GaN ultraviolet light emitting diodes - Semantic Scholar

28.12.2011 - Yan, P. Rinke, M. Scheffler, and C. G. Van de Walle, Appl. Phys. Lett. 95, 121111 (2009). 19A. Knauer, H. Wenzel, T. Kolbe, S. Einfeldt, M. Weyers, M. Kneissl, and. G. Tränkle, Appl. Phys. Lett. 92, 191912 (2008). 20T. Kolbe, T. Sembdner, A. Knauer, V. Kueller, H. Rodriguez, S. Einfeldt,. P. Vogt, M. Weyers, ...
519KB Größe 2 Downloads 189 Ansichten
Effect of temperature and strain on the optical polarization of (In)(Al)GaN ultraviolet light emitting diodes Tim Kolbe, Arne Knauer, Chris Chua, Zhihong Yang, Viola Kueller et al. Citation: Appl. Phys. Lett. 99, 261105 (2011); doi: 10.1063/1.3672209 View online: http://dx.doi.org/10.1063/1.3672209 View Table of Contents: http://apl.aip.org/resource/1/APPLAB/v99/i26 Published by the American Institute of Physics.

Related Articles Silicon nanoparticle-ZnS nanophosphors for ultraviolet-based white light emitting diode J. Appl. Phys. 112, 074313 (2012) Strong light-extraction enhancement in GaInN light-emitting diodes patterned with TiO2 micro-pillars with tapered sidewalls Appl. Phys. Lett. 101, 141105 (2012) Electron injection in magnesium-doped organic light-emitting diodes Appl. Phys. Lett. 101, 141102 (2012) Localized surface plasmon-enhanced ultraviolet electroluminescence from n-ZnO/i-ZnO/p-GaN heterojunction light-emitting diodes via optimizing the thickness of MgO spacer layer Appl. Phys. Lett. 101, 142101 (2012) The reduction of efficiency droop by Al0.82In0.18N/GaN superlattice electron blocking layer in (0001) oriented GaN-based light emitting diodes Appl. Phys. Lett. 101, 131113 (2012)

Additional information on Appl. Phys. Lett. Journal Homepage: http://apl.aip.org/ Journal Information: http://apl.aip.org/about/about_the_journal Top downloads: http://apl.aip.org/features/most_downloaded Information for Authors: http://apl.aip.org/authors

Downloaded 04 Oct 2012 to 13.1.101.37. Redistribution subject to AIP license or copyright; see http://apl.aip.org/about/rights_and_permissions

APPLIED PHYSICS LETTERS 99, 261105 (2011)

Effect of temperature and strain on the optical polarization of (In)(Al)GaN ultraviolet light emitting diodes Tim Kolbe,1,a) Arne Knauer,2 Chris Chua,3 Zhihong Yang,3 Viola Kueller,2 Sven Einfeldt,2 Patrick Vogt,1 Noble M. Johnson,3 Markus Weyers,2 and Michael Kneissl1,2 1

Institute of Solid State Physics, Technische Universita¨t Berlin, Hardenbergstraße 36, 10623 Berlin, Germany Ferdinand-Braun-Institut, Leibniz-Institut fu¨r Ho¨chstfrequenztechnik, Gustav-Kirchhoff-Straße 4, 12489 Berlin, Germany 3 Palo Alto Research Center, 3333 Coyote Hill Road, Palo Alto, California 94304, USA 2

(Received 10 October 2011; accepted 5 December 2011; published online 28 December 2011) The temperature and strain dependence of the polarization of the in-plane electroluminescence of (0001) orientated (In)(Al)GaN multiple quantum well light emitting diodes in the ultraviolet spectral range has been investigated. For light emitting diodes with emission wavelength shorter than 300 nm the transversal-electric polarized emission intensity increases relative to the transversal-magnetic emission with increasing temperature, whereas it decreases for ultraviolet light emitting diodes with longer emission wavelength. This effect can be attributed to occupation of deeper valence bands with increasing temperature. In addition, strain also strongly influence the in-plane light polarization of near ultraviolet light emitting diodes. The transversal-magnetic polarized emission becomes more dominant with decreasing in-plane tensile strain of the InGaN/ C 2011 American Institute of Physics. (In)(Al)GaN multiple quantum well active region. V [doi:10.1063/1.3672209] Deep ultraviolet (UV) light emitting diodes (LEDs) based on group-III nitrides are of great interest for a number of applications such as water purification, UV curing, photospectroscopy, and bioanalysis.1,2 However, despite the enormous progress that group-III nitride-based UV LEDs have recently made,3–8 state of the art UV LEDs exhibit external quantum efficiencies (EQE) in the range of only a few percent.9 The limited light extraction efficiency significantly contributes to these low EQEs. For visible and near-infrared LEDs, it was observed that the emission is mainly transverse-electric (TE) polarized (E\c).10 Therefore, the photons can easily escape through the substrate or the surface.10 However, for deep UV LEDs with a high aluminum mole fraction in the (In)(Al)GaN active regions, the main emission is transverse-magnetic (TM) polarized (Ejjc) because of the rearrangement of the valence bands at the C point of the Brillouin zone. With shorter wavelength, the split-off hole band moves closer to the conduction band relative to the heavy and light hole bands, and as a consequence, the TM polarized emission becomes more dominant.11–16 Therefore, less photons can be extracted from the light escape cone via the surface or the bottom of the LEDs (Ref. 8) which results in a relatively poor extraction efficiency. In previous publications,15,16 we have shown that LEDs with an emission wavelength around 300 nm have zero polarization meaning that the TE and TM polarized in-plane emission intensities were equal. For LEDs with shorter emission wavelengths, it was found that the in-plane emitted light is mainly TM polarized and for LEDs with longer wavelength TE polarized, respectively. In this paper, we have investigated the temperature and strain dependence of the optical polarization of UV LEDs. In order to investigate a)

Author to whom correspondence should be addressed. Electronic mail: [email protected].

0003-6951/2011/99(26)/261105/4/$30.00

these effects, the degree of polarization of ultraviolet-A (UV-A) and ultraviolet-B (UV-B) LEDs has been measured for temperatures ranging between 20  C and 170  C. Because the strain has a strong influence on the electronic band structure of LEDs (Refs. 17 and 18), also the strain dependence of the polarization of the in-plane emission of LEDs was investigated. Therefore, the in-plane polarization of LEDs emitting near 380 nm has been investigated. By changing the aluminum and indium mole fraction in the InAlGaN barriers, near UV LEDs with differently strained InGaN/(In)(Al)GaN multi-quantum well (MQW) active regions were compared. The LED heterostructures were grown by metalorganic vapor phase epitaxy (MOVPE) on (0001) c-plane sapphire substrates. For the temperature dependent measurements LEDs with InGaN, InAlGaN, and AlGaN MQW active regions emitting in the UV-A and UV-B spectral ranges have been compared. Details of the LED heterostructures and the growth process can be found in Refs. 19–21. For the strain dependent measurements near UV LEDs with an emission wavelength of around 380 nm were used. The active region of these LEDs consists of five 3.5 nm thick In0.02Ga0.07N quantum wells (QWs) surrounded by 7 nm thick GaN or InxAl0.16Ga0.84xN barriers. The aluminum and indium mole fractions were determined by high resolution x-ray diffraction (HRXRD) from calibration samples. GaN and three different InxAl0.16Ga0.84xN barrier layers with an indium mole fraction x of 0, 0.02, and 0.03 were used in this part of the study. More details can be found in Ref. 19. After MOVPE growth, the samples were annealed in nitrogen ambient in order to activate the p-type conductivity. LEDs were fabricated with standard chip-processing technologies. 100 lm  100 lm mesa structures were defined by inductively coupled plasma etching in order to expose the n-AlGaN surface. A Pd/Au p-contact and a Ti/Al n-contact were deposited to form the p-electrode and the n-electrode, respectively.

99, 261105-1

C 2011 American Institute of Physics V

Downloaded 04 Oct 2012 to 13.1.101.37. Redistribution subject to AIP license or copyright; see http://apl.aip.org/about/rights_and_permissions

261105-2

Kolbe et al.

The electrical and optical characteristics of the LEDs were measured on-wafer at room temperature (barrier series) under direct current (DC) operating conditions without any active cooling. For the temperature dependent measurements, the wafer was placed on a Peltier element. A schematic illustration of the experimental setup is shown in Ref. 15. The in-plane emitted light was collected with a lens and analyzed after an iris diaphragm by a Glan-Taylor prism and an optical fiber spectrometer. The light was then focused with a second lens on the optical fiber. For calibration of the experimental setup, an unpolarized light source with a second polarizer was used. In a first step, the temperature dependence of the polarization of the in-plane emitted light of 288 nm, 303 nm, 320 nm, and 379 nm LEDs has been investigated. The measurements were performed at a fixed DC injection current of 20 mA. Figure 1 shows the change of the peak energy of the electroluminescence spectrum for increasing temperature of the Peltier element between 20  C and 170  C. A nearly linear decrease of the energy with increasing Peltier element temperature can be observed for all samples. This indicates that the sample temperature increases linearly with the temperature of the Peltier element, if we assume a linear band gap vs. temperature relationship in the investigated temperature range.22 Furthermore, the ratio of the peak shift is nearly the same for all samples which means that the temperature change for all samples is nearly identically. Therefore, the temperature dependencies of the in-plane polarization of the individual samples can be compared to each other. Figure 2 shows the change of the degree of polarization of the in-plane electroluminescence as a function of the temperature for LEDs with emission wavelength ranging between 288 nm and 379 nm. The degree of polarization (P) is defined as P ¼ (ITE  ITM)/(ITE þ ITM), where ITE and ITM are the integrated intensities of the in-plane emitted TE and TM polarized electroluminescence, respectively. At room temperature, the degree of polarization is 0.15 for the 288 nm LEDs, 0.03 for 303 nm, 0.27 for 320 nm, and 0.54 for the 379 nm LEDs. The degree of polarization was found to increase with temperature for the 288 nm LEDs, whereas it remains constant for the 303 nm LEDs and even decreases for the 320 nm and 379 nm LEDs. These observations can be explained by the different arrangements of the valance bands

FIG. 1. (Color online) Change of the peak energy of the electroluminescence spectrum of 288 nm, 303 nm, 320 nm, and 379 nm LEDs operated at 20 mA with increasing temperature of the Peltier element.

Appl. Phys. Lett. 99, 261105 (2011)

FIG. 2. (Color online) Change of the degree of polarization of the in-plane electroluminescence as a function of the temperature for 288 nm, 303 nm, 320 nm, and 379 nm LEDs operated at 20 mA.

(due to different crystal-field splitting) at the C point of the Brillouin zone for the different LEDs. In our previous publications,15,16 we have shown that at 20  C LEDs with emission wavelength of around 300 nm have zero polarization meaning that the intensity of the TE and TM polarized part of the in-plane emission is the same. In this case, the maxima of the heavy hole (HH), light hole (LH), and split-off hole (SO) valance bands are nearly at the same energetic position as shown schematically in Fig. 3(b). For LEDs with shorter wavelengths, it was found that the emitted in-plane light is mainly TM polarized, because the topmost valence band is the split-off hole band, and the main transition takes place between this band and the conduction band (see Fig. 3(c)). Finally, for LEDs with longer wavelength, it was observed that the emitted in-plane light is mainly TE polarized. For these LEDs, the heavy and light hole bands are the topmost valence bands (see Fig. 3(a)). Therefore, the main transition takes place between the conduction band and these two valence bands. A heating of the LEDs increases the thermal energy of the carriers. Therefore, the holes increasingly occupy deeper valence bands with an increasing sample temperature. Thus, the distribution of the transitions between the conduction band and the heavy, light, and split-off hole bands changes. For LEDs with an emission wavelength shorter than 300 nm the holes occupy more and more the deeper heavy and light hole bands relative to the split-off hole band. Therefore, the TE polarized part of the in-plane emission increases with an increasing sample temperature. For the LEDs with longer emission wavelength than 300 nm, the situation is opposite. When increasing the temperature of these LEDs the holes occupy more and more the deeper

FIG. 3. (Color online) Schematic of band structures at the C point of the Brillouin zone for: (a) dominant transversal-electric, (b) balanced transversal-electric/transversal-magnetic, and (c) dominant transversal-magnetic in-plane light emission.

Downloaded 04 Oct 2012 to 13.1.101.37. Redistribution subject to AIP license or copyright; see http://apl.aip.org/about/rights_and_permissions

261105-3

Kolbe et al.

split-off hole band relative to the heavy and light hole bands. For this reason, the TM polarized part of the in-plane emission increases with increasing sample temperature. For the 303 nm LEDs, the maxima of the heavy, light, and split-off hole bands are nearly at the same energetic position. Therefore, the degree of polarization is almost independent of the sample temperature. These results are in a good agreement with our previous publications15,16 and confirm a cross over of the different valence bands for LEDs with an emission wavelength of around 300 nm. Another effect which can also result in a change of the degree of polarization with temperature is the thermally induced strain based on the different thermal expansion coefficients of the LED template and the active region. However, for an increase of the temperature from 20  C to 170  C, the thermally induced in-plane strain can be estimated to a change by less than 0.04%. Therefore, this effect plays only a minor role for the change of the degree of polarization with increasing temperature which can be also seen in the next paragraph. In a next step, the degree of polarization has been investigated as a function of the in-plane strain of the MQW barriers for LEDs with an emission wavelength of around 380 nm. Therefore, LEDs with InGaN QWs surrounded by 7 nm thick GaN or InxAl0.16Ga0.84xN barriers with indium mole fraction x of 0, 0.02, and 0.03 have been compared. HRXRD measurements on calibration samples were performed to determine the in-plane strain of the InGaN/ (In)(Al)GaN MQW active region. The X2h scans of the (0002) reflection can be found in Ref. 23. It was found that the zero order x-ray diffraction peak of the active region shifts towards the GaN main peak when the indium mole fraction of the barriers increases. This indicates a decrease of the in-plane tensile strain in the active region with respect to the GaN template. Figure 4 shows the degree of polarization as a function of the in-plane tensile strain of the (In)(Al)GaN MQW barriers. It was found that the degree of polarization decreases from 0.67 to 0.43 when the in-plane tensile strain of the QW barriers is reduced from 0.38% for the LEDs with the Al0.16Ga0.84N barriers to 0% for the LEDs with the GaN barriers. This observation shows that the in-plane strain of the active region has a strong influence on the degree of polarization of UV LEDs. The change of the polarization of the

FIG. 4. Degree of polarization of the in-plane electroluminescence as a function of the in-plane tensile strain of the multi-quantum well barriers of 380 nm LEDs operated at 20 mA.

Appl. Phys. Lett. 99, 261105 (2011)

in-plane emission should be related to a change of the electronic band structure of the active region and so a rearrangement of the valence bands at the C point of the Brillouin zone. The measurements suggest that the energy difference between the heavy and light hole bands and the split-off hole band increases with increasing tensile strain of the active region. Therefore, the transition between the conduction, heavy, and light hole bands increases in intensity which enhances the TE polarized emission. The change of the electronic band structure of the quantum wells with changing barrier material is not totally clear at the moment. Therefore, further investigations of this effect are necessary. In order to enhance the TE emission intensity and consequently the light extraction efficiency particularly of UV LEDs, these results indicate that the strain state of the active region can be instrumental to control and increase the TE polarized portion of the emitted light. In summary, we have investigated the temperature and strain dependence of the polarization of the in-plane electroluminescence of (0001) orientated (In)(Al)GaN MQW LEDs in the UV spectral range. It was found that with increasing temperature, the TE polarized emission increases relative to the TM emission for LEDs with an emission wavelength shorter than 300 nm and decreases for LEDs with a longer emission wavelength. This effect can be attributed to the temperature dependence of the occupation of the different valence bands, e.g., the heavy, light, and split-off hole band. The in-plane strain of the active region has a strong influence on the in-plane light polarization of near UV LEDs. It was found that the TM polarized part of the in-plane emitted light becomes more dominant relative to the TE polarized part for MQWs with decreasing tensile in-plane strain. Therefore, the strain state of the active region is one parameter that needs to be optimized to improve the light extraction efficiency of UV LEDs. We thank T. Tessaro for technical assistance regarding the operation of the MOVPE growth system. This work was partially supported by the German Federal Ministry of Education and Research within the “Deep UV LED” project under Contract Nos. 13N9933 and 13N9934 and by the Investitionsbank Berlin under ProFIT No. 10135613. 1

M. A. Wu¨rtele, T. Kolbe, M. Lipsz, A. Ku¨lberg, M. Weyers, M. Kneissl, and M. Jekel, Water Res. 45, 1481 (2011). 2 K. Davitt, Y. Song, W. R. Patterson III, A. Nurmikko, M. Gherasimova, J. Han, Y. Pan, and R. K. Chang, Opt. Express 13, 9548 (2005). 3 A. Khan, K. Balakrishnan, and T. Katona, Nat. Photonics 2, 77 (2008). 4 J. R. Grandusky, S. R. Gibb, M. C. Mendrick, and L. J. Schowalter, Appl. Phys. Express 3, 072103 (2010). 5 H. Tsuzuki, F. Mori, K. Takeda, T. Ichikawa, M. Iwaya, S. Kamiyama, H. Amano, I. Akasaki, H. Yoshida, M. Kuwabara et al., Phys. Status Solidi A 206, 1199 (2009). 6 V. Adivarahan, Q. Fareed, M. Islam, T. Katona, B. Krishnan, and A. Khan, Jpn. J. Appl. Phys. Part 1 46, 877 (2007). 7 W. Sun, M. Shatalov, J. Deng, X. Hu, J. Yang, A. Lunev, Y. Bilenko, M. Shur, and R. Gaska, Appl. Phys. Lett. 96, 061102 (2010). 8 H. Hirayama, Y. Tsukada, T. Maeda, and N. Kamata, Appl. Phys. Express 3, 031002 (2010). 9 M. Kneissl, T. Kolbe, C. Chua, V. Kueller, N. Lobo, J. Stellmach, A. Knauer, H. Rodriguez, S. Einfeldt, Z. Yang et al., Semicond. Sci. Technol. 26, 014036 (2011). 10 H. Morkoc, S. Strite, G. B. Gao, M. E. Lin, B. Sverdlov, and M. Burns, J. Appl. Phys. 76, 1363 (1994).

Downloaded 04 Oct 2012 to 13.1.101.37. Redistribution subject to AIP license or copyright; see http://apl.aip.org/about/rights_and_permissions

261105-4 11

Kolbe et al.

K. B. Nam, J. Li, M. L. Nakarmi, J. Y. Lin, and H. X. Jiang, Appl. Phys. Lett. 84, 5264 (2004). 12 R. G. Banal, M. Funato, and Y. Kawakami, Phys. Rev. B 79, 121208(R) (2009). 13 A. A. Yamaguchi, Appl. Phys. Lett. 96, 151911 (2010). 14 J. Zhang, H. Zhao, and N. Tansu, Appl. Phys. Lett. 97, 111105 (2010). 15 T. Kolbe, A. Knauer, C. Chua, Z. Yang, S. Einfeldt, P. Vogt, N. M. Johnson, M. Weyers, and M. Kneissl, Appl. Phys. Lett. 97, 171105 (2010). 16 T. Kolbe, A. Knauer, J. Stellmach, C. Chua, Z. Yang, H. Rodrigues, S. Einfeldt, P. Vogt, N. M. Johnson, M. Weyers et al., Proc. SPIE 7939, 79391G (2011). 17 L. Dong, S. K. Yadav, R. Ramprasad, and S. P. Alpay, Appl. Phys. Lett. 96, 202106 (2010).

Appl. Phys. Lett. 99, 261105 (2011) 18

Q. Yan, P. Rinke, M. Scheffler, and C. G. Van de Walle, Appl. Phys. Lett. 95, 121111 (2009). A. Knauer, H. Wenzel, T. Kolbe, S. Einfeldt, M. Weyers, M. Kneissl, and G. Tra¨nkle, Appl. Phys. Lett. 92, 191912 (2008). 20 T. Kolbe, T. Sembdner, A. Knauer, V. Kueller, H. Rodriguez, S. Einfeldt, P. Vogt, M. Weyers, and M. Kneissl, Phys. Status Solidi A 207, 2198 (2010). 21 M. Kneissl, Z. Yang, M. Teepe, C. Knollenberg, N. M. Johnson, A. Usikovl, and V. Dmitriev, Jpn. J. Appl. Phys. Part 1 45, 3905 (2006). 22 N. Nepal, J. Li, M. L. Nakarmi, J. Y. Lin, and H. X. Jiang, Appl. Phys. Lett. 87, 242104 (2005). 23 T. Kolbe, A. Knauer, H. Wenzel, S. Einfeldt, V. Kueller, P. Vogt, M. Weyers, and M. Kneissl, Phys. Status Solidi C 6, S889 (2009). 19

Downloaded 04 Oct 2012 to 13.1.101.37. Redistribution subject to AIP license or copyright; see http://apl.aip.org/about/rights_and_permissions