search for




 

Post-Heat Treatment on Cu(In,Ga)Se2 Solar Cells with CBD-ZnS Buffer Layers as a Function of ITO Growth Temperature
Applied Science and Convergence Technology 2018;27:189-193
Published online November 30, 2018;  https://doi.org/10.5757/ASCT.2018.27.6.189
© 2018 The Korean Vacuum Society.

Woo-Jung Leea, Dae-Hyung Choa, Hye-Jung Yua, Won Seok Hana, and Yong-Duck Chunga,b,*

aICT Materials Technology Research Laboratory, Electronics and Telecommunications Research Institute, Daejeon 34129, Republic of Korea, bDepartment of Advanced Device Technology, Korea University of Science and Technology, Daejeon 34113, Republic of Korea
Correspondence to: *Corresponding author: E-mail: ydchung@etri.re.kr
Received November 12, 2018; Accepted November 27, 2018.
Abstract

We fabricated Cu(In,Ga)Se2 (CIGS) solar cells with a the chemical bath deposition-ZnS buffer layer by varying the substrate temperature for indium tin oxide (ITO) growth (Tsub.ITO) from room temperature (RT) to 200 °C. The CIGS solar cell efficiency increased with increasing Tsub.ITO. After light soaking (LS), the CIGS solar cell performance improved noticeably, with a rise in fill factor, except for Tsub.ITO at RT, due to the high resistivity of ITO film. Post heat treatment (PHT) was carried out on the CIGS solar cells, in ambient air at 200 °C, with increasing annealing time, from 10 min to 1 h. After PHT for more than 10 min, cell performance was superior to that after LS at Tsub.ITO ≤ 100 °C, with substantially increased cell efficiency. This was due to simultaneously enhancing the quality of the ITO film with supplementing thermal energy, and curing a defect at the p-n junction. At Tsub.ITO ≥ 150 °C, cell performance improved after LS, compared to after PHT, regardless of the annealing time. After LS, photoexcited carriers were generated, which was beneficial for curing defects at the p-n junction, resulting in elevating cell performance. However, after PHT, excessive thermal energy was injected into the solar cell, which induced Zn diffusion into the CIGS absorber layer, forming different defect states, such as ZnCu and Zni, from the defect located at the p-n junction.

Keywords : Cu(In,Ga)Se2, ITO growth temperature, Chemical bath deposition-ZnS, Light soaking effect, Post heat treatment effect
I. Introduction

The Cu(In,Ga)Se2 (CIGS) solar cell has attracted substantial interest as a next-generation photovoltaic device, and its characteristics have been investigated for decades, due to several advantages, such as superior power conversion efficiency, controllable bandgap engineering and high absorption coefficient [13]. Recently, CIGS solar cell efficiency of up to ~22.9 % was reported by Solar Frontier [4], and there continues to be effort applied up to the present day in the development of CIGS solar cells with improved efficiency.

The typical structure of a CIGS solar cell involves a soda-lime-glass (SLG)/Mo/CIGS/CdS/transparent conducting oxide (TCO). The CIGS absorber layer is deposited by a 3-stage co-evaporation method, and includes various defect states [57]. As a buffer layer, chemical bath deposition (CBD)-CdS is widely used, although the use of toxic and harmful CdS should be avoided, for environmentally-friendly and renewable energy devices, and thus, Zn-based materials are needed, as an alternative buffer layer. In the CIGS solar cell, the choice of TCO material for the top electrode is also considered as an important issue, in terms of cell efficiency, and for commercial products, although ZnO:Al has been adopted, indium tin oxide (ITO) is well known for its outstanding physical and electrical properties, such as wide bandgap range, from 3.5 eV to 4.3 eV, high electrical conductivity and good thermal stability [8,9].

In a previous study, we fabricated CIGS solar cells with CBD–CdS buffer layers by varying the ITO growth temperature, and investigated the relationship between cell performance and ITO film property, as a function of growth temperature [10]. The efficiency of the CIGS solar cells gradually increased with each increment of ITO growth temperature, from 25 °C to 200 °C, mainly due to improved ITO film quality, and this implied that the ITO layer had a decisive influence on cell performance.

On the other hand, the most important arbiter of cell performance is p-n junction quality, among the several CIGS solar cell interface layers. Where CBD-ZnS has been applied as a buffer layer to CIGS solar cells, there are metastable defect states at the p–n junction, which cause cell degradation owing to the lowered fill factor (FF). However, notably, FF is compensated by light soaking (LS) and post-annealing treatments, after which cell efficiency noticeably increased [1113]. The study of this LS effect, of raising cell efficiency in CIGS solar cells with CBD-ZnS buffer layers, is therefore important.

In this study, we fabricated CIGS solar cells with a CBD-ZnS buffer layer, by varying the ITO growth temperature from room temperature (RT) to 200 °C, to examine the effect of ITO film quality on CIGS solar cells, and then sequentially carried out LS and post-heat treatment (PHT) on completed solar cells, to test cell performance enhancement. Cell performances were measured, after LS, under a solar simulator with air mass (AM) 1.5 spectrum, for 1 h, and PHT in an ambient air at 200 °C, with annealing time increased from 10 min to 1 h. From the results, we found that CIGS solar cell efficiency noticeably increased after LS or PHT, with the increase closely related to the growth temperature of the ITO.

II. Experiment

A 2.2 μm CIGS layer was deposited by the 3-stage co-evaporation method onto 900 nm, Mo-coated SLG [5,14]. CBD-ZnS, with a thickness of ~50 nm, was synthesized onto the CIGS layer in an aqueous solution, at 85 °C for 80 min, by dipping into a home-made water bath system [15]. The CBD processing conditions were as follows: mole concentrations of zinc sulfate, thiourea, and ammonia were 0.02 M, 0.17 M, and 2 M, respectively. ITO film was prepared on the CBD–ZnS buffer layer by radio frequency (RF) magnetron sputtering, using a sintered target (In2O3:SnO2=90:10 wt.%) [10]. A 150 nm ITO film was deposited at various growth temperatures (Tsub.ITO), from RT to 200 °C, under fixed conditions of RF power of 100 W, Ar flow of 50 sccm, and operating pressure of 0.67 Pa. The completed solar cell consisted of Mo/CIGS/CBD-ZnS/i-ZnO/ITO/Ni/Al, and had an active area of 0.47 cm2.

The CIGS solar cell parameters were deduced by measuring the current density (J)-voltage (V) curve, with a source meter (Keithley 2400), under a global AM 1.5 spectrum for 1000 W/m2 at RT. To examine the LS and PHT effect on CIGS solar cells, as a function of Tsub.ITO, we carried out LS using a solar simulator, with AM 1.5 spectrum, for 1 h at RT, and PHT in an ambient air, at 200 °C, with annealing time increased from 10 min to 1 h. As CIGS solar cell efficiency can return to its initial status, after LS, if it is kept in dark conditions for several hours, PHT was carried out the day after LS. To extract the photocurrent dependence of the CIGS solar cells as a function ITO growth temperature, J-V curves were repeatedly measured, after LS and PHT, under dark and white light.

III. Results and discussion

The performance of the CIGS solar cells, as a function of Tsub.ITO, was measured before and after LS for 1 h, as shown in Fig. 1. Cell efficiency increased with increasing Tsub.ITO. It was an interesting finding that cell efficiency was near zero in the case of Tsub.ITO at RT, which was a result distinct from the cell efficiency of approximately 7 % achieved with a CdS buffer layer, under the same condition of Tsub.ITO at RT [10]. After LS for 1 h, cell efficiency increased, except for Tsub.ITO at RT, maintaining its change tendency, which was mainly due to the substantial rise in FF, rather than open circuit voltage (VOC) or short circuit current (JSC).

It is known that the irradiation of blue photons in the solar spectrum was the origin of the increased cell efficiency after LS [16]. In our previous study, we actually demonstrated this effect, showing increased cell efficiency with the CBD-ZnS buffer layer, by only soaking UV light into the solar cell [15,17]. According to reported results, the CIGS solar cells with Zn–based buffer layers showed noticeably increased cell efficiency after post-annealing treatment of the completed solar cells [1821]. Unlike LS treatment, which is reversible and reproducible after several hours, post-annealing treatment is irreversible, allowing retention of improved cell performance.

In this experiment, we carried out PHT on CIGS solar cells, as a function of Tsub.ITO, by gradually increasing annealing time from 10 min to 1 h, while repeatedly measuring cell performance, as indicated in Fig. 2. Compared to as-fabricated cells, cell efficiency increased as soon as PHT was conducted for 10 min, in all cases, and became saturated after PHT for 20 min or more. For Tsub.ITO at RT and 100 °C, the increased cell efficiency is ascribed to the increased JSC and VOC, while for Tsub.ITO of 150 °C and 200 °C, it is attributed to the increased FF, which was similar to results achieved after LS; i.e., we discovered that the PHT had a beneficial effect on CIGS solar cells with ITO film grown at Tsub.ITO ≤ 100 °C.

Figure 3 shows the different CIGS solar cell efficiencies, after LS for 1 h (LS), PHT for 1 h (PHT) and then additional LS for 1 h after PHT for 1 h (PHT+LS). At Tsub.ITO at RT, after LS, cell efficiency was near zero, while after PHT, cell efficiency increased, up to ~6 %. At Tsub.ITO at 100 °C, the increased cell efficiencies were similar for LS, PHT and PHT+LS treatments. At Tsub.ITO ≥ 150 °C, the highest cell efficiency was achieved after LS, and then slightly decreased after PHT and PHT+LS. From these results, we concluded that the increase of cell efficiency at Tsub.ITO at RT was due to the supplementary thermal energy changing the ITO film quality during PHT at 200 °C. Based on previous results on the electrical properties of ITO film, as a function of Tsub.ITO, Sn ions substituted into the In2O3 lattice can release more electrons and thereby increase carrier concentration, at Tsub.ITO ≥ 150 °C, which facilitated the stability of the ITO film, rendering it suitable for the TCO layer [10]. In other words, ITO film quality is too poor to apply to solar cells, when grown at Tsub.ITO ≤ 100 °C, and thus, PHT performed at 200 °C, for 1h, makes better quality ITO film, resulting in increased cell efficiency, similarly to the LS effect. The detailed CIGS solar cell parameters (efficiency, JSC, VOC and FF), after various post-treatments (LS, PHT and PHT+LS), as a function of Tsub.ITO, have been summarized in Table I.

To analyze differences between the LS and PHT effects on CIGS solar cells, as a function of Tsub.ITO, CIGS solar cell J-V curves corresponding to each cell of Fig. 3 were measured, under the dark and white light conditions, as shown in Fig. 4. The dark current is expressed with a black line and there was no photocurrent, as expected. With irradiation by white light, photocurrent occurred in as-fabricated solar cells, and was clearly related to Tsub.ITO. For Tsub.ITO at RT, there was little photocurrent, due to the high resistivity of the ITO film. As Tsub.ITO increased, although photocurrent also increased with improved ITO film quality, distorted J-V curves, called ‘kinks’ [22,23], were clearly revealed at Tsub.ITO ≥ 100 °C. After LS and PHT, the kink disappeared and the FF increased, in all cases.

In a previous study, we interpreted this LS effect, raising cell efficiency in CIGS solar cells with CBD-ZnS buffer layers, as a result of the introduction of a deep acceptor (VCu-VSe)3−, located near the p-n junction, as follows; in as-fabricated solar cells, the J-V curve was distorted, due to the number of electrons trapped in deep acceptor states, resulting in FF loss. After LS for 1 h, the deep acceptor states were already filled with electrons excited from the p-n junction by the blue photon injection, thus lowering the barrier height for electron transport, which enhanced cell efficiency [15]. The recovered J-V curves, after LS and PHT, signified that PHT at 200 °C for more than 10 min was sufficient to fill the deep acceptor states with thermally excited electrons, resulting in improved cell efficiency.

In this experiment, after LS, cell efficiency at Tsub.ITO, at RT, was almost zero, because the ITO film wasn’t conducting with high resistivity, blocking carrier transport to the electrode. After PHT, however, cell efficiency for Tsub.ITO at RT increased up to ~ 6 %, with conducting ITO film, and did not drop FF. Thus, PHT on CIGS solar cells at Tsub.ITO, at RT, improved cell performance through both ITO film quality and defect curing at the p-n junctions. These positive effects after PHT occurred in CIGS solar cells at the Tsub.ITO of 100 °C. On the other hand, cell efficiencies were slightly less, after PHT, compared to those after LS, from 14.72 % to 13.25 % for Tsub.ITO at 150 °C, and from 14.89 % to 12.62 % for Tsub.ITO at 200 °C, respectively. This seemed to be due to the Zn atoms, with their monovalent atomic character, becoming easily diffused into the absorber, through the abundant VCu sites in the CIGS, which become defect point of ZnCu, occupying VCu and interstitial Zni [24]. The supply of additional thermal energy can cause a Zn diffusion into the CIGS layer, at Tsub.ITO ≥ 150 °C, by PHT at 200 °C for 1 h, bringing about negative effects on cell performance. Actually, as Tsub.ITO became higher, cell degradation became worse; i.e., excessive post-annealing processes should be curtailed, and optimized conditions for improved cell performance should be based on the CIGS solar cell.

IV. Summary

We fabricated CIGS solar cells with a CBD-ZnS buffer layer by increasing the Tsub.ITO from RT to 200 °C, and then investigated the effects of LS and PHT on these. As Tsub.ITO increased, cell performance improved. After LS, cell efficiency increased in all cases with the rise of FF, excluding Tsub.ITO at RT, which remained near zero after LS, due to the high resistivity ITO film blocking photocurrent flow. As soon as PHT time reached 10 min, at Tsub.ITO ≤ 100 °C, cell efficiency improved, while at Tsub.ITO ≥ 150 °C, the improvement was similar to that from the LS process. PHT was carried out sequentially on the CIGS solar cells, with PHT time increased up to 1 h, and it was found that cell efficiency was saturated in all cases after PHT time of 20 min. Compared to cell efficiency after LS, at Tsub.ITO ≤ 100 °C, it clearly increased, whereas at Tsub.ITO ≥ 150 °C it slightly decreased, after PHT for 1 h.

From these results, we discovered that PHT at 200 °C had an effect on CIGS solar cells similar to the LS effect, in terms of enhanced cell performance. The change of cell performance brought about by PHT and LS can be explained by the supply of additional thermal energy. At Tsub.ITO ≤ 100 °C, thermal energy by PHT facilitated both improved ITO film quality and curing of the deep acceptor defect at the p-n junction, while at Tsub.ITO ≥ 150 °C, excessive thermal energy caused Zn diffusion into the CIGS layer, resulting in a degree of cell degradation.

Acknowledgments

This work was supported by the Technology Development Program to Solve Climate Changes of the National Research Foundation (NRF) funded by the Ministry of Science, ICT & Future Planning (2016M1A2A2936754) and ‘New & Renewable Energy’ project of the Korea Institute of Energy Technology Evaluation and Planning (KETEP) grant funded by the Korean government Ministry Of Trade, Industry & Energy (20183010014310).

Figures
Fig. 1. Performance of CIGS solar cells, as a function of Tsub.ITO, before and after LS for 1 h.

Fig. 2. Performance of CIGS solar cells, as a function PHT time, from 10 min to 1 h, at different Tsub.ITO. PHT temperature was maintained at 200 °C.

Fig. 3. Comparison of CIGS solar cell efficiency, depending on Tsub.ITO, after LS for 1 h, PHT for 1 h and additional LS for 1h after PHT for 1h (PHT+LS).

Fig. 4. J-V curves for CIGS solar cells at different Tsub.ITO, after LS for 1 h, PHT for 1 h, and additional LS for 1 h after PHT for 1 h (PHT+LS), under conditions of dark and white light.

Tables

Parameters of CIGS solar cells with a CBD–ZnS buffer layer as a function of Tsub.ITO, from RT to 200 °C, after different post treatments (LS, PHT, and PHT+LS).

Tsub. (°C)As-fabricatedLSPHTPHT+LS

10 min20 min40 min60 min
RTVOC (V)0.200.180.390.460.490.500.49
JSC (mA/cm2)18.5322.7928.9730.3831.2330.5729.96
FF (%)29.9030.0037.9039.3041.4039.7038.70
Efficiency (%)1.091.214.335.496.336.085.68
100VOC (V)0.350.510.530.550.550.560.57
JSC (mA/cm2)30.5234.5934.4235.2834.9235.1734.90
FF (%)35.0049.2046.1047.9048.5048.0051.30
Efficiency (%)3.698.628.439.279.349.3610.17
150VOC (V)0.540.580.570.580.570.570.58
JSC (mA/cm2)34.9436.2735.0536.6036.3536.4936.16
FF (%)40.4069.5063.5064.1063.5064.0065.30
Efficiency (%)7.5414.7212.7013.5013.1813.2513.70
200VOC (V)0.570.600.580.580.580.570.58
JSC (mA/cm2)35.0436.0535.9736.3436.0436.1135.79
FF (%)41.5068.5059.2060.8061.2060.9062.00
Efficiency (%)8.3414.8912.3512.8912.7212.6212.93

References
  1. Nakada, T (2012). Electron Mater Lett. 8, 179.
    CrossRef
  2. Repins, I, Contreras, MA, Egaas, B, DeHart, C, Scharf, J, Perkins, CL, To, B, and Noufi, R (2008). Prog Photovolt: Res Appl. 16, 235.
    CrossRef
  3. Wei, SH, Zhang, SB, and Zunger, A (1998). Appl Phys Lett. 72, 3199.
    CrossRef
  4. Green, MA, Hishikawa, Y, Dunlop, ED, Levi, DH, Hohl-Ebinger, J, and Ho-Baillie, AWY (2018). Prog Photovolt: Res Appl. 26, 427.
    CrossRef
  5. Chung, YD, Cho, DH, Han, WS, Park, NM, Lee, KS, and Kim, JH (2010). J Kor Phys Soc. 57, 1826.
    CrossRef
  6. Rudmann, D, da Cunha, AF, Kaelin, M, Kurdesau, F, Zogg, H, Tiwari, AN, and Bilger, G (2004). Appl Phys Lett. 84, 1129.
    CrossRef
  7. Cho, DH, Lee, KS, Chung, YD, Kim, JH, Park, SJ, and Kim, J (2012). Appl Phys Lett. 101, 023901.
    CrossRef
  8. Negami, T, Hashimoto, Y, and Nishiwaki, S (2001). Sol Energy Mater Sol Cells. 67, 331.
    CrossRef
  9. Lee, H, and Cho, SW (2016). Appl Sci Converg Technol. 25, 128.
    CrossRef
  10. Cho, DH, Chung, YD, Lee, KS, Park, NM, Kim, KH, Choi, HW, and Kim, J (2012). Thin Solid Films. 520, 2115.
    CrossRef
  11. Naghavi, N, Temgoua, S, Hildebrandt, T, Guillemoles, JF, and Lincot, D (2015). Prog Photovolt: Res Appl. 23, 1820.
    CrossRef
  12. Kobayashi, T, Yamaguchi, H, and Nakada, T (2014). Prog Photovolt: Res Appl. 22, 115.
    CrossRef
  13. Kushiya, K, and Yamase, O (2000). Jpn J Appl Phys. 39, 2577.
    CrossRef
  14. Cho, DH, Lee, WJ, Wi, JH, Han, WS, Kim, TG, Kim, JW, and Chung, YD (2016). ETRI J. 38, 265.
    CrossRef
  15. Lee, WJ, Yu, HJ, Wi, JH, Cho, DH, Han, WS, Yoo, J, Yi, Y, Song, JH, and Chung, YD (2016). ACS Appl Mater Interfaces. 8, 22151.
    Pubmed CrossRef
  16. Eisgruber, I, Granata, J, Sites, J, Hou, J, and Kessler, J (1998). Sol Energy Mater Sol Cells. 53, 367.
    CrossRef
  17. Yu, HJ, Lee, WJ, Wi, JH, Cho, DH, Han, WS, Chung, YD, Kim, TS, and Song, JH (2016). Phys Chem Chem Phys. 18, 33211.
    Pubmed CrossRef
  18. Wi, JH, Kim, TG, Kim, JW, Lee, WJ, Cho, DH, Han, WS, and Chung, YD (2015). ACS Appl Mater Interfaces. 7, 17425.
    Pubmed CrossRef
  19. Wi, JH, Lee, WJ, Cho, DH, Han, WS, Yun, JH, and Chung, YD (2014). Phys Status Solidi A. 211, 2172.
    CrossRef
  20. Nakada, T, Kobayashi, T, Kumazawa, T, and Yamaguchi, H (2013). IEEE J Photovol. 3, 461.
    CrossRef
  21. Kobayashi, T, Kumazawa, T, Kao, ZJL, and Nakada, T (2014). Sol Energy Mater Sol Cells. 123, 197.
    CrossRef
  22. Pudov, A, Sites, J, Contreras, M, Nakada, T, and Schock, HW (2005). Thin Solid Films. 480, 273.
    CrossRef
  23. Köntges, M, Reineke-Koch, R, Nollet, P, Beier, J, Schäffler, R, and Parisi, J (2002). Thin Solid Films. 403, 280.
    CrossRef
  24. Lee, CS, Kim, S, Al-Ammar, EA, Kwon, H, and Ahn, BT (2014). ECS J Solid State Sci Technol. 3, Q99.
    CrossRef


January 2019, 28 (1)