Open Access

Investigation of GaInAs strain reducing layer combined with InAs quantum dots embedded in Ga(In)As subcell of triple junction GaInP/Ga(In)As/Ge solar cell

  • Senlin Li1, 2, 3,
  • Jingfeng Bi1, 2,
  • Mingyang Li2,
  • Meijia Yang2,
  • Minghui Song1, 2,
  • Guanzhou Liu1, 2,
  • Weiping Xiong2,
  • Yang Li3,
  • Yanyan Fang3,
  • Changqing Chen3,
  • Guijiang Lin1, 2Email author,
  • Wenjun Chen2,
  • Chaoyu Wu2 and
  • Duxiang Wang2
Nanoscale Research Letters201510:111

Received: 12 January 2015

Accepted: 14 February 2015

Published: 7 March 2015


The InAs/GaAs quantum dots structure embedded in GaInP/Ga(In)As/Ge triple junction solar cell with and without Ga0.90In0.10As strain reducing layer was investigated. Conversion efficiency of 33.91% at 1,000 suns AM 1.5D with Ga0.90In0.10As strain reducing layer was demonstrated. A 1.19% improvement of the conversion efficiency was obtained via inserting the Ga0.90In0.10As strain reducing layer. The main contribution of this improvement was from the increase of the short-circuit current, which is caused by the reduction of the Shockley–Read–Hall recombination centers. Consequently, there was a decrease in open circuit voltage due to the lower thermal activation energy of confined carriers in Ga0.9In0.1As than GaAs and a reduction in the effective band gap of quantum dots.


InAs quantum dotsTriple junction solar cellStrain reducing layerMetal organic chemical vapor deposition


Today’s state-of-the-art conversion efficiency of matured lattice-matched GaInP/Ga(In)As/Ge solar cell is 41.6% ± 2.5% at 364 suns concentration and 32.0% ± 1.5% at one sun [1], which have been widely used for terrestrial photovoltaic [2] and space applications [3]. Experimentally, the outdoor module performance had strong relations with the Ga(In)As subcell in CPV power plants. In this case, for further efficiency enhancement, quantum structures such as quantum well [4] and quantum dots (QDs) [5-14] are usually introduced to extend the absorption spectrum to lower energy photons, thus increasing the current density. Additionally, the QDs enhanced devices may improve the radiation tolerance for space application, temperature coefficients, and spectral response under concentration [14-16]. However, the incorporation of quantum structures is usually accompanied by a significant decrease in open-circuit voltage (Voc). For these reasons, numerous efforts have been made inducing a high-potential barriers fence [5-7] and strain compensation technique [9-11]. As reported, the Voc could recover via the decrease in trapping and recombination of carriers in the QDs and wetting layer, which is due to resonant tunneling through a new electronic state created by the fence layer. For the strain compensation technique, the Voc might maintain by the decrease of Shockley-Read-Hall (SRH) recombination that benefit from the less defects. Almost all these mentioned above quantum dots solar cell are based on a GaAs substrate. When InAs/GaAs QD structure is grown on a Ge substrate, GaAs become the natural compensation materials for its lattice constant between InAs and Ge. Nevertheless, the larger lattice mismatch between InAs QDs and GaAs space layer usually induces defects. These defects become the trap center for photon generated carries, thus deteriorating the performance of device. In this paper, a Ga0.90In0.10As strain reducing layer (SRL) was applied to reduce these defects in the QD solar cell. The effects of Ga0.90In0.10As SRL on InAs QDs embedded in GaInP/Ga(In)As/Ge triple junction solar cell (TJSC) were investigated.


The QD TJSC was fabricated on a 4 in. Ge (001) substrate 9° offcut toward <111> direction via metal organic chemical vapor deposition (MOCVD). As shown in Figure 1, the Ge subcell and p-i-n Ga(In)As subcell are connected by the n++GaAs/p++GaAs tunnel junction (TJ), and the p-i-n Ga(In)As subcell and GaInP subcell are connected by the n++GaAs/p++AlGaAs TJ. A five-layer InAs QDs with and without Ga0.90In0.10As SRL embedded in the i-region of p-i-n Ga(In)As subcell was studied.
Figure 1

The QD TJSC image. (a) The structure of QD TJSC, QD structure of (b) sample A1 with Ga0.90In0.10As SRL, (c) sample A2 without Ga0.90In0.10As SRL.

Prior to the InAs QD structure growth, Ge subcell, n++GaAs/p++GaAs TJ, AlGaAs back surface field, and Ga(In)As base were formed on the Ge substrate, and then a 15 nm intrinsic GaAs was deposited at 620°C. Thereafter, the substrate temperature was lowed 510°C for the InAs QDs growth via Stranski-Krastanov growth mode with a coverage of 1.8 monolayer, followed by a 3 nm Ga0.90In0.10As SRL, an 1 nm CaP strain compensation layer, and an 11 nm GaAs layer, and then the substrate temperature was ramped to 620°C for a high temperature GaAs growth with a thickness of 25 nm. By repeating this sequence, a five periods of InAs QDs were fabricated in the intrinsic region. For the surface and strain analysis, the fifth layer InAs QDs was uncapped, denoted as sample A1. For a comparison, the 3 nm Ga0.90In0.10As SRL in sample A1 was replaced by a 3 nm GaAs layer with all the other growth parameters remained the same, denoted as sample A2. Then, the Ga(In)As emitter and AlGaAs window were deposited, followed by a n++GaAs/p++AlGaAs TJ, GaInP subcell, and GaAs contact layer, completing the QD TJSC growth. The QD TJSC that based on sample A1 is denoted as sample B1, and that based on sample A2 is denoted as sample B2.

Post MOCVD growth, the structure was processed into chips with size of 1 × 1 cm2. The front grids were patterned by photolithography and chemical etch with a shadow ratio of 5%, and then the contact metal was deposited by thermal evaporation. Subsequently, an Al2O3/Ti3O5 dual layer anti-reflection coating (ARC) was evaporated on the front surface of the cells forming the front-side passivation of the top cell to improve the spectral response. All the solar cells were isolated by a chemical etch.

The InAs QD densities and morphologies of sample A1 and A2 were characterized by atomic force microscopy (AFM). The optical performance was measured by a room temperature (RT) photoluminescence (PL). The structural characterization like the strain condition was analyzed by high resolution X-ray diffraction (HRXRD). The external quantum efficiency (EQE) of QD TJSC was measured with a commercial Enlitech system. The current-voltage (I-V) characteristics were measured using a solar simulator at RT under 1,000 suns, low-AOD, AM 1.5D conditions.

Results and discussion

As illustrated in Figure 2, the Ga(In)As and Ge are lattice matched which were confirmed by the single Bragg peak. In both HRXRD spectra, the five period InAs/GaAs QD superlattice (SL) fringe peaks are visible. The zeroth-order peak of sample A1 is buried underneath the Ga(In)As and Ge Bragg peak, but a shift of 140 arc-second is observed in sample A2, meaning that the five period InAs/GaAs QD SL is under a tensile strain from the substrate. The out-of-plane strain ε is 1.045‰, which is determined from a differentiated formulation of Bragg’s law shown as followed [17]:
Figure 2

HRXRD of InAs/GaAs QD structure. (a) Sample A1 with Ga0.90In0.10As SRL, (b) sample A2 without Ga0.90In0.10As SRL.

$$ {\varepsilon}_{\perp }=\varDelta a/a=\varDelta {\theta}_{sl} cot{\theta}_b $$
Where Δθsl is the difference in angle between the substrate Bragg peak and the zeroth order SL peak, θb is the value of the substrate Bragg angle, and Δa/a is the fractional lattice mismatch representing the out-of-plane strain of the SL. Corresponding the in-plane strain ε // (1.136‰) of InAs QDs can be estimated by the equation [18]
$$ {\varepsilon}_{\perp }=-2\frac{C_{12}}{C_{11}}{\varepsilon}_{//} $$

These results suggested that more defects may be formed in the QD TJSC without Ga0.90In0.10As SRL.

Figures 3a and b show the 2 × 2 μm AFM images of QDs with and without Ga0.90In0.10As SRL. The corresponding room temperature PL spectra are plotted in Figure 4. By inserting an Ga0.90In0.10As SRL, the InAs QD density of sample A1 is increased from 8.1 × 109 cm−2 (sample A2) to 1.2 × 1010 cm−2 (sample A1), which is consistent with the PL intensity. As is well known, the strain fields created by the QDs in the first layer strongly influence the QD nucleation on the second layer [19]. The increasing of the QD density of sample A1 with Ga0.90In0.10As SRL can be ascribed to the reduction of the accumulated strain field, which is assumed by the strain as interpreted by the XRD results. Correspondingly, the PL intensity of QDs is improved by the inserting of Ga0.90In0.10As SRL. For one reason, the QD density is increased. Secondly, less defects may be existed in sample A1 due to the decrease of lattice mismatch, leading to less SRH recombination centers [20]. These two reasons induced a better optical performance of sample A1. In addition, the PL peak of sample A1 exhibits a red shift of about 3 meV with respect to samples A2. The following reasons may need to be considered: Firstly, the relief of the compressive strain conducted a 1.94 meV red shift to the band energy, which is calculated by the following equation [21],
Figure 3

The 2 × 2 μm AFM images of the fifth layer InAs QDs. (a) Sample A1 with Ga0.90In0.10As SRL, (b) sample A2 without Ga0.90In0.10As SRL.

Figure 4

PL spectra of InAs/GaAs QDs with and without Ga 0.90 In 0.10 As SRL.

$$ \varDelta {E}_0=1.71\varepsilon (eV) $$

Secondly, the Ga0.90In0.10As SRL could suppress the In/Ga inter-diffusion [22] compared to GaAs, thus conducting a less blue-shift. At last, the less quantum confinement of GaInAs [23] than GaAs will result in a red shift to the QDs.

As shown in Figure 5, the top cell in the QD TJSC with and without a Ga0.90In0.10As SRL yields EQE of 12.65 and 12.55 mA/cm2, respectively. The relative lower EQE of the top cell without Ga0.90In0.10As s SRL may be caused by the deteriorated crystalline quality of GaInP [24], which was affected by the strained InAs/GaAs QD SL in the middle cell. The EQE of the middle cell is 11.16 and 10.51 mA/cm2 in sample B1 and sample B2, respectively, and exhibits absorption edge at 950 nm. This extended absorption range is due to the QDs absorption, whose threshold wavelength is in agreement with the QD ground state transition. The spectral response of the wetting layer and QD ground state transitions is approximately 5% and approximately 1%, respectively. Remarkably, the middle cell with Ga0.90In0.10As SRL yields higher EQE response (about 75%) than that without Ga0.90In0.10As SRL (about 70%) in the spectral region below 880 nm, which can be attributed to the less defects in sample B1 and thus reduced the SRH recombination of photon generated carriers. The EQE of the bottom cell in the QD TJSC with and without Ga0.90In0.10As SRL shows an equal value at 17.2 mA/cm2 and a spectral response of approximately 73%. In comparison with top cell, the Ge bottom cell has a relative lower spectral response. Because the Ge bottom cell is indirect band gap semiconductor that has a short minority carrier diffusion lengths and short minority carrier lifetimes [25]. In a word, the short-circuit current (Isc) of the QD TJSC in this case is limited by the middle cell, suggesting that the effects of Ga0.90In0.10As SRL on current of the middle cell is significant to the final Isc of the QD TJSC.
Figure 5

EQE of QD TJSC with and without Ga 0.90 In 0.10 As SRL.

Figure 6a shows the I-V characteristics of the QD TJSC under 1,000 suns. The values of open-circuit voltage (Voc), fill factor (FF), short-circuit current (Isc), and convention efficiency (η) are listed in Table 1. The Isc of sample B1 is 12.79 A, and a 2.73% higher than that of sample B2. As mentioned above, the crystalline quality is improved by inserting of a Ga0.90In0.10As SRL with reduced the defects. So the improvement of Isc is assumed due to the reduction in SRH recombination that originated from the defects, which is in well agreement with the EQE characterization as shown in Figure 5. To further clarify this speculation, the dark I-V was measured by Keithley system and displayed in Figure 6b. As we can see that the two I-V curves are almost overlapped with each other. The only difference is the hump that appears at the sample B2 with respecting the different epitaxial structure of Ga0.90In0.10As SRL layer. Know to all, the corresponding ideality factor (n) as a function of applied voltage is described as Equation 4:
Figure 6

The I-V measurements and dark I-V curves. (a) I-V measurements at 1,000 suns AM 1.5D of QD TJSC with and without Ga0.90In0.10As SRL, (b) dark I-V curves and the corresponding ideality factor of QD TJSC with and without Ga0.90In0.10As SRL.

Table 1

The open-circuit voltage (V oc ), fill factor (FF), short-circuit current (I sc ), and conversion efficiency (η) of QD TJSC


I sc (A)

V oc (V)



Sample B1





Sample B2





$$ \frac{1}{n}=\frac{k_BT}{q}\frac{d(lnJ)}{d(V)} $$

Where the kB is the Boltzmann constant, q is the elementary charge, and J is the current density. As illustrated in Figure 6b, the ideality factor of the sample without Ga0.90In0.10As SRL is varied from 2 to 16 at a low forward voltage bias (1.2~1.9 V hump area), but the one from sample B1 almost remains constant. The large ideality factor (n > 10) in the sample without Ga0.90In0.10As SRL suggests that more SRH recombination is occurred [26]. Notably, the Voc of sample B1 is 30 mV lower than that of sample B2. The difference in Voc between samples with (B1) and without (B2) Ga0.90In0.10As SRL is larger than that in InAs ground state energy observed in PL. So the fact that the In0.1Ga0.9As SRL has lower band energy relative to GaAs and consequently decreases the thermal activation energy of confined carriers need to be considered in the case. Nevertheless, by inserting Ga0.90In0.10As SRL, conversion efficiency of GaInP/Ga(In)As/Ge TJSC embedded with InAs QDs is improved by 1.19% relatively and arrive at 33.91% under 1,000 suns AM 1.5D.


In summary, the InAs/GaAs QDs structures embedded in GaInP/Ga(In)As/Ge TJSC with and without Ga0.90In0.10As SRL were investigated. The QD density was increased and a red shift of the QD illumination peak was observed by inserting a Ga0.90In0.10As SRL. Also, the strain was reduced by the inserting of Ga0.90In0.10As SRL. Although a reduction in Voc was observed in the sample B1 due to the decrease of thermal activation energy of confined carriers of Ga0.90In0.10As SRL and a reduction in the total energy band gap, the conversion efficiency of QD TJSC is improved by 1.19% relatively via inserting the Ga0.90In0.10As SRL. The cell with Ga0.90In0.10As SRL showed an improved device quality and reduced SRH recombination of photon generated carriers, thus resulted in a large increase of Isc. Finally, the QD TJSC with Ga0.90In0.10As SRL has demonstrated a conversion efficiency of 33.91% at 1,000 suns AM 1.5D.



atomic force microscope


anti-reflection coating




external quantum efficiency




fill factor


high resolution x-ray diffraction


metal organic chemical vapor deposition


quantum dots


room temperature


short-circuit current




strain reducing layer


super lattice


tunnel junction


triple junction solar cell


open-circuit voltage



This work was supported by a grand from the National High Technology Research and Development Program of China (863 programs) (No. 2012AA051402).

Authors’ Affiliations

Xiamen San’an Optoelectronics Co., Ltd, Xiamen, China
Tianjin San’an Optoelectronics Co., Ltd, Tianjin, China
Wuhan National Laboratory for Optoelectronics, School of Optical and Electronic Information, Huazhong University of Science and Technology, Wuhan, China


  1. Green MA, Emery K, Hishikawa Y, Warta W. Solar cell efficiency tables (version 37). Prog Photovolt Res Appl. 2011;19:84–92.View ArticleGoogle Scholar
  2. Bosi M, Pelosi C. The potential of III-V semiconductors as terrestrial photovoltaic devices. Prog Photovolt Res Appl. 2007;15:51–68.View ArticleGoogle Scholar
  3. Takamoto T, Kaneiwa M, Imaizumi M, Yamaguchi M. InGaP/GaAs-based multijunction solar cells. Prog Photovolt Res Appl. 2005;13:495–511.View ArticleGoogle Scholar
  4. Fujii H, Toprasertpong K, Wang Y, Watanabe K, Sugiyama M, Nakano Y. 00-period, 1.23-eV bandgap InGaAs/GaAsP quantum wells for high-efficiency GaAs solar cells: toward current-matched Ge-based tandem cells. Prog Photovolt Res Appl. 2014;22:784–95.View ArticleGoogle Scholar
  5. Wei G, Forrest SR. Intermediate-band solar cells employing quantum dots embedded in an energy fence barrier. Nano Lett. 2007;7:218–22.View ArticleGoogle Scholar
  6. Sablon KA, Little JW, Olver KA, Wang ZM, Dorogan VG, Mazur YI, et al. Effects of AlGaAs energy barriers on InAs/GaAs quantum dot solar cells. J Appl Phys. 2010;108:074305.View ArticleGoogle Scholar
  7. Tutu FK, Lam P, Wu J, Miyashita N, Okada Y, Lee K-H, et al. InAs/GaAs quantum dot solar cell with an AlAs cap layer. Appl Phys Lett. 2013;102:163907.View ArticleGoogle Scholar
  8. Linares PG, Martí A, Antolín E, Farmer CD, Ramiro Í, Stanley CR, et al. Voltage recovery in intermediate band solar cells. Sol Energ Mat Sol C. 2012;98:240–4.View ArticleGoogle Scholar
  9. Hubbard SM, Cress CD, Bailey CG, Raffaelle RP, Bailey SG, Wilt DM. Effect of strain compensation on quantum dot enhanced GaAs solar cells. Appl Phys Lett. 2008;92:123512.View ArticleGoogle Scholar
  10. Laghumavarapu RB, El-Emawy M, Nuntawong N, Moscho A, Lester LF, Huffaker DL. Improved device performance of InAs/GaAs quantum dot solar cells with GaP strain compensation layers. Appl Phys Lett. 2007;91:243115.View ArticleGoogle Scholar
  11. Oshima R, Takata A, Okada Y. Strain-compensated InAs/GaNAs quantum dots for use in high-efficiency solar cells. Appl Phys Lett. 2008;93:083111.View ArticleGoogle Scholar
  12. Bailey CG, Forbes DV, Raffaelle RP, Hubbard SM. Near 1 V open circuit voltage InAs/GaAs quantum dot solar cells. Appl Phys Lett. 2011;98:163105.View ArticleGoogle Scholar
  13. Tanabe K, Guimard D, Bordel D, Arakawa Y. High-efficiency InAs/GaAs quantum dot solar cells by metalorganic chemical vapor deposition. Appl Phys Lett. 2012;100:193905.View ArticleGoogle Scholar
  14. Hubbard SM, Bailey CG, Aguinaldo R, Polly S, Forbes DV, Raffaelle RP. Characterization of quantum dot enhanced solar cells for concentrator photovoltaics. In: Photovoltaic Specialists Conference (PVSC), 2009 34th IEEE. 2009. p. 000090–5.View ArticleGoogle Scholar
  15. Cress CD, Hubbard SM, Landi BJ, Raffaelle RP, Wilt DM. Quantum dot solar cell tolerance to alpha-particle irradiation. Appl Phys Lett. 2007;91:183108.View ArticleGoogle Scholar
  16. Sogabe T, Shoji Y, Ohba M, Yoshida K, Tamaki R, Hong HF, et al. Intermediate-band dynamics of quantum dots solar cell in concentrator photovoltaic modules. Sci Rep. 2014;4:4792.View ArticleGoogle Scholar
  17. Bailey CG, Hubbard SM, Forbes DV, Raffaelle RP. Evaluation of strain balancing layer thickness for InAs/GaAs quantum dot arrays using high resolution x-ray diffraction and photoluminescence. Appl Phys Lett. 2009;95:203110.View ArticleGoogle Scholar
  18. Ekins-Daukes NJ, Kawaguchi K, Zhang J. Strain-balanced criteria for multiple quantum well structures and its signature in X-ray rocking curves. Cryst Growth Des. 2002;2:287–92.View ArticleGoogle Scholar
  19. Howe P. Competition between strain-induced and temperature-controlled nucleation of InAs/GaAs quantum dots. J Appl Phys. 2004;95:2998.View ArticleGoogle Scholar
  20. Ng JT, Bangert U, Missous M. Formation and role of defects in stacked large binary InAs/GaAs quantum dot structures. Semicond Sci Tech. 2007;22:80–5.View ArticleGoogle Scholar
  21. Torchynska TV. Photoluminescence peculiarities in InGaAs/GaAs structures with different InAs quantum dot densities. J Lumin. 2013;136:75–9.View ArticleGoogle Scholar
  22. Litvinov D, Blank H, Schneider R, Gerthsen D, Vallaitis T, Leuthold J, et al. Influence of InGaAs cap layers with different In concentration on the properties of InGaAs quantum dots. J Appl Phys. 2008;103:083532.View ArticleGoogle Scholar
  23. Yin Z, Tang X, Liu W, Zhang D, Du A. Effects of InxGa1−xAs matrix layer on InAs quantum dot formation and their emission wavelength. J Appl Phys. 2006;100:033109.View ArticleGoogle Scholar
  24. Kerestes C, Polly S, Forbes D, Bailey C, Podell A, Spann J, et al. Fabrication and analysis multijunction solar cell with an InGaAs Quantum dots junction. Prog Photovolt Res Appl. 2014;22:1172–9.View ArticleGoogle Scholar
  25. Dimroth F, Kurtz S. High-efficiency multijunction solar cells. MRS Bull. 2007;32:230–5.View ArticleGoogle Scholar
  26. Kim H, Park MH, Park SJ, Kim HS, Song JD, Kim SH, et al. Influence of InAs quantum dots on the transport properties of GaAs-based solar cell devices. Curr Appl Phys. 2014;14:192–5.View ArticleGoogle Scholar


© Li et al.; licensee Springer. 2015

This is an Open Access article distributed under the terms of the Creative Commons Attribution License (, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly credited.