Photoluminescence of a Single-Crystal Silicon Quantum Well

  • PDF / 510,113 Bytes
  • 6 Pages / 414.72 x 648 pts Page_size
  • 94 Downloads / 221 Views

DOWNLOAD

REPORT


oxide layer, with slightly greater variation in layer thicknesses. The primary advantage of the tripleimplant wafer was the absence of silicon islands in the buried oxide near the substrate 6 . All wafers had (100) orientation. Pieces of SIMOX were oxidized in a quartz tube furnace (dry 02, 1 atm., 11000 C). Hightemperature oxidation in 02 was chosen because it produces flat interfaces with a low density of midgap states. 7 We found it useful to prepare layers with gradually tapered thickness. This permitted us to measure the photoluminescence spectrum and efficiency vs. average layer thickness, d, after each process (oxidation, anneal, etc.), thereby separating the effects of layer thickness and surface chemistry. Tapered samples were produced by performing the initial oxidation in a spatial temperature gradient. The resulting structure is illustrated schematically in the left inset of Fig. 1. The silicon layer thickness was characterized by near-normal incidence reflection spectra (beam size 1.5 x 2 mm), which were fitted to a three-layer model using bulk values for the index of refraction of c-Si 8 and a-SiO 2 .9 Spectra from standard wafers were significantly distorted by the silicon islands near the back of the buried oxide, making accurate determination of the layer thickness impossible. The triple-implant wafer yielded excellent fits, as illustrated in Fig. 1,which shows reflection spectra taken at several positions, z, along a tapered sample. The inset at the lower right of Fig. 1 shows d(z) for this sample. From the smooth, monotonic character of the curve and its agreement with the calculated profile based on the temperature-dependent oxidation constants for Si(100), we conclude that the reflection spectra and 3-layer fits give accurate average values of layer thickness.

800 nm

500 nm

200 nm

300 nm

10-1 10-2

10-3 U

w

10-4

1-5

1

2

3

4

5

6

7

hv (eV)

Figure 1 Reflectivity spectra of a silicon quantum well layer at five positions along direction of taper. Discrete symbols are data; smooth curves are fits to a three-layer model using bulk dielectric functions. Successive curves are offset for clarity. The layer thickness, d, obtained from the fits is indicated at the right, and summarized in the right inset. 982

Information about the uniformity of the layer thickness on the micrometer scale was obtained by scanning a HeNe laser beam focused with a 100x microscope objective (N.A. = 0.9, spot size -0.9 Mm) across the sample at positions for which d < 5 nm and recording reflectivity vs. position over a -25-Mm range. Calculated reflectivity changes are linear with layer thickness variation for d < 5 nm, from which a thickess spread of -1.5 nm peak-to-peak was estimated. Atomic force microscope (AFM) images of the surface made after removal of the upper oxide with buffered oxide etch had a rms roughness of 2 nm, and showed significantly larger min-to-max thickness variations on in-plane length scales of 100-1000 nm. RESULTS Room temperature cw luminescence spectra were taken at three different