Open Access
Add to BibSonomy
Abstract
The effect of rapid thermal annealing (RTA) on the optical properties of InGaAsP with band-gap energy of around 1.05 eV for quadruple-junction solar cells grown by molecular beam epitaxy (MBE) has been investigated. The photoluminescence (PL) spectrum of InGaAsP film annealed at 800 °C has strong integrated intensity and low activation energy of band-tail states. The time-resolved PL measurement shows that the decay time of the InGaAsP annealed at 800 °C and as-grown one are 11.6 ns and 3.0 ns at 10 K, respectively. An S-shape PL decay time as a function of temperature for the InGaAsP annealed at 800 °C is observed and is explained by the carrier relaxation dynamics. The RTA process induces reorganization of In and Ga inside the alloy due to the existence of miscibility gap in InGaAsP grown by MBE owing to the Be diffusion at high temperature and results in an increased composition uniformity and an improved PL intensity.
© 2017 Optical Society of America
1. Introduction
The multi-junction solar cell (MJSC) has been one of the most promising options to photoelectric conversion efficiency [1–3]. Especially, III–V MJSC is the most efficient photovoltaic device due to superior properties in terms of material quality, stability, miscibility, and tunability of bandgap. Triple-junction InGaP/GaAs/InGaAs concentrator cell has reached an efficiency of 44.4% at 302 suns [4]. The efficiency also can be improved by adding additional junctions. The best bandgap matching and corresponding efficiency limit about multi-junction cell was estimated through detailed balance principle [5,6]. The solar cell with 4–6 junctions shows enormous potential in terms of higher efficiency. And semiconductor material with bandgap of near 1 eV is required in all such designs. The theoretical optimum for quadruple-junction solar cell under AM 1.5D spectral conditions is reached for a bandgap combination of 1.9/1.4/1.0/0.5 eV [7]. Compared with ternary compound semiconductor material such as InGaAs or InGaP, the InGaAsP quaternary alloy, lattice-matched to InP substrates, has been identified as a promising candidate for multi-junction solar cell [8] due to its wide range of band gap energies (0.75 eV~1.35 eV) [9] and the independent regulation of the band gap energy with the lattice constant. GaInP/GaAs/GaInAsP/GaInAs quadruple junction solar cell has been reported the highest efficiency of 46.0% at 508 suns [4].
Many successful cases of InGaAsP-based solar cell have been grown by metal-organic chemical vapor deposition(MOCVD) [7,10]. Molecular beam epitaxy (MBE) has the potential to grow high quality InGaAsP-based materials-due to its ultra-high vacuum environment and ultra-high purity metal sources. However, the growth of high quality InGaAsP with 1.0 eV band gap energy on InP substrate by MBE has been a challenge in the photovoltaic cell field [11–13]. We have achieved the growth of high quality InGaAsP with 1.0eV and InGaAsP-based single cell with the efficiency of 18.8% [14]. Considering the low growth temperature of MBE epitaxy, high temperature annealing might contribute to the further improvement of material quality. However, effect of annealing temperature on optical properties of the MBE-grown InGaAsP has been rarely reported.
In this work, Be-doped InGaAsP samples with bandgap energy of around 1.05 eV at room temperature were grown by MBE and then the rapid thermal annealing (RTA) was performed at 700 °Cand 800 °C, respectively. The continuous-wave photoluminescence (CWPL) and time-resolved photoluminescence (TRPL) spectra measurements were employed to study the optical properties. The InGaAsP film annealed at 800 °C has strong integrated intensity and low activation energy of band-tail states. The time-resolved PL measurement shows that the decay time of the InGaAsP annealed at 800 °C and as-grown one are 11.6 ns and 3.0 ns at 10 K, respectively. An S-shape PL decay time as a function of temperature for the InGaAsP annealed at 800 °C is observed and is explained by the carrier relaxation dynamics. The RTA process induces reorganization of In and Ga inside the alloy due to the existence of miscibility gap in InGaAsP grown by MBE owing to the Be diffusion at high temperature and results in an increased composition uniformity and an improved PL intensity.
2. Experimental details
2.1 Grown and RTA of InGaAsP material
The InGaAsP material under investigation was grown using all solid-state MBE equipped with a valved phosphorous (P) cracking cell and a valved arsenic (As) cracking cell. A 1000 nm InGaAsP layer with Be-doped concentration of 2 × 1017 cm−3 was grown on InP substrate at 489 °C, and the growth rate was about 1 μm/h. During the process of InGaAsP growth, the reflection high energy electron diffraction (RHEED) image real-time monitored surface reconstruction. The beam equivalent pressure ratio of As and P was calibrated by a combination of beam flux gauge and X-ray diffraction (XRD) measurements for a lattice-matched InGaAsP. The RTA measurement was carried out in a flowing N2 gas ambient in a RTP-500 thermal processor system at the temperature of 700 °C and 800 °C for 1 s, respectively. As-grown sample and RTA samples were denoted as Sample A (as-grown), Sample B (700 °C) and Sample C (800 °C) .
2.2 Measurements by CWPL and TRPL
CWPL was excited by tunable Ti-sapphire lasers with an excitation wavelength of 773 nm. TRPL decay curves were measured using an 800 nm laser line from a Ti-sapphire laser with a pulse width of 100 fs and a petition frequency of 80 MHz, and detected with a Hamamatsu C4334-04 synchroscan streak camera, with a time resolution of 15 ps.
3. Results and discussion
Figure 1 shows the PL spectra at 10 K of sample A, B and C with an excitation power of 5 mW, which is represented by the dash line and solid line, respectively. As shown in the figure, the PL peak energy is around 1.1 eV. While the PL peak energy of sample C is about 10 meV higher than those of sample A and sample B, which is indicated that a blue shift of PL energy occurred after RTA at 800 °C. The same behavior was also observed in other semiconductor materials [15]. Furthermore, the integrated intensity of the PL peaks of sample C is approximately 15 times stronger than that of sample A, while sample B has a little promotion.
The enhanced PL intensity in the case of RTA is attributed to the removal of deep non-radiative traps [16], which indicates that influence of non-radiative recombination inside of InGaAsP material has been greatly reduced under 800 °C annealing.
In order to further explain the change of optical performances in InGaAsP material, temperature-dependent PL measurements were performed at the excitation power of 5 mW, as shown in Fig. 2(a) of sample C. With increasing temperature, the PL peak energy shows a clear red-shift towards the lower energy direction and the corresponding full width at half maximum (FWHM) increases. For comparison, Fig. 2(b) shows the PL energy variation as a function of temperature of the three samples. There is a deviation caused by Stokes shift at low temperature below 77 K. The deviations of sample A, B and C are in descending order, and it is very close between sample A and sample B. The small Stokes shift in sample C indicates its good material quality. The Varshni empirical relationship [17] calculation was performed by using the parameters of E0 (1.1eV) and (eV/K) of 4.657 × 10−4, 4.150 × 10−4 and 2.977 × 10−4, (K) of 368.69, 312.97 and 240.61 for samples A, B and C, respectively. The fitted value of sample C is very similar to the published data of InGaAsP.
Fig. 2 (a) Temperature-dependent PL spectra of sample C and (b) PL peak energies as function of temperature of the three samples.
Furthermore, it is important to note that the PL spectrum of sample A has almost disappeared above 200 K. The occurrence of PL quenching at higher temperature is attributed to an enhancement of non-radiative recombination which suppresses the emission of InGaAsP material [18]. Figure 3 presents the variation of PL intensity as a function of temperature for sample A, sample B and sample C. Activation energy of the band-tail states in InGaAsP material could be determined by the variable-temperature PL spectral. The temperature dependence of the integrated intensity of the PL spectral is usually described by a phenomenological expression [19]:
Fig. 3 Integrated PL intensities as a function of reciprocal temperature for InGaAsP (a) Sample A, (b) Sample B and (c) Sample C.
Where is the PL integrated intensity at 0 K, C is the ratio between non- radiative recombination and the radiative recombination, is the activation energy of band-tail states and is the Boltzmann constant.
When the temperature is high enough (T>100K),, the formula could be simplified into the following form:
Simplification and logarithmic processing on both sides, finally the formula could be expressed as:
From Eq. (4), the activation energies of band tail states in sample A and C as shown in Figs. 3(a), 3(b) and 3(c) are 33.3 meV, 32.6 meV and 26.5 meV, respectively. Obviously, the activation energy of sample C is much smaller than that of sample A and sample B, while sample A and sample B have similar values. The presence of impurities and defects in InGaAsP causes a smearing of band edges and formation of tails in the density of states extending into the band gap [20]. Photo-created carriers can be trapped by these localized states at the band tails at low temperature, which would result in deterioration of the optical properties of materials. The above experimental results show that high temperature annealing is in a position to suppress the band-tail effect. This leads to a decrease in the activation energy of band-tail states for InGaAsP after 800 °C annealing. Obviously, 700 °C is too low to improve the intrinsic properties of the InGaAsP materials.
Time-resolved PL measurements were made to measure the PL decay time which is a quite important performance index for the quality of semiconductor materials and devices. The decay curves of InGaAsP samples at 10K are shown in Fig. 4.
The PL decay time of sample C after 800 °C annealing has longer decay time of 11.6 ns, whereas the PL decay times of sample A and sample B are 3.0 ns and 3.3 ns at 10 K, respectively. This behavior shows that there exit non-radiative recombination centers in as-grown InGaAsP material, and the concentration of these non-radiative recombination centers decreases significantly after RTA at 800 °C due to the removal of all kinds of crystal defects. The reduction of non-radiative recombination leads to the increase of the PL decay time. However, RTA at 700 °C has slightly effect to the sample of InGaAsP material on PL behavior.
Figure 5(a) shows the PL decay curves for 800 °C-annealed InGaAsP sample at different temperatures. Figure 5(b) shows the PL decay time as a function of measurement temperature. The decay times remain at a high level of 10.4 ns at room temperature, which demonstrates that radiative recombination plays a dominant role in the recombination process and the quality of InGaAsP material has been improved significantly after 800 °C annealing.
Fig. 5 (a) PL decay curves of sample C at different temperature from 50 K to 300 K;(b) PL decay time as a function of temperature about sample C.
The decay time shows an S-shape variation with increasing temperature: it increases at temperatures lower than 50 K, and then decreases between 50 K~150 K, and increases again when temperature is over 150 K. The phenomenon would be explained by the carrier relaxation dynamics. When injecting non-equilibrium carrier, the decay times of semiconductor could be expressed as:
Where is concentration of non-radiative recombination center, and are the capture coefficient of minority carriers, and are equilibrium carrier concentration, respectively. For p-typed semiconductor, ,, and is related to as:
The concentration of carrier at low injection, so it can be ignored. and is about the same in general recombination. We can simplify Eq. (5) to:
Fermi level of P type semiconductor could be expressed as:
Finally the relationship of decay time , recombination center concentration and temperature T was further simplified:
Where a and b are two positive constants. At low temperature below 50 K, the concentration of non-radiative recombination center remains constant because the carriers are most localized, so decay time increases with the increase of temperature. However, in the temperature range of 50 K ~150 K, more non-radiative recombination centers thermally activated and accelerate carrier recombination. The decay time tends to decrease with the increase of temperature. In the case of higher temperature above 150 K, all non-radiative recombination centers has been activated, and then temperature becomes the single factor in decay time . A good agreement of experiment and theory indicates that RTA process effectively improve the material quality.
The high quality InGaAsP material with a 1.05eV bandgap was difficult to grow due to the existence of miscibility gap in InGaAsP grown by MBE. The miscibility gap has been known in a wide composition range of InGaAsP materials grown by liquid phase epitaxy (LPE) [21,22]. In this compositional range, the solid solution is energetically unstable and InGaAsP crystals show no uniform growth and decompose into spatial regions that are InAs-or GaP-rich [23]. The increase in PL linewidth is widely used as an index of phase separation in miscibility. Figure 6 shows the PL linewidth as a function of PL photon energy. The high PL linewidth increases as the composition approaches the 1.0-1.05 eV may be related to the immiscibility of InGaAsP. The phase separation causes the degradation of the optical characteristics of InGaAsP epitaxial layers and results in poor performance of solar cells. By employing RTA treatment under a proper condition, it can be effectively remove the non-radiative defects and therefore improve the crystal quality. The enhancement of PL intensity and the increase of PL decay time for annealed sample may be attributed to the conversion of Be from interstitial to substitutional atoms and reorganization of In and Ga inside InGaAsP alloy during 800 °C annealing process. Because of Be diffusion and more composition uniformity, the defects decrease and higher quality of InGaAsP material can be obtained. And annealing at 700 °C, this phenomena hardly occurred in the interior of the InGaAsP material due to the low annealing temperature. RTA consumes neutral vacancy and generates self-interstitial of III group through the diffusion of Be atoms, and then converts most of interstitial atoms into substitutional atoms. Meanwhile, Be diffusion during annealing process has caused the redistribution of III atoms and the uniformity of material composition. And RTA can effectively reduce the concentration of non-radiative recombination centers such as crystal defect and impurity and improve the quality of InGaAsP [24,25].
4. Conclusions
The effect of high annealing temperatures on optical properties of InGaAsP material with band-gap energy of 1.05 eV grown by MBE has been investigated. The InGaAsP sample after annealing at 800 °C has a markedly stronger PL integrated intensity and a longer PL decay time than the sample before annealing. These behaviors indicate that rapid thermal annealing can effectively improve the quality of InGaAsP and enhance the optical properties, by reducing the non-radiative recombination centers attributed to the conversion of Be from interstitial to substitutional atoms and reorganization of In and Ga inside InGAsP alloy during 800 °C-annealing process. TRPL spectra suggest that the decay time is 10.4 ns at room temperature. The low annealing temperature was difficult to drive rearrangement of atoms and improve the quality of InGaAsP. It is extremely important to study the optimal annealing process in the future for the existence of miscibility gap in InGaAsP around 1.05 eV.
Funding
National Natural Science Foundation of China (Grant Nos. 61704186, 61774165, 61534008, and 61376081); Key Frontier Scientific Research Program of the Chinese Academy of Sciences (Grant No. QYZDB-SSW-JSC014); Science and Technology Service Network Initiative of the Chinese Academy of Sciences, Key R&D Program of Jiangsu Province (Grant No. BE2016085); External Cooperation Program of BIC, Chinese Academy of Sciences (Grant No. 121E32KYSB20160071).
Acknowledgments
We are thankful for the technical support from Nano Fabrication Facility, Platform for Characterization & Test of SINANO, CAS.
References and links
1. D. J. Friedman, S. R. Kurtz, K. A. Bertness, A. E. Kibbler, C. Kramer, J. M. Olson, D. L. King, B. R. Hansen, and J. K. Snyder, “30.2% Eficient GaInP/GaAs monolithic two-terminal tandem concentrator cell,” Prog. Photovolt. Res. Appl. 3(1), 47–50 (1995).
2. M. Yamaguchi, “III-V compound multi-junction solar cells: present and future,” Sol. Energy Mater. Sol. Cells 75(1), 261–269 (2003).
3. F. Dimroth, R. Beckert, M. Meusel, U. Schubert, and A. W. Bett, “Metamorphic GayIn1-yP/Ga1-xInxAs tandem solar cells for space and for terrestrial concentrator applications at C>1000 suns,” Prog. Photovolt. Res. Appl. 9(3), 165–178 (2001).
4. M. A. Green, K. Emery, Y. Hishikawa, and W. Warta, “Solar cell efficiency tables (version 49),” Prog. Photovolt. Res. Appl. 25(1), 3–13 (2017).
5. A. Marti and G. L. Araujo, “Limiting efficiencies for photovoltaic energy conversion in multigap systems,” Sol. Energy Mater. Sol. Cells 43(2), 203–222 (1996).
6. W. Shockley and H. Queisser, “Detailed balance limit of efficiency of p-n junction solar cells,” J. Appl. Phys. 32(3), 510–519 (1961).
7. F. Dimroth, M. Grave, P. Beutel, U. Fiedeler, C. Karcher, T. N. D. Tibbits, E. Oliva, G. Siefer, M. Schachtner, A. Wekkeli, A. W. Bett, R. Krause, M. Piccin, N. Blanc, C. Drazek, E. Guiot, B. Ghyselen, T. Salvetat, A. Tauzin, T. Signamarcheix, A. Dobrich, T. Hannappel, and K. Schwarzburg, “Wafer bonded four-junction GaInP/GaAs//GaInAsP/GaInAs concentrator solar cells with 44.7% efficiency,” Prog. Photovolt. Res. Appl. 22(3), 277–282 (2014).
8. B. C. Chung, F. G. Virshup, M. L. Ristow, M. W. Wanlass, and M. Klausmeier, “25.2%-efficiency (1-Sun, air mass 0) AlGaAs/GaAs/InGaAsP three-junction, two-terminal solar cell,” Proc. 22th IEEE Photovoltaic Specialists Conf. (IEEE,New York, 1991) pp. 54–57.
9. P. R. Sharp, M. L. Timmons, R. Venkatasubramanian, R. Pickett, J. S. Hills, J. Hancock, and J. Hutchby, “Development of 20% efficient GaInAsP solar cells,” Proc. 23th IEEE Photovoltaic Specialists Conf. (IEEE, New York, 1993) pp. 633–638.
10. H. J. Schimper, Z. Kollonitsch, K. Moller, U. Seidel, U. Bloeck, K. Schwarzburg, F. Willing, and T. Hannappel, “Material studies regarding InP-based high-efficiency solar cells,” J. Cryst. Growth 287(2), 642–646 (2006).
11. N. Dharmarasu, M. Yamaguchi, A. Khan, T. Yamada, T. Tanabe, S. Takagishi, T. Takamoto, T. Ohshima, H. Itoh, M. Imaizumi, and S. Matsuda, “High-radiation-resistant InGaP, InGaAsP, and InGaAs solar cells for multijuction solar cells,” Appl. Phys. Lett. 79(15), 2399–2401 (2001).
12. J. N. Baillargeon, A. Y. Cho, F. A. Thiel, R. J. Fischer, P. J. Pearah, and K. Y. Cheng, “Reproducibility studies of lattice matched GaInAsP on (100) InP grown by molecular beam epitaxy using solid phosphorus,” Appl. Phys. Lett. 65(2), 207–209 (1994).
13. J. N. Baillargeon, A. Y. Cho, and K. Y. Cheng, “Incorporation of arsenic and phosphorus in GaxIn1−xAsyP1−y alloys grown by molecular beam epitaxy using solid phosphorus and arsenic valved cracking cells,” J. Appl. Phys. 79(10), 7652–7657 (1996).
14. L. Ji, M. Tan, C. Ding, K. Honda, R. Harasawa, Y. Yasue, Y. Y. Wu, P. Dai, A. Tackeuchi, L. F. Bian, S. L. Lu, and H. Yang, “The striking influence of rapid thermal annealing on InGaAsP grown by MBE: material and photovoltaic device,” J. Cryst. Growth 458, 110–114 (2017).
15. S. L. Lu, J. N. Wang, J. S. Huang, L. F. Bian, D. S. Jiang, C. L. Yang, J. M. Dai, W. K. Ge, Y. Q. Wang, J. Y. Zhang, and D. Z. Shen, “The effects of rapid thermal annealing on the optical properties of Zn1−xMnxSe epilayer grown by MOCVD on GaAs substrate,” J. Cryst. Growth 249(3), 538–543 (2003).
16. L. Qian, S. D. Benjamin, P. W. E. Smith, B. J. Robinson, and D. A. Thompson, “Subpicosecond carrier lifetime in beryllium-doped InGaAsP grown by He-plasma-assisted molecular beam epitaxy,” Appl. Phys. Lett. 71(11), 1513–1515 (1997).
17. Y. P. Varshni, “Temperature dependence of the energy gap in semiconductors,” Physica 34(1), 149–154 (1967).
18. C. F. Li, D. Y. Lin, Y. S. Huang, Y. F. Chen, and K. K. Tiong, “Temperature dependence of quantized states in an In0.86Ga0.14As0.3P0.7/InP quantum well heterostructure,” J. Appl. Phys. 81(1), 400–405 (1997).
19. S. L. Hong, “Time-resolved and temperature-dependent photoluminescence studies on CdTe/ZnTe quantum dots with different ZnTe capping layer thicknesses,” Thin Solid Films 547, 272–275 (2013).
20. H. B. Bebb and E. W. Williams, “Photoluminescence I: Theory,” in Semiconductors and Semimetals, R. K. Willardson, A. C. Beer, ed. (Academic, 1972).
21. M. Quillec, C. Daguet, J. L. Benchimol, and H. Launois, “InxGa1-xAsyP1-y alloy stabilization by the InP substrate inside an unstable region in liquid phase epitaxy,” Appl. Phys. Lett. 40(4), 325–326 (1982).
22. A. G. Norman and G. R. Booker, “Transmission electron-microscope and transmission electron-diffraction observations of alloy clustering in liquid-phase epitaxial (001) GaInAsP layers,” J. Appl. Phys. 57(10), 4715–4720 (1985).
23. K. Non and M. Takemi, “Investigation of anomalous optical characteristics of InGaAsP layers on GaAs substrates grown by metalorganic vapor phase epitaxy,” Jpn. J. Appl. Phys. 47(3R), 1484–1490 (2008).
24. S. Koumetz and C. Dubois, “Be diffusion behavior in InGaAs, InGaAsP and InGaAs/InGaAsP GSMBE structures,” J. Cryst. Growth 252(1), 14–18 (2003).
25. L. Ji, S. L. Lu, Y. Y. Wu, P. Dai, L. F. Bian, M. Arimochi, T. Watanabe, N. Asaka, M. Uemura, A. Tackeuchi, S. Uchida, and H. Yang, “Carrier recombination dynamics of MBE grown InGaAsP layers with 1eV bandgap for quadruple-junction solar cells,” Sol. Energy Mater. Sol. Cells 127, 1–5 (2014).
