Effect of Hydrogen Dilution on the Intrinsic a-Si:H Film of the
Heterojunction Silicon-Based Solar Cell
Jui-Chung Hsiao,
a,bChien-Hsun Chen,
b,zChao-Cheng Lin,
bDer-Ching Wu,
band
Peichen Yu
a,zaDepartment of Photonics and Institute of Electro-Optical Engineering, National Chiao Tung University, HsinChu 30010, Taiwan
bGreen Energy and Environment Research Laboratories, Industrial Technology Research Institute, HsinChu, Taiwan
In this work, effects of a hydrogen dilution ratio on the intrinsic amorphous hydrogenated silicon (i-a-Si:H) film of heterojunction silicon-based (HJS) solar cells were systematically studied. Long lifetime samples were obtained for R 5 5, indicating a good a-Si:H/c-Si interface. The dark conductivity was drastically decreased for R = 2, indicating a good film quality. Consequently, an optimized power conversion efficiency of the HJS solar cells was obtained at a moderate R between 2 and 5. In contrast to the pre-vious emphasis on long lifetime, the results indicate that both the interface and film qualities are correlated to the hydrogen dilu-tion, which are important to achieve high-efficiency HJS solar cells. We show that the most optimized HJS solar cell exhibits a marked efficiency of 17.27%.
VC2011 The Electrochemical Society. [DOI: 10.1149/1.3607981] All rights reserved.
Manuscript submitted December 16, 2010; revised manuscript received April 26, 2011. Published July 13, 2011.
The heterojunction silicon-based (HJS) solar cell has recently attracted more attention because of its very high efficiency and sim-ple structure which is composed of hydrogenated amorphous silicon (a-Si:H) layers and a crystalline silicon (c-Si) substrate.1–3All of the processes required to produce of the heterojunction solar cells can be implemented below 250C using plasma enhanced chemical
vapor deposition (PECVD) system and HJS solar cells with a high energy conversion efficiency (>23%) and record high open circuit voltage (Voc) of >740 mV can thus be obtained.
2
The success of heterojunction solar cells is due to the insertion of an intrinsic a-Si:H (i-a-a-Si:H) layer between an emitter layer and a c-Si substrate. This i-a-Si:H layer effectively passivates the silicon surface, reduc-ing the surface recombination velocity.
Numerous deposition parameters control the quality of an i-a-Si:H layer.4–8The key considerations in setting the deposition parameters are the need to passivate the surface of a c-Si substrate; to prevent the initiation of the localized epitaxial growth at the a-Si:H/c-Si interface and to reduce the defects in a-Si:H films.4,9,10 Hydrogen dilution of the silane gas mixture is extensively used in the deposition of a-Si:H films in thin film solar cells to improve performance and sta-bility.11–14Furthermore, some studies have discussed the effects of the H2 dilution ratio (R¼ SiH4/H2) in HJS solar cells. Kim et al.
pointed out that an R of 2–4 can effectively improve the efficiencies of p-type HJS solar cells compared to those without H2 dilution.
6
More recently, Jeon et al. and Dao et al. used radio frequency PECVD and inductive coupled plasma CVD, respectively, to deposit a-Si:H films on c-Si substrates with various R.7,8The optimum R values to obtain the highest effective lifetime were very different (R¼ 15 for Jeon et al. and R¼ 1 for Dao et al.), perhaps because different deposi-tion systems were used. Note that both groups studied the quality of the a-Si:H layer without the fabrication of HJS solar cells. Das et al. fully produced HJS solar cells with R values from 0 to 10.5 They found a high effective lifetime does not ensure a high HJS solar cell efficiency because of the low fill factor (FF).
In this work, the a-Si:H films were deposited using very high fre-quency (80 MHz) PECVD system at a low temperature of200C
with R values from 0 to 8. Surface passivation quality of i-a-Si:H films on c-Si substrates, microstructure at the a-Si:H/c-Si interface and the dark conductivity of i-a-Si:H films were investigated. We found that the quality at the a-Si:H/c-Si interface and the quality inside the a-Si:H film are both important to obtain high efficiency HJS solar cells.
The n-type CZ Si(100) substrates were cut into 2 2 cm. The re-sistivity and thickness were 1–5 X cm and 170 m, respectively. Samples were dipped in 5% HF to remove the native oxide layer
and were rinsed in ionized water. The a-Si:H thin films were de-posited using a 80 MHz PECVD system. An i-a-Si:H layer with the a thickness of5 nm was deposited on a c-Si substrate, and then a B-doped a-Si:H film was deposited as a p-type emitter layer with a thickness of20 nm using B2H6as the precursor. A Ga-doped ZnO
(GZO) film with a thickness of 80 nm was sputtered on a p-type a-Si:H layer as an anti-reflection coating and conductive layer, and then an Ag grid layer using a shadow metal mask to define the grid pattern with a thickness of200 nm. To simplify and elucidate the effect of the H2dilution ratio (R¼ H2/SiH4), Ag/GZO layers were
directly sputtered on the back of a c-Si substrate as metal contact layers without a back surface field (BSF) layer. The performance of each HJS solar cell was characterized under standard test conditions (25C, 1000 W/m2
, AM 1.5 G). The dark current-voltage (I-V) measurements were done to extract the conductivities of the i-a-Si:H films. The i-a-i-a-Si:H films were symmetrically deposited on the c-Si substrate to evaluate the quality of surface passivation, which was determined by the micro photo conductance decay (u-PCD, SemiLab WT-2000) to extract the effective lifetime. The micro-structure at the a-Si:H/c-Si interface was investigated by high-reso-lution transmission electron microscopy (HR-TEM, JEOL- 2100F).
Figure1shows the PV characteristic parameters, including open circuit voltage (Voc), short current density (Jsc), fill factor (FF),
se-ries resistance (Rs), and solar cell efficiency of HJS solar cells as
functions of R in the intrinsic a-Si:H layers. There are some interest-ing findinterest-ings can be extracted from Fig.1. First, both Vocand Jscare
almost unchanged when R is smaller than 5. They decrease drasti-cally to 0.55 V and 22 mA/cm2, respectively, when R is larger than 5. Second, the series resistance decreases rapidly from 5 to 2 X as R is 2 and further decreases to 1 as R rises to 8. Third, the FFs increase with R. Fourth, the efficiencies of the HJS solar cells increase to 8.4% at R¼ 2, and remain almost the same as R increases from 2 to 8. Notably, the typical “S” curves are observed in the I-V measure-ment at R < 2.
The relationship between lifetime and Vocis given by the
follow-ing equation: Voc¼ kT q ln Jsc J0 þ 1 [1] and J0¼ qn2 iW NDs [2]
Here, K is Boltzmann’s constant, T is the absolute temperature, q is the electric charge, W is the thickness of the silicon substrate and s is the effective lifetime. NDand nidenote the donor concentration
z
E-mail: [email protected]; [email protected]
Journal of The Electrochemical Society, 158 (9) H876-H878 (2011)
0013-4651/2011/158(9)/H876/3/$28.00VCThe Electrochemical Society
H876
) unless CC License in place (see abstract). ecsdl.org/site/terms_use
address. Redistribution subject to ECS terms of use (see 140.113.38.11
and intrinsic concentration of the silicon substrate, respectively. Figure2shows the effective lifetimes of i-a-Si:H films symmetri-cally deposited on n-Si substrates as a function of R. Two regions are clearly distinguished. All of the lifetimes are almost the same (100 s) when R is smaller than 5, whereas decreases rapidly to 70 s as R rises to 8. These results suggest that H2dilution in the
a-Si:H films plays a critical role in the passivation of the c-Si sub-strates. This tendency is wholly consistent with the relation between Vocand R in Fig.1, which is consistent with Eq.1. As mentioned
earlier, localized epitaxy on a c-Si substrate seriously affects the passivation effect of a-Si:H films on a c-Si substrate. HRTEM was
utilized to investigate the micro-structure of an interface of an a-Si:H/c-Si, as shown in Fig.3. The a-SI:H/c-Si interface with R¼ 0 is abrupt while that R¼ 8 exhibits localized partial epitaxial growth and is rough . In this study, the lifetime is reduced if the a-Si:H/c-Si interface is roughened. The localized epitaxial growth with an aver-aged size of2.5 nm scattered on the c-Si substrate suggests more dangling bonds which act as recombination centers in the midgap states at an interface of an a-Si:H/c-Si, leading to a poorer passivation.4,15
Reducing the defect density of an intrinsic a-Si:H film is very important if the film is to be applied in an HJS solar cell. The dark conductivity of an a-Si:H film, which implies the quantities of defects, is widely used to evaluate the quality of an a-Si:H film.16–19 Carriers can transport in an a–Si film through the localized defect states by a defect-assisted hopping mechanism.20 Figure 4 shows the dark conductivity of intrinsic a-Si:H thin films as a function of R. For 8 R 2, the dark conductivity remains constant at about 1.6 1011S/cm. However, the dark conductivity increases rapidly to about 6 1010S/cm at R¼ 0, suggesting that hydrogen is criti-cal to the conductivity of i-a-Si:H films. Very probably, these defects are passivated by hydrogen, which further hamper the trans-portation of carriers, reducing the dark conductivity. Notably, these defects act as recombination centers when the HJS cell is under illu-mination. Very recently, Rahmouni et al.10 showed by simulation that FFs decreased from 0.769 to 0.721 as the defect density in the i-a-Si:H film increased from 9 1014
to 9 1017
cm3, indicating how defects inside the i-a-Si:H film affect the HJS cell performance. This can explain the low FF at R¼ 0 in our study.
As mentioned earlier, the typical “S” curves are observed in the I-V measurement at R < 2. In an HJS cell, an i-a-Si:H film is sand-wiched between an emitter layer and a base substrate. An i-a-Si:H
Figure 1. Photovoltaic characteristics of (a) series resistance, Rs(b) conver-sion efficiency, Eff. (c) fill factor, FF (d) short circuit current density, Jscand (e) open circuit voltage, Vocfor HJS solar cells as a function of the hydrogen dilution ratio, R.
Figure 2. Effective lifetimes of i-a-Si:H films symmetrically deposited on n-Si substrates as a function of the hydrogen dilution ratio, R.
Figure 3. HRTEM images of the a-Si:H/c-Si interface at (a) R¼ 0, (b) R¼ 8.
Figure 4. The measured dark conductivity values as a function of the hydro-gen dilution ratio, R for the i-a-Si:H thin films.
Journal of The Electrochemical Society, 158 (9) H876-H878 (2011) H877
) unless CC License in place (see abstract). ecsdl.org/site/terms_use
address. Redistribution subject to ECS terms of use (see 140.113.38.11
film in an HJS cell must be a good passivation layer to reduce the density of states on the surface of the silicon substrate. It also acts as a conducting layer through which minority carriers can pass under illumination. For example, an i-a-Si:H film with a high lifetime can be obtained at R¼ 0, suggesting a good surface passivation. How-ever, the dark conductivity results in Fig.4reveal too many defects inside an i-a-Si:H film, which make it too resistive to conduct mi-nority carriers and further are responsible for a high series resist-ance, as shown in Fig.1. Meanwhile, an i-a-Si film with a low dark conductivity can be obtained at R¼ 8, suggesting high film quality. It had a low lifetime, indicating a worse interface property between the a-Si:H and the c-Si substrate, as supported by the microstructural observation in Fig.3. The optimization of the i-a-Si:H films for the HJS solar cells is the trade-off between the series resistance and life-time. Consequently, optimal i-a-Si:H films will be obtained at an R of 2–5. They have both a favorable film quality and a favorable interface property, which explain the HJS solar cell efficiencies at R¼ 2–5.
An abrupt a-Si:H/c-Si interface(or a high effective lifetime) does not guarantee a high HJS cell efficiency, because the defects inside the i-a-Si:H film can seriously affect the HJS solar cell performance. Only when the defects inside an a-Si:H film are passivated (or reduced) and the a-Si:H/c-Si interface is abrupt, can a high HJS cell efficiency be excepted. In this work, the spatial distribution of the defects of the a-Si:H film from the surface to the bulk is controlled by introducing the hydrogen during deposition. At R¼ 0, a well-passivated surface with an abrupt a-Si:H/c-Si interface is obtained, but because of the insufficiency of the hydrogen, the defects in the i-a-Si:H film cannot be effectively passivated. As the amount of the
hydrogen is increased, the defects inside an a-Si:H film are passi-vated at R¼ 2–5. As the amount of the hydrogen is increased further to R¼ 8, localized epitaxy occurs at the a-Si:H/c-Si interface, detri-mentally affecting the transportation of the minority carriers. The HJS solar cell with an optimized i-a-Si:H film and a back surface field layer showed an efficiency of 17.27% (aperture area) in Fig.5.
Heterojunction silicon-based solar cells were fabricated by vary-ing R from 0 to 8. The i-a-Si:H films with excellent passivation were obtained at R 5 5. The i-a-Si:H films that were deposited at R=2 contained fewer defects. Optimum HJS cell efficiencies were achieved at 2 5 R 5 5. This result indicates not only an abrupt a-Si:H/c-Si interface but also a high a-Si:H film quality are required to optimize HJS cell efficiency.
Acknowledgment
The authors thank Dr. Yu –Hung Chen, Dr. Chiung-Nan Jim Lee and Mr. Shih-Ting Liao for valuable discussion. This work was sup-ported by Bureau of Energy under the contract No. 9455DI1110.
References
1. Q. Wang, M. R. Page, E. Iwaniczko, Y. Xu, L. Roybal, R. Bauer, B. To, H.-C. Yuan, A. Duda, F. Hasoon et al.,Appl. Phys. Lett., 96, 013507 (2010).
2. T. Mishima, M. Taguchi, H. Sakata, and E. Maruyama,Sol. Energy Mater. Sol. Cells (2010).
3. D. Mun˜oz, A. S. Ozanne, S. Harrison, A. Danel, F. Souche, C. Denis, A. Favier, T. Desrues, S. M. de Nicola´s, N. Nguyen et al.,35rd IEEE Photovoltaic Specialists Conference (2010).
4. H. Fujiwaraa and M. Kondo,Appl. Phys. Lett., 90, 013503 (2007).
5. U. K. Das, M. Z. Burrows, M. Lu, S. Bowden, and R. W. Birkmire,Appl. Phys. Lett., 92, 063504 (2008).
6. S.-K. Kim, J. C. Lee, S.-J. Park, Y.-J. Kim, and K. H. Yoon,Sol. Energy Mater. Sol. Cells, 92, 298 (2008).
7. M. Jeon, S. Yoshiba, and K. Kamisako,J. Korean Phys. Soc., 54, 194 (2009). 8. V. A. Dao, N. V. Duy, J. Heo, H. Choi, Y. Kim, Lakshminarayan, and J. Yi,Jpn. J.
Appl. Phys., 48, 066509 (2009).
9. T. H. Wang, E. Iwaniczko, M. R. Page, D. H. Levi, Y. Yan, H. M. Branz, and Q. Wang,Thin Solid Films, 501, 284 (2006).
10. M. Rahmouni, A. Datta, P. Chatterjee, J. Damon-Lacoste, C. Ballif, and P. Roca i Cabarrocas,J. Appl. Phys., 107, 054521 (2010).
11. F. Meillaud, A. Shah, E. Vallat-Sauvain, X. Niquille, M. Dubey, and C. Ballif, 20th European Photovoltaic Solar Energy Conference, p. 6 (2005).
12. R. Platz, S. Wagner, C. Hof, A. Shah, S. Wieder, and B. Rech,J. Appl. Phys., 84, 3949 (1998).
13. W. M. M. Kessels, R. J. Severens, A. H. M. Smets, B. A. Korevaar, G. J. Adriaens-sens, D. C. Schram, and M. C. M. van de Sanden,J. Appl. Phys., 89, 2404 (2001). 14. S. HIZA, A. YAMADA, and M. KONAGAI,Jpn. J. Appl. Phys. Part 1, 47, 6222
(2008).
15. H. Angermann, L. Korte, J. Rappich, E. Conrad, I. Sieber, M. Schmidt, K. Hu¨b-ener, and J. Hauschild,Thin Solid Films, 516, 6775 (2008).
16. S. Tsuda, T. Takahama, Y. Hishikawa, H. Tarui, H. Nishiwaki, K. Wakisaka and S. Nakano,J. Non-Cryst. Solids, 164–166, 679 (1993).
17. F. Gaspari, S. K. O’Leary, and S. Zukotynski,J. Non-Crystal. Solids, 155, 149 (1993).
18. M. C. Ozturk and M. G. Thompson,Appl. Phys. Lett., 44, 916 (1984).
19. S. Inthisang, K. Sriprapha, S. Miyajima, A Yamada, and M. Konagai,Jpn. J. Appl. Phys., 48, 122402 (2009).
20. T. F. Schulze, L. Korte, E. Conrad, M. Schmidt, and B. Rech,J. Appl. Phys., 107, 023711 (2010).
Figure 5. Illuminated current density-voltage (J-V) curve of the optimized HJS solar cell.
Journal of The Electrochemical Society, 158 (9) H876-H878 (2011) H878
) unless CC License in place (see abstract). ecsdl.org/site/terms_use
address. Redistribution subject to ECS terms of use (see 140.113.38.11
