The Use of Spatial Analysis Techniques in Defect and Nanostructure Studies

Abstract

Spatial analysis techniques were used to investigate defects and nanostructures in the III-nitride system. Dislocations in GaN films were distributed nonrandomly, forming both short-scale and large-scale linear arrays aligned along h11 " 20i: Both low-density InGaN/GaN quantum dots (QDs) and In droplets on the surface of InAlN films were in spatially random positions and showed no spatial autocorrelation, but high-density InGaN/GaN QDs showed a tendency toward short-range ordering and all features showed a nonrandom size distribution. The spatial arrangements of defects and nanostructures relate to their generation processes and may also affect device properties.

The effects of annealing on non-polar (1 1 2¯ 0) a-plane GaN films

“…In Fig. 7, the black lines show a Poissonian distribution, as expected for a random distribution of BSF lengths [26]. The measurements of the BSF lengths were limited by the size of the electron-transparent specimen area; some BSF lengths were necessarily underestimated.…”

“…The data for the as-grown film show no statistically significant deviation from a spatially random distribution, but the data for the annealed film show statistically significant spatial clustering (po0.01) on length scales up to 200 nm (Fig. 4) [26]. Fig.…”

“…2(d). This is shown clearly in the radial distribution plots of the dislocation positions, which were obtained using plan-view images taken from multiple TEM specimens to ensure good statistics [16,25,26]. The data for the as-grown film show no statistically significant deviation from a spatially random distribution, but the data for the annealed film show statistically significant spatial clustering (po0.01) on length scales up to 200 nm (Fig.…”

Microscopy of defects in semiconductors

“…The clustering indicated by the L(r) is probably due to these lines of dislocations. For more information on statistical analysis of the distribution of dislocations the reader is referred to the work of Moram et al who have published detailed studies on the use of spatial analysis techniques for the analysis of defects observed in AFM and CL images, see for example [105,106]. Note that Moram et al employed the AFM techniques for dislocation identification described in Section 10.3.1.2.…”

The effects of varying metal precursor fluxes on the growth of InAlN by metal organic vapour phase epitaxy

“…(However it should be noted that droplet formation is not a significant destination for indium, very roughly 10% of the indium put into the reactor will incorporate into the InAlN epilayer and only 0.01% of it will be left on the surface as droplets.) The positions of these droplets were also measured on all of these samples, and analysed using spatial analysis techniques; they were in spatially random positions and showed no spatial autocorrelation [33].…”