after a roundtrip. A calculated intensity pattern of such a ring mode in the cross section is shown in Fig. 3 c. The video shows the calculated time evolution of the light wave in a microtube resonator that is emitted by a dipole located inside the top right wall of the microtube.
In order to avoid losses into the substrate microtubes have to freestanding. Applying optical lithography a U-shaped mesa is defined that result in free-standing microtube bridges after rolling-up. This is shown in Fig. 1a and sketched in Fig. 2b.
To couple colloidal nanoemitters that cannot be inserted inside the microtube wall during the growth we use the unique properties of rolled-up microtube resonators. The basic idea (sketched in Fig. 4 a) is as follows: We prepare an “empty” resonator that has no emitters incorporated in the tube walls. Then, a solution of light emitting particles, such as chemically synthesized nanocrystals, is filled into the hollow core of the microtube. The light emission from nanoemitters locate close to the wall can couple to the long into the core ranging evanescent fields of the resonant modes. In the computer simulations shown in Fig. 4 b the areas close to the sidewalls of the microtube where nanoemitters can couple their light into the microtube resonator are shown in red.
In the following two videos the emission of a nanoemitter that is positioned close to the sidewall (left video) or moved slightly into the inside (right video) is shown. One observes that only the emitter close to the sidewall can excite the ring mode inside the microtube wall.
Due to the long-ranging evanescent fields into microtube paired with the property of being hollow microtubes are ideal candidates to be combined with colloidal nanoemitters.
Experiments with Colloidal Nanocrystals
Utilizing a micro syringe mounted in a micro positioning system a solution of PbS nanocrystals is filled into the microtube as shown in Fig. 6 a. The filling level is marked by a black arrow. After fluid filling the microtube is investigated by micro fluorescence spectroscopy. The results are shown in Fig. 6 b. One observes groups of sharp lines representing the resonator modes that are superimpose on the broad emission spectrum of the PbS nanocrystals ensemble. For every group an integer number m of wavelengths fits into circumference of the microtube. The sharp lines within the groups represent modes with different antinodes along the microtube axis. To observe sharp resonator modes one has to constrict the light running out of the microtube in axial direction. Therefore we prepared microtubes as sketched in Fig 7a. By etching small stripes into the U-shaped mesa as marked by the yellow boxes in Fig. 7 b we achieve a thicker ring between two thinner rings at a special axial position of the microtube. Between these two thinner rings the light is confined in axial direction. A spatially and energy resolved fluorescence measurement between the rings is shown in Fig. 7c. One observes an increasing number of axial antinodes between the rings with increasing energy for every group.
The modes form standing waves between the rings like particle waves in a one-dimensional potential. By the technique with the etched rings a full controlled and well localized confinement of light is achieved which is important for further experiments and for applications.