The 25th International Conference on Amorphous and Nano-crystalline Semiconductors
August 18–23, 2013 Toronto, Ontario Canada
a-Si: Electronic Structure, Defects, Metastability I
Chair: Rana Biswas, Iowa State University
Evidence of Two-step Processes in the Structural Relaxation of Amorphous Silicon
1. Department of Electronic Materials Engineering, Research School of Physics and Engineering, Australian National University, Canberra, 0200, Australia
2. Lehrstuhl für Experimentalphysik IV, Universität Augsburg, 86135, Germany
Studies of (pure) ion-implanted amorphous silicon (a-Si) show that, during thermal annealing, the Si-Si network undergoes extensive restructuring toward a state of lowest free energy, a process called structural relaxation. However, despite this being the focus of much research, the mechanism(s) that govern the kinetics of the structural relaxation still remain unclear. The work presented here addresses this issue.
Annealing of a-Si leads to a change in its mechanical, vibrational, and electrical properties, whereby these properties are sensitive to the degree of relaxation. The effect of annealing temperatures up to 550°C on the mechanical behavior of a-Si was studied using nanoindentation by measuring the probability of a pressure-induced phase transition. The changes in the covalent Si-Si network were further tracked using Raman microspectroscopy and electrical measurements.
The mechanical behavior of a-Si is very sensitive to the "state" (relaxed or unrelaxed) of the amorphous network, whereby unrelaxed (as-implanted) a-Si deforms plastically via flow and relaxed (thermally annealed) a-Si via a pressure-induced phase transformation. The probability of phase transformations is found to change significantly in a very narrow temperature range from 300°C to 350°C during thermal annealing. In contrast, the reduction in the average tetrahedral bond-angle distortion (Δθb) from 10.8° to 9.4° as observed by Raman microspectroscopy occurs at slightly higher temperatures, namely over a temperature range centered at 370°C.
Previous experiments have shown that the electrical conduction in amorphous semiconductors occurs via variable-range hopping of the charge carriers in localized states, e.g. dangling bonds, near the Fermi level. This is characterized by Mott's variable-range hopping relation. Thus, conductivity measurements in a-Si as a function of annealing temperature provide additional information on the evolution of the density of the dangling bonds and the role of defect annihilation during structural relaxation. The conductivity, and hence the density of dangling bonds, is found to progressively decrease from 1021 eV–1 cm–3 to 1019 eV–1 cm–3 with annealing temperature. A major decrease in the density of the dangling bonds was observed around an annealing temperature of 250°C, significantly lower than the temperatures where a reduction in Δθb or the probability for phase transformations occurs.
This strongly suggests that structural relaxation is at least a two-step process involving first a reduction in defects and dangling bond density followed by a reduction in the average tetrahedral bond-angle distortion as second step.
Keywords: amorphous silicon, structural relaxation, defects
Improved Modulated Photocarrier Grating Technique and Determination of the Density of Acceptor States in Hydrogenated Amorphous Silicon
1. Laboratoire de Génie Electrique de Paris (CNRS UMR 8507), Plateau de Moulon, 11 rue Joliot Curie, 91190 Gif sur Yvette, France
2. INTEC (UNL-CONICET), Güemes 3450, 3000 Santa Fe, Argentina, and FIQ (UNL), Santiago del Estero 2829, 3000 Santa Fe, Argentina
In this communication we present a new version of the Modulated Photocarrier Grating (MPG) technique, originally proposed by Hattori et al. in 1993 . Moreover, we will show measurements of the density of states (DOS) in the conduction band tail of hydrogenated amorphous silicon thin films, using the proposal of Schmidt et al.  which uses a combination of MPG and modulated photoconductivity (MPC) measurements.
The MPG technique emerged as an evolution of the steady-state photocarrier grating (SSPG) technique, and has been used to estimate the ambipolar diffusion length of amorphous silicon samples. Another attractive fact about MPG is that it can test the hypothesis of ambipolar transport, which is often used without knowing if that condition is actually held. These two important facts make MPG an interesting technique to characterize amorphous semiconductors.
In this work we analyze the original setup proposed by Hattori et al. , also used by Morgado et al. . We have found some issues regarding the rotation of the linear polarization of the light needed to establish the modulating intensity pattern required in MPG. One of the requisites for this pattern is to change the contrast of the interference while keeping at fixed positions the nodes and antinodes of the light intensity. We have found that using the original setup proposed in  may lead to a movement of nodes and antinodes. To avoid this problem, we present a new experimental setup that gives the correct intensity pattern. To check this new setup we perform measurements of both MPG and MPC varying the absolute temperature T and the generation rate G0. Then we compare these measurements with numerical simulations to check all the observed behaviors.
Having checked this new setup, the second part of the present communication is related to the expression proposed by Schmidt et al.  to estimate the DOS in the forbidden gap, using measurements of both MPG (ΔjMPG) and MPC (ΔjMPC) techniques under the same experimental conditions.
where N(Efn) is the DOS at the electrons quasi-Fermi level, μn is the electron mobility in extended states, q the electron charge, cn the capture coefficient for electrons, kbT the thermal energy and σ0 the steady-state photoconductivity. We show how the approximations used to derive Eq. (1) put restrictions on the frequency range and on the dc photoconductivity that can be used in the experiment. Finally, we show estimations of the conduction band tail DOS using Eq. (1).
 K. Hattori, Y. Koji, S. Fukuda, W. Ma, H. Okamoto, and Y. Hamakawa, Journal of Applied Physics 73, 3846 (1993)
 J. A. Schmidt, N. Budini, F. Ventosinos, and C. Longeaud, Physica Status Solidi (A) 207, 556 (2010)
 E. Morgado, J. Díez, R. Schwarz, A. Maçarico, and S. Koynov, Journal of Non-Crystalline Solids, 266 290 (2000)
Keywords: amorphous silicon, photoconductivity, density of states
Correlation between Preparation Condition and Recombination Rates at Radiative Defects in a-Si:H
1. Department of Applied Science, Yamaguchi University, Ube 755-8611, Japan
2. Department of Electrical-System Engineering, Hiroshima Institute of Technology, Miyake, Saeki-ku, Hiroshima 731-5193, Japan
3. Present address: C-305, 2-12 Wakabadai, Inagi, Tokyo 206-0824, Japan
Light-induced creation of defects in a-Si:H, which is known as the Staebler-Wronski effect, has been received much attention in physics and application of amorphous semiconductors. The dominant defects in a-Si:H are the dangling bonds of silicon which act as non-radiative recombination centres . The observation of defect photo-luminescence (PL) and its increase in intensity after illumination  indicate the existence of defects which act as radiative recombination centres and their light-induced creation. The study of the radiative defects is also important to understand the light-induced phenomena in a-Si:H.
We have recently reported the temperature variation of recombination rates at radiative defects . The increase of the radiative recombination rate with increasing temperature has been attributed to thermal excitation of holes from deep tail states to shallow tail states. We have also reported  that the temperature variations of the non-radiative recombination rates are fitted by the theoretical calculation by Englman and Jortner  for the case of strong electron-phonon coupling.
The comparison of the results obtained for various a-Si:H films including those after illumination has suggested that the temperature dependences of the radiative and non-radiative recombination rates are possibly affected by the amorphous network and the nature of native and photo-created defects. The amorphous network is affected by the preparation condition, especially, the substrate temperature. In order to clarify how the nature of the amorphous network affects the results, we have to compare the results for the a-Si:H films as grown where the influence of photo-created defects is neglected. In our previous studies, such results have been obtained only for an a-Si:H film of high defect density prepared at a substrate temperature of 150°C. In this paper, the results in the a-Si:H of lower defect density as grown, prepared at a substrate temperature higher than 180°C, are presented. We discuss how the preparation condition affects the temperature variation of recombination rates.
 K. Morigaki, Physics of Amorphous Semiconductors, World Scientific Publ. Singapore, 1999
 J. I. Pankove and J. E. Berkeyheiser, Appl. Phys. Lett., 37, 705 (1980)
 C. Ogihara, Y. Inagaki, A. Taketa, and K. Morigaki, J. of Non-Cryst. Solids, 358, 2004 (2012)
 R. Englman and J. Jortner, Mol. Phys., 18, 145 (1970)
Keywords: amorphous silicon, defects, photoluminescence, recombination, lifetime
The Network Environment of Light Induced Defects in Hydrogenated Amorphous Silicon Revealed: The Role of Hydrogenated Volume Deficiencies
Photovoltaic Materials and Devices, Delft University of Technology, Mekelweg 4, 2628 CD Delft, The Netherlands
In state-of-the-art thin film Si (TF Si) double- and triple-junctions solar cells, the junctions based on hydrogenated amorphous silicon (a-Si:H) and/or silicon-germanium (a-SiGe:H) contribute 50%–75% to the total power generation of the multi-junctions. Consequently, stable efficiencies above 15% for TF Si based multi-junctions can only be achieved by tackling the meta-stability of its amorphous junctions. Progress in understanding of the nature and mechanisms of the light induced defects (LIDs) is a very slow process. The main reasons for this slow progress are i) the lack of tools to define the nanostructure of the various types of a-Si:H due to its complex nature of the network, ii) the nature of the dominant native and meta-stable defects and their local environments are not yet identified, iii) the overall lack of empirical correlations between the amorphous material nanostructure, defect entities and performance in solar cells.
In this contribution we present the recent breakthroughs made in tackling the issues addressed above. For the first time ever, clear experimental relations have been demonstrated between 1) various nanostructures of a-Si:H, 2) the degradation of the performance of a-Si:H solar cells under light soaking, and 3) the measured increase in defect densities in the a-Si:H absorber in solar cells. These results indisputably show that the "fast" LIDs created in the first (most important) 10 hours of light soaking reside at hydrogenated volume deficiencies such as the hydrogenated voids and are not related to bulk-like coordination defects. The additional "slow" LIDs created in the time frame of 10 up to 1000 hours of light soaking (in which the LID-generation saturates), are on the contrary independent of the nanostructure. The "fast" LIDs can be thermally annealed out at relative low temperatures (<130°C), whereas significantly higher temperatures (>170°C) are required to anneal out the "slow" LIDs.
The progress has been established using systematic and detailed experimental studies using state-of-the-art diagnostic tools. First, the nanostructure of the a-Si:H is proposed to be described by the size distribution of hydrogenated volume deficiencies (ranging from divacancies, multi-vacancies up to nano-sized voids) using both Doppler Broadening Positron Annihilation Spectroscopy (DBPAS) and Fourier Transform Infrared absorption Spectroscopy (FTIR). Secondly, a newly developed processing window using higher pressure, allows us to process "device grade" a-Si:H with a wide variety of size distributions of the volume deficiencies. This set of high quality a-Si:H cells allows us to build up a clear statistics on the relation between nanostructure and metastability. Thirdly, a novel analyses tool based on Fourier Transform Photoconductivity Spectroscopy (FTPS) has been used to measure in-situ accurately the increase of sub gap absorption related to LIDs in the a-Si:H absorber of the solar cell. Finally, the results are related to the degradation of the external parameters and external-quantum-efficiencies of the solar cells studied.
The impact of the above described new results on the understanding of the nature of LIDs in a-Si:H based solar cells will be discussed and processing routes to minimize the metastability effect will be presented.
Keywords: a-Si:H, nanostructure, solar cells, light induced defects, volume deficiencies
Th-C1.5 9:40–10:00 cancelled
Relation Between Hydrogen Bonding Environment and High Frequency Si–H Stretch Vibrations in Hydrogenated Amorphous Silicon