Open Access
Add to BibSonomyAbstract
The exciton localization effect in nonpolar a-plane GaN/AlGaN multiquantum wells (MQWs) structures with different quantum confinements (different well thicknesses or Al molar fractions of barrier) has been investigated by temperature dependent photoluminescence (PL). An “S-shaped” PL peak energy variation is observed in the spectra, indicating the existence of localized states. The exciton localization energy is larger in the MQWs with stronger quantum confinement. A good agreement of the localization energy is obtained by theoretical calculation assuming ± 5% fluctuation of well thickness, which demonstrates that potential minima caused by well thickness fluctuation are the major origin of exciton localization. In addition, the internal quantum efficiency shows more than 3 times enhancement with decreasing the well width from 8.1 to 2.7 nm due to the strong exciton localization which can prevent the carriers from trapping into the nonradiative recombination centers.
© 2015 Optical Society of America
1. Introduction
Recently, a great deal of attention has been attracted to the study of GaN based quantum well (QW) structures grown along nonpolar crystallographic axis [1–9 ]. Such orientations provide a means of eliminating the polarization induced electric field across the QW, resulting in a short radiative recombination lifetime compared with QW structures grown along the polar c-axis [1,10 ]. However, since the defect and dislocation densities are very high in nonpolar films grown on foreign substrates, the quantum efficiency is low and far from the expectation which should be much higher than that of polar counterpart. In InGaN/GaN MQWs light emitting devices, it is known that exciton localization is an important reason for the high quantum efficiency since it can prevent carriers from trapping into the nonradiative recombination centers [11]. Therefore, it makes sense to study the exciton localization in nonpolar GaN films and MQWs. Paskov et al. first investigated the carrier localization on the basal stacking faults (BSFs) in a-plane GaN films and attributed to the hole localization caused by the polarization-field-induced potential in the vicinity of the BSFs [12]. Corfdir et al. studied the exciton localization on BSFs in a-plane epitaxial lateral overgrown GaN in detail [13,14 ], and suggested that either quantum coupling between BSFs or a randomly distributed donors in the vicinity of the BSFs could be the origins of the exciton localization [13]. Later, the same group reported the exciton localization in a-plane GaN/AlGaN MQWs grown on bulk GaN, and proposed that the exciton localization was caused by well-width variation induced potential fluctuation [15]. Recently, Huang et al. showed strong carrier localization with energy as large as 51 meV in a-plane GaN/AlGaN MQWs and indicated that the BSFs were the origins of localized states [16]. Considering the various origins of the exciton localization, a further understanding of the exciton behavior is essential. Moreover, no report has been made to investigate the quantum confinement dependence of the exciton localization in a-plane GaN/AlGaN MQWs with high Al molar fraction of barrier grown on r-plane sapphire.
In this paper, we carry out a temperature dependent PL study of a-plane GaN/AlGaN MQWs with different quantum confinements (i.e. different quantum well widths or Al molar fractions of barrier). We then analyze the temperature dependence of the PL peak energy by Varshni fitting and discuss the origins of the exciton localization according to the influence of quantum confinement on the exciton localization energy. Finally, theoretical calculations of proposed MQWs model based on the well thickness fluctuation are performed and a good fit of the localization energy is obtained.
2. Experimental details
Four a-plane GaN/AlGaN MQWs, denoted as sample A, B, C, and D, were grown on r-plane sapphire by low pressure metal-organic chemical vapor deposition (MOCVD). Figure 2(a) shows the schematic diagram of the MQWs structure. First, a high quality AlN buffer layer was grown, followed by the deposition of a 3 μm thick unintentional doped Al0.1Ga0.9N template layer, and then 20 pairs doped GaN/AlGaN MQWs were grown on the template. Only the QW was doped with Si density about 1 × 1019 cm−3. The growth conditions were identical for all the samples except for the growth time of GaN well and flow of the metal-organic sources to achieve different well thickness and barrier height. The barrier thickness was 5 nm for all samples, while the well thicknesses of sample A, B, C, and D were 2.7, 5.4, 8.1, and 3.5 nm, respectively. The Al molar fraction of barriers of samples was 0.5 except for sample D with 0.4 Al molar fraction of barrier. All the structure parameters were confirmed by high-resolution X-ray diffraction (HR-XRD) measurements. Cross-section transmission electron microscopy (TEM) images were obtained on a FEI Tecnai G20 operating at 200 kV. PL measurements were performed by using a 325 nm He-Cd laser as the excitation source. The sample was placed inside a closed-cycle He cryostat with the temperature range of 11-300 K controlled by a Lakeshore temperature controller. PL signals were dispersed by a 0.75 m spectrometer and then detected by a high sensitivity photomultiplier tube for UV-visible wavelengths using the lock-in technique.
3. Results and discussion
Figure 1 shows the cross-section TEM images of sample A as a representative. The BSFs can be clearly seen as thin lines aligning in parallel with [0001] c-axis orientation and propagate from below the AlGaN template layer towards the GaN/AlGaN MQWs, as indicated by the arrow in Fig. 1(a). The interfaces of the GaN/AlGaN MQWs are clear, but the width of GaN well has little fluctuation, as shown in Fig. 1(b).
Fig. 1 (a) Cross-section TEM image of sample A taken along the [ ] zone axis. (b) The enlarged TEM image of the MQWs marked by the dash circle in (a).
Figure 2(b) shows the normalized PL spectra of four samples measured at 11 K. The PL spectrum of sample A is centered at 3.6036 eV and presents a PL linewidth of 114 meV. When the well thickness increases from 2.7 to 8.1 nm, the PL peak energy decreases from 3.6036 to 3.4892 eV, which agrees well with the interband transition trend expected for the polarization free MQWs [17,18 ]. Meanwhile, it is found that the thicker the QW, the narrower the emission spectrum, which is consistent with previous reports by other groups [15,16,18 ]. It should be noticed that the PL linewidths of samples in this work are much larger than previous values obtained in other reports [15,18 ]. This can be interpreted by the inhomogeneous broadening caused by the high density donor and large Al molar fraction of barrier. One may feel puzzled at first that the BSFs-related emissions, which are usually observed in a-plane GaN films [12–14 ], and MQWs [8,16 ], are seemingly absent in our samples. However, the density of BSFs in our samples is about 4 × 105 cm−1 estimated by the TEM images as shown in Fig. 1(a), comparable to previous reports [6,14,16,18 ], but much higher than that reported by Corfdir et al. In Corfdir’s work, BSFs-related emissions are not observed due to the high crystal quality of the MQWs grown on bulk GaN [15,19 ]. Carefully analyzing the profiles of the PL spectra, we find that all the spectra are asymmetrical line shape with a shoulder lying at lower energy side of the PL peak position. The shoulder and the main PL peak can be extracted separately by Gaussian fitting, as shown in Fig. 3(a) . The fitted results show that the energy separation of the two peaks is in the range of 40-60 meV, much smaller than the LO-Phonon energy of GaN (92 meV), but close to previous reported energy separation of the BSFs and exciton emission peaks [16,18 ]. This demonstrates that the shoulder peak cannot come from the phonon replica of the main emission peak. Moreover, the shoulder peak is located at 3.4686 eV at low temperature, comparable to previous BSFs location reported by other researchers [16,18 ]. Therefore, the shoulder and main peaks can be assigned to the BSFs and exciton emissions, respectively. Meanwhile, as shown in Fig. 3(b), the intensity of the BSFs emission is much weaker than that of exciton emission and quenches faster with increasing temperature. The strong exciton emission can be interpreted by the large quantum confinement of carriers in QWs with high barriers. Based on this, it can be concluded that the BSF emission has little effect on the measured PL peak position in our work. Thus, in the following discussion, the PL peak is treated as the exciton emission peak. To investigate the emission spectra variation with temperature, the PL spectra are measured at different temperature and shown in Fig. 2(c) (sample B as a representative). A “red-blue-red” shift of PL peak energy can be seen in the spectra when increasing the temperature from 11 to 300 K. It is well known that the “S-shaped” PL peak variation with temperature is a typical indication of the exciton localization behavior. At low temperature, the excitons are localized at potential minima, and when temperature increases, excitons delocalize into quasi-continuous states of GaN well, resulting in the blue-shift of the emission peak. The redshift of PL peak energy is a result of the temperature induced bandgap shrinkage.
Fig. 2 (a) Schematic diagram of the a-plane GaN/AlGaN MQWs structure. (b) Normalized PL spectra of four samples measured at 11 K. (c) PL spectra of sample B measured at temperature ranging from 11 to 300 K.
Fig. 3 (a) PL spectrum for sample B measured at 11 K with two Gaussian fitting curves representing the exciton emission and BSF emission, respectively. (b) Normalized integrated emission intensity for exciton emission and BSF emission, respectively.
To figure out the origins of the exciton localization, temperature dependence of the PL peak energy for the four samples is fitted by Varshni’s empirical formula as follows [20]:
where Eg (0) is the bandgap energy of GaN QW at 0 K, α and β are the Varshni fitting parameters. The fitted results with experimental data are shown in Fig. 4 . It is obvious that the PL peak energy with increasing temperature for the four samples all show an “S-shaped” variation and the variation becomes stronger when the well thickness decreases from 8.1 to 2.7 nm, indicating the deeper exciton localization in MQWs with narrower GaN QW. To describe the degree of exciton localization, exciton localization energy Eloc is adopted, which is defined as the difference between fitted and experimental emission peak energy at lowest temperature, as indicated by the double arrows in Fig. 4. The exciton localization energies Eloc are 25, 11.3, 8.7, and 14.4 meV for samples A, B, C, and D, respectively, a little larger than that obtained by Corfdir et al [15]. It should be pointed out that the only difference among the four samples is the quantum confinement strength. Here, we define the strength energy as the separation of the QW emission peak energy and the bulk GaN bandgap energy. Figure 5(a) shows the relationship of localization energy and quantum confinement energy. It is found that the stronger the quantum confinement, the larger the exciton localization energy, which is just opposite to the results reported by Huang et al [16]. Having in mind that the localization energy here is much smaller than that obtained in Huang’s work, it then can be concluded that BSFs is not the origins of the localized states in our samples. Actually, exciton localization has been observed previously in polar GaN/AlGaN MQWs in [21], and interpreted by the perturbation of confinement energy caused by well width fluctuation or alloy disorder. Moreover, our findings are coincident with their conclusions. Therefore, here we prefer the well-thickness fluctuation induced potential minima as the origin of exciton localization. To confirm our assumption, theoretical calculations of the emission peak energy of a-plane GaN/AlGaN MQWs are performed under a consideration of ± 5% well-thickness fluctuation. This fluctuation of well thickness is appropriate and can be demonstrated by the TEM image in Fig. 1(b). The detailed calculation procedures can be found in our previous study [22]. Figure 5(c) shows the calculated and experimental room temperature (RT) peak energies for the four samples. A great fit of the experimental data is obtained, which indicates that our calculation procedures are valid and feasible. Then, the peak energy fluctuations are calculated by assuming ± 5% well-thickness fluctuation, as plotted in Fig. 5(d). Interestingly, the calculated values are close to the exciton localization energies, especially for sample A and B. It is worth noting that inhomogeneity of AlGaN barrier alloy can also bring out the fluctuation of peak energy as reported in [21]. However, by calculating, we find that a serious inhomogeneity with ± 20% to ± 60% fluctuations of the Al molar fraction of the barrier is needed to account for the exciton localization energies for all samples. Obviously, this large inhomogeneity of Al molar fraction is not practical in our samples according to the results of the HR-XRD measurement (not shown here). Thus, we believe that fluctuation of well thickness is the major origin of the exciton localization. In addition, it is found that the internal quantum efficiency increases with localization energy and shows more than 3 times enhancement when the well thickness decreases from 8.1 to 2.7 nm, as shown in Fig. 5(b). This can be attributed to the larger exciton localization energy which can prevent carriers from trapping into the nonradiative recombination centers more effectively.Fig. 4 Temperature dependence of PL peak energy (solid line with closed circles) for four samples. The solid line represents the least-squares fit of the peak energy with Varshni’s empirical formula.
Fig. 5 (a) Localization energy Eloc as a function of quantum confinement energy. (b) Internal quantum efficiency as a function of localization energy Eloc. (c) Experimental (circles) and calculated (triangles) peak energy of four samples. (d) Localization energy Eloc (circles) and calculated peak energy fluctuation (triangles) of four samples.
Figure 6 shows the Arrhenius plot of the integrated PL intensity for the four samples. It is found that the PL intensity variations cannot be described by a thermal quenching process with single activation energy. Therefore, two different thermal quenching channels with two activation energies are introduced as follows [23]:
where I(0) is the integrated PL intensity at 0 K, A and B are two adjustable parameters, Ea and Eb are two activation energies relative to the two different thermal quenching channels, and kB is Boltzmann constant. The fitted activation energies Ea and Eb are summarized in Table 1 . According to the previous studies [24], the thermal quenching process can be related to the delocalization or dissociation of excitons. However, the two activation energies are found to have no trends or relationships with localization energy for all samples. Considering the heavy doping of GaN well, which may result in serious defects, we predict that the strange thermal quenching process may be related to the doping-induced defects in the localization region. To make clear the origins of the two thermal quenching channels, further investigations of the PL quenching process are needed and work is ongoing.Fig. 6 Arrhenius plot of the integrated PL intensity for the four samples. The solid line represents the least-squares fit of experimental data using Eq. (2).
Table 1. Summary of the structure parameters and optical properties of the four samples
4. Conclusion
In conclusion, the relationships between quantum confinement and exciton localization in a-plane GaN/AlGaN MQWs grown on r-plane sapphire have been studied by temperature dependent PL. It is found that the exciton localization energy is larger in the MQWs with stronger quantum confinement (narrower well width or higher Al molar fraction of barrier), which results in a higher internal quantum efficiency. By theoretical calculations, inhomogeneity of AlGaN barrier alloy is excluded in the origins of the exciton localization, and ± 5% fluctuation of well thickness is accounted for the formation of that localization.
Acknowledgments
This work was supported by the National Natural Science Foundation of China (NSFC) (Grant No. 2012CB619302, 2010CB923204), the National Basic Research Program (Grant No. 60976042, 51002058, 10990102), the Science and Technology Bureau of Wuhan City (Grant No. 2014010101010003).
References and links
1. P. Waltereit, O. Brandt, A. Trampert, H. T. Grahn, J. Menniger, M. Ramsteiner, M. Reiche, and K. H. Ploog, “Nitride semiconductors free of electrostatic fields for efficient white light-emitting diodes,” Nature 406(6798), 865–868 (2000). [CrossRef] [PubMed]
2. T. Koida, S. F. Chichibu, T. Sota, M. D. Craven, B. A. Haskell, J. S. Speck, S. P. DenBaars, and S. Nakamura, “Improved quantum efficiency in nonpolar (11-20) AlGaN/GaN quantum wells grown on GaN prepared by lateral epitaxial overgrowth,” Appl. Phys. Lett. 84(19), 3768 (2004). [CrossRef]
3. H. Teisseyre, C. Skierbiszewski, B. Lucznik, G. Kamler, A. Feduniewicz, M. Siekacz, T. Suski, P. Perlin, I. Grzegory, and S. Porowski, “Free and bound excitons in GaN/AlGaN homoepitaxial quantum wells grown on bulk GaN substrate along the nonpolar (11-20) direction,” Appl. Phys. Lett. 86(16), 162112 (2005). [CrossRef]
4. T. M. Al Tahtamouni, A. Sedhain, J. Y. Lin, and H. X. Jiang, “Growth and photoluminescence studies of a-plane AlN/AlxGa1−xN quantum wells Appl. Phys. Lett.90(22), 221105 (2007). [CrossRef]
5. T. J. Badcock, P. Dawson, M. J. Kappers, C. McAleese, J. L. Hollander, C. F. Johnston, D. V. Sridhara Rao, A. M. Sanchez, and C. J. Humphreys, “Optical polarization anisotropy of a-plane GaN/AlGaN multiple quantum well structures grown on r-plane sapphire substrates,” J. Appl. Phys. 105(12), 123112 (2009). [CrossRef]
6. T. J. Badcock, P. Dawson, M. J. Kappers, C. McAleese, J. L. Hollander, C. F. Johnston, D. V. Sridhara Rao, A. M. Sanchez, and C. J. Humphreys, “Optical properties of GaN/AlGaN quantum wells grown on nonpolar substrates,” Appl. Phys. Lett. 93(10), 101901 (2008). [CrossRef]
7. P. Corfdir, J. Levrat, A. Dussaigne, P. Lefebvre, H. Teisseyre, I. Grzegory, T. Suski, J.-D. Ganiere, N. Grandjean, and B. Deveaud-Pledran, “Intrinsic dynamics of weakly and strongly confined excitons in nonpolar nitride-based heterostructures,” Phys. Rev. B 83(24), 245326 (2011). [CrossRef]
8. P. Corfdir, P. Lefebvre, L. Balet, S. Sonderegger, A. Dussaigne, T. Zhu, D. Martin, J.-D. Ganiere, N. Grandjean, and B. Deveaud-Pledran, “Exciton recombination dynamics in a-plane (Al,Ga)N/GaN quantum wells probed by picosecond photo and cathodoluminescence,” J. Appl. Phys. 107(4), 043524 (2010). [CrossRef]
9. D. Rosales, B. Gil, T. Bretagnon, B. Guizal, F. Zhang, S. Okur, M. Monavarian, N. Izyumskaya, V. Avrutin, Ü. Özgür, H. Morkoç, and J. H. Leach, “Excitonic recombination dynamics in non-polar GaN/AlGaN quantum wells,” J. Appl. Phys. 115(7), 073510 (2014). [CrossRef]
10. E. Kuokstis, C. Q. Chen, M. E. Gaevski, W. H. Sun, J. W. Simin, M. Asif Khan, H. P. Maruska, D. W. Hill, M. M. C. Chou, J. J. Gallagher, and B. Chai, “Polarization effects in photoluminescence of C- and M-plane GaN/AlGaN multiple quantum wells,” Appl. Phys. Lett. 81(22), 4130 (2002). [CrossRef]
11. K. P. O’Donnell, R. W. Martin, and P. G. Middleton, “Origin of Luminescence from InGaN Diodes,” Phys. Rev. Lett. 82(1), 237–240 (1999). [CrossRef]
12. P. P. Paskov, R. Schifano, B. Monemar, T. Paskova, S. Figge, and D. Hommel, “Emission properties of a-plane GaN grown by metal-organic chemical-vapor deposition,” J. Appl. Phys. 98(9), 093519 (2005). [CrossRef]
13. P. Corfdir, P. Lefebvre, J. Levrat, A. Dussaigne, J. D. Ganiere, D. Martin, J. Ristic, T. Zhu, N. Grandjean, and B. Deveaud-Pledran, “Exciton localization on basal stacking faults in a-plane epitaxial lateral overgrown GaN grown by hydride vapor phase epitaxy,” J. Appl. Phys. 105(4), 043102 (2009). [CrossRef]
14. P. Corfdir, J. Ristic, P. Lefebvre, T. Zhu, D. Martin, A. Dussaigne, J. D. Ganiere, N. Grandjean, and B. Deveaud-Pledran, “Low-temperature time-resolved cathodoluminescence study of exciton dynamics involving basal stacking faults in a-plane GaN,” Appl. Phys. Lett. 94(20), 201115 (2009). [CrossRef]
15. P. Corfdir, A. Dussaigne, H. Teisseyre, T. Suski, I. Grzegory, P. Lefebvre, E. Giraud, J. D. Ganiere, N. Grandjean, and B. Deveaud-Pledran, “Thermal carrier emission and nonradiative recombinations in nonpolar (Al,Ga)N/GaN quantum wells grown on bulk GaN,” J. Appl. Phys. 111(3), 033517 (2012). [CrossRef]
16. H.-M. Huang, C.-Y. Chang, Y.-P. Lan, T.-C. Lu, H.-C. Kuo, and S.-C. Wang, “Ultraviolet emission efficiency enhancement of a-plane AlGaN/GaN multiple-quantum-wells with increasing quantum well thickness,” Appl. Phys. Lett. 100(26), 261901 (2012). [CrossRef]
17. M. D. Craven, P. Waltereit, J. S. Speck, and S. P. DenBaars, “Well-width dependence of photoluminescence emission from a-plane GaN/AlGaN multiple quantum wells,” Appl. Phys. Lett. 84(4), 496 (2004). [CrossRef]
18. A. Dussaigne, P. Corfdir, J. Levrat, T. Zhu, D. Martin, P. Lefebvre, J. D. Ganiere, R. Butte, B. Deveaud-Pledran, N. Grandjean, Y. Arroyo, and P. Stadelmann, “One-dimensional exciton luminescence induced by extended defects in nonpolar GaN/(Al,Ga)N quantum wells,” Semicond. Sci. Technol. 26(2), 025012 (2011). [CrossRef]
19. P. Corfdir, A. Dussaigne, H. Teisseyre, T. Suski, I. Grzegory, P. Lefebvre, E. Giraud, M. Shahmohammadi, R. T. Phillips, J. D. Ganiere, N. Grandjean, and B. Deveaud, “Temperature-Dependence of Exciton Radiative Recombination in (Al,Ga)N/GaN Quantum Wells Grown on a-Plane GaN Substrates,” Jpn. J. Appl. Phys. 52(8S), 08JC01 (2013). [CrossRef]
20. Y. P. Varshni, “Temperature dependence of the energy gap in semiconductors,” Physica 34(1), 149–154 (1967). [CrossRef]
21. I. Friel, C. Thomidis, and T. D. Moustakas, “Well width dependence of disorder effects on the optical properties of AlGaN/GaN quantum wells,” Appl. Phys. Lett. 85(15), 3068 (2004). [CrossRef]
22. F. Wu, W. Tian, W. Y. Yan, J. Zhang, S. C. Sun, J. N. Dai, Y. Y. Fang, Z. H. Wu, and C. Q. Chen, “Terahertz intersubband transition in GaN/AlGaN step quantum well,” J. Appl. Phys. 113(15), 154505 (2013). [CrossRef]
23. E. Monroy, N. Gogneau, F. Enjalbert, F. Fossard, D. Jalabert, E. Bellet-Amalric, L. S. Dang, and B. Daudin, “Le Si Dang, and B. Daudin, “Molecular-beam epitaxial growth and characterization of quaternary III–nitride compounds,” J. Appl. Phys. 94(5), 3121 (2003). [CrossRef]
24. H. Murotani, Y. Yamada, T. Taguchi, A. Ishibashi, Y. Kawaguchi, and T. Yokogawa, “Temperature dependence of localized exciton transitions in AlGaN ternary alloy epitaxial layers,” J. Appl. Phys. 104(5), 053514 (2008). [CrossRef]
