The 25th International Conference on Amorphous and Nano-crystalline Semiconductors
August 18–23, 2013 Toronto, Ontario Canada
a-Si:H/c-Si Interface I
Chair: Rene van Swaaij, Delft University of Technology
Fr-B1.1 (invited) 8:10–8:40
Atomic Structure of Interface States in a-Si:H / c-Si Heterojuction Solar Cells
1. Institut für Silizium-Photovoltaik, Helmholtz-Zentrum Berlin für Materialien und Energie, Kekuléstraße 5, D-12489 Berlin, Germany
2. Fachbereich Physik, Freie Universität Berlin, Arnimallee 14, D-14195 Berlin, Germany
3. Lehrstuhl für Theoretische Physik, Universität Paderborn, Warburger Straße 100, D-33098 Paderborn, Germany
In an hydrogenated amorphous/crystalline silicon (a-Si:H/c-Si) heterojunction solar cell passivation of the c-Si surfaces is realized with an intrinsic and/or doped thin (3–10 nm) amorphous Silicon (a-Si:H) layer. The structure-function relationship of this technologically very important interface is still in dispute. Open questions concern (i) the morphology of the interface region under nm-thin a-Si:H layers, (ii) the electronic structure of the heterointerface and (iii) of its native interface defects as well as (iv) the prevailing charge carrier transport and recombination mechanisms at the p/n junction. If interface states are paramagnetic the tool of choice for structural characterization is electrically detected magnetic resonance (EDMR). EDMR detects paramagnetic states through their typical electron spin resonance (ESR) signature in the photocurrent of the solar cell. From the EDMR spectra the different states that participate in transport and recombination can be distinguished through their characteristic g tensors. However, to extract the microscopic structure from the magnetic interaction parameters (g tensor, hyperfine (hf) splitting), advanced computational methods such as density functional theory (DFT) are mandatory which are capable of reproducing g tensor and hf parameters. So far, however, such an analysis has been restricted to localized defects. Extended or weakly localized states such as band tail states in a-Si:H have not been tackled in this context yet.
In this paper we will show that through their specific anisotropy two different dangling bond defects can be identified at the -oriented Si interface. Their g tensors are very similar to the well-known Pb defects known from the SiO2/c-Si interface. In addition an isotropic EDMR signal at g = 2.004 is identified. By a DFT analysis of the magnetic Berry phases we reproduce well the measured g tensors of the two types of dangling bond defects which are shown to exist at the boundary between the two phases of silicon. In addition we calculate a nearly isotropic signal with g = 2.004 which is connected to a weakly localized states in the conduction band of the a-Si:H tissue, located only within a few atomic distances from the c-Si interface. From the data analysis we arrive at the conclusion that this delocalized state in the a-Si:H tissue is populated through a tunneling-like transition from the c-Si phase even at very low temperatures. After its localization in the a-Si:H tissue this state induces electron-hole recombination with the dangling bond-type defects directly located at the interface of the crystalline phase.
In summary, by combining high-resolution EDMR measurements on real devices with state-of-the-art DFT calculations we were able to locate device-limiting paramagnetic states in the a-Si:H/c-Si interface structure, assign their orbital configuration and extension into the Si tissue, and locate these states in the solar cell band structure. Our work links the structure and function of the interface to the orbital symmetry of the recombination active defects. In particular, the important role of extended states in noncrystalline materials is shown.
Keywords: defects, interface, ESR
Interface Defect Monitoring using Surface Photovoltage Spectroscopy in Amorphous/Crystalline Silicon Heterojunction Solar Cell
State Key Laboratory on Integrated Optoelectronics, Institute of Semiconductors, Chinese Academy of Sciences, Beijing, China
We proposed that the surface photovoltage spectroscopy (SPS) can be used to characterize interface states between hydrogenated amorphous and crystalline silicon (a-Si:H/c-Si). A thin a-Si:H layer was deposited on three kinds of substrate (n-type c-Si, stainless steel and TCO/Ag/glass). The three kinds of samples were analyzed by SPS. We found that only the SPS of a-Si:H/c-Si structure exhibits a peak in 1.2 eV. That implies that the SPS signal is not caused by amorphous silicon film but can be ascribed to the interface states of a-Si:H/c-Si. Using hydrogen plasma treatment (HPT) on c-Si surface we observed a systematic variation in the SPS signal upon HPT time, which was associated to the change of interface state density. In the meantime, the measurement about minority carrier life of samples with the different HPT time proved the improvement of the interface passivation in amorphous/crystalline silicon. Our results show that the SPS can be used to optimize the HIT solar cell preparation process.
Keywords: solar cells, surface photovoltage, interface states, heterojunction, minority carrier life
Amorphous/Crystalline Silicon Interfaces: Correlation between Infrared Spectroscopy and Electronic Passivation Properties
1. École Polytechnique Fédérale de Lausanne (EPFL), Institute of Microengineering (IMT), Photovoltaics and Thin Film Electronics Laboratory, Rue A.-L. Breguet 2, CH-2000 Neuchâtel, Switzerland
2. Institute of Physics ASCR, v. v. i., Cukrovarnická 10, 162 00 Prague, Czech Republic
The ultrathin layers of hydrogenated amorphous silicon passivating surface of crystalline silicon have recently gained a lot of interest, due to their application in high-efficiency silicon heterojunction solar cells. The full understanding of the device is however not yet complete. Currently several fundamental questions about the microstructure of good passivation layer are in the process of being answered. While in the case of thin-film p-i-n devices a high material density and low concentration of voids is commonly accepted to be beneficial, in the case of thin passivation layer of heterojunction solar cell the situation is more complex. The amorphous layer should not only have a low defect density, but should also allow for hydrogen movement toward crystalline/amorphous interface to passivate crystalline silicon surface. We have previously shown that this process, activated by temperature annealing or plasma treatment can dramatically improve open circuit voltage of heterojunction solar cell up to 725 mV .
In this contribution we show that attenuated total reflectance (ATR) Fourier transform infrared (FTIR) spectroscopy can be a sensitive probe for such ultrathin layers or even sub-monolayer amounts of surface atoms and can be applied relatively easily in heterojunction solar cell research [1–3]. Combined with effective lifetime measurements of passivated silicon wafers, this offers a unique combination of chemical, morphological and transport characterization method sensitive to surface and ultrathin amorphous silicon layers.
The microstructure parameter, corresponding to concentration of voids, can be accurately calculated from the ratio of higher hydride (2100 cm–1) and monohydride (2000 cm–1) vibrational peaks. We observed that material with more voids obtained by introducing hydrogen plasma steps during deposition leads to a better passivation. We observed that the hydrogen plasma treatment affects several nanometers of the layer, changing significantly the microstructure and eventually leading to etching of the amorphous layer. The effect of the hydrogen plasma on the interface with crystalline silicon is however detrimental, but can be shielded by several nanometers thick layer of amorphous silicon.
The new requirements for passivating layers, generally different from those for thin-film devices, give a new motivation to study the light induced degradation / annealing effects. The annealing has the same activation energy as in the case of bulk amorphous silicon and follows the stretched exponential behavior indicating hydrogen movement. It restores most of the electronic defects created either by light soaking  or sputtering damage , but at the same time changes permanently the microstructure, as can be revealed by high sensitive ATR-FTIR.
 A. Descoeudres, L. Barraud, S. De Wolf, B. Strahm, D. Lachenal, C. Guérin, Z. C. Holman, F. Zicarelli, B. Demaurex, J. Seif, J. Holovsky, and C. Ballif, Applied Physics Letters, vol. 99, no. 12, p. 123506, 2011
 B. Demaurex, S. De Wolf, A. Descoeudres, Z. Charles Holman, and C. Ballif, Applied Physics Letters, vol. 101, no. 17, p. 171604, 2012
 E. El Mhamdi, J. Holovsky, S. De Wolf, B. Demaurex, and C. Ballif, presented at the Silicon PV, Hamelin, Germany, 2013
Keywords: vibrational spectroscopy, heterojunction, passivation, stability
Facile Grown Oxide Based Passivation for Silicon Heterojunction PV Cells
1. Department of Electrical and Computer Engineering, University of Toronto, Toronto, Ontario, Canada
2. Department of Material Science and Engineering, University of Toronto, Toronto, Ontario, Canada
High efficiency (~25%) hydrogenated amorphous silicon (aSi:H) - crystalline silicon (cSi) heterojunction photovoltaic (PV) devices have been fabricated using all low temperature processing. The low thermal budget based fabrication process is deemed suitable for ultra-thin silicon wafers which would further reduce the material cost and potentially make this technology a subsidy free energy resource. Nevertheless, the passivating hydrogenated amorphous silicon layers deposited on cSi surfaces are a source of parasitic optical absorption loss and do exhibit light induced degradation.
We have recently developed alternative low temperature passivation and fabrication schemes based on facile grown oxide and PECVD silicon nitride dual layers. Specifically, we have attained a surface recombination velocity as low as 8 cm/sec which is comparable to state-of-the-art passivation schemes used for crystalline silicon. We have successfully implemented the new passivation scheme in aSi:H-cSi heterojunction devices. A maximum cell efficiency of 16.7% has been achieved for proof-of-concept cells using untextured crystalline silicon under AM 1.5 solar irradiation.
The paper will be present a detailed study of the passivation characteristics as a function of the passivation material parameters. Also, various techniques that can be used to equivalently produce high quality facile grown oxide based passivation will be discussed. Experimental results on the stability of the passivation materials against light exposure will be reported.
Keywords: passivation, amorphous silicon, crystalline silicon, native oxide, PECVD silicon nitride
a-Si:H/c-Si Interface Degradation upon ITO Sputtering: Influence of the Doping
1. TOTAL New Energies, Paris La Défense, France
2. LPICM, CNRS, Ecole Polytechnique, Palaiseau, France
Passivation of c-Si wafers has become a crucial technological challenge within the past few years with the race for higher efficiencies of heterojunction solar cells  and the development of alternative front and rear passivation layers for homojunction solar cells. Passivation quality is directly related to the defect density at the interface between the substrate and the passivating layer.
In some cases, sputtering processes are needed to finalize a device as it is the case for heterojunction solar cells, for which ITO is sputtered on both sides of the sample by PVD (Physical Vapor Deposition) to lower the reflection and enhance the carrier collection. It has been shown that such sputtering processes can greatly degrade the cell minority carrier lifetime, and this degradation can apparently be recovered by thermal annealing [2–4]. Despite these results, little is known about the mechanisms leading to such degradation.
In this work, we address the issue of ITO sputtering on (p)a-Si:H passivating layers and its impact on the sample carrier effective lifetime. We find that ITO magnetron sputtering deteriorates the minority carrier lifetime of symmetrical (p)a-Si:H/(i)a-Si:H/(n)c-Si/(i)a-Si:H/(p)a-Si:H samples (p+i/i+p stacks) more significantly than in other structures such as (n)a-Si:H/(i)a-Si:H/(n)c-Si/(i)a-Si:H/(n)a-Si:H (n+i/i+n stacks) or (i)a-Si:H/(i)a-Si:H/(n)c-Si/(i)a-Si:H/(i)a-Si:H (i/i stacks). For n+i/i+n stacks, we measured close to 30% decrease in lifetime, going from 2866 μs down to 1973 μs after ITO deposition, whether it be at 180°C or RT. Despite this degradation, several steps of thermal annealing under ambient atmosphere led us back to the as-deposited lifetime value. Moreover, in the case of p+i/i+p stacks, the lifetime value drastically dropped by 66%. The same annealing steps would lead us to a lifetime lower by 39% from the initial value. In other words, in the case of a (p)a-Si:H layer, the original lifetime could not be recovered by thermal annealing. Therefore, the (p)a-Si:H layer is less resilient to ITO sputtering than an (n) or (i)a-Si:H layer and is responsible for the major part of the lifetime drop in a cell structure (p+i/i+n stack). We think that (p)a-Si:H is more prone to bond-breaking and atom displacement upon atomic impacts. Indeed, upon sputtering, the ion energy seems to be more easily transferred from the (p)a-Si:H surface layer down to the (i)a-Si:H layer and to the interface with the crystalline wafer eventually, leading to the creation of defects at the interface and therefore to more significant effective lifetime drops. This lower resilience of the (p)a-Si:H layer was confirmed by experiments on cell structures where the ITO deposited on the (n)-side led to a decrease of 5% only whereas the same deposition process on the (p)-side would further degrade our passivation by 42% leading to a 45% overall lifetime degradation compared to the initial carrier effective lifetime prior ITO deposition. Microwave PCD allowed for local estimations of the lifetime drop, confirming that damages within the material occurred on the regions exposed to the ITO sputtering process.
For the experiments, 4" (n)-type (FZ)-Si double-side polished and textured wafers, with a resistivity ranging from 2.6 to 20 Ωcm were used. Intrinsic and (p)-doped a-Si:H layers were deposited in a single-chamber 13.56 MHz RF-PECVD reactor at low temperature (200°C) whereas ITO was deposited in a RF-magnetron PVD reactor. Effective lifetime measurements were made on a Sinton Consulting WCT-120 equipment. More localized lifetime analyses were carried out on a microwave PCD setup with the microwave excitation (@ 30 GHz) on one side and the monochromatic light pulse excitation (@ 532 nm) on the other.
 T. Kinoshita, D. Fujishima, A. Yano, A. Ogane, S. Tohoda, K. Matsuyama, Y. Nakamura, N. Tokuoka, H. Kanno, H. Sakata, M. Taguchi, and E. Maruyama, "The approaches for high efficiency HIT solar cell with very thin (<100 μm) silicon wafer over 23%," 26th EUPVSC Proceeding, pp. 871–874, 2011
 B. Demaurex, S. De Wolf, A. Descoeudres, Z. C. Holman, and C. Ballif, "Damage at hydrogenated amorphous/crystalline silicon interfaces by indium-tin oxide overlayer sputtering," Appl. Phys. Lett., vol. 101, p. 171604, 2012
 A. Illiberi, P. Kudlacek, A. H. M. Smets, M. Creatore, and M. C. M. van de Sanden, "Effect of ion bombardment on the a-Si:H based surface passivation of c-Si surfaces," Appl. Phys. Lett., vol. 98, p. 242115, 2011
 D. Zhang, A. Tavakoliyaraki, Y. Wu, R. van Swaaij, and M. Zeman, "Influence of ITO deposition and post annealing on HIT solar cell structures," Energy Procedia, vol. 8, pp. 207–213, 2011
Keywords: passivation, degradation, a-Si:H, ITO, sputtering