{"id":1200,"date":"2013-07-25T18:27:16","date_gmt":"2013-07-25T18:27:16","guid":{"rendered":"https:\/\/mtlsites.mit.edu\/annual_reports\/2013\/?p=1200"},"modified":"2013-08-05T18:02:05","modified_gmt":"2013-08-05T18:02:05","slug":"understanding-the-light-emission-mechanisms-in-ingan-by-correlating-its-structural-and-optical-properties","status":"publish","type":"post","link":"https:\/\/mtlsites.mit.edu\/annual_reports\/2013\/understanding-the-light-emission-mechanisms-in-ingan-by-correlating-its-structural-and-optical-properties\/","title":{"rendered":"Understanding the Light Emission Mechanisms in InGaN by Correlating its Structural and Optical Properties"},"content":{"rendered":"

Solid state light-emitting diodes (LEDs) containing InxGa1-xN (0 \u2264 x \u2264 1)\/GaN hetrostructures as active components are of particular importance because of their ability to emit intense light in the entire visible spectrum. However, the cause of their high light-emission efficiency is a matter of debate. Scanning transmission electron microscopy (STEM) and low-loss electron energy loss spectroscopy (EELS) studies conducted at voltages \u2265 200 kV have attributed strong emission to the formation of \u201clarge-scale\u201d1-3-nm In-rich clusters[1<\/a>]<\/sup>. However, this finding was questioned in a high resolution transmission electron microscopy (HRTEM) study, which argued that observed spinodal decomposition was caused due to knock-on damage by the 200 kV electron beam[2<\/a>]<\/sup>.<\/p>\n

To understand whether or not InGaN alloy in these hetrostructures undergoes spinodal decomposition, we employed atomically resolved aberration (CS)-corrected STEM at 120 kV to demonstrate that this voltage is below the knock-on displacement threshold for InGaN as no discernible electron-beam damage was observed after a continuous exposure for extended periods of time (see Figure 1). To verify whether small contrast variations in the InGaN quantum wells (QWs) seen in Figure 1 correspond to a variation in In composition, \u2206x, we employed low-loss EELS at 120 kV. By scanning the electron beam in QWs, identified as InGaN 1 and InGaN 2 in Figure 2, the energy of the plasmon peak (Ep), which is characteristically composition-dependent, is measured with a spatial resolution of 0.3 nm; Ep is plotted in Figure 2. By measuring \u2206x, for each of the QWs separately, we determined that the measured changes in plasmon energies correspond to compositional variation of 5.4% or less, which is too small to be attributed to the existence of large-scale In-rich clusters. The next step is to correlate these structural and electronic properties with the optical properties with sub-wavelength resolution by employing a custom cathodoluminescence set-up incorporated in a STEM.<\/p>\n\n\t\t