Comment on “Direct evidence of nanocluster-induced luminescence in InGaN epifilms” [Appl. Phys. Lett. 86, 021911 (2005)

Abstract

In a recent letter, Chang et al. claim to have obtained “Direct evidence of nanocluster-induced luminescence in InGaN epifilms.” The work comprises measurements by x-ray diffraction XRD , scanning electron microscopy SEM , energy dispersive x-ray spectrometry EDS , and cathodoluminescence CL of a 0.2 m thick In0.186Ga0.814N layer on a GaN buffer layer grown by metal organic chemical vapor deposition MOCVD . The authors present an XRD scan on a symmetric 0002 reflection where the InGaN-related diffraction peak is shown to have two components. This observation is interpreted as “clear and direct evidence of phase separation.” Vegard’s law is used directly to obtain the “two” compositions in the InxGa1−xN film, allegedly corresponding to x=0.125 and x=0.223. Subsequently, the authors present an SEM/CL surface image of the InxGa1−xN film where they identify brighter and darker regions on a 1 m length scale. In addition, EDS measurements at electron beam energies of 15 keV in the so-called “nanoclusters” were performed. By comparing the intensity of the In and Ga characteristic x-ray lines the authors argue that the “nanoclusters” observed in the SEM image correspond to In-rich regions. Finally, the paper presents spatially resolved CL results showing a two component CL peak whose relative intensities vary with the beam position in the sample surface. Besides an overstretching of the term “nano,” which is clear even for an audience outside the field see the length scale of 1 m in Fig. 2 , the interpretation and conclusions inferred from the experimental results in Ref. 1 have the potential to mislead unwary readers. Consequently, a more careful examination is mandatory. This comment highlights a few ambiguous issues and attempts to provide some clarification. Starting from the XRD results, it should be stressed that the observation of two out-of-plane lattice constants, c1 and c2, cannot be interpreted as evidence of two phases. This “direct” deduction of phase segregation is, in fact, based on an incorrect application of Vegard’s neglecting the strain effects of in a strained epitaxial system. This common mis-