Hostname: page-component-7b7567c8c-w7t8q Total loading time: 0 Render date: 2026-11-17T23:13:50.756Z Has data issue: false hasContentIssue false

Temperature Distribution in InGaN-MQW LEDs under Operation

Published online by Cambridge University Press:  13 June 2014

Veit Schwegler
Affiliation:
Dept. of Optoelectronics, University of Ulm, 89069 Ulm, Germany
Matthias Seyboth
Affiliation:
Dept. of Optoelectronics, University of Ulm, 89069 Ulm, Germany
Sven Schad
Affiliation:
Dept. of Optoelectronics, University of Ulm, 89069 Ulm, Germany
Marcus Scherer
Affiliation:
Dept. of Optoelectronics, University of Ulm, 89069 Ulm, Germany
Cristoph Kirchner
Affiliation:
Dept. of Optoelectronics, University of Ulm, 89069 Ulm, Germany
Markus Kamp
Affiliation:
Dept. of Optoelectronics, University of Ulm, 89069 Ulm, Germany
Ulrich Stempfle
Affiliation:
Dept. of Semiconductor Physics, University of Ulm , 89069 Ulm, Germany
Wolfgang Limmer
Affiliation:
Dept. of Semiconductor Physics, University of Ulm , 89069 Ulm, Germany
Rolf Sauer
Affiliation:
Dept. of Semiconductor Physics, University of Ulm , 89069 Ulm, Germany

Abstract

The temperature distribution in InGaN-MQW light emitting diodes was examined during operation with spatially resolved micro-Raman and micro-Electroluminescence measurements. The experimental results were compared to finite element simulations. A good agreement between the different experimental and calculated data is found. Maximum operation temperatures up to 140 °C at a moderate forward currents of 30 mA are revealed by all three independent methods. Influences of substrate thickness, different substrates, and even bond-wires are shown.

Information

Type
Research Article
Copyright
Copyright © 1996 Materials Research Society
Figure 0

Figure 1:Temperature dependence of output power showing the almost exponential decrease of light output with increased temperature.

Figure 1

Figure 2.Temperature distribution derived at different points from μ-EL (current density j=320 A/cm2), the redshift beneath the p-contact is due to increased temperature resulting from higher current density, the reduced redshift beneath the bond wire is due to heat drain effect. For an extended discussion of measuring method and results cf. [4]

Figure 2

Figure 3.Temperature of LEDs under operation, LEDs mounted on teflon heat sink. Comparison of temperatures derived from EL and Raman measurements. In good agreement, both methods yield temperatures as high as 415 K even at moderate currents of 30 mA.

Figure 3

Figure 4.Cross section of temperature distribution in a LED under operational current of 63 mA mounted on metal heat sink obtained with finite element calculation. The elevated temperature beneath the p-contact is clearly visible.

Figure 4

Figure 5.Comparison of spatially resolved temperature profiles obtained from Raman measurements and finite element calculations. Operation conditions as in Figure 4. The data show good agreement of both methods. The impact of a bond-wire on device temperature is demonstrated.

Figure 5

Figure 6.Device simulations with varied substrate thickness or material show the impact of the thermal conductivity of the substrate. Advantages of SiC are demonstrated.