Nano Electronic Materials Research Group
Novel technique
One of the major barriers in studying the optical properties of semiconductor QDs is to isolate the spectral features of a single QD out of a large inhomogeneous ensemble. This barrier has been overcome by either using near field probes or micron scale apertures to reduce the number of QD under observation. Although near-field techniques provide very high spatial resolution, their poor collection efficiency and technological complexity severely limit overall effectiveness and flexibility to combine with other existing optical spectroscopic techniques. In far-field optics based techniques, micron scale apertures have been successfully utilized to the study the spectroscopic properties an individual QD in low density SAQD's and naturally formed QD system. However, for a high density QD ensemble, an aperture of 1 micron in diameter can only reduced the number of probed QD's to ~ a hundred. The spectra collected through the aperture usually composed of a group of atomically sharp spectral lines. It is essential to further isolate the spectral feature of a single QD out of that jungle of peaks.
Spectroscopic Imaging
We accomplished this task by exploiting the fact that the position of the diffraction limited image of a point light source can be determined with nanometer scale much below the diffraction limit. We used conventional low temperature microphoto-luminescence (µPL) imaging system with Ti:Sapphire laser beam incident at the 30º inclined to the sample surface. The images of QD's collected by the imaging optics overlapped each other in the real space. However, since each quantum dot usuallyemits at different wavelengths, we can dispersed their PL peak images without distorting their spatial positions along the axis perpendicular to the spectral axis by using imaging spectrograph. Liquid nitrogen cooled charged coupled device detector (CCD) mounted at the exit plane of the spectrograph can collect spectral images containing both spatial and spectral information. Spatial mis-alignment of PL peak in the magnified spectral image shown in Fig(b) (peak #1-#4) clearly demonstrated this fact. The exact spatial origin of these peaks can be determined by fitting their intensity distribution along the lines parallel to x-axis with the point-spread function as shown in Fig.(c). In this way, we can determine the position of the PL peak with precision of ±0.1 pixel (±40nm). This method is most effective in locating spectrally isolated PL peak.
2 Dimensional Mapping of QDs at Nano-meter Scale.
Two Dimensional Mapping of QDs under an aperture with 2 mm diameter is achieved by scanning the real space image of the aperture across the slit of the imaging spectrograph with 200nm steps. A spectral image like the one shown in the previous figure was taken at each scanned position. Two dimensional map was reconstructed from the spectra taken at each position. Fitting the intensity in x and y direction gives the center position of the QDs within the aperture. Center position and size of the ellipses in the right figure correspond to the position of the QDs and uncertainty in position given by fitting process.
Cross Sectional NanoPL.
Cross-sectional nano-PL is yet another scheme to perform single QD spectroscopy. As the name suggests, we perform the PL measurements on the cleaved edge of MBE grown sample. The sample can be arranged in two geometries. Each has its own advantages. In the first case we aligned the sample so that the QD layer is perpendicular to the spectrograph slit as shown in the top schematic. By narrowing the spectrograph slit, one can limit the observation region down to a few hundred nanometers (slit size divided by the magnification). This significantly reduced the effective probing volume as the other lateral dimension is automatically confined by the sample structure and the probing depth along the z-direction is limited by the adsorption length (<1 micron). The spatial location of QDs along one dimension can be further separated by scanning either the sample or the focusing lens mounted in front of the imaging spectrograph as the one shown in the top figure. This scheme of cross-sectional nano-PL can be adopted with other types of detectors such as linear array or single element detectors. Alternatively, when CCD is available, one can orient the QD layer along the slit of imaging spectrograph as shown in the bottom schematic and to use one of the dimensions of the CCD array to determine the spatial location of the QD in the same manner as the top-view nano-PL [bottom figure]. In either geometry of cross-sectional PL, one eliminates the need to use micron/submicron apertures to reduce the effective probing volume. By using this new method, we studied the quantum confined Stark effect in single stacked and vertically coupled double stacked SAQDs. The results are submitted to be published in APL.
Nano Electronic Material Research Group
