GaN multiple quantum wells

Superlattices and Microstructures 71 (2014) 38–45

Contents lists available at ScienceDirect

Superlattices and Microstructures journal homepage: www.elsevier.com/locate/superlattices

Characteristics of nanoporous InGaN/GaN multiple quantum wells W.J. Wang a, G.F. Yang a,c,⇑, P. Chen a,b, Z.G. Yu a, B. Liu a, Z.L. Xie a, X.Q. Xiu a, Z.L. Wu b, F. Xu b, Z. Xu b, X.M. Hua a, H. Zhao a, P. Han a, Y. Shi a, R. Zhang a, Y.D. Zheng a a Jiangsu Provincial Key Laboratory of Advanced Photonic and Electronic Materials and School of Electronic Science and Engineering, Nanjing University, Nanjing 210093, China b Institute of Optoelectronics, Nanjing University and Yangzhou, Yangzhou 225009, China c School of Science, Jiangnan University, Wuxi 214122, China

a r t i c l e

i n f o

Article history: Received 23 December 2013 Received in revised form 1 March 2014 Accepted 11 March 2014 Available online 22 March 2014 Keywords: Nanostructure fabrication Nanoporous structure InGaN/GaN multiple quantum wells Photoluminescence

a b s t r a c t The nanoporous InGaN/GaN multiple quantum wells (MQWs) has been fabricated through rapid thermal annealing (RTA) and inductively coupled plasma (ICP) dry etching process using self-assembled Ni nanoporous masks. In comparison with the as-grown planar InGaN/GaN MQWs, both internal quantum efficiency and light extraction efficiency for nanoporous InGaN/GaN MQWs are increased, which can be concluded from the photoluminescence (PL) measurements. The thermal activation energy of nanoporous structure (107.44 meV) is significantly higher than that of the asgrown sample (33.02 meV) from temperature-dependent PL measurement, indicating that carriers are well confined and the nonradiative recombination caused by the dislocations and other defects has been reduced. Besides, enhanced light scattering in the disordered nanoporous system can further increase the output emission intensity. The enhanced performance of nanoporous InGaN/GaN MQWs reveals its promising applications for high-efficiency light-emitting devices. Ó 2014 Elsevier Ltd. All rights reserved.

⇑ Corresponding author at: Jiangsu Provincial Key Laboratory of Advanced Photonic and Electronic Materials and School of Electronic Science and Engineering, Nanjing University, Nanjing 210093, China. Tel.: +86 25 83685251; fax: +86 25 83685476. E-mail address: [email protected] (G.F. Yang). http://dx.doi.org/10.1016/j.spmi.2014.03.012 0749-6036/Ó 2014 Elsevier Ltd. All rights reserved.

W.J. Wang et al. / Superlattices and Microstructures 71 (2014) 38–45

39

1. Introduction III-nitride-based semiconductors have attracted a lot of attention for the fabrication of solar-blind ultraviolet detectors, laser diodes, and high power electronic devices in the last two decades. InGaN/ GaN multiple quantum wells (MQWs) have been used as active layers for blue, green, and white light emitting diodes (LEDs) and laser diodes (LDs) [1–5]. Due to the large lattice mismatch and the difference in thermal expansion coefficients between the epilayer and the substrate, high density (108– 1010 cm2) of threading dislocations and other crystalline defects originating at the film/substrate (or film/buffer layer) interface and propagating through the entire InGaN/GaN MQW structure. These threading dislocations and other crystalline defects are acted as nonradiative centers [6,7] to reduce the internal quantum efficiency. Recently, in an attempt to minimize the density of dislocations and other defects for improving internal quantum efficiency for InGaN based MQWs, a lot of methods such as homoepitaxy growth on GaN single crystal substrates [8], epitaxial layer overgrowth [9,10], multiple intermediate layers [11] have been demonstrated. However, these techniques are extremely expensive and difficult to reduce the dislocation density to be less than 106 cm2. In addition, the fact that they are highly nonuniform and small in size also makes them unsuitable for mass production. In order to obtain high performance and highly reliable nitride-based optoelectronic devices, effective, affordable, large area, and low threading dislocation (TD) density materials or device structures are demanded. Fabrication of nanostructured InGaN/GaN based optoelectronic devices is one of a promising approaches to reduce the dislocation density for further improving the devices’ performance, such as fabrication of InGaN/GaN based LEDs into nanorod arrays [12,13] and nanotips [14]. In addition, by reducing the size of the emitting area to nanostructures with a higher surface area are also helpful for increasing the light extraction efficiency. As a result, low-dislocation density nanostructures can improve both internal quantum efficiency (gint) and light extraction efficiency (gext). In this letter, we report on the fabrication and characterization of the nanoporous InGaN/GaN MQWs. The nanoporous InGaN/GaN MQWs are fabricated by inductively coupled plasma (ICP) topdown etching using self-assembled Ni nanoparticle masks on a planar InGaN/GaN MQW structure. The optical properties of the nanoporous InGaN/GaN MQWs and as-grown planar samples are investigated by means of photoluminescence (PL). From the results, enhanced performance of nanoporous InGaN/GaN MQWs is obtained compared with the as-grown MQW sample, attributing to the effects of drastic reduction of dislocations and defects. 2. Experimental details InGaN/GaN MQW structure was grown on a (0 0 0 1) sapphire substrate in a metalorganic chemical vapor deposition (MOCVD) system. On top of the substrate, a 30 nm thick low-temperature GaN buffer layer and a 2-lm thick Si-doped GaN epitaxial layer were grown, followed by a MQW structure with five periods of In0.15Ga0.85N/GaN MQWs (2 nm wells and 8 nm quantum barriers) as the emitting layers. The barrier layers were grown at 825 °C and the wells were grown at 745 °C. Finally, the MQW structure was terminated with the deposition of a 10-nm thick GaN cap layer. After the epitaxial growth, the nanoporous InGaN/GaN MQWs were fabricated by using self-assembled Ni nanomasks. For the nanoporous fabrication process, firstly, a 6 nm Ni layer was deposited on top of the InGaN/GaN MQWs by electron beam evaporation. Next, the Ni-coated MQW sample was subsequently subjected to rapid temperature annealing under N2 atmosphere at 800 °C for 30 s to form self-assembled nickel nanomask. And then, the sample was etched down by the ICP system (Oxford Plasma System 100, ICP 180) using Cl2/BCl3 gas. Finally, the nanoporous MQWs were immersed in 50% HCl solvent to minor the damages caused by the dry etching, which has been reported previously in Ref. [15]. The surface morphology was examined by field emission scanning electron microscopy (FE-SEM). The optical properties of the samples were investigated by temperature dependent photoluminescence measurements. The samples to be measured have been placed in an evacuated cryostat for

40

W.J. Wang et al. / Superlattices and Microstructures 71 (2014) 38–45

carrying out PL studies ranging from 15 to 300 K using a 266-nm cw laser. Room temperature photoluminescence measurements were carried out by Renishaw inVia microscope system, a cw He–Cd laser with a wavelength of 325 nm was used for PL excitation. In order to accurately study the strainrelaxation influences in the nanoporous InGaN/GaN MQWs, reciprocal space mapping (RSM) measurements have been performed on all samples using a PANalytical X’Pert Pro MRD high-resolution X-ray diffractometer. 3. Results and discussion The top-view and cross-sectional SEM images of the etched nanoporous InGaN/GaN MQWs are shown schematically in Fig. 1(a) and (b), respectively. It can be seen that the fabricated sample is composed of coalesced nanoporous structure instead of normal nanocolumns or nanopillars, and the area of the nanoporous accounts for 40.2% of the total area. The average pore size of the nanoporous structure is between 400 nm and 900 nm, while the grain boundaries are in the range of 100  300 nm. As shown in Fig. 1(b), the etched depth of the porous structure is about 230 nm with the inclination angle of the sidewalls more than 85°, indicating the MQW structure has been etched through to the GaN epitaxial layer. Fig. 2 shows the normalized PL intensities of the nanoporous MQWs and as-grown planar MQW sample at room temperature. The PL peak energies of the film and nanoporous samples are located at 2.942 eV and 2.965 eV, respectively. The PL peak width of the nanoporous MQWs is slightly

Fig. 1. SEM images of top-view (a) and side-view (b) of nanoporous InGaN/GaN MQWs after ICP etching.

41

W.J. Wang et al. / Superlattices and Microstructures 71 (2014) 38–45

Fig. 2. PL spectra of the nanoporous MQWs (solid line) and as-grown planar MQWs (dashed line) at room temperature.

narrower than that of the planar MQWs. The spectral full width at half maximum (FWHM) of the nanoporous and planar MQWs are 135 meV and 141 meV, respectively. The FWHM of 135 meV for nanoporous MQWs is comparable [16] or narrower [17] than those of other nanostructured InGaN/ GaN MQWs. The observed narrow bandwidth for our nanoporous MQWs may be attributed to the reduction of band tails instead of the influence of strain relaxation. As we know, the density of dislocations in GaN layer ordinarily is 108–109 cm2. Therefore, the average distance between the dislocations is about 1 lm  300 nm, which is much larger than the grain boundaries of the nanoporous MQWs in this work. The dislocation density in the nanoporous structure is lower than that of planar MQWs after the ICP top-down etching process. The band tail states induced by dislocations in the nanoporous can be significant reduced. More importantly, Fig. 2 exhibits that the PL intensity of InGaN/GaN nanoporous MQWs is about 2.75 times than that of the as-grown MQWs, which can be attributed to the drastically reduced non-radiative recombination caused by dislocations and other defects in nanoporous MQWs and the light scattering enhancement on the disordered nanoporous surface. For an MQW sample, its PL intensity is determined by the light emitting area, the absorption efficiency of the pumping laser (gabs), gint and gext. The PL intensity can be expressed by the following equation: [18]

PL intensity / ðEmission AreaÞ  gabs  gint  gext :

ð1Þ

We assume that gabs for the as-grown MQWs and the nanoporous MQWs can be equal for the effective united area, which has been proved by Kuo et al. [18]. Thus, gint  gext of the nanoporous MQWs reaches 6.84 times than that of the as-grown planar MQW sample. In order to exclude the strain relaxation influence on the nanoporous sample, detailed RSM measurements have been performed along the asymmetric (1 0 5) direction on the two samples. Fig. 3(a) and (b) show the (1 0 5) RSM for planar and nanoporous MQWs, respectively. Based on a standard RSM theory [19], fully strained lines (red1 vertical lines) can be drawn as shown in Fig. 3(a) and (b). It can be seen that the centre position of the 0th order satellite peak of nanoporous sample is overlap the fully strained line, which suggests that nanoporous MQWs is nearly fully strained. The RSM analysis indicates that the strain in InGaN/GaN MQWs is not relaxed after the nanoporous structure has been fabricated. The results exhibit that the blue shift is induced by the reduction of dislocations and other defects, rather than the strain-relaxation of fabricated nanoporous structure. To clarify the influence of dislocation reduction on the MQW nanoporous quality, we estimated the internal quantum efficiency of the InGaN MQW samples using the temperature dependence of the integrated PL intensity. Temperature dependent PL main peak energy for as-grown MQW and nanoporous has been shown in Fig. 4, over a temperature range from 15 to 300 K with a 266 nm laser. 1

For interpretation of color in Fig. 3, the reader is referred to the web version of this article.

42

W.J. Wang et al. / Superlattices and Microstructures 71 (2014) 38–45

Fig. 3. (a) (1 0 5) XRD RSMs for planar MQWs and (b) (1 0 5) XRD RSMs for nanoporous, and fully stained line indicated by red vertical lines.

Fig. 4. Temperature dependent PL main peak energy for the nanoporous MQWs and as-grown planar MQWs from 15 K to 300 K.

We can find that the energy band gap of these samples tend to decrease with increasing the temperature. This can be explained by the thermally induced band-gap reduction. Moreover, S-shaped PL shift of two samples was observed. The MQW nanoporous sample has stable main peak energy below 180 K, and a smaller red shift with increasing the temperature, indicating that carriers in nanoporous MQWs are well confined with little influence of temperature.

W.J. Wang et al. / Superlattices and Microstructures 71 (2014) 38–45

43

It is well known that the S-shaped temperature induced blue-shift of emission peak of InGaN, due to the presence of deeply confined localized states within the tail density of states. For as-grown (T > 80 K) and nanoporous (T > 180 K) MQWs, the PL peak experiences a general red shift which is satisfactorily explained from the Varshni’s formula [20]

EðTÞ ¼ Eð0Þ 

aT 2 bþT



d2 ; kB T

ð2Þ

where E(0) is the energy band gap at zero temperature, a and b are Varshni’s fitting thermal coefficients. The last term on the right-hand side comes from the localization effect, and d is the characteristic parameter describing the shape of the energy band-tail, and kB is Boltzmann’s constant. Similar Sshape behavior has been reported previously for III-nitride ternary or QW system [21–22]. The estimated tail parameter d = 11.86 meV (as-grown MQWs) and 10.57 meV (nanoporous MQWs) are of the same order as the reported values for InGaN/GaN quantum well nanostructures (30 meV) [22], and the coefficients a and b appear to be loosely correlated by b  a 106 K2/eV, which is in agreement with the values usually reported [23]. Meanwhile, compared with the as-grown MQW sample, our nanoporous MQWs have a red-shift at low temperature, which may be due to band tail localization. After the ICP etching, the variation of roughness at interfaces, and the fluctuations of alloy and well thickness in the nanoporous MQWs result in the change of excitonic band tail states. The localization of excitons by potential fluctuations in the nanoporous MQWs causes the PL peak red shift at low temperature [24]. The PL behavior at room temperature shows a blue-shift for the InGaN/GaN nanoporous MQWs in Fig. 4. This result reveals that the nanoporous structure affects the PL behavior because the density of the dislocations and other defects are reduced in the nanoporous MQWs. Therefore, the transition levels of the nanoporous MQWs with fewer defect states and tail states are more concentrated at the band edge compared to the wider transition in as-grown planar MQWs, which is in agreement with the blue-shift and narrower bandwidth in the PL spectrum observed in Fig. 2. Fig. 5 shows the Arrhenius plot of the integrated PL intensity obtained from the main emission over the temperature range from 15 to 300 K. In general, the internal quantum efficiency value at low temperatures 15 K can be regarded as 100% when neglecting the nonradiative recombination process. At room temperature, the internal quantum efficiency values were about 32% and 59.8% for the conventional as-grown and nanoporous MQWs, respectively. Although the gint for the nanoporous MQWs reaches 59.8%, but it is still far from perfect. This indicates that the nanoporous cannot avoid all dislocations and some surface defects may still remain in our structure, which act as the non-radiative centers to reduce the gint. The calculation of the activation energy in the thermal activated processes is generally given as [25]

Fig. 5. Arrhenius plots of the integrated PL intensities of the nanoporous MQWs and as-grown planar MQWs as a function of temperature from 15 to 300 K, where the solid lines represent the fitted results.

44

W.J. Wang et al. / Superlattices and Microstructures 71 (2014) 38–45

Table 1 The fitted results with respect to the experimental results of temperature dependent PL integrated intensities for planar and nanoporous MQWs. Samples

A1

E1/meV

A2

E2/meV

Planar Nanoporous

0.17 0.22

4.66 3.90

6.78 32.74

33.02 107.44

h i X IðTÞ ¼ I0 = 1 þ Ai expðEi =K B TÞ ;

ð3Þ

where Ei are the activation energies of the corresponding nonradiative recombination centers, Ai are rate constants related to the density of these centers, and kB is Boltzmann’s constant. The best fitted parameters of the samples are obtained if two nonradiative channels are included, and the parameters are presented in Table 1. The value of activation energies E1 (4 meV) of the two samples for the first thermal quenching channels at low temperature changes little, and the rate constant is very small. The nonradiative recombination center with activation energy of 4 meV can be related to thermal quenching of bound excitions [26]. The main quenching process of the nanoporous sample is related to the thermal escape of carriers out of confining potential with the large activation energy of 107.44 meV correlated with the depth of the confining potential. It is worth noting that this activation energy value is of the same order as the reported value for InGaN MQWs nanorods [27] much larger than that of InGaN planar MQWs (33.02 meV). Additionally, a larger rate constant of 32.74 is observed for the nanoporous MQWs. This may primarily results from surface damage induced by the ICP etching. Since the activation energy E2 obtained for planar InGaN MQWs is much smaller than the band offset and bandgap energy difference between the wells and the barriers, the thermal quenching of the InGaN emission in the planar MQWs is attributed to the thermal excitation of carriers out of the potential minima caused by energy fluctuations in the band edge induced by alloy disorder [22], and the thermal quenching of the InGaN emission in the nanoporous MQWs with large thermal activation energy of the nanoporous MQWs is due to the thermal activation of carriers from the InGaN wells into the GaN barriers. Large thermal activation energy of the nanoporous MQWs with increasing temperature from 15 to 300 K indicating that carriers are well confined by the GaN barrier and with little influence of non-radiative recombinations on the dislocations and other defects. Therefore, we conclude that the internal quantum efficiency and light extraction efficiency of nanoporous MQWs are reasonably increased to be 1.87 times and 3.66 times than that of the as-grown planar MQWs.

4. Conclusions In summary, nanoporous InGaN/GaN MQWs exhibits significant improvement of the internal quantum efficiency and light extraction efficiency compared with the as-grown planar MQWs. Detailed Xray RSM measurement and temperature dependent PL measurements have confirmed that a drastic reduction of dislocations and other defects in the nanoporous structure, which effectively reduces non-radiative recombination and band-tail in the active region. And thus, higher internal quantum efficiency, higher thermal activation energy, and more stable optical properties with temperature for the nanoporous MQWs are obtained. We believe that the nanoporous MQW structure has great potential to improve the devices’ performance. Acknowledgments This work is supported by Special Funds for Major State Basic Research Project (2011CB301900), the National Nature Science Foundation of China (61176063, 60990311, 60820106003, 60906025, 60936004), the Fundamental Research Funds for the Central Universities (JUSRP11408), the nature science foundation of Jiangsu province (BK2008019, BK2010385, BK2009255, BK2010178, BK2011436), and the Research Funds from NJU-Yangzhou Institute of Optoelectronics.

W.J. Wang et al. / Superlattices and Microstructures 71 (2014) 38–45

45

References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27]

S. Nakamura, G. Fasol, The Blue Laser Diode: The Complete Story, Springer, Berlin, 2000. J. Piprek, Z.M. Simon Li, Appl. Phys. Lett. 102 (2013) 023510. W.C. Peng, Y.S. Wu, Appl. Phys. Lett. 88 (2006) 181117. Y.K. Kuo, T.H. Wang, J.Y. Chang, J.D. Chen, IEEE Photon. Tech. Lett. 24 (2012) 1506. Y.K. Kuo, T.H. Wang, J.Y. Chang, Appl. Phys. Lett. 100 (2012) 031112. S.J. Ronser, E.C. Carr, M.J. Luydowise, G. Girolami, H.I. Erikson, Appl. Phys. Lett. 70 (1997) 420. T. Hino, S. Tomiya, T. Miyajima, K. Yanashima, S. Hashimoto, M. Ikeda, Appl. Phys. Lett. 76 (2000) 3421. X.A. Cao, J.M. Teetsov, M.P. D’Evelyn, D.W. Merfeld, C.H. Yan, Appl. Phys. Lett. 85 (2004) 7. A. Bell, R. Liu, F.A. Ponce, H. Amano, I. Akasaki, D. Cherns, Appl. Phys. Lett. 82 (2003) 349. S. Nagahama, N. Iwasa, M. Senoh, T. Matsusgita, Y. Sugimoto, H. Kiyoku, T. Kozaki, M. Sano, H. Matsumura, H. Umemoto, K. Chocho, T. Mukai, Jpn. J. Appl. Phys. 39 (2000) L647. A. Kikuchi, T. Yamada, S. Nakamura, K. Kusakabe, D. Sugihara, K. Kishino, Jpn. J. Appl. Phys. 39 (2000) L330. Y.R. Wu, C.H. Chiu, C.Y. Chang, P.C. Yu, H.C. Kuo, IEEE J. Sel. Top. Quantum Electron. 15 (2009) 1226. V. Ramesh, A. Kikuchi, K. Kishino, M. Funato, Y. Kawakami, J. Appl. Phys. 107 (2010) 114303. H.M. Kim, Y.H. Cho, H. Lee, S.I. Kim, S.R. Ryu, D.Y. Kim, T.W. Kang, K.S. Chung, Nano Letter. 4 (2004) 1059. W.J. Wang, P. Chen, Z.G. Yu, B. Liu, Z.L. Xie, X.Q. Xiu, Z.L. Wu, F. Xu, Z. Xu, X.M. Hua, P. Han, Y. Shi, R. Zhang, Y.D. Zheng, Chin. Phys. Lett. 30 (2013) 078502. H.S. Chen, D.M. Yeh, Y.C. Lu, C.Y. Chen, C.F. Huang, T.Y. Tang, C.C. Yang, C.S. Wu, C.D. Chen, Nanotechnol. 17 (2006) 1454. Y. Kawakami, S. Suzuki, A. Kaneta, M. Funato, A. Kikuchi, K. Kishino, Appl. Phys. Lett. 89 (2006) 163124. M.L. Kuo, Y.J. Lee, T.C. Shen, S.Y. Lin, Opt. Lett. 34 (2009) 2078. H. Heinke, S. Einfeldt, B. Kuhn-Heinrich, G. Plahl, M.O. Möller, G. Landwehr, J. Phys. D: Appl. Phys. 28 (1995) A104. X.H. Zheng, H. Chen, Z.B. Yan, D.S. Li, H.B. Yu, Q. Huang, J.M. Zhou, J. Appl. Phys. 96 (2004) 1899. Y.S. Park, C.M. Park, D.J. Fu, T.W. Kang, J.E. Oh, Appl. Phys. Lett. 85 (2004) 5718. Y.H. Cho, G.H. Gainer, A.J. Fischer, J.J. Song, S. Keller, U.K. Mishra, S.P. DenBaars, Appl. Phys. Lett. 73 (1998) 1370. W. Shan, B.D. Little, J.J. Song, Z.C. Feng, M. Schurman, R.A. Stall, Appl. Phys. Lett. 69 (1996) 3315. A. Aït-Ouali, R.Y.-F. Yip, J.L. Brebner, R.A. Masut, J. Appl. Phys. 83 (1998) 3153. D. Bimberg, M. Sondergeld, E. Grobe, Phys. Rev. B 4 (1971) 3451. A. Yasan, R. McClintock, K. Mayes, D.H. Kim, P. Kung, M. Razeghi, Appl. Phys. Lett. 83 (2003) 4083. Y. Sun, Y.H. Cho, H.M. Kim, T.W. Kang, Appl. Phys. Lett. 87 (2005) 093115.