- Email: [email protected]
GaN multiple quantum well solar cells performance
Accepted Manuscript Optimization of growth and fabrication techniques to enhance the InGaN/GaN multiple quantum well solar cells performance
Shiming Liu, Quan Wang, Hongling Xiao, Kun Wang, Cuimei Wang, Xiaoliang Wang, Weikun Ge, Zhanguo Wang PII:
S0749-6036(17)30502-5
DOI:
10.1016/j.spmi.2017.05.014
Reference:
YSPMI 4992
To appear in:
Superlattices and Microstructures
Received Date:
28 February 2017
Revised Date:
03 May 2017
Accepted Date:
03 May 2017
Please cite this article as: Shiming Liu, Quan Wang, Hongling Xiao, Kun Wang, Cuimei Wang, Xiaoliang Wang, Weikun Ge, Zhanguo Wang, Optimization of growth and fabrication techniques to enhance the InGaN/GaN multiple quantum well solar cells performance, Superlattices and Microstructures (2017), doi: 10.1016/j.spmi.2017.05.014
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT Highlights
A new structure of InGaN/GaN MQW solar cell with thin quantum wells was grown. The carrier lifetime and peak EQE of the new sample are 7.8 ns and 61%, respectively. The conversion efficiency of 3.56% is among the best for InGaN/GaN solar cells.
1
ACCEPTED MANUSCRIPT Optimization of growth and fabrication techniques to enhance the InGaN/GaN multiple quantum well solar cells performance Shiming Liu1,3, Quan Wang1, Hongling Xiao1,3, Kun Wang1, Cuimei Wang1,3, Xiaoliang Wang1,2,3,*, Weikun Ge1, Zhanguo Wang1,2,3 1Key
Lab of Semiconductor Materials Science, Institute of Semiconductors, Chinese
Academy of Sciences, Beijing 100083, People’s Republic of China 2Beijing
Key Laboratory of Low Dimensional Semiconductor Materials and Devices,
Beijing 100083, People’s Republic of China 3University
of Chinese Academy of Sciences, Beijing 100049, People’s Republic of
China * Corresponding author. Tel: +86-010-82304170; fax: +86-010-82304140. E-mail: [email protected] (X.L. Wang)
Abstract : Two structures of InGaN/GaN multiple quantum well solar cells are grown and fabricated in this work, which are 30 periods In0.14Ga0.86N/GaN (1.72 nm/4.14 nm, sample A) and In0.19Ga0.81N/GaN (2.76 nm/4.14 nm, sample B), respectively. The peak external quantum efficiency of sample A can reach over 60%, which is considered to be related with the longer carrier lifetime. The current density-voltage characteristics show that sample A exhibited an open-circuit voltage of 2.13 V, a short-circuit current density of 2.55 mA/cm2, a fill-factor of 65.7% and a conversion efficiency of 3.56%. Those indicators are among the best up to date as far as we know for InGaN/GaN solar cells.
Keywords: InGaN; multiple quantum well; solar cell; conversion efficiency
1
ACCEPTED MANUSCRIPT 1. Introduction Ⅲ-nitride materials attract comprehensive attention in both electronic and optoelectronic fields [1,2]. InGaN materials have shown superior performance in laser diodes (LD) and light-emitting diodes (LED) [3,4]. Since the band gap of InN was proved to be 0.7eV [5-7], InGaN-based alloys could absorb a broad range of solar spectrum ranging from ultraviolet to infrared, they had thus become very promising materials for full spectrum solar cells. Theoretical studies suggest that the conversion efficiency of InGaN-based solar cells can reach to 60% [8-13]. InGaN materials have also high absorption coefficients of about 105 cm-1 to solar light and superior radiation resistance compared with Si and GaAs [14,15]. Based on those outstanding features, InGaN-based alloys should be extremely suitable for fabrication of space solar cells. However, the conversion efficiency of the fabricated InGaN-based solar cells remains so far much lower than theoretical predictions [16-19], which is mainly due to the difficulty of growing high crystalline quality InGaN layer with high indium composition. It is also hard to grow thick InGaN layers due to the large lattice mismatch between InN and GaN [20,21]. In order to overcome the above obstacles, the research of InGaN solar cells has been focused on InGaN/GaN multiple quantum well (MQW) solar cells in recent years [22-27]. For example, Y.J. Lee et al. have used patterned sapphire substrates (PSS) to improve the crystalline quality in MQW, the external quantum efficiency (EQE) and the performance of InGaN/GaN MQW solar cells [28]. S.B. Choi et al. have discussed the effect of indium composition with 17% and 25%, the results show that the solar cell with lower indium composition of 17% has a higher conversion efficiency due to better carrier escape [29]. The effect of the quantum well thickness with 1.3, 3.1 and 5.4 nm is investigated by L. Redaelli et al., and their results indicate that even though thicker QW thickness can absorb more light, it cannot be fully exploited to improve the efficiency of solar cells, due to the reduced overlap of the ground state electron and hole wavefunctions caused by the quantum confined Stark effect [30]. J.J. Wierer et al. have studied the influence of barrier thickness of 3.0, 6.3 and 10.0 nm on the performance of InGaN/GaN MQW solar cells, and found that the 6.3 nm barrier sample exhibits the best overall 2
ACCEPTED MANUSCRIPT performance [31]. In this letter, all of the above achievement have been taken into account. Considering the structure of 30 periods In0.20Ga0.80N/GaN (3 nm/4 nm) MQW solar cell which can achieve a high performance of 3.33% and the discussions about the effects of substrates, indium composition, and quantum well and barrier thicknesses [28-32], we designed a new structure of 30 periods In0.15Ga0.85N/GaN (1.8 nm/4 nm) grown on PSS as sample A, while a solar cell with the above mentioned MQW structure grown on PSS as sample B. Through the current density-voltage (J-V) characteristics which is measured under Air Mass 1.5 Global (AM 1.5G) illumination (100 mW/cm2), sample A exhibited good performance with conversion efficiency of 3.56%.
2. Experimental procedure Metal organic chemical vapor deposition (MOCVD) was employed to grow the two InGaN/GaN MQW structures on PSS. The substrates were cleaned by H2 at 1100 °C, followed by the growth of a 1 μm unintentionally doped GaN buffer layer, and a 2.5 μm Si-doped n-type (5×1018 cm-3) GaN layer at 1055 °C with the Ⅴ/Ⅲ ratio of ～1200. Then, the MQW structures were grown as the light absorbing region. The growth temperatures of the InGaN quantum wells are about 690 °C for sample A and 680 °C for sample B, and the Ⅴ/Ⅲ ratio is ～2000. The growth temperature of GaN barriers is 810 °C, with theⅤ/Ⅲ ratio of ～12000. After that, a 150 nm Mgdoped p-type (5×1017 cm-3) GaN layer was grown at 920 °C with the Ⅴ/Ⅲ ratio of ～5000. A Ni/Au (5 nm/5 nm) metal system was deposited as a semitransparent pcontact layer and the Ohmic contact for n-GaN was defined by Ti/Al/Ti/Au (30 nm/100 nm/30 nm/60 nm) metal system. No antireflection layer was coated on the solar cells. Figure 1 shows the top view of a fabricated device under an optical microscope together with the cross section of a typical InGaN/GaN MQW solar cell.
3
ACCEPTED MANUSCRIPT
Fig. 1. The top view of a fabricated device and the cross section of the InGaN/GaN MQW solar cell structure.
3. Results and discussions In the process of growing InGaN/GaN MQW solar cell structures, the thicknesses of the quantum wells and barriers are usually slightly different to the designed values. Thus, cross-sectional high-resolution transmission electron microscopy (HRTEM) images were implemented for samples A and B as depicted in Fig. 2, in order to obtain the exact MQW thicknesses for the two samples. The HRTEM is conducted on Tecnai G2 F20 S-Twin at 200 kV. Figure 2 (a) and (b) show the micrographs with large areas from samples A and B, respectively. Discernible threading dislocations (TDs) in sample B can be observed, whereas large TDs in sample A are almost unobservable. In Fig. 2 (b), it can be seen that the dislocation pits are formed when the TDs pass through the MQWs, and the TDs will extend to the surface. These TDs may act as leakage passages which can worsen the performance of solar cells. Detailed MQWs structures of the two samples are shown in Fig. 2 (c) and (d). The interfaces between GaN barrier layers and InGaN quantum well layers in sample A are clearer than that in sample B. The average thickness of the quantum wells are measured as 1.72 nm for sample A and 2.76 nm for sample B, and the barriers are 4
ACCEPTED MANUSCRIPT about 4.14 nm for both samples.
Fig. 2. (a) and (b) are the cross-sectional HRTEM images with large areas of samples A and B,
respectively, (c) and (d) are the detailed images of MQWs for the two
samples. High-resolution x-ray diffraction (HRXRD) ω-2θ scans of (0002) reflection measured by PANalytical X’Pert3 MRD and the simulation curves performed by the X’Pert Epitaxy software for sample A and sample B are shown in Fig. 3. Sharp satellite peaks can be observed from the MQWs, indicating good structural quality. The full width at half maximum (FWHM) of the satellite (i.e. 0th order satellite) from sample A is much lower than that of sample B, reflecting the crystalline quality of sample A is better than that of sample B. This is mainly due to the lower indium composition and lower thickness of quantum well for the former. The thicknesses of one period calculated by the distance between the satellites are 5.8 nm for sample A and 6.8 nm for sample B, both are in accordance with the results measured by HRTEM.
5
ACCEPTED MANUSCRIPT
Fig. 3. HRXRD ω-2θ scans of (0002) reflections and the corresponding simulated curves for the two samples. The XRD reciprocal space mapping (RSM) of the asymmetric (-1015) reflection are measured to obtain the information of strain in MQW as well as the indium composition for samples A and B. In Fig. 4 (a), the vertical dashed line labeled as “S” is from the GaN pseudo-substrate. The Qx of GaN pseudo-substrate peak and InGaN/GaN MQWs satellite peaks from sample A are in the same place, while for sample B, a very small shift of InGaN/GaN MQWs satellite peaks to the GaN pseudosubstrate peak can be detected as that shown in Fig. 4 (b). It indicates that the structure of InGaN/GaN MQWs is fully strained in sample A, but partial relaxed in sample B. The degree of relaxation is defined as R = (aMQW - aGaN)/(aMQW0 - aGaN) [30], and it is estimated to be 7.4% for sample B. Besides, according to the reference [33], the indium composition of InGaN quantum wells can be calculated as 0.14 and 0.19 for samples A and B, respectively. The results conform to the simulated curves in Fig. 3, the actual MQW structures of samples A and B are therefore obtained.
6
ACCEPTED MANUSCRIPT
Fig. 4. (a) RSM of the (-1015) reflection for sample A. (b) RSM of the (-1015) reflection for sample B. Figure 5 shows the time-resolved photoluminescence (TRPL) decay curves measured for sample A and sample B at room temperature using a 375 nm pulsed laser source with 20 MHz. The rate equation of the carrier density N(t) in QW can be written as [34] 𝑑𝑁/𝑑𝑡 =‒ 𝑁/𝜏 + (1 ‒ 𝑅)𝛼𝑃pump/(ℎ𝜐L𝐴L)
(1) (2)
1/𝜏 = 1/𝜏nr + 1/𝜏r
where R is the reflectivity on the semiconductor surface, is the absorption coefficient in QWs, Ppump and hυL are the time averaged power and photon energy of the excitation source, respectively, and AL is the net illuminated area. τ is the carrier lifetime, which can be separated into non-radiative one (τnr) and radiative one (τr). τnr and τr are given by: 1/𝜏final = 2/𝜏nr
(3)
1/𝜏initial = 2(1/𝜏nr + 1/𝜏r)
(4)
where τinitial and τfinal can be calculated based on the slope of the initial and final stages of the decay curves. As is shown in Fig. 5, each TRPL curve can be well fitted with the dashed curve which is determined by the summation of the two exponential functions. According to Eqs. (3) and (4), the values of τnr and τr are calculated to be 72.8 ns and 8.7 ns for sample A, and 66.6 ns and 1.6 ns for sample B, respectively. 7
ACCEPTED MANUSCRIPT The carrier lifetime of sample A and sample B can then be obtained as7.8 ns and 1.6 ns based on Eq. (2), respectively. Compared to sample B, longer carrier lifetime suggests that the photon-generated carriers of sample A will have more time to escape from MQWs before being recombined, and hence leading to higher conversion efficiency.
Fig. 5. TRPL decay curves and the fitted curves of sample A and sample B. In order to study the tendency of whether the photogenerated carriers escape from MQWs to the electrodes or recombine in MQWs, external quantum efficiency (EQE) are measured by Enlitech QE-R solar cell analysis system. As is shown in Fig. 6, samples A and B show zero efficiency over 450 nm and 470 nm, respectively. The redshift of the photoresponse is due to the higher indium composition in sample B. As is known, sample with higher indium composition can absorb more light and may have a higher Jsc. However, under the wavelength of 410 nm, the EQE of sample B is much lower than that of sample A, and the peak EQE of sample B is only 32% at 400 nm compared to that of 61% at 380 nm for sample A. It can be deduced that compared with sample B, the photogenerated carriers in sample A were apt to escape from MQWs, resulting from the higher carrier lifetime in sample A.
8
ACCEPTED MANUSCRIPT
Fig. 6. EQE of samples A and B measured from 300 nm to 480 nm. Energy band diagram can provide guidance for understanding the tendency of carrier transportation. Fig. 7 shows the energy band and electric field simulations of single QW in samples A and B. The simulated models and material parameters are described by references [35-37]. It can be observed that the energy levels of electrons and holes in QW of sample B are closer than that of sample A, and the reverse electric field of QW in sample B is higher than that in sample A. Therefore, the recombination of photogenerated carriers in sample A will be reduced due to the lower indium composition and well thickness, implying a better performance.
Fig. 7. Simulated energy band diagrams and electric fields of single QW in samples A and B. Figure 8(a) displays the dark current-voltage (I-V) curves of sample A and sample B which are measured using Agilent B2902A source/measure unit. The leakage 9
ACCEPTED MANUSCRIPT current of sample A is visibly lower than that of sample B under a reverse bias voltage of -5 V. While the current density-voltage (J-V) characteristics of the two samples are measured under AM 1.5G illumination in Fig. 8(b). Compared with sample B, the values of open-circuit voltage (Voc) and short-circuit current density (Jsc) for sample A are much enhanced due to the lower leakage current and longer carrier lifetime. Finally, the summary of the measured solar cell parameters and the InGaN/GaN MQW structures are shown in Table 1.
Fig. 8. (a) The dark I-V curves of sample A and sample B. (b) The J-V characteristics of sample A and sample B under AM1.5G illumination. Table 1 Summary of InGaN/GaN MQW structures and solar cell parameters of the samples.
Sample
Indium composition
InGaN well
GaN barrier
thickness
thickness
(nm)
(nm)
Voc
Jsc
FF
(V)
(mA/cm2)
(%)
η (%)
A
0.14
1.72
4.14
2.13
2.55
65.7
3.56
B
0.19
2.76
4.14
1.16
1.82
49.0
1.04
4. Summary In summary, we fabricated an InGaN/GaN MQW solar cell with new structure and studied the characteristics in comparison with the general structure. It can be seen that although the indium composition and quantum well thickness of sample A are lower than that of sample B, which indicates the structure of sample A has no advantage in absorbing more light compared with sample B, a much enhanced Jsc and 10
ACCEPTED MANUSCRIPT Voc of sample A are however obtained. Confirmed by HRTEM, HRXRD and RSM measurements, the crystalline quality of sample A is surely better than sample B, which results in a decrease of the leakage current and a higher Voc for sample A. In addition, as measured by TRPL, the carrier lifetime of sample A (7.8 ns) is much longer than that of sample B (1.6 ns), which can assist the carrier transportation and enhance Jsc. Moreover, the peak EQE of sample A is 61%, which is larger than the 32% of sample B. Consequently, sample A has a better performance with conversion efficiency of 3.56%, being three times better than sample B.
Acknowledgments This work was supported by the Beijing Municipal Science and Technology Project [grant number Z161100000416002]; and the State Key Development Program for Basic Research of China [grant number 2012CB619303].
11
ACCEPTED MANUSCRIPT References [1] S.J. Pearton, F. Ren, GaN Electronics, Adv. Mater. 12 (2000) 1571-1580. [2] H.X. Jiang, S.X. Jin, J. Li, J. Shakya, J. Y. Lin, III-nitride blue microdisplays, Appl. Phys. Lett. 78 (2001) 1303-1305. [3] T.C. Lu, C.C. Kao, H.C. Kuo, G.S. Huang, S.C. Wang, CW lasing of current injection blue GaN-based vertical cavity surface emitting laser, Appl. Phys. Lett. 92 (2008) 141102. [4] S. Nakamura, The roles of structural imperfections in InGaN-based blue lightemitting diodes and laser diodes, Science 281(1998) 956-961. [5] J. Wu, W. Walukiewicz, K.M. Yu, W. Shan, J.W. Ager III, E.E. Haller, Hai Lu, William J. Schaff, W.K. Metzger, Sarah Kurtz, Superior radiation resistance of In1−xGaxN alloys: Full-solar-spectrum photovoltaic material system, J. Appl. Phys. 94 (2003) 6477-6482. [6] H.L. Xiao, X.L. Wang, J.X. Wang, N.H. Zhang, H.X. Liu, Y.P. Zeng, J.M. Jin, Z.G. Wang, Growth and characterization of InN on sapphire substrate by RFMBE, J. Cryst. Growth 276 (2005) 401-406. [7] V.Y. Davydov, A.A. Klochikhin, V.V. Emtsev, S.V. Ivanov, V.V. Vekshin, F. Bechstedt, J. Furthmuller, H. Harima, A.V. Mudryi, A. Hashimoto, A. Yamamoto, J. Aderhold, J. Graul, E.E. Haller, Band Gap of InN and In-Rich InxGa1-xN alloys (0.36 < x < 1), Phys. Status Solidi B 230 (2002) R4-R6. [8] A. Luque, A. Martí, Increasing the efficiency of ideal solar cells by photon induced transitions at intermediate levels, Phys. Rev. Lett. 78 (1997) 5014-5017. [9] B. Wen, J.J. Zhou, R.L. Jiang, Z.L. Xie, D.J. Chen, X.L. Ji, P. Han, R. Zhang, Y.D. Zheng, Theoretical calculation of conversion efficiency of InGaN solar cells, J. Semicond. 28 (2007) 1392-1395. [10]X.B. Zhang, X.L. Wang, H.L. Xiao, C.B. Yang, J.X. Ran, C.M. Wang, Q.F. Hou, J.M. Li, Z.G. Wang, Theoretical design and performance of InxGa1−xN twojunction solar cells, J. Phys. D: Appl. Phys. 41 (2008) 245106. [11]T. Nozawa, Y. Arakawa, Detailed balance limit of the efficiency of multilevel intermediate band solar cells, Appl. Phys. Lett. 98 (2011) 171108. 12
ACCEPTED MANUSCRIPT [12]Q.W. Deng, X.L. Wang, H.L. Xiao, C.M. Wang, H.B. Yin, H. Chen, Q.F. Hou, D.F. Lin, J.M. Li, Z.G. Wang, X. Hou, An investigation on InxGa1−xN/GaN multiple quantum well solar cells, J. Phys. D: Appl. Phys. 44 (2011) 265103. [13]A. Adaine, S.O. Hamady, N. Fressengeas, Simulation study of a new InGaN player free Schottky based solar cell, Superlattices Microstruct. 96 (2016) 121-133. [14]J.F. Muth, J.H. Lee, I.K. Shmagin, R.M. Kolbas, H.C. Casey, B.P. Keller, U.K. Mishra, S.P. DenBaars, Absorption coefficient, energy gap, exciton binding energy, and recombination lifetime of GaN obtained from transmission measurements, Appl. Phys. Lett. 71 (1997) 2572-2574. [15]J. Wu, When group-III nitrides go infrared: New properties and perspectives, J. Appl. Phys. 106 (2009) 011101. [16]C.B. Yang, X.L. Wang, H.L. Xiao, J.X. Ran, C.M. Wang, G.X. Hu, X.H. Wang, X.B. Zhang, J.P. Li, J.M. Li, Photovoltaic effects in InGaN structures with p–n junctions, Phys. Status Solidi A 204 (2007) 4288-4291. [17]J.N. Carl, G.T. Nikholas, C.C. Samantha, I. Michael, P.D. Steven, K.M. Umesh, High quantum efficiency InGaN/GaN solar cells with 2.95 eV band gap, Appl. Phys. Lett. 93 (2008) 143502. [18]H.W. Wang, H.C. Chen, Y.A. Chang, C.C. Lin, H.W. Han, M.A. Tsai, H.C. Kuo, Conversion efficiency enhancement of GaN/InGaN solar cells with nano patterned sapphire and biomimetic surface antireflection process, Photonic Tech. L. 23 (2011) 1304-1306. [19]L. Jing, H.L. Xiao, X.L. Wang, C.M. Wang, Q.W. Deng, Z.D. Li, J.Q. Ding, Z.G. Wang, X. Hou, Enhanced performance of InGaN/GaN multiple quantum well solar cells with patterned sapphire substrate, J. Semicond. 34 (2013) 124004. [20]A. Wakahara, T. Tokuda, X.Z. Dang, S. Noda, A. Sasaki, Compositional inhomogeneity and immiscibility of a GaInN ternary alloy, Appl. Phys. Lett. 71 (1997) 906-908. [21]I.H. Kim, H.S. Park, Y.J. Park, T.K. Dahal, Formation of V-shaped pits in InGaN/GaN multiquantum wells and bulk InGaN films, Appl. Phys. Lett. 73 (1998) 1634-1636. 13
ACCEPTED MANUSCRIPT [22]R. Dahal, J. Li, K. Aryal, J. Y. Lin, H. X. Jiang, InGaN/GaN multiple quantum well concentrator solar cells, Appl. Phys. Lett. 97 (2010) 073115. [23]J.H. Song, J.H. Oh, J.P. Shim, J.H. Min, D.S. Lee, T.Y. Seong, Improved efficiency of InGaN/GaN-based multiple quantum well solar cells by reducing contact resistance, Superlattices Microstruct. 52 (2012) 299-305. [24]N.G. Young, E.E. Perl, R.M. Farrell, M. Iza, S. Keller, J.E. Bowers, S. Nakamura, S.P. DenBaars, J.S. Speck, High-performance broadband optical coatings on InGaN/GaN solar cells for multijunction device integration, Appl. Phys. Lett. 104 (2014) 163902. [25]S. Valdueza-Felip, A. Mukhtarova, L. Grenet, C. Bougerol, C. Durand, J. Eymery, E. Monroy, Improved conversion efficiency of as-grown InGaN/GaN quantum-well solar cells for hybrid integration, Appl. Phys. Express 7 (2014) 032301. [26]S.M. Liu, H. L. Xiao, Q. Wang, J. D. Yan, X. M. Zhan, J. M. Gong, X. L. Wang, Z. G. Wang, InxGa1− xN/GaN multiple quantum well solar cells with conversion efficiency of 3.77%, Chin. Phys. Lett. 32 (2015) 088401. [27]A. Mukhtarova, S. Valdueza-Felip, L. Redaelli, C. Durand, C. Bougerol, E. Monroy, J. Eymery, Dependence of the photovoltaic performance of pseudomorphic InGaN/GaN multiple-quantum-well solar cells on the active region thickness, Appl. Phys. Lett. 108 (2016) 161907. [28]Y.J. Lee, M.H. Lee, C.M. Cheng, C.H. Yang, Enhanced conversion efficiency of InGaN multiple quantum well solar cells grown on a patterned sapphire substrate, Appl. Phys. Lett. 98 (2011) 263504. [29]S.B. Choi, J.P. Shim, D.M. Kim, H.L. Jeong, Y.D. Jho, Y.H. Song, D.S. Lee, Effect of indium composition on carrier escape in InGaN/GaN multiple quantum well solar cells, Appl. Phys. Lett. 103 (2013) 033901. [30]L. Redaelli, A. Mukhtarova, S. Valdueza-Felip, A. Ajay, C. Bougerol, C. Himwas, J. Faure-Vincent, C. Durand, J. Eymery, E. Monroy, Effect of the quantum well thickness on the performance of InGaN photovoltaic cells, Appl. Phys. Lett. 105 (2014) 131105. 14
ACCEPTED MANUSCRIPT [31]J. J. Wierer, Jr., D. D. Koleske, S. R. Lee, Influence of barrier thickness on the performance of InGaN/GaN multiple quantum well solar cells, Appl. Phys. Lett. 100 (2012) 111119. [32]N. G. Young, R. M. Farrell, Y. L. Hu, Y. Terao, M. Iza, S. Keller, S. P. DenBaars, S. Nakamura, J. S. Speck, High performance thin quantum barrier InGaN/GaN solar cells on sapphire and bulk (0001) GaN substrates, Appl. Phys. Lett. 103 (2013) 173903. [33]M.A. Moram, M.E. Vickers, X-ray diffraction of III-nitrides, Rep. Prog. Phys. 72 (2009) 036502. [34]H. Kim, D.S. Shin, H.Y. Ryu, J.I. Shim, Analysis of Time-resolved photoluminescence of InGaN quantum wells using the carrier rate equation, Jpn. J. Appl. Phys. 49 (2010) 112402. [35]T. Tao, T. Zhi, B. Liu, M.x. Li, Z. Zhuang, J.P. Dai, Y. Li, F.L. Jiang, W.j. Luo, Z.L. Xie, D.J. Chen, P. Chen, Z.S. Li, Z.G. Zou, R. Zhang, and Y.D. Zheng, Significant improvements in InGaN/GaN nano-photoelectrodes for hydrogen generation by structure and polarization optimization, Sci. Rep. 6 (2016) 20218. [36]G.F. Brown, J.W. Ager III, W. Walukiewicz and J. Wu, “Finite element simulations of compositionally graded InGaN solar cells,” Sol. Energy Mater. Sol. Cells 94 (2010) 478. [37]I. Vurgaftman and J. R. Meyer, “Band parameters for nitrogen-containing semiconductors”, J. Appl. Phys. 94 (2007) 073705.
15
Shiming Liu, Quan Wang, Hongling Xiao, Kun Wang, Cuimei Wang, Xiaoliang Wang, Weikun Ge, Zhanguo Wang PII:
S0749-6036(17)30502-5
DOI:
10.1016/j.spmi.2017.05.014
Reference:
YSPMI 4992
To appear in:
Superlattices and Microstructures
Received Date:
28 February 2017
Revised Date:
03 May 2017
Accepted Date:
03 May 2017
Please cite this article as: Shiming Liu, Quan Wang, Hongling Xiao, Kun Wang, Cuimei Wang, Xiaoliang Wang, Weikun Ge, Zhanguo Wang, Optimization of growth and fabrication techniques to enhance the InGaN/GaN multiple quantum well solar cells performance, Superlattices and Microstructures (2017), doi: 10.1016/j.spmi.2017.05.014
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT Highlights
A new structure of InGaN/GaN MQW solar cell with thin quantum wells was grown. The carrier lifetime and peak EQE of the new sample are 7.8 ns and 61%, respectively. The conversion efficiency of 3.56% is among the best for InGaN/GaN solar cells.
1
ACCEPTED MANUSCRIPT Optimization of growth and fabrication techniques to enhance the InGaN/GaN multiple quantum well solar cells performance Shiming Liu1,3, Quan Wang1, Hongling Xiao1,3, Kun Wang1, Cuimei Wang1,3, Xiaoliang Wang1,2,3,*, Weikun Ge1, Zhanguo Wang1,2,3 1Key
Lab of Semiconductor Materials Science, Institute of Semiconductors, Chinese
Academy of Sciences, Beijing 100083, People’s Republic of China 2Beijing
Key Laboratory of Low Dimensional Semiconductor Materials and Devices,
Beijing 100083, People’s Republic of China 3University
of Chinese Academy of Sciences, Beijing 100049, People’s Republic of
China * Corresponding author. Tel: +86-010-82304170; fax: +86-010-82304140. E-mail: [email protected] (X.L. Wang)
Abstract : Two structures of InGaN/GaN multiple quantum well solar cells are grown and fabricated in this work, which are 30 periods In0.14Ga0.86N/GaN (1.72 nm/4.14 nm, sample A) and In0.19Ga0.81N/GaN (2.76 nm/4.14 nm, sample B), respectively. The peak external quantum efficiency of sample A can reach over 60%, which is considered to be related with the longer carrier lifetime. The current density-voltage characteristics show that sample A exhibited an open-circuit voltage of 2.13 V, a short-circuit current density of 2.55 mA/cm2, a fill-factor of 65.7% and a conversion efficiency of 3.56%. Those indicators are among the best up to date as far as we know for InGaN/GaN solar cells.
Keywords: InGaN; multiple quantum well; solar cell; conversion efficiency
1
ACCEPTED MANUSCRIPT 1. Introduction Ⅲ-nitride materials attract comprehensive attention in both electronic and optoelectronic fields [1,2]. InGaN materials have shown superior performance in laser diodes (LD) and light-emitting diodes (LED) [3,4]. Since the band gap of InN was proved to be 0.7eV [5-7], InGaN-based alloys could absorb a broad range of solar spectrum ranging from ultraviolet to infrared, they had thus become very promising materials for full spectrum solar cells. Theoretical studies suggest that the conversion efficiency of InGaN-based solar cells can reach to 60% [8-13]. InGaN materials have also high absorption coefficients of about 105 cm-1 to solar light and superior radiation resistance compared with Si and GaAs [14,15]. Based on those outstanding features, InGaN-based alloys should be extremely suitable for fabrication of space solar cells. However, the conversion efficiency of the fabricated InGaN-based solar cells remains so far much lower than theoretical predictions [16-19], which is mainly due to the difficulty of growing high crystalline quality InGaN layer with high indium composition. It is also hard to grow thick InGaN layers due to the large lattice mismatch between InN and GaN [20,21]. In order to overcome the above obstacles, the research of InGaN solar cells has been focused on InGaN/GaN multiple quantum well (MQW) solar cells in recent years [22-27]. For example, Y.J. Lee et al. have used patterned sapphire substrates (PSS) to improve the crystalline quality in MQW, the external quantum efficiency (EQE) and the performance of InGaN/GaN MQW solar cells [28]. S.B. Choi et al. have discussed the effect of indium composition with 17% and 25%, the results show that the solar cell with lower indium composition of 17% has a higher conversion efficiency due to better carrier escape [29]. The effect of the quantum well thickness with 1.3, 3.1 and 5.4 nm is investigated by L. Redaelli et al., and their results indicate that even though thicker QW thickness can absorb more light, it cannot be fully exploited to improve the efficiency of solar cells, due to the reduced overlap of the ground state electron and hole wavefunctions caused by the quantum confined Stark effect [30]. J.J. Wierer et al. have studied the influence of barrier thickness of 3.0, 6.3 and 10.0 nm on the performance of InGaN/GaN MQW solar cells, and found that the 6.3 nm barrier sample exhibits the best overall 2
ACCEPTED MANUSCRIPT performance [31]. In this letter, all of the above achievement have been taken into account. Considering the structure of 30 periods In0.20Ga0.80N/GaN (3 nm/4 nm) MQW solar cell which can achieve a high performance of 3.33% and the discussions about the effects of substrates, indium composition, and quantum well and barrier thicknesses [28-32], we designed a new structure of 30 periods In0.15Ga0.85N/GaN (1.8 nm/4 nm) grown on PSS as sample A, while a solar cell with the above mentioned MQW structure grown on PSS as sample B. Through the current density-voltage (J-V) characteristics which is measured under Air Mass 1.5 Global (AM 1.5G) illumination (100 mW/cm2), sample A exhibited good performance with conversion efficiency of 3.56%.
2. Experimental procedure Metal organic chemical vapor deposition (MOCVD) was employed to grow the two InGaN/GaN MQW structures on PSS. The substrates were cleaned by H2 at 1100 °C, followed by the growth of a 1 μm unintentionally doped GaN buffer layer, and a 2.5 μm Si-doped n-type (5×1018 cm-3) GaN layer at 1055 °C with the Ⅴ/Ⅲ ratio of ～1200. Then, the MQW structures were grown as the light absorbing region. The growth temperatures of the InGaN quantum wells are about 690 °C for sample A and 680 °C for sample B, and the Ⅴ/Ⅲ ratio is ～2000. The growth temperature of GaN barriers is 810 °C, with theⅤ/Ⅲ ratio of ～12000. After that, a 150 nm Mgdoped p-type (5×1017 cm-3) GaN layer was grown at 920 °C with the Ⅴ/Ⅲ ratio of ～5000. A Ni/Au (5 nm/5 nm) metal system was deposited as a semitransparent pcontact layer and the Ohmic contact for n-GaN was defined by Ti/Al/Ti/Au (30 nm/100 nm/30 nm/60 nm) metal system. No antireflection layer was coated on the solar cells. Figure 1 shows the top view of a fabricated device under an optical microscope together with the cross section of a typical InGaN/GaN MQW solar cell.
3
ACCEPTED MANUSCRIPT
Fig. 1. The top view of a fabricated device and the cross section of the InGaN/GaN MQW solar cell structure.
3. Results and discussions In the process of growing InGaN/GaN MQW solar cell structures, the thicknesses of the quantum wells and barriers are usually slightly different to the designed values. Thus, cross-sectional high-resolution transmission electron microscopy (HRTEM) images were implemented for samples A and B as depicted in Fig. 2, in order to obtain the exact MQW thicknesses for the two samples. The HRTEM is conducted on Tecnai G2 F20 S-Twin at 200 kV. Figure 2 (a) and (b) show the micrographs with large areas from samples A and B, respectively. Discernible threading dislocations (TDs) in sample B can be observed, whereas large TDs in sample A are almost unobservable. In Fig. 2 (b), it can be seen that the dislocation pits are formed when the TDs pass through the MQWs, and the TDs will extend to the surface. These TDs may act as leakage passages which can worsen the performance of solar cells. Detailed MQWs structures of the two samples are shown in Fig. 2 (c) and (d). The interfaces between GaN barrier layers and InGaN quantum well layers in sample A are clearer than that in sample B. The average thickness of the quantum wells are measured as 1.72 nm for sample A and 2.76 nm for sample B, and the barriers are 4
ACCEPTED MANUSCRIPT about 4.14 nm for both samples.
Fig. 2. (a) and (b) are the cross-sectional HRTEM images with large areas of samples A and B,
respectively, (c) and (d) are the detailed images of MQWs for the two
samples. High-resolution x-ray diffraction (HRXRD) ω-2θ scans of (0002) reflection measured by PANalytical X’Pert3 MRD and the simulation curves performed by the X’Pert Epitaxy software for sample A and sample B are shown in Fig. 3. Sharp satellite peaks can be observed from the MQWs, indicating good structural quality. The full width at half maximum (FWHM) of the satellite (i.e. 0th order satellite) from sample A is much lower than that of sample B, reflecting the crystalline quality of sample A is better than that of sample B. This is mainly due to the lower indium composition and lower thickness of quantum well for the former. The thicknesses of one period calculated by the distance between the satellites are 5.8 nm for sample A and 6.8 nm for sample B, both are in accordance with the results measured by HRTEM.
5
ACCEPTED MANUSCRIPT
Fig. 3. HRXRD ω-2θ scans of (0002) reflections and the corresponding simulated curves for the two samples. The XRD reciprocal space mapping (RSM) of the asymmetric (-1015) reflection are measured to obtain the information of strain in MQW as well as the indium composition for samples A and B. In Fig. 4 (a), the vertical dashed line labeled as “S” is from the GaN pseudo-substrate. The Qx of GaN pseudo-substrate peak and InGaN/GaN MQWs satellite peaks from sample A are in the same place, while for sample B, a very small shift of InGaN/GaN MQWs satellite peaks to the GaN pseudosubstrate peak can be detected as that shown in Fig. 4 (b). It indicates that the structure of InGaN/GaN MQWs is fully strained in sample A, but partial relaxed in sample B. The degree of relaxation is defined as R = (aMQW - aGaN)/(aMQW0 - aGaN) [30], and it is estimated to be 7.4% for sample B. Besides, according to the reference [33], the indium composition of InGaN quantum wells can be calculated as 0.14 and 0.19 for samples A and B, respectively. The results conform to the simulated curves in Fig. 3, the actual MQW structures of samples A and B are therefore obtained.
6
ACCEPTED MANUSCRIPT
Fig. 4. (a) RSM of the (-1015) reflection for sample A. (b) RSM of the (-1015) reflection for sample B. Figure 5 shows the time-resolved photoluminescence (TRPL) decay curves measured for sample A and sample B at room temperature using a 375 nm pulsed laser source with 20 MHz. The rate equation of the carrier density N(t) in QW can be written as [34] 𝑑𝑁/𝑑𝑡 =‒ 𝑁/𝜏 + (1 ‒ 𝑅)𝛼𝑃pump/(ℎ𝜐L𝐴L)
(1) (2)
1/𝜏 = 1/𝜏nr + 1/𝜏r
where R is the reflectivity on the semiconductor surface, is the absorption coefficient in QWs, Ppump and hυL are the time averaged power and photon energy of the excitation source, respectively, and AL is the net illuminated area. τ is the carrier lifetime, which can be separated into non-radiative one (τnr) and radiative one (τr). τnr and τr are given by: 1/𝜏final = 2/𝜏nr
(3)
1/𝜏initial = 2(1/𝜏nr + 1/𝜏r)
(4)
where τinitial and τfinal can be calculated based on the slope of the initial and final stages of the decay curves. As is shown in Fig. 5, each TRPL curve can be well fitted with the dashed curve which is determined by the summation of the two exponential functions. According to Eqs. (3) and (4), the values of τnr and τr are calculated to be 72.8 ns and 8.7 ns for sample A, and 66.6 ns and 1.6 ns for sample B, respectively. 7
ACCEPTED MANUSCRIPT The carrier lifetime of sample A and sample B can then be obtained as7.8 ns and 1.6 ns based on Eq. (2), respectively. Compared to sample B, longer carrier lifetime suggests that the photon-generated carriers of sample A will have more time to escape from MQWs before being recombined, and hence leading to higher conversion efficiency.
Fig. 5. TRPL decay curves and the fitted curves of sample A and sample B. In order to study the tendency of whether the photogenerated carriers escape from MQWs to the electrodes or recombine in MQWs, external quantum efficiency (EQE) are measured by Enlitech QE-R solar cell analysis system. As is shown in Fig. 6, samples A and B show zero efficiency over 450 nm and 470 nm, respectively. The redshift of the photoresponse is due to the higher indium composition in sample B. As is known, sample with higher indium composition can absorb more light and may have a higher Jsc. However, under the wavelength of 410 nm, the EQE of sample B is much lower than that of sample A, and the peak EQE of sample B is only 32% at 400 nm compared to that of 61% at 380 nm for sample A. It can be deduced that compared with sample B, the photogenerated carriers in sample A were apt to escape from MQWs, resulting from the higher carrier lifetime in sample A.
8
ACCEPTED MANUSCRIPT
Fig. 6. EQE of samples A and B measured from 300 nm to 480 nm. Energy band diagram can provide guidance for understanding the tendency of carrier transportation. Fig. 7 shows the energy band and electric field simulations of single QW in samples A and B. The simulated models and material parameters are described by references [35-37]. It can be observed that the energy levels of electrons and holes in QW of sample B are closer than that of sample A, and the reverse electric field of QW in sample B is higher than that in sample A. Therefore, the recombination of photogenerated carriers in sample A will be reduced due to the lower indium composition and well thickness, implying a better performance.
Fig. 7. Simulated energy band diagrams and electric fields of single QW in samples A and B. Figure 8(a) displays the dark current-voltage (I-V) curves of sample A and sample B which are measured using Agilent B2902A source/measure unit. The leakage 9
ACCEPTED MANUSCRIPT current of sample A is visibly lower than that of sample B under a reverse bias voltage of -5 V. While the current density-voltage (J-V) characteristics of the two samples are measured under AM 1.5G illumination in Fig. 8(b). Compared with sample B, the values of open-circuit voltage (Voc) and short-circuit current density (Jsc) for sample A are much enhanced due to the lower leakage current and longer carrier lifetime. Finally, the summary of the measured solar cell parameters and the InGaN/GaN MQW structures are shown in Table 1.
Fig. 8. (a) The dark I-V curves of sample A and sample B. (b) The J-V characteristics of sample A and sample B under AM1.5G illumination. Table 1 Summary of InGaN/GaN MQW structures and solar cell parameters of the samples.
Sample
Indium composition
InGaN well
GaN barrier
thickness
thickness
(nm)
(nm)
Voc
Jsc
FF
(V)
(mA/cm2)
(%)
η (%)
A
0.14
1.72
4.14
2.13
2.55
65.7
3.56
B
0.19
2.76
4.14
1.16
1.82
49.0
1.04
4. Summary In summary, we fabricated an InGaN/GaN MQW solar cell with new structure and studied the characteristics in comparison with the general structure. It can be seen that although the indium composition and quantum well thickness of sample A are lower than that of sample B, which indicates the structure of sample A has no advantage in absorbing more light compared with sample B, a much enhanced Jsc and 10
ACCEPTED MANUSCRIPT Voc of sample A are however obtained. Confirmed by HRTEM, HRXRD and RSM measurements, the crystalline quality of sample A is surely better than sample B, which results in a decrease of the leakage current and a higher Voc for sample A. In addition, as measured by TRPL, the carrier lifetime of sample A (7.8 ns) is much longer than that of sample B (1.6 ns), which can assist the carrier transportation and enhance Jsc. Moreover, the peak EQE of sample A is 61%, which is larger than the 32% of sample B. Consequently, sample A has a better performance with conversion efficiency of 3.56%, being three times better than sample B.
Acknowledgments This work was supported by the Beijing Municipal Science and Technology Project [grant number Z161100000416002]; and the State Key Development Program for Basic Research of China [grant number 2012CB619303].
11
ACCEPTED MANUSCRIPT References [1] S.J. Pearton, F. Ren, GaN Electronics, Adv. Mater. 12 (2000) 1571-1580. [2] H.X. Jiang, S.X. Jin, J. Li, J. Shakya, J. Y. Lin, III-nitride blue microdisplays, Appl. Phys. Lett. 78 (2001) 1303-1305. [3] T.C. Lu, C.C. Kao, H.C. Kuo, G.S. Huang, S.C. Wang, CW lasing of current injection blue GaN-based vertical cavity surface emitting laser, Appl. Phys. Lett. 92 (2008) 141102. [4] S. Nakamura, The roles of structural imperfections in InGaN-based blue lightemitting diodes and laser diodes, Science 281(1998) 956-961. [5] J. Wu, W. Walukiewicz, K.M. Yu, W. Shan, J.W. Ager III, E.E. Haller, Hai Lu, William J. Schaff, W.K. Metzger, Sarah Kurtz, Superior radiation resistance of In1−xGaxN alloys: Full-solar-spectrum photovoltaic material system, J. Appl. Phys. 94 (2003) 6477-6482. [6] H.L. Xiao, X.L. Wang, J.X. Wang, N.H. Zhang, H.X. Liu, Y.P. Zeng, J.M. Jin, Z.G. Wang, Growth and characterization of InN on sapphire substrate by RFMBE, J. Cryst. Growth 276 (2005) 401-406. [7] V.Y. Davydov, A.A. Klochikhin, V.V. Emtsev, S.V. Ivanov, V.V. Vekshin, F. Bechstedt, J. Furthmuller, H. Harima, A.V. Mudryi, A. Hashimoto, A. Yamamoto, J. Aderhold, J. Graul, E.E. Haller, Band Gap of InN and In-Rich InxGa1-xN alloys (0.36 < x < 1), Phys. Status Solidi B 230 (2002) R4-R6. [8] A. Luque, A. Martí, Increasing the efficiency of ideal solar cells by photon induced transitions at intermediate levels, Phys. Rev. Lett. 78 (1997) 5014-5017. [9] B. Wen, J.J. Zhou, R.L. Jiang, Z.L. Xie, D.J. Chen, X.L. Ji, P. Han, R. Zhang, Y.D. Zheng, Theoretical calculation of conversion efficiency of InGaN solar cells, J. Semicond. 28 (2007) 1392-1395. [10]X.B. Zhang, X.L. Wang, H.L. Xiao, C.B. Yang, J.X. Ran, C.M. Wang, Q.F. Hou, J.M. Li, Z.G. Wang, Theoretical design and performance of InxGa1−xN twojunction solar cells, J. Phys. D: Appl. Phys. 41 (2008) 245106. [11]T. Nozawa, Y. Arakawa, Detailed balance limit of the efficiency of multilevel intermediate band solar cells, Appl. Phys. Lett. 98 (2011) 171108. 12
ACCEPTED MANUSCRIPT [12]Q.W. Deng, X.L. Wang, H.L. Xiao, C.M. Wang, H.B. Yin, H. Chen, Q.F. Hou, D.F. Lin, J.M. Li, Z.G. Wang, X. Hou, An investigation on InxGa1−xN/GaN multiple quantum well solar cells, J. Phys. D: Appl. Phys. 44 (2011) 265103. [13]A. Adaine, S.O. Hamady, N. Fressengeas, Simulation study of a new InGaN player free Schottky based solar cell, Superlattices Microstruct. 96 (2016) 121-133. [14]J.F. Muth, J.H. Lee, I.K. Shmagin, R.M. Kolbas, H.C. Casey, B.P. Keller, U.K. Mishra, S.P. DenBaars, Absorption coefficient, energy gap, exciton binding energy, and recombination lifetime of GaN obtained from transmission measurements, Appl. Phys. Lett. 71 (1997) 2572-2574. [15]J. Wu, When group-III nitrides go infrared: New properties and perspectives, J. Appl. Phys. 106 (2009) 011101. [16]C.B. Yang, X.L. Wang, H.L. Xiao, J.X. Ran, C.M. Wang, G.X. Hu, X.H. Wang, X.B. Zhang, J.P. Li, J.M. Li, Photovoltaic effects in InGaN structures with p–n junctions, Phys. Status Solidi A 204 (2007) 4288-4291. [17]J.N. Carl, G.T. Nikholas, C.C. Samantha, I. Michael, P.D. Steven, K.M. Umesh, High quantum efficiency InGaN/GaN solar cells with 2.95 eV band gap, Appl. Phys. Lett. 93 (2008) 143502. [18]H.W. Wang, H.C. Chen, Y.A. Chang, C.C. Lin, H.W. Han, M.A. Tsai, H.C. Kuo, Conversion efficiency enhancement of GaN/InGaN solar cells with nano patterned sapphire and biomimetic surface antireflection process, Photonic Tech. L. 23 (2011) 1304-1306. [19]L. Jing, H.L. Xiao, X.L. Wang, C.M. Wang, Q.W. Deng, Z.D. Li, J.Q. Ding, Z.G. Wang, X. Hou, Enhanced performance of InGaN/GaN multiple quantum well solar cells with patterned sapphire substrate, J. Semicond. 34 (2013) 124004. [20]A. Wakahara, T. Tokuda, X.Z. Dang, S. Noda, A. Sasaki, Compositional inhomogeneity and immiscibility of a GaInN ternary alloy, Appl. Phys. Lett. 71 (1997) 906-908. [21]I.H. Kim, H.S. Park, Y.J. Park, T.K. Dahal, Formation of V-shaped pits in InGaN/GaN multiquantum wells and bulk InGaN films, Appl. Phys. Lett. 73 (1998) 1634-1636. 13
ACCEPTED MANUSCRIPT [22]R. Dahal, J. Li, K. Aryal, J. Y. Lin, H. X. Jiang, InGaN/GaN multiple quantum well concentrator solar cells, Appl. Phys. Lett. 97 (2010) 073115. [23]J.H. Song, J.H. Oh, J.P. Shim, J.H. Min, D.S. Lee, T.Y. Seong, Improved efficiency of InGaN/GaN-based multiple quantum well solar cells by reducing contact resistance, Superlattices Microstruct. 52 (2012) 299-305. [24]N.G. Young, E.E. Perl, R.M. Farrell, M. Iza, S. Keller, J.E. Bowers, S. Nakamura, S.P. DenBaars, J.S. Speck, High-performance broadband optical coatings on InGaN/GaN solar cells for multijunction device integration, Appl. Phys. Lett. 104 (2014) 163902. [25]S. Valdueza-Felip, A. Mukhtarova, L. Grenet, C. Bougerol, C. Durand, J. Eymery, E. Monroy, Improved conversion efficiency of as-grown InGaN/GaN quantum-well solar cells for hybrid integration, Appl. Phys. Express 7 (2014) 032301. [26]S.M. Liu, H. L. Xiao, Q. Wang, J. D. Yan, X. M. Zhan, J. M. Gong, X. L. Wang, Z. G. Wang, InxGa1− xN/GaN multiple quantum well solar cells with conversion efficiency of 3.77%, Chin. Phys. Lett. 32 (2015) 088401. [27]A. Mukhtarova, S. Valdueza-Felip, L. Redaelli, C. Durand, C. Bougerol, E. Monroy, J. Eymery, Dependence of the photovoltaic performance of pseudomorphic InGaN/GaN multiple-quantum-well solar cells on the active region thickness, Appl. Phys. Lett. 108 (2016) 161907. [28]Y.J. Lee, M.H. Lee, C.M. Cheng, C.H. Yang, Enhanced conversion efficiency of InGaN multiple quantum well solar cells grown on a patterned sapphire substrate, Appl. Phys. Lett. 98 (2011) 263504. [29]S.B. Choi, J.P. Shim, D.M. Kim, H.L. Jeong, Y.D. Jho, Y.H. Song, D.S. Lee, Effect of indium composition on carrier escape in InGaN/GaN multiple quantum well solar cells, Appl. Phys. Lett. 103 (2013) 033901. [30]L. Redaelli, A. Mukhtarova, S. Valdueza-Felip, A. Ajay, C. Bougerol, C. Himwas, J. Faure-Vincent, C. Durand, J. Eymery, E. Monroy, Effect of the quantum well thickness on the performance of InGaN photovoltaic cells, Appl. Phys. Lett. 105 (2014) 131105. 14
ACCEPTED MANUSCRIPT [31]J. J. Wierer, Jr., D. D. Koleske, S. R. Lee, Influence of barrier thickness on the performance of InGaN/GaN multiple quantum well solar cells, Appl. Phys. Lett. 100 (2012) 111119. [32]N. G. Young, R. M. Farrell, Y. L. Hu, Y. Terao, M. Iza, S. Keller, S. P. DenBaars, S. Nakamura, J. S. Speck, High performance thin quantum barrier InGaN/GaN solar cells on sapphire and bulk (0001) GaN substrates, Appl. Phys. Lett. 103 (2013) 173903. [33]M.A. Moram, M.E. Vickers, X-ray diffraction of III-nitrides, Rep. Prog. Phys. 72 (2009) 036502. [34]H. Kim, D.S. Shin, H.Y. Ryu, J.I. Shim, Analysis of Time-resolved photoluminescence of InGaN quantum wells using the carrier rate equation, Jpn. J. Appl. Phys. 49 (2010) 112402. [35]T. Tao, T. Zhi, B. Liu, M.x. Li, Z. Zhuang, J.P. Dai, Y. Li, F.L. Jiang, W.j. Luo, Z.L. Xie, D.J. Chen, P. Chen, Z.S. Li, Z.G. Zou, R. Zhang, and Y.D. Zheng, Significant improvements in InGaN/GaN nano-photoelectrodes for hydrogen generation by structure and polarization optimization, Sci. Rep. 6 (2016) 20218. [36]G.F. Brown, J.W. Ager III, W. Walukiewicz and J. Wu, “Finite element simulations of compositionally graded InGaN solar cells,” Sol. Energy Mater. Sol. Cells 94 (2010) 478. [37]I. Vurgaftman and J. R. Meyer, “Band parameters for nitrogen-containing semiconductors”, J. Appl. Phys. 94 (2007) 073705.
15