Probing alloy formation using different excitonic species: The particular case of InGaN

Abstract

Since the early 1960s, alloys are commonly grouped into two classes, featuring bound states in the bandgap (I) or additional, non-discrete, band states (II). As a consequence, one can observe either a rich and informative zoo of excitons bound to isoelectronic impurities (I), or the typical bandedge emission of a semiconductor that shifts and broadens with rising isoelectronic doping concentration (II). Microscopic material parameters for class I alloys can directly be extracted from photoluminescence (PL) spectra, whereas any conclusions drawn for class II alloys usually remain indirect and limited to macroscopic assertions. Nonetheless, here, we present a comprehensive spectroscopic study on exciton localization in a so-called mixed crystal alloy (class II) that allows us to access microscopic alloy parameters. In order to exemplify our experimental approach we study bulk InxGa1−xN epilayers at the onset of alloy formation (0 ≤ x ≤ 2.4%) in order to understand the material's particular robustness to point and structural defects. Based on an in-depth PL analysis it is demonstrated how different excitonic complexes (free, bound, and complex bound excitons) can serve as a probe to monitor the dilute limit of class II alloys. From an x-dependent linewidth analysis we extract the length scales at which excitons become increasingly localized, meaning that they convert from a free to a bound particle upon alloy formation. Already at x = 2.4% the average exciton diffusion length is reduced to 5.7±1.3 nm at a temperature of 12 K, thus, detrimental exciton transfer mechanisms towards non-radiative defects are suppressed. In addition, the associated low temperature luminescence data suggests that a single indium atom does not suffice in order to permanently capture an exciton. The low density of silicon impurities in our samples even allows studying their local, indium-enriched environment at the length scale of the exciton Bohr radius based on impurity bound excitons. The associated temperature-dependent PL data reveals an alloying dependence for the exciton-phonon coupling. Thus, the formation of the random alloy can not only directly be monitored by the emission of various excitonic complexes, but also more indirectly via the associated coupling(s) to the phonon bath. Micro-PL spectra even give access to a forthright probing of silicon bound excitons embedded in a particular environment of indium atoms, thanks to the emergence of a hierarchy of individual, energetically sharp emission lines (full width at half maximum ≈ 300 µeV). Consequently, the present spectroscopic study allows us to extract first microscopic alloy properties formerly only accessible for class I alloys. * gordon.callsen@epfl.ch tronic centers (II) evokes the formation of mixed crystal alloys like, e.g., SiGe [14], GaAsP [15], InGaAs [16], AlGaAs [17], InGaN [18], AlGaN [19], CdSSe [20, 21], ZnSeTe [21], and MgZnO [22]. In these cases, no new electronic levels are formed in the bandgap, but rather in the bands themselves, a process often described a...

Experimental quantification of atomically-resolved HAADF-STEM images using EDX

“…Examples of such clusters can be seen in the areas pointed out by white arrows in mapping (i) of Figure 2. The clusters in the same well are far apart enough to create localized emission centers, given the low hole diffusion lengths typically observed in InGaN alloys[27,31], without any other phase separation or clustering mechanism to have to be taken into account.Between the two aforementioned peaks of GaN and InGaN, the histogram in Figure2 (a)is not typical of homogeneously graded interfaces between InGaN and GaN. Instead of continuous tail, the histogram displays additional peaks at distinct indium compositions.To obtain a good fit to the histogram, a total of six more peaks needs to be added to the sum of Gaussian curves used to fit the peaks of the cores of the GaN barriers and the InGaN wells.…”