- Email: [email protected]
Characterization of SiNx:H thin film as a hydrogen passivation layer for silicon solar cells with passivated contacts
Thin Solid Films 675 (2019) 109–114
Contents lists available at ScienceDirect
Thin Solid Films journal homepage: www.elsevier.com/locate/tsf
Characterization of SiNx:H thin film as a hydrogen passivation layer for silicon solar cells with passivated contacts
T
Jae Eun Kima,1, Se Jin Parka,1, Ji Yeon Hyuna, Hyomin Parka, Soohyun Baea, Kwang-sun Jib, ⁎ Hyunho Kima, Kyung Dong Leea, Yoonmook Kangc, Hae-Seok Leec, Donghwan Kima,c, a
Department of Materials Science and Engineering, Korea University, 145 Anam-ro, Seongbuk-gu, Seoul 02841, Republic of Korea Energy & Environment Materials & Devices Team, Materials & Devices Advanced Research Institute, LG Electronics, 38 Baumoe-ro, Seocho-gu, Seoul 06763, Republic of Korea c KU-KIST Green School, Graduate School of Energy and Environment, Korea University, 145 Anam-ro, Seongbuk-gu, Seoul 02841, Republic of Korea b
A R T I C LE I N FO
A B S T R A C T
Keywords: Silicon nitride Passivated contact Passivation Solar cells
Silicon nitride (SiNx:H) films are generally used as passivation and anti-reflection layers in solar cells, and they are usually made by plasma-enhanced chemical vapor deposition (PECVD). Silicon nitride could act as a hydrogen diffusion source, and it also plays a role in chemical passivation. In this study, we investigated the improvement of the passivation characteristics of the passivated contact structure by a PECVD SiNx:H hydrogenation process and the characteristics of SiNx:H for improving the passivation characteristics. It was confirmed that the passivation characteristics cannot be predicted only by the mass density of the SiNx:H film, and the chemical bonding ratio in the SiNx:H thin film is also important. In addition, higher passivation characteristics can be obtained when SiNx:H thin films with higher SeH bond concentration and dominant N2SieH2 bonds are used.
1. Introduction Hydrogenated silicon nitride (SiNx:H) layer deposited by plasmaenhanced chemical vapor deposition (PECVD) is used as a hydrogen source for improving the performance of semiconductor-based electronic and electric devices. The diffused hydrogen from SiNx:H by heat treatment associates with impurities and defects in bulk and passivates them [1,2]. There are two mechanisms proposed to describe the hydrogen desorption mechanism from SiNx:H films, either through the diffusion of atomic hydrogen, and/or molecular hydrogen [3,4]. Benoit et al. proposed that the hydrogen desorption reaction occurs mainly through dissociation reactions between SieH and NeH bonds giving free molecular hydrogen and generating new SieSi or SieN bonds, which are observed by the bonding and mass measurements of SiNx:H before and after annealing [5]. Therefore, the chemical bonding configuration in SiNx:H can be used as an indicator of the hydrogen release potential, while the mass density of SiNx:H is proposed as another criterion. Low density (< 2.5 g/cm3) films desorb high amount of H (over 1 × 1017 at./cm2) and are permeable to H, while higher density (≥2.5 g/cm3) films act as barriers and do not allow H to pass through them [5]. However, several studies have shown that a higher hydrogen
desorption from SiNx:H by the annealing process does not simply lead to a higher degree of bulk passivation in the multicrystalline silicon solar cells [6–8]. Rather, the degree of bulk passivation increases as the density of the SiNx:H film increases. Therefore, additional research is needed to improve passivation quality by hydrogenation. The stagnant efficiency of silicon solar cells has increased in recent years with the development of various structures such as heterojunction back contact, polycrystalline on oxide passivating contact and tunnel oxide passivated contact structures [9–11]. This is accomplished by applying a passivated contact structure onto the previously proposed back contact and heterojunction structures [12]. The passivated contact structure provides interface passivation by placing a few-nanometerthick layers of thin intrinsic amorphous silicon or oxide thin films between the bulk and the heavily doped layer. Furthermore, the highly doped layers provide band-bending in the bulk, resulting in a field passivation effect [12,13]. In addition, it is possible to eliminate complicated processes such as contact patterning and local doping formation process for minimizing the recombination area and improving the contact resistance. The passivated contact structure further enhances the passivation effect through the hydrogenation process which can passivate grain boundaries in the high-doped layer and the pin-holes
⁎
Corresponding author at: Department of Materials Science and Engineering, Korea University, 145 Anam-ro, Seongbuk-gu, Seoul 02841, Republic of Korea. E-mail address: [email protected] (D. Kim). 1 These authors have equally contributed to this work. https://doi.org/10.1016/j.tsf.2019.02.016 Received 1 August 2018; Received in revised form 19 January 2019; Accepted 10 February 2019 Available online 11 February 2019 0040-6090/ © 2019 Elsevier B.V. All rights reserved.
Thin Solid Films 675 (2019) 109–114
J.E. Kim, et al.
generated at the silicon oxide (SiOx) thin film/bulk silicon (Si) interface [14–16]. PECVD-deposited SiNx:H is used for hydrogenation of silicon solar cells. However, previous studies have focused on optimizing surface passivation and antireflective film functions together [17–19]. Further studies on SiNx:H are needed to improve and understand the passivation properties of silicon solar cells with the passivated contacts. In this study, we investigated the correlation between the quality of hydrogen passivation, mass density, and the chemical bonding of SiNx:H layers for passivated contact solar cells. Various SiNx:H were evaluated at densities over the 2.55–2.75 g/cm3 range. The changes in the passivation qualities were confirmed from the implied Voc by quasisteady-state photoconductance (QSSPC) and chemical bonding was detected by Fourier transformed infrared (FT-IR) measurements.
Table 1 The deposition conditions for SiNx:H film with different mass densities. The mass density and the composition of the deposited films are measured by XRR and RBS-ERD, respectively. SiNx:H film Name
Film Film Film Film Film Ref.
A B C D E
SiH4/NH3
1.5 1 1 0.6 0.3 1.5
RF power
Pressure
Composition ratio
Mass density
W
Pa
Si
N
H
g/cm3
40 40 25 25 25 60
9.8 9.8 29.4 29.4 29.4 9.8
3.0 3.0 3.0 3.0 3.0
3.8 3.1 4.0 4.3 4.3 –
0.92 1.4 1.5 1.5 1.7
2.75 2.55 2.70 2.70 2.70 2.05
2. Experimental passivation results with SiNx:H films of higher density [6]. Therefore, we have fabricated SiNx:H films with various densities to confirm the correlation between the passivation characteristics and mass density, as shown in Table 1. The passivation characteristics according to the density difference were confirmed by the implied Voc values obtained by QSSPC measurements after the SiNx:H layer deposition and heat treatment process on the passive contact structure. Here, i-poly-Si is used instead of the heavily doped poly-Si to avoid the influence of field passivation by the highly doped region and the in-diffusion of the dopant into the bulk during the passivated contact manufacturing process. Fig. 1 shows the passivation characteristics according to different densities measured using QSSPC. The passivated contact structure prior to hydrogenation of SiNx:H exhibited an iVoc value of 500 mV and increased iVoc in all samples after hydrogenation, although this varied with the mass density. The SiNx:H layers with a density of 2.5 g/cm3 or higher exhibited hydrogenation characteristics of 670 mV or higher. The passivation characteristic of silicon nitride with a density of 2.05 g/ cm3 was < 600 mV. The mass density and passivation trends showed similar behavior to previously reported results [6]. However, the hydrogenation trend could not be explained by the mass density over 2.5 g/cm3. The iVoc was maintained even after removing the SiNx:H film. This means that the passivated contact structure is passivated by the hydrogen from the SiNx:H thin film. To explain this, we measured the hydrogen concentration in a passivated contact structure and the chemical bonding in silicon nitrides. Firstly, SIMS measurements were performed to confirm that the changes of passivation properties were due to hydrogen. Fig. 2a shows the SIMS data of the hydrogen content of poly-Si/thin oxide/c-Si on the heat-treated samples after the deposition of SiNx:H. The hydrogen content between the samples is different, and it is confirmed that hydrogen in particular is abundant in the oxide/Si interface. Hydrogen can passivate pin-holes at the SiOx/Si interface [14–16]. The relationship between the concentration of hydrogen at the interface and the implied Voc is shown in Fig. 2b, which indicates high passivation properties with high hydrogen content. The
2.1. Sample preparation A passivated contact structure was fabricated on both sides of low resistivity (0.5–3 Ω·cm), n-type (100)-oriented, 200 um, Czochralski (Cz) monocrystalline Si substrate wafer. Firstly, the wafer was subjected to saw-damage etching with potassium hydroxide solution, followed by RCA cleaning. Thereafter, a thin oxide layer was deposited on both sides of the silicon wafer by dipping the wafer in a 30% hydrogen peroxide solution at 80 °C for 10 min, and then the wafer underwent an annealing process at 600 °C under N2 atmosphere to stabilize the thin oxide layer [16]. Then, intrinsic polycrystalline silicon (i-poly-Si) was deposited on both sides of the silicon wafer using a low-pressure chemical vapor deposition process. After the i-poly-Si deposition, the wafer was annealed at 900 °C for 60 min under N2 atmosphere. The SiNx:H films were grown by PECVD at a temperature of ~380 °C. The gases used for PECVD were SiH4, NH3, N2, and H2. The gas ratio, deposition pressure, and RF power were varied to create silicon nitrides with various mass densities. The hydrogenation performance of SiNx:H was measured after it underwent annealing in a rapid thermal processing (RTP) chamber at 600 °C for 15 min at a heating rate of 5 °C/min, under ambient N2 pressure. The temperature of the samples was monitored using a thermocouple placed on the front side. 2.2. Characterization The implied voltage (iVoc) values were obtained from the QSSPC measurements in order to compare the passivation characteristics of the annealed samples. X-ray reflectivity (XRR), Rutherford backscattering spectrometry-elastic recoil detection (RBS-ERD), and secondary ion mass spectrometry (SIMS) measurements were performed to analyze the mass density, composition of the films, and the hydrogen concentration on the tunnel oxide, respectively. The SieH and NeH bonds of SiNx:H were detected by FT-IR spectroscopy and their concentrations were deduced from the stretching mode absorbance peak areas using the IR-scattering cross-sections of Lanford and Rand [5], as follows: (1)
CX − Y = A/(σX − Y × t) −3
where CX-Y, A, and t are the concentration of the XeY bond in cm , the area of the XeY bond stretching mode absorption peak, and the thickness of the film, respectively. σSi−H and σN−H are the absorption cross-sections of the bonds, and the values for these are 7.4 × 10−18 cm2 and 5.3 × 10−18 cm2, respectively. 3. Results and discussion In the literature, the mass density of the SiNx:H film was reported as one of the factors influencing the probability of hydrogen release by the film after the heat treatment. Higher hydrogenation desorption properties have been reported for low mass density SiNx:H films [5]. On the other hand, multicrystalline silicon solar cells showed better
Fig. 1. Implied Voc of the passivated contact samples hydrogenated using the SiNx:H as functions of the mass density of the as-deposited SiNx:H films. 110
Thin Solid Films 675 (2019) 109–114
J.E. Kim, et al.
Fig. 2. (a) The H concentration profile of the passivated contact samples hydrogenated using the SiNx:H films. The arrow indicates the position of the oxide layer in passivated contact structure (b) Correlation between H concentration in the oxide layer and passivation characteristics.
passivated contact structure is passivated by hydrogen diffused from SiNx:H, and the passivation characteristics cannot be predicted only by the mass density of the SiNx:H film. Next, the chemical bonding of as-deposited SiNx:H with a mass density of > 2.5 g/cm3 was analyzed by FT-IR measurements as shown in Fig. 3a. Two ranges of wavenumbers from 2005 to 2220 cm−1 and from 3300 to 3460 cm−1 imply SieH stretching bonds and NeH stretching bonds, respectively. The most cited hydrogen desorption reactions from SiNx:H films are shown in Eq. (2) and Eq. (3) [4,5].
Si▬H + Si▬H ↔ Si▬Si + H2
(2)
Si▬H + N▬H ↔ Si▬N + H2
(3)
Table 2 Detailed bonding structures consisting of SieH and NeH stretching bands and their respective peak positions. Stretching band
Bonding structure
Wavenumbers(cm−1)
SieH
Si3SieH Si2SieH2 NSi2SieH NSiSieH2, N2SiSieH N2SieH2 N3SieH Si2NeH⋯H Si2NeH SiNeH2
2005 2065 2082 2140 2175 2220 3300 3350 3460
NeH
It is expected that the frequency and occurrence of reactions in Eq. (2) and Eq. (3) will be different depending on the distribution of the chemical bonds in silicon nitrides. Previous studies have insisted that the passivation quality after the firing process depended strongly on the SieH bonding density [20]. In order to compare the passivation characteristics according to the bonding ratios, the concentration of SieH and NeH bonding in the SiNx:H thin film was calculated from the FT-IR results using Eq. (1). As shown in Fig. 3b, the SiNx:H film with high SieH ratio showed high iVoc characteristics. The chemical bonding peaks differed not only in concentration but also in peak position. The results of FT-IR can be interpreted according to Table 2 [20,21]. SiNx:H films have their peak position near 2175–2220 cm−1 and 3300–3350 cm−1 wavenumbers. In addition, N2SieH2 bonding and N3SieH bonding are the predominant SieH bonding structures in SiNx:H films. At the same time, in the case of NeH stretching band, Si2NeH⋯H bonding and Si2N-H bonding are the predominant NeH bonding structures in SiNx:H films. Deconvolution was performed to obtain a more accurate analysis of the internal bonding. Figs. 4 and 5 show the results of the SieH and NeH bonding deconvolution, respectively, according to the SiNx:H films A to E. We introduced a factor R(SieH) and R(NeH) to explain the amount of N2SieH2 bonding and
Si2N-H⋯H bonding in the SiNx:H films. R(SieH) means the ratio of AN2Si−H2 to the sum of AN2Si−H2 and AN3Si−H. R(NeH) is the ratio of ASi2N−H⋯H to the sum of ASi2N−H⋯H and ASi2N−H. From Fig. 6, it is confirmed that as the number of N2SieH2 bonds in SiNx:H increased, the iVoc improved. For NeH bonding, the measured values were deconvoluted in the same manner as for SieH. In the case of NeH bonding, the R factor was not significantly different, varying from 0.43 to 0.50, between the samples. Therefore, it is confirmed that higher passivation characteristics can be obtained when SiNx:H thin films with a higher SeH bond concentration and dominant N2SieH2 bonds in the sample are used. This is consistent with the results of Mäckel, H. et al. that the formation of = Si-H2 is important for a high-quality silicon surface passivation [20]. Finally, the change of the hydrogen content in SiNx:H was qualitatively confirmed by the FT-IR values before and after its heat treatment. The atomic concentration of [H] has been deduced by the sum of SieH and NeH bonding concentrations. In Fig. 7, the highest degree of change was observed for silicon nitrides with high R(SieH) values for hydrogenation. This supports the mechanism for hydrogen diffusion of
Fig. 3. (a) FT-IR absorption spectra of as-deposited SiNx:H films (b) Implied Voc of the passivated contact samples hydrogenated using the SiNx:H as functions of the SieH bonding ratio in as-deposited SiNx:H films, which are calculated from the FT-IR results (a). 111
Thin Solid Films 675 (2019) 109–114
J.E. Kim, et al.
Fig. 4. Deconvolution of SieH absorption peaks using N2SieH2 and N3Si-H bonding structure peaks.
Fig. 5. Deconvolution of NeH absorption peaks using Si2N-H⋯H and Si2N-H bonding structure peaks. 112
Thin Solid Films 675 (2019) 109–114
J.E. Kim, et al.
Fig. 6. (a) Correlation between the ratio of N2SieH2 in the SieH chemical bond peak and the implied Voc (b) Correlation between the ratio of Si2N-H⋯H in the NeH chemical bond peak and the implied Voc.
References [1] D. Benoit, J. Regolini, P. Morin, Hydrogen desorption and diffusion in PECVD silicon nitride. Application to passivation of CMOS active pixel sensors, Microelectron. Eng. 84 (2007) 2169–2172, https://doi.org/10.1016/j.mee.2007. 04.071. [2] B. Sopori, Y. Zhang, R. Reedy, K. Jones, Y. Yanfa, M. Al-Jassim, B. Bathey, J. Kalejs, A comprehensive model of hydrogen transport into a solar cell during silicon nitride processing for fire-through metallization, Proceedings of the 31st IEEE Photovoltaic Specialists Conference, 2005, pp. 1039–1042, , https://doi.org/10.1109/PVSC. 2005.1488311. [3] J. Robertson, Defects and hydrogen in amorphous silicon nitride, Philos. Mag. B 69 (1994) 307–326, https://doi.org/10.1080/01418639408240111. [4] C. Boehme, G. Lucovsky, H loss mechanism during anneal of silicon nitride: chemical dissociation, J. Appl. Phys. 88 (2000) 6055–6059, https://doi.org/10.1063/ 1.1321730. [5] D. Benoit, P. Morin, J. Regolini, Determination of silicon nitride film chemical composition to study hydrogen desorption mechanisms, Thin Solid Films 519 (2011) 6550–6553, https://doi.org/10.1016/j.tsf.2011.04.130. [6] J. Hong, W.M.M. Kessels, W.J. Soppe, A.W. Weeber, W.M. Arnoldbik, M.C.M. van de Sanden, Influence of the high-temperature "firing" step on high-rate plasma deposited silicon nitride films used as bulk passivating antireflection coatings on silicon solar cells, J. Vac. Sci. Technol. B 21 (2003) 2123–2132, https://doi.org/10. 1116/1.1609481. [7] F. Chen, I. Romijn, A. Weeber, J. Tan, B. Hallam, J. Cotter, Relationship between PECVD silicon nitride film composition and surface and edge passivation, Proceedings of the 22nd European Photovoltaic Solar Energy Conference and Exhibition, 2007, pp. 1053–1060. [8] V. Verlaan, C.H.M. Van der Werf, Z.S. Houweling, I.G. Romijn, A.W. Weeber, H.F.W. Dekkers, H.D. Goldbach, R.E.I. Schropp, Multi-crystalline Si solar cells with very fast deposited (180 nm/min) passivating hot-wire CVD silicon nitride as antireflection coating, Prog. Photovolt. 15 (2007) 563–573, https://doi.org/10.1002/ pip.760. [9] K. Yoshikawa, H. Kawasaki, W. Yoshida, T. Irie, K. Konishi, K. Nakano, T. Uto, D. Adachi, M. Kanematsu, H. Uzu, K. Yamamoto, Silicon heterojunction solar cell with interdigitated back contacts for a photoconversion efficiency over 26%, Nat. Energy 2 (2017) 17032, https://doi.org/10.1038/nenergy.2017.32. [10] F. Haase, C. Hollemann, S. Schäfer, A. Merkle, M. Rienäcker, J. Krügener, R. Brendel, R. Peibst, Laser contact openings for local poly-Si-metal contacts enabling 26.1%-efficient POLO-IBC solar cells, Sol. Energy Mater. Sol. Cells 186 (2018) 184–193, https://doi.org/10.1016/j.solmat.2018.06.020. [11] A. Richter, J. Benick, F. Feldmann, A. Fell, M. Hermle, S.W. Glunz, n-Type Si solar cells with passivating electron contact: identifying sources for efficiency limitations by wafer thickness and resistivity variation, Sol. Energy Mater. Sol. Cells 173 (2017) 96–105, https://doi.org/10.1016/j.solmat.2017.05.042. [12] J. Melskens, B.W.H. van de Loo, B. Macco, L.E. Black, S. Smit, W.M.M. Kessels, Passivating contacts for crystalline silicon solar cells: from concepts and materials to prospects, IEEE J. Photovolt. 8 (2018) 373–388, https://doi.org/10.1109/ JPHOTOV.2018.2797106. [13] F. Feldmann, M. Simon, M. Bivour, C. Reichel, M. Hermle, S.W. Glunz, Efficient carrier-selective p- and n-contacts for Si solar cells, Sol. Energy Mater. Sol. Cells 131 (2014) 100–104, https://doi.org/10.1016/j.solmat.2014.05.039. [14] G. Yang, A. Ingenito, N. van Hameren, O. Isabella, M. Zeman, Design and application of ion-implanted polySi passivating contacts for interdigitated back contact c-Si solar cells, Appl. Phys. Lett. 108 (2016) 033903, https://doi.org/10.1063/1. 4940364. [15] N.H. Nickel, N.M. Johnson, W.B. Jackson, Hydrogen passivation of grain boundary defects in polycrystalline silicon thin films, Appl. Phys. Lett. 62 (1993) 3285–3287, https://doi.org/10.1063/1.109101. [16] H. Kim, S. Bae, K.S. Ji, S.M. Kim, J.W. Yang, C.H. Lee, K.D. Lee, S. Kim, Y. Kang, H.S. Lee, D. Kim, Passivation properties of tunnel oxide layer in passivated contact silicon solar cells, Appl. Surf. Sci. 409 (2017) 140–148, https://doi.org/10.1016/j. apsusc.2017.02.195.
Fig. 7. Change of hydrogen content in SiNx:H thin films with different SieH bonding ratio before and after annealing.
SiNx:H by Eq. (2) and Eq. (3). 4. Conclusions In this study, we conducted hydrogenation of the passivated contact structure using PECVD SiNx:H as a hydrogen source and observed the correlation between SiNx:H mass density, chemical bonding of the film, and the passivation results. Passivated contact structure showed improved passivation characteristics after hydrogenation using the SiNx:H thin films. After hydrogenation, the hydrogen profile shows an accumulated distribution at the SiOx/Si interface, which is inferred to have resulted in an improved interface passivation. In addition, the SiNx:H films with more SieH bonding showed higher iVoc even though they had a similar mass density. From this study, we conclude that SiNx:H can be used as a good hydrogen source for the passivated contact structure. In addition, the passivation characteristics cannot be predicted only by the mass density of the SiNx:H film, and the chemical bonding ratio of SieH in the thin film is also important. In particular, N2SieH2 bonding in SieH contributes to hydrogen passivation. Acknowledgements This work was supported by the Korea Institute of Energy Technology Evaluation and Planning (KETEP) and the Ministry of Trade, Industy & Energy (MOTIE) of the Republic of Korea (No. 20163010012430 and 20163030014020). This work was also supported by the "Human Resources Program in Energy Technology" of the Korea Institute of Energy Technology Evaluation and Planning (KETEP), with financial support from the Ministy of Trade, Industry and Energy, Republic of Korea (No. 20154030200760). 113
Thin Solid Films 675 (2019) 109–114
J.E. Kim, et al.
the composition of hydrogenated silicon nitride SiNx:H for efficient surface and bulk passivation of silicon, Sol. Energy Mater. Sol. Cells 93 (2009) 1281–1289, https://doi.org/10.1016/j.solmat.2009.01.023. [20] H. Mäckel, R. Lüdemann, Detailed study of the composition of hydrogenated SiNx layers for high-quality silicon surface passivation, J. Appl. Phys. 92 (2002) 2602–2609, https://doi.org/10.1063/1.1495529. [21] F. Ay, A. Aydinli, Comparative investigation of hydrogen bonding in silicon based PECVD grown dielectrics for optical waveguides, Opt. Mater. 26 (2004) 33–46, https://doi.org/10.1016/j.optmat.2003.12.004.
[17] F. Giorgis, F. Giuliani, C.F. Pirri, E. Tresso, C. Summonte, R. Rizzoli, R. Galloni, A. Desalvo, P. Rava, Optical, structural and electrical properties of device-quality hydrogenated amorphous silicon-nitrogen films deposited by plasma-enhanced chemical vapour deposition, Philos. Mag. B 77 (1998) 925–944, https://doi.org/10. 1080/13642819808206395. [18] F. Duerinckx, J. Szlufcik, Defect passivation of industrial multicrystalline solar cells based on PECVD silicon nitride, Sol. Energy Mater. Sol. Cells 72 (2002) 231–246, https://doi.org/10.1016/S0927-0248(01)00170-2. [19] J.F. Lelièvre, E. Fourmond, A. Kaminski, O. Palais, D. Ballutaud, M. Lemiti, Study of
114
Contents lists available at ScienceDirect
Thin Solid Films journal homepage: www.elsevier.com/locate/tsf
Characterization of SiNx:H thin film as a hydrogen passivation layer for silicon solar cells with passivated contacts
T
Jae Eun Kima,1, Se Jin Parka,1, Ji Yeon Hyuna, Hyomin Parka, Soohyun Baea, Kwang-sun Jib, ⁎ Hyunho Kima, Kyung Dong Leea, Yoonmook Kangc, Hae-Seok Leec, Donghwan Kima,c, a
Department of Materials Science and Engineering, Korea University, 145 Anam-ro, Seongbuk-gu, Seoul 02841, Republic of Korea Energy & Environment Materials & Devices Team, Materials & Devices Advanced Research Institute, LG Electronics, 38 Baumoe-ro, Seocho-gu, Seoul 06763, Republic of Korea c KU-KIST Green School, Graduate School of Energy and Environment, Korea University, 145 Anam-ro, Seongbuk-gu, Seoul 02841, Republic of Korea b
A R T I C LE I N FO
A B S T R A C T
Keywords: Silicon nitride Passivated contact Passivation Solar cells
Silicon nitride (SiNx:H) films are generally used as passivation and anti-reflection layers in solar cells, and they are usually made by plasma-enhanced chemical vapor deposition (PECVD). Silicon nitride could act as a hydrogen diffusion source, and it also plays a role in chemical passivation. In this study, we investigated the improvement of the passivation characteristics of the passivated contact structure by a PECVD SiNx:H hydrogenation process and the characteristics of SiNx:H for improving the passivation characteristics. It was confirmed that the passivation characteristics cannot be predicted only by the mass density of the SiNx:H film, and the chemical bonding ratio in the SiNx:H thin film is also important. In addition, higher passivation characteristics can be obtained when SiNx:H thin films with higher SeH bond concentration and dominant N2SieH2 bonds are used.
1. Introduction Hydrogenated silicon nitride (SiNx:H) layer deposited by plasmaenhanced chemical vapor deposition (PECVD) is used as a hydrogen source for improving the performance of semiconductor-based electronic and electric devices. The diffused hydrogen from SiNx:H by heat treatment associates with impurities and defects in bulk and passivates them [1,2]. There are two mechanisms proposed to describe the hydrogen desorption mechanism from SiNx:H films, either through the diffusion of atomic hydrogen, and/or molecular hydrogen [3,4]. Benoit et al. proposed that the hydrogen desorption reaction occurs mainly through dissociation reactions between SieH and NeH bonds giving free molecular hydrogen and generating new SieSi or SieN bonds, which are observed by the bonding and mass measurements of SiNx:H before and after annealing [5]. Therefore, the chemical bonding configuration in SiNx:H can be used as an indicator of the hydrogen release potential, while the mass density of SiNx:H is proposed as another criterion. Low density (< 2.5 g/cm3) films desorb high amount of H (over 1 × 1017 at./cm2) and are permeable to H, while higher density (≥2.5 g/cm3) films act as barriers and do not allow H to pass through them [5]. However, several studies have shown that a higher hydrogen
desorption from SiNx:H by the annealing process does not simply lead to a higher degree of bulk passivation in the multicrystalline silicon solar cells [6–8]. Rather, the degree of bulk passivation increases as the density of the SiNx:H film increases. Therefore, additional research is needed to improve passivation quality by hydrogenation. The stagnant efficiency of silicon solar cells has increased in recent years with the development of various structures such as heterojunction back contact, polycrystalline on oxide passivating contact and tunnel oxide passivated contact structures [9–11]. This is accomplished by applying a passivated contact structure onto the previously proposed back contact and heterojunction structures [12]. The passivated contact structure provides interface passivation by placing a few-nanometerthick layers of thin intrinsic amorphous silicon or oxide thin films between the bulk and the heavily doped layer. Furthermore, the highly doped layers provide band-bending in the bulk, resulting in a field passivation effect [12,13]. In addition, it is possible to eliminate complicated processes such as contact patterning and local doping formation process for minimizing the recombination area and improving the contact resistance. The passivated contact structure further enhances the passivation effect through the hydrogenation process which can passivate grain boundaries in the high-doped layer and the pin-holes
⁎
Corresponding author at: Department of Materials Science and Engineering, Korea University, 145 Anam-ro, Seongbuk-gu, Seoul 02841, Republic of Korea. E-mail address: [email protected] (D. Kim). 1 These authors have equally contributed to this work. https://doi.org/10.1016/j.tsf.2019.02.016 Received 1 August 2018; Received in revised form 19 January 2019; Accepted 10 February 2019 Available online 11 February 2019 0040-6090/ © 2019 Elsevier B.V. All rights reserved.
Thin Solid Films 675 (2019) 109–114
J.E. Kim, et al.
generated at the silicon oxide (SiOx) thin film/bulk silicon (Si) interface [14–16]. PECVD-deposited SiNx:H is used for hydrogenation of silicon solar cells. However, previous studies have focused on optimizing surface passivation and antireflective film functions together [17–19]. Further studies on SiNx:H are needed to improve and understand the passivation properties of silicon solar cells with the passivated contacts. In this study, we investigated the correlation between the quality of hydrogen passivation, mass density, and the chemical bonding of SiNx:H layers for passivated contact solar cells. Various SiNx:H were evaluated at densities over the 2.55–2.75 g/cm3 range. The changes in the passivation qualities were confirmed from the implied Voc by quasisteady-state photoconductance (QSSPC) and chemical bonding was detected by Fourier transformed infrared (FT-IR) measurements.
Table 1 The deposition conditions for SiNx:H film with different mass densities. The mass density and the composition of the deposited films are measured by XRR and RBS-ERD, respectively. SiNx:H film Name
Film Film Film Film Film Ref.
A B C D E
SiH4/NH3
1.5 1 1 0.6 0.3 1.5
RF power
Pressure
Composition ratio
Mass density
W
Pa
Si
N
H
g/cm3
40 40 25 25 25 60
9.8 9.8 29.4 29.4 29.4 9.8
3.0 3.0 3.0 3.0 3.0
3.8 3.1 4.0 4.3 4.3 –
0.92 1.4 1.5 1.5 1.7
2.75 2.55 2.70 2.70 2.70 2.05
2. Experimental passivation results with SiNx:H films of higher density [6]. Therefore, we have fabricated SiNx:H films with various densities to confirm the correlation between the passivation characteristics and mass density, as shown in Table 1. The passivation characteristics according to the density difference were confirmed by the implied Voc values obtained by QSSPC measurements after the SiNx:H layer deposition and heat treatment process on the passive contact structure. Here, i-poly-Si is used instead of the heavily doped poly-Si to avoid the influence of field passivation by the highly doped region and the in-diffusion of the dopant into the bulk during the passivated contact manufacturing process. Fig. 1 shows the passivation characteristics according to different densities measured using QSSPC. The passivated contact structure prior to hydrogenation of SiNx:H exhibited an iVoc value of 500 mV and increased iVoc in all samples after hydrogenation, although this varied with the mass density. The SiNx:H layers with a density of 2.5 g/cm3 or higher exhibited hydrogenation characteristics of 670 mV or higher. The passivation characteristic of silicon nitride with a density of 2.05 g/ cm3 was < 600 mV. The mass density and passivation trends showed similar behavior to previously reported results [6]. However, the hydrogenation trend could not be explained by the mass density over 2.5 g/cm3. The iVoc was maintained even after removing the SiNx:H film. This means that the passivated contact structure is passivated by the hydrogen from the SiNx:H thin film. To explain this, we measured the hydrogen concentration in a passivated contact structure and the chemical bonding in silicon nitrides. Firstly, SIMS measurements were performed to confirm that the changes of passivation properties were due to hydrogen. Fig. 2a shows the SIMS data of the hydrogen content of poly-Si/thin oxide/c-Si on the heat-treated samples after the deposition of SiNx:H. The hydrogen content between the samples is different, and it is confirmed that hydrogen in particular is abundant in the oxide/Si interface. Hydrogen can passivate pin-holes at the SiOx/Si interface [14–16]. The relationship between the concentration of hydrogen at the interface and the implied Voc is shown in Fig. 2b, which indicates high passivation properties with high hydrogen content. The
2.1. Sample preparation A passivated contact structure was fabricated on both sides of low resistivity (0.5–3 Ω·cm), n-type (100)-oriented, 200 um, Czochralski (Cz) monocrystalline Si substrate wafer. Firstly, the wafer was subjected to saw-damage etching with potassium hydroxide solution, followed by RCA cleaning. Thereafter, a thin oxide layer was deposited on both sides of the silicon wafer by dipping the wafer in a 30% hydrogen peroxide solution at 80 °C for 10 min, and then the wafer underwent an annealing process at 600 °C under N2 atmosphere to stabilize the thin oxide layer [16]. Then, intrinsic polycrystalline silicon (i-poly-Si) was deposited on both sides of the silicon wafer using a low-pressure chemical vapor deposition process. After the i-poly-Si deposition, the wafer was annealed at 900 °C for 60 min under N2 atmosphere. The SiNx:H films were grown by PECVD at a temperature of ~380 °C. The gases used for PECVD were SiH4, NH3, N2, and H2. The gas ratio, deposition pressure, and RF power were varied to create silicon nitrides with various mass densities. The hydrogenation performance of SiNx:H was measured after it underwent annealing in a rapid thermal processing (RTP) chamber at 600 °C for 15 min at a heating rate of 5 °C/min, under ambient N2 pressure. The temperature of the samples was monitored using a thermocouple placed on the front side. 2.2. Characterization The implied voltage (iVoc) values were obtained from the QSSPC measurements in order to compare the passivation characteristics of the annealed samples. X-ray reflectivity (XRR), Rutherford backscattering spectrometry-elastic recoil detection (RBS-ERD), and secondary ion mass spectrometry (SIMS) measurements were performed to analyze the mass density, composition of the films, and the hydrogen concentration on the tunnel oxide, respectively. The SieH and NeH bonds of SiNx:H were detected by FT-IR spectroscopy and their concentrations were deduced from the stretching mode absorbance peak areas using the IR-scattering cross-sections of Lanford and Rand [5], as follows: (1)
CX − Y = A/(σX − Y × t) −3
where CX-Y, A, and t are the concentration of the XeY bond in cm , the area of the XeY bond stretching mode absorption peak, and the thickness of the film, respectively. σSi−H and σN−H are the absorption cross-sections of the bonds, and the values for these are 7.4 × 10−18 cm2 and 5.3 × 10−18 cm2, respectively. 3. Results and discussion In the literature, the mass density of the SiNx:H film was reported as one of the factors influencing the probability of hydrogen release by the film after the heat treatment. Higher hydrogenation desorption properties have been reported for low mass density SiNx:H films [5]. On the other hand, multicrystalline silicon solar cells showed better
Fig. 1. Implied Voc of the passivated contact samples hydrogenated using the SiNx:H as functions of the mass density of the as-deposited SiNx:H films. 110
Thin Solid Films 675 (2019) 109–114
J.E. Kim, et al.
Fig. 2. (a) The H concentration profile of the passivated contact samples hydrogenated using the SiNx:H films. The arrow indicates the position of the oxide layer in passivated contact structure (b) Correlation between H concentration in the oxide layer and passivation characteristics.
passivated contact structure is passivated by hydrogen diffused from SiNx:H, and the passivation characteristics cannot be predicted only by the mass density of the SiNx:H film. Next, the chemical bonding of as-deposited SiNx:H with a mass density of > 2.5 g/cm3 was analyzed by FT-IR measurements as shown in Fig. 3a. Two ranges of wavenumbers from 2005 to 2220 cm−1 and from 3300 to 3460 cm−1 imply SieH stretching bonds and NeH stretching bonds, respectively. The most cited hydrogen desorption reactions from SiNx:H films are shown in Eq. (2) and Eq. (3) [4,5].
Si▬H + Si▬H ↔ Si▬Si + H2
(2)
Si▬H + N▬H ↔ Si▬N + H2
(3)
Table 2 Detailed bonding structures consisting of SieH and NeH stretching bands and their respective peak positions. Stretching band
Bonding structure
Wavenumbers(cm−1)
SieH
Si3SieH Si2SieH2 NSi2SieH NSiSieH2, N2SiSieH N2SieH2 N3SieH Si2NeH⋯H Si2NeH SiNeH2
2005 2065 2082 2140 2175 2220 3300 3350 3460
NeH
It is expected that the frequency and occurrence of reactions in Eq. (2) and Eq. (3) will be different depending on the distribution of the chemical bonds in silicon nitrides. Previous studies have insisted that the passivation quality after the firing process depended strongly on the SieH bonding density [20]. In order to compare the passivation characteristics according to the bonding ratios, the concentration of SieH and NeH bonding in the SiNx:H thin film was calculated from the FT-IR results using Eq. (1). As shown in Fig. 3b, the SiNx:H film with high SieH ratio showed high iVoc characteristics. The chemical bonding peaks differed not only in concentration but also in peak position. The results of FT-IR can be interpreted according to Table 2 [20,21]. SiNx:H films have their peak position near 2175–2220 cm−1 and 3300–3350 cm−1 wavenumbers. In addition, N2SieH2 bonding and N3SieH bonding are the predominant SieH bonding structures in SiNx:H films. At the same time, in the case of NeH stretching band, Si2NeH⋯H bonding and Si2N-H bonding are the predominant NeH bonding structures in SiNx:H films. Deconvolution was performed to obtain a more accurate analysis of the internal bonding. Figs. 4 and 5 show the results of the SieH and NeH bonding deconvolution, respectively, according to the SiNx:H films A to E. We introduced a factor R(SieH) and R(NeH) to explain the amount of N2SieH2 bonding and
Si2N-H⋯H bonding in the SiNx:H films. R(SieH) means the ratio of AN2Si−H2 to the sum of AN2Si−H2 and AN3Si−H. R(NeH) is the ratio of ASi2N−H⋯H to the sum of ASi2N−H⋯H and ASi2N−H. From Fig. 6, it is confirmed that as the number of N2SieH2 bonds in SiNx:H increased, the iVoc improved. For NeH bonding, the measured values were deconvoluted in the same manner as for SieH. In the case of NeH bonding, the R factor was not significantly different, varying from 0.43 to 0.50, between the samples. Therefore, it is confirmed that higher passivation characteristics can be obtained when SiNx:H thin films with a higher SeH bond concentration and dominant N2SieH2 bonds in the sample are used. This is consistent with the results of Mäckel, H. et al. that the formation of = Si-H2 is important for a high-quality silicon surface passivation [20]. Finally, the change of the hydrogen content in SiNx:H was qualitatively confirmed by the FT-IR values before and after its heat treatment. The atomic concentration of [H] has been deduced by the sum of SieH and NeH bonding concentrations. In Fig. 7, the highest degree of change was observed for silicon nitrides with high R(SieH) values for hydrogenation. This supports the mechanism for hydrogen diffusion of
Fig. 3. (a) FT-IR absorption spectra of as-deposited SiNx:H films (b) Implied Voc of the passivated contact samples hydrogenated using the SiNx:H as functions of the SieH bonding ratio in as-deposited SiNx:H films, which are calculated from the FT-IR results (a). 111
Thin Solid Films 675 (2019) 109–114
J.E. Kim, et al.
Fig. 4. Deconvolution of SieH absorption peaks using N2SieH2 and N3Si-H bonding structure peaks.
Fig. 5. Deconvolution of NeH absorption peaks using Si2N-H⋯H and Si2N-H bonding structure peaks. 112
Thin Solid Films 675 (2019) 109–114
J.E. Kim, et al.
Fig. 6. (a) Correlation between the ratio of N2SieH2 in the SieH chemical bond peak and the implied Voc (b) Correlation between the ratio of Si2N-H⋯H in the NeH chemical bond peak and the implied Voc.
References [1] D. Benoit, J. Regolini, P. Morin, Hydrogen desorption and diffusion in PECVD silicon nitride. Application to passivation of CMOS active pixel sensors, Microelectron. Eng. 84 (2007) 2169–2172, https://doi.org/10.1016/j.mee.2007. 04.071. [2] B. Sopori, Y. Zhang, R. Reedy, K. Jones, Y. Yanfa, M. Al-Jassim, B. Bathey, J. Kalejs, A comprehensive model of hydrogen transport into a solar cell during silicon nitride processing for fire-through metallization, Proceedings of the 31st IEEE Photovoltaic Specialists Conference, 2005, pp. 1039–1042, , https://doi.org/10.1109/PVSC. 2005.1488311. [3] J. Robertson, Defects and hydrogen in amorphous silicon nitride, Philos. Mag. B 69 (1994) 307–326, https://doi.org/10.1080/01418639408240111. [4] C. Boehme, G. Lucovsky, H loss mechanism during anneal of silicon nitride: chemical dissociation, J. Appl. Phys. 88 (2000) 6055–6059, https://doi.org/10.1063/ 1.1321730. [5] D. Benoit, P. Morin, J. Regolini, Determination of silicon nitride film chemical composition to study hydrogen desorption mechanisms, Thin Solid Films 519 (2011) 6550–6553, https://doi.org/10.1016/j.tsf.2011.04.130. [6] J. Hong, W.M.M. Kessels, W.J. Soppe, A.W. Weeber, W.M. Arnoldbik, M.C.M. van de Sanden, Influence of the high-temperature "firing" step on high-rate plasma deposited silicon nitride films used as bulk passivating antireflection coatings on silicon solar cells, J. Vac. Sci. Technol. B 21 (2003) 2123–2132, https://doi.org/10. 1116/1.1609481. [7] F. Chen, I. Romijn, A. Weeber, J. Tan, B. Hallam, J. Cotter, Relationship between PECVD silicon nitride film composition and surface and edge passivation, Proceedings of the 22nd European Photovoltaic Solar Energy Conference and Exhibition, 2007, pp. 1053–1060. [8] V. Verlaan, C.H.M. Van der Werf, Z.S. Houweling, I.G. Romijn, A.W. Weeber, H.F.W. Dekkers, H.D. Goldbach, R.E.I. Schropp, Multi-crystalline Si solar cells with very fast deposited (180 nm/min) passivating hot-wire CVD silicon nitride as antireflection coating, Prog. Photovolt. 15 (2007) 563–573, https://doi.org/10.1002/ pip.760. [9] K. Yoshikawa, H. Kawasaki, W. Yoshida, T. Irie, K. Konishi, K. Nakano, T. Uto, D. Adachi, M. Kanematsu, H. Uzu, K. Yamamoto, Silicon heterojunction solar cell with interdigitated back contacts for a photoconversion efficiency over 26%, Nat. Energy 2 (2017) 17032, https://doi.org/10.1038/nenergy.2017.32. [10] F. Haase, C. Hollemann, S. Schäfer, A. Merkle, M. Rienäcker, J. Krügener, R. Brendel, R. Peibst, Laser contact openings for local poly-Si-metal contacts enabling 26.1%-efficient POLO-IBC solar cells, Sol. Energy Mater. Sol. Cells 186 (2018) 184–193, https://doi.org/10.1016/j.solmat.2018.06.020. [11] A. Richter, J. Benick, F. Feldmann, A. Fell, M. Hermle, S.W. Glunz, n-Type Si solar cells with passivating electron contact: identifying sources for efficiency limitations by wafer thickness and resistivity variation, Sol. Energy Mater. Sol. Cells 173 (2017) 96–105, https://doi.org/10.1016/j.solmat.2017.05.042. [12] J. Melskens, B.W.H. van de Loo, B. Macco, L.E. Black, S. Smit, W.M.M. Kessels, Passivating contacts for crystalline silicon solar cells: from concepts and materials to prospects, IEEE J. Photovolt. 8 (2018) 373–388, https://doi.org/10.1109/ JPHOTOV.2018.2797106. [13] F. Feldmann, M. Simon, M. Bivour, C. Reichel, M. Hermle, S.W. Glunz, Efficient carrier-selective p- and n-contacts for Si solar cells, Sol. Energy Mater. Sol. Cells 131 (2014) 100–104, https://doi.org/10.1016/j.solmat.2014.05.039. [14] G. Yang, A. Ingenito, N. van Hameren, O. Isabella, M. Zeman, Design and application of ion-implanted polySi passivating contacts for interdigitated back contact c-Si solar cells, Appl. Phys. Lett. 108 (2016) 033903, https://doi.org/10.1063/1. 4940364. [15] N.H. Nickel, N.M. Johnson, W.B. Jackson, Hydrogen passivation of grain boundary defects in polycrystalline silicon thin films, Appl. Phys. Lett. 62 (1993) 3285–3287, https://doi.org/10.1063/1.109101. [16] H. Kim, S. Bae, K.S. Ji, S.M. Kim, J.W. Yang, C.H. Lee, K.D. Lee, S. Kim, Y. Kang, H.S. Lee, D. Kim, Passivation properties of tunnel oxide layer in passivated contact silicon solar cells, Appl. Surf. Sci. 409 (2017) 140–148, https://doi.org/10.1016/j. apsusc.2017.02.195.
Fig. 7. Change of hydrogen content in SiNx:H thin films with different SieH bonding ratio before and after annealing.
SiNx:H by Eq. (2) and Eq. (3). 4. Conclusions In this study, we conducted hydrogenation of the passivated contact structure using PECVD SiNx:H as a hydrogen source and observed the correlation between SiNx:H mass density, chemical bonding of the film, and the passivation results. Passivated contact structure showed improved passivation characteristics after hydrogenation using the SiNx:H thin films. After hydrogenation, the hydrogen profile shows an accumulated distribution at the SiOx/Si interface, which is inferred to have resulted in an improved interface passivation. In addition, the SiNx:H films with more SieH bonding showed higher iVoc even though they had a similar mass density. From this study, we conclude that SiNx:H can be used as a good hydrogen source for the passivated contact structure. In addition, the passivation characteristics cannot be predicted only by the mass density of the SiNx:H film, and the chemical bonding ratio of SieH in the thin film is also important. In particular, N2SieH2 bonding in SieH contributes to hydrogen passivation. Acknowledgements This work was supported by the Korea Institute of Energy Technology Evaluation and Planning (KETEP) and the Ministry of Trade, Industy & Energy (MOTIE) of the Republic of Korea (No. 20163010012430 and 20163030014020). This work was also supported by the "Human Resources Program in Energy Technology" of the Korea Institute of Energy Technology Evaluation and Planning (KETEP), with financial support from the Ministy of Trade, Industry and Energy, Republic of Korea (No. 20154030200760). 113
Thin Solid Films 675 (2019) 109–114
J.E. Kim, et al.
the composition of hydrogenated silicon nitride SiNx:H for efficient surface and bulk passivation of silicon, Sol. Energy Mater. Sol. Cells 93 (2009) 1281–1289, https://doi.org/10.1016/j.solmat.2009.01.023. [20] H. Mäckel, R. Lüdemann, Detailed study of the composition of hydrogenated SiNx layers for high-quality silicon surface passivation, J. Appl. Phys. 92 (2002) 2602–2609, https://doi.org/10.1063/1.1495529. [21] F. Ay, A. Aydinli, Comparative investigation of hydrogen bonding in silicon based PECVD grown dielectrics for optical waveguides, Opt. Mater. 26 (2004) 33–46, https://doi.org/10.1016/j.optmat.2003.12.004.
[17] F. Giorgis, F. Giuliani, C.F. Pirri, E. Tresso, C. Summonte, R. Rizzoli, R. Galloni, A. Desalvo, P. Rava, Optical, structural and electrical properties of device-quality hydrogenated amorphous silicon-nitrogen films deposited by plasma-enhanced chemical vapour deposition, Philos. Mag. B 77 (1998) 925–944, https://doi.org/10. 1080/13642819808206395. [18] F. Duerinckx, J. Szlufcik, Defect passivation of industrial multicrystalline solar cells based on PECVD silicon nitride, Sol. Energy Mater. Sol. Cells 72 (2002) 231–246, https://doi.org/10.1016/S0927-0248(01)00170-2. [19] J.F. Lelièvre, E. Fourmond, A. Kaminski, O. Palais, D. Ballutaud, M. Lemiti, Study of
114