Symposium Organizers
Bhushan L. Sopori National Renewable Energy Laboratory
Bernhard Dimmler Würth Solar GmbH & Co. KG
Jeffrey Yang United Solar Ovonic LLC
Thomas Surek Surek PV Consulting
Q1: Crystalline Silicon Technologies
Session Chairs
Monday PM, November 30, 2009
Room 306 (Hynes)
9:30 AM - **Q1.1
Hydrogen Passivation for Crystalline Silicon Solar Cells.
Michael Stavola 1
1 Department of Physics, Lehigh University, Bethlehem, Pennsylvania, United States
Show AbstractThe Si substrates that are often used for the fabrication of solar cells to reduce cost give rise to defect issues that must be addressed. Hydrogen is commonly introduced into silicon solar cells to reduce the deleterious effects of defects and to increase cell efficiency [1]. A process that is used by industry to introduce hydrogen is by the post-deposition annealing of a hydrogen-rich SiNx layer that is used as an antireflection coating [2]. A number of questions about this hydrogen introduction process and hydrogen’s subsequent interactions with defects have proved difficult to address because of the low concentration of hydrogen that is introduced into the Si bulk.Fundamental studies of hydrogen-containing defects in silicon provide a foundation for addressing issues of interest to the Si solar-cell community. Strategies have been developed by which hydrogen in silicon can be detected by IR spectroscopy with high sensitivity [3,4]. The introduction of hydrogen into Si by the post-deposition annealing of a SiNx coating has been investigated to reveal hydrogen’s concentration, diffusivity, and reactions with defects. The effect of processing variations on the concentration of hydrogen that is introduced into the Si bulk has also been studied. The contributions of F. Jiang, S. Kleekajai, V. Yelundur, A. Rohatgi, L. Carnel, J. Kalejs, and G. Hahn to our studies are gratefully acknowledged. This work has been supported by the Silicon Solar Research Center SiSoC Members through NCSU Subaward No. 2008-0519-02 and NSF Grant No. DMR 0802278.[1] J. I. Hanoka, C. H. Seager, D. J. Sharp, and J. K. G. Panitz, Appl. Phys. Lett. 42,618 (1983).[2] F. Duerinckx and J. Szlufcik, Sol. Energy Mater. Sol. Cells 72, 231 (2002).[3] F. Jiang et al., Appl. Phys. Lett. 83, 931 (2003).[4] S. Kleekajai et al., J. Appl. Phys. 100, 093517 (2006).
10:00 AM - Q1.2
A New, Ultrafast Technique for Mapping Dislocation Density in Large-area, Single-crystal and Multicrystalline Si Wafers.
Bhushan Sopori 1 , Przemyslaw Rupnowski 1 , Mathew Albert 2 , Chandra Khattak 2 , Mike Seacrist 3
1 , National Renewable Energy Laboratory, Golden, Colorado, United States, 2 , GT Solar, Merrimack, New Hampshire, United States, 3 , MEMC, St. Peters, Missouri, United States
Show AbstractAverage dislocation density and spatial distribution of dislocations are routinely used as a measure of crystal quality of single- and multicrystalline Si (mc-Si) wafers. A variety of techniques have been developed to generate dislocation maps, including X-ray imaging, Cu decoration, and chemical delineation. The most common method is to defect etch the wafer with a suitable chemical etchant and then count the etch pits using an optical microscope. Commercial camera systems, with image analysis software, are available as microscope attachments that can count etch pits within the field of view and combine that information to produce maps of dislocation distribution over a wafer. An improved technique uses light scattered by etch pits to statistically count dislocations. The wafer is illuminated by a laser beam and the total scattered light, which is proportional to the number of etch pits in the illuminated region, is measured. An instrument based on this technique takes 30–60 minutes to map a 6-in x 6-in wafer.This paper describes a new technique that uses scattering from a defect-etched wafer to map dislocation distribution of the entire wafer in a single image. The measurement is very fast and compatible with large-area wafers. In this technique, the single- or multicrystalline wafer is polished to produce a damage-free polished surface. The wafer is then defect etched using Sopori etch (HF:CH3COOH:HNO3 in a 36:15:1 ratio) for 30 s to produce etch pits at dislocation sites. The shape of the etch pit depends on the direction of dislocation at the surface and does not depend on the orientation of the wafer or grain (in mc-Si). The wafer is then placed in a reflectometer where a set of lights, symmetrically placed around the wafer, illuminate it at an oblique incidence. The light scattered normal to the wafer is collected by a camera and imaged. The image corresponds to the local reflectance of the defect-etched wafer. Because local scattering is proportional to the density of etch pits, the camera image is proportional to the local variation in the dislocation density of the wafer. The system is calibrated by using a reference sample to convert the reflectance map into a dislocation map. This technique allows a fast (< 1 s) mapping of dislocations. An interesting feature of this etch is that the scattering cross-section of all dislocations (which can have circular, elliptical, or comet shapes) is the same. Thus, all dislocations are counted. An instrument based on this technique is now commercially available. We will show results that demonstrate: (i) repeatability of defect etching of large mc-Si wafers, (ii) variation of dislocation patterns over selected parts of a mc-Si ingot, (iii) a correlation between defect maps and photocurrent maps of commercial Si solar cells, and (iv) a correlation between defect distribution and the solar cell performance.This abstract is subject to government rights.
10:15 AM - Q1.3
Low-cost, High Efficiency Solar Cells on Scrapped CMOS Silicon.
Daniel Inns 1 , Joel de Souza 1 , K. Saenger 1 , H. Hovel 1 , D. Sadana 1
1 T. J. Watson Research Centre, IBM, Yorktown Heights, New York, United States
Show AbstractThe cost of scrapped Si from the CMOS industry is extremely low which makes it an attractive material for solar industry. However, the minority carrier lifetime of this material is very low and variable, typically ~ 1 µs compared to the lifetime of the original prime-Si wafer which is > 500 µs. Solar cells made on scrapped wafers therefore result in efficiencies which are inferior to that from a prime CMOS grade Si. We have developed a novel and effective low-cost metal gettering anneal process which allows the minority lifetime of the scrapped wafer to recover to close to its original value, a 100-500 fold increase. Since the efficiency of a solar cell is directly impacted by the minority carrier lifetime, cell efficiency of the improved scrapped Si is nearly equivalent to that from a prime-Si wafer. In order to erase the processing history of the wafer, surface etching is performed to remove ~ 20 µm of surface Si. Following this is a unique impurity gettering step that is performed at > 1300°C with chlorine-containing gas to enable efficient gettering of metals out of the substrate. An efficiency of ~15% has been demonstrated on both prime and improved scrapped wafers using rudimental device design to study the validity of our unique metal gettering process. This efficiency is being improved to much higher values by refinements in device design, anti-reflection coating(s) and surface passivation schemes.
10:30 AM - **Q1.4
Crystalline Silicon Technology for Solar Applications.
Aditya Deshpande 1 , Mike Seacrist 1 , Steve Kimbel 1 , Gang Shi 1 , Jihong Chen 1
1 , MEMC Electronic Materials, St Peters, Missouri, United States
Show AbstractThe use of crystalline silicon in solar applications exceeds the silicon consumed in semiconductor applications. Further, the growth rate of silicon use in solar applications has been higher than the growth rate of use in semiconductor applications over the past several years. Many similarities and synergies exist between manufacturing silicon for semiconductor and solar applications. These include producing polysilicon raw material, growing silicon crystals, and converting crystals into silicon wafers by wire slicing. For these reasons there is a strong motivation for silicon suppliers to participate in the crystalline silicon solar market. Crystalline silicon solar cells are the workhorse of the photovoltaic industry and have a significant portion of the market share of the world production of solar cells. The key driver and challenge for crystalline silicon in solar is cost which is influenced by both the silicon material cost and silicon performance. For silicon to maintain and improve on solar cell market share, further reductions in production cost as well as improvements in solar cell efficiency are necessary. The approach of a vertically integrated silicon supplier to the challenge of improving solar cell efficiency performance while also improving silicon manufacturing productivity and reducing cost will be described. A technical roadmap for crystalline silicon material will be presented and discussed. Key components of the cost are silicon feedstock, crystallization, and slicing. The approaches for commercial production of all these steps will be contrasted with other available methods. The use of directional solidification (DS) methods to grow multi-crystalline silicon (mc-Si) is a large fraction of the crystalline silicon market. The efficiency of mc-Si solar cells is usually lower than for single crystal silicon because of a high degree of material defects that include dislocations, random grain orientations, grain boundaries, impurity precipitates, and inclusions. Typical defects and impurities in mc-Si wafers and their influence on the device performance are reviewed. Detailed characterization of these defects is not straightforward. Methods developed for characterization of these defects will be presented.
11:30 AM - **Q1.5
Developments in Crystalline Silicon-based Photovoltaic Product Architecture and Manufacturing.
Juris Kalejs 1
1 , American Solar Technologies, Chelmsford, Massachusetts, United States
Show AbstractSolar electric (Photovoltaic) crystalline silicon (c-Si) product diversity has changed very little over three decades of development, including the last decade of unprecedented expansion of the industry. The dominant module product comprising over 90% of cumulative installations, which exceed 15 GW worldwide, still employs an ubiquitous configuration, a platform based on a planar laminate. This paper will review trends in module architecture and manufacturing methods for this currently dominant PV c-Si commodity module platform. The commodity flat-plate module contains typically 60-72 solar cells cut from multicrystalline blocks as 156 mm square areas, or 156 mm dimension pseudo-squares cut from single crystal boules. New module design and manufacturing approaches different from those of the commodity PV product are now in development and piloting. Developments which use innovations in manufacturing processes, i.e., stringing of cells and packaging in a laminate, will be discussed.
12:00 PM - **Q1.6
Contactless Measurement of Carrier Lifetime on As-Grown or Shaped Ingots, Sections, and Blocks.
Ronald Sinton 1 , Tanaya Mankad 1 , M. Forsyth 1 , James Swirhun 1
1 , Sinton Instruments, Inc., Boulder, Colorado, United States
Show AbstractThis work will describe recent developments in measurement techniques for assessing the bulk lifetime of ingots, sections, and blocks without surface preparation. This permits detailed characterization of the materials as they exist in the production environment. Prior to sawing into wafers, it is possible to characterize the quality of the feedstock and growth parameters of bulk silicon. This allows quick feedback in order to fully optimize the growth. This can also be used to qualify the quality and suitability of the crystalline silicon for particular solar cell processes. Measurement at this stage, compared to after wafering, is extremely useful and cost effective. The measurements are more sensitive to true bulk parameters before wafering, the entire ingot can be assessed quickly, and the cost of wafering can be adverted or modified if all or a portion of the piece “fails”. Therefore measurements in the ingots or blocks prior to sawing into wafers present an unusual combination of industrial and scientific advantages compared to measurements of wafers. After sawing, the unpassivated surfaces of the wafers can compromise electronic material measurements until at least the phosphorus-diffusion step in the process which acts to passivate the surface recombination.The parameters that are determined by Quasi-Steady-State Photoconductance, QSSPC, or transient photoconductance measurements on as-grown or shaped material are the bulk lifetime, the resistivity, and the “trapping” which can be a measure of crystalline quality. Patterns in any of these parameters, from top to bottom of the grown piece or across the diameter of a sectioned CZ ingot give valuable information concerning feedstock, external contamination during growth (from the crucible or growth furnace), and the thermal growth conditions.Three special cases will be described in some detail.1) Boron-doped CZ. The special characteristics of this material are the strongly injection-level-dependent lifetime and the B, O, and Fe spatial dependences that gives can give rise to strong lifetime variations in both the growth direction and radially. Typical ranges of bulk lifetime are 10-500 microseconds.2)B-doped multicrystalline silicon. This material, like B-CZ, has strong spatial dependence of the lifetime, ohm-cm, and trapping as a function of both the growth direction and the position of the block relative to the crucible.3)The highest efficiency cells in the industry use n-type CZ or FZ silicon. The desired sensitivity for these processes requires the accurate discrimination of differences between silicon in the 1-10 ms range. In this range, a significantly different measurement and analysis technique will be presented in detail.
12:30 PM - Q1.7
Commercial Production of Silicon Solar Cell Feedstock by Upgrade of Metallurgical Grade Silicon.
John Mott 1 , Julio Bragagnolo 1 , Michael Hayes 1
1 , Ohio Solar Energy, LLC, Alliance, Ohio, United States
Show AbstractIntroduction. The relationship between impurity content in Solar Grade Silicon (SGS) and solar cell quality is the subject of intensive research. The PV industry has developed around the use of silicon made by the Siemens process for the semiconductor industry, with impurity levels typically in the parts per billion by weight (ppbw) range. There is a growing consensus that SGS with impurities in the parts per million range (ppmw) can be obtained cost effectively from Metallurgical Grade Silicon (MGS) and used to yield solar cells with comparable performance (see for example ‘Beneficial Effects of Dopant Compensation on Carrier Lifetime in Upgraded Metallurgical Silicon’ by S. Dubois et al. in the 23rd European Photovoltaic Solar Energy Conference, Valencia, September, 2008). This provides insight on the success encountered by Timminco, an early SGS market entrant, in commercializing silicon material with [P] levels of the order of 2 ppmw. Current Work. Analysing data from 16 UDS runs on samples taken from the melt, before and after UDS, and a solid sample taken from the silicon frozen on the cold silicon collection surface, we note that the average values of [P] in the molten silicon samples increase from 11.9 ppmw before UDS to 15.9 ppmw after UDS. The average value of [P] in the solid silicon sample is 4.9 ppmw. This demonstrates an effective refining ratio of 0.41, even at a 50% solid fraction. This is important as UDS, by its nature, implies a loss of silicon, while little or no silicon is lost in B reduction. Performing a secondary UDS on silicon obtained from these primary UDS runs yields [P] around 2 ppmw.In addition to P and B reduction, in this paper we also discuss the hardware designed to implement this process in commercial production in volumes exceeding 4,000 MT per year. MB Scientific, the original process developer, and NC Consulting, an engineering company, have developed a plant design that can produce SGS at an estimated cost that will allow for profitable large scale production, and have joined in a new company, Ohio Solar Energy, to commercialize the large-scale production technology. Future Work. While the UDS equipment design is completed, we have so far succeeded in decreasing the B concentration to 30% of the initial value by using glass slagging, wherein molten glass devoid of boron is vigorously mixed with the silicon metal and made to ‘getter’ the boron in the silicon and is then removed from the silicon metal. Repeated with new glass each time, the number of steps is dependent on the starting concentration of the boron in the silicon. The difficulty with reducing B is related to the P levels in the glass constituents due to back-contamination with successive washes. A new furnace, with more powerful agitation and designed to prevent recontamination of the UDS-processed silicon, and purer glass will enable B removal to ≤1ppmw target levels.
12:45 PM - Q1.8
Efficient Single-crystal Black Silicon Solar Cells with Anti-reflection by a Nanocatalyzed One-step Etch.
Hao-Chih Yuan 1 , Vernon Yost 1 , Matthew Page 1 , Howard Branz 1
1 , National Renewable Energy Lab, Golden, Colorado, United States
Show AbstractWithout using silicon nitride or another dielectric anti-reflection (AR) layer, we have fabricated confirmed 16.8%-efficient prototype solar cells on 2.7 ohm-cm, 300 um p-type single-crystal Si (100) substrates. Aside from the use of an inexpensive single-step nanocatalyzed liquid etch [Branz, Appl. Phys. Lett. 94, 231121 (2009)] to produce a nanoporous black silicon surface layer, processing of these cells is nearly identical to the present practice in PV silicon manufacturing, including a POCl3-diffused emitter and aluminum back-surface field. Open-circuit voltage (612 mV) and fill factor (80%) of the single-crystal black silicon cells is comparable to a planar control. The weighted average reflectance from 350 to 1000 nm of the single-crystal black silicon cells is below 2% compared with 34.3% of the planar control with no AR. As a result, the short-circuit current of the black silicon solar cells is 38% higher than the planar control. Nonetheless, the 34.7 mA/cm2 short-circuit current density of the black silicon solar cells is about 3 mA/cm2 below that predicted by the reflectance reduction alone. Our modeling shows that the current deficit is due to high recombination in the nanoporous layer, which impacts the short-wavelength spectral response. We also study the optical properties of the nanoporous black silicon surface layer and our measurements reveal some scattering in the nanoporous surface layer, but little internal reflection (light trapping). The studies on the optoelectronic and optical properties enable us to describe not only possible improvements to our solar cells, but general design considerations for high-efficiency solar cells based on density-graded black-silicon surfaces. Finally, we estimate the potential cost advantage of eliminating the vacuum-coated silicon-nitride anti-reflection equipment from the PV manufacturing line.
Q2: CdTe and GaAs Based Technologies
Session Chairs
Monday PM, November 30, 2009
Room 306 (Hynes)
2:30 PM - **Q2.1
Present Status of Research and Industrial Development of CdTe/CdS Solar Cells.
Ramesh Dhere 1 , David Albin 1 , Xiaonan Li 1 , Timothy Gessert 1
1 , National Renewable Energy Lab, Golden, Colorado, United States
Show AbstractThin-film solar cells have attracted considerable attention due to their potential for low-cost production. CdTe has been one of the main contenders in this arena because its bandgap of 1.5 eV is ideally matched to the solar spectrum and the binary compound allows the freedom to choose a variety of fabrication techniques. This presentation will highlight key developments during the last forty years that have been stepping stones for progress in device performance. Industrial activity began in the early 1980s when Matsushita introduced screen-printed modules, and there were several players in the field. The field has expanded tremendously in the last five years with First Solar leading the way. First Solar is already the largest producers of photovoltaics in the United States, and with a planned expansion to 1 GW by the end of 2009, they will be contending for the top spot worldwide. We will present an overview of different industries involved in the field and their approaches. In addition, we will present the ongoing research at our laboratory (NREL) and others and will analyze the status of the research. The efficiency of the champion CdTe cell is 16.5%, and module efficiency based on present knowledge is expected to reach around 12.5%, which is well below its potential. We will present the analysis of the device performance and the main parameters affecting the performance. Further improvement in module performance is unlikely without better understanding these parameters. Other area of emphasis is the long-term reliability and accelerated life testing that is necessary to understand the effect of processing changes on product reliability. NREL is developing the Process Development and Integration Laboratory (PDIL) to facilitate interaction among industry and research groups. The presentation will provide an overview of the CdTe tool being developed and provide some details about the capabilities of the tool, in particular, and PDIL facility, in general. This abstract is subject to government rights.
3:00 PM - Q2.2
Study of the Electrical Properties of Cross Sections of CdTe/CdS Solar Cells Measured with Scanning Kelvin Probe Microscopy.
Helio Moutinho 1 , Ramesh Dhere 1 , Chun-Sheng Jiang 1 , Mowafak Al-Jassim 1
1 , National Renewable Energy Laboratory, Golden, Colorado, United States
Show AbstractWe apply scanning Kelvin probe microscopy (SKPM) to analyze cross sections of working CdTe/CdS solar cells under different bias conditions. This technique is performed inside a scanning probe microscope (SPM), and it provides the distribution of the electrical potential inside the device with high spatial resolution. The SKPM and topographic images are compared to associate variations of the potential with topographic features. For instance, it is possible to compare the position of the p-n junction with the metallurgical junction between CdTe and CdS. In SKPM, we apply AC and DC bias between the tip of the SPM and the sample. The signal measures the difference between the work functions of the tip and sample surface, and it is proportional to the surface potential of the sample. By biasing the cell during measurement (reverse and forward polarizations), we avoided artifacts such as Fermi-level pinning, and we were able to investigate the distribution of the electrical potential inside a live device polarized under different conditions. By taking the derivative of the potential, we determined the distribution of the electric field, and by locating the maximum of the electric field, we located the position of the p-n junction.This work complements our other work presented in the 2009 Spring MRS Meeting that investigated different ways to prepare cross sections of samples, compared two different SKPM measurement procedures (using the first and second resonance cantilever peaks), and determined the position of the junction for a standard CdTe/CdS solar cell. In this work, we investigate the change in the electrical potential using different bias, forward and reverse, showing the change in the width of the depletion region. By calculating the derivative of the potential, we observed the distribution of the electric field inside the device and noticed the following: the field has a strong value concentrated on a thin layer of the device at the junction, and a much smaller value moving away from the junction, showing that there is a change on the electrical properties of the device at the interface. To investigate whether interdiffusion of Te and S is responsible for this effect, we analyzed solar cells without CdCl2 heat treatment, as well as cells produced without the CdS layer. In both cases, the distribution of the electrical potential and electric field was different than for a standard device. In this work, we will also present results of the distribution of the electrical potential on the back contact of the solar cells. In our case, we use graphite paste deposited on the CdTe film followed by the deposition of Ag film. We will show that, in general, the potential drop between the CdTe/graphite interface is much smaller than in the junction region. This abstract is subject to government rights.
3:15 PM - Q2.3
Finite Element Model to Understand the Effect of O2 on Closed Space Sublimation of CdTe.
Nirav Vora 1 , Ramesh Dhere 1
1 National Center for Photovoltaics, National Renewable Energy laboratory, Golden, Colorado, United States
Show AbstractIncorporating O2 in the closed space sublimation (CSS) of CdTe thin film has resulted in improved cell efficiencies. Many studies have been undertaken to understand this effect on cell efficiency. In this work we study the effect of oxygen on lateral uniformity of the deposited CdTe film. A finite element model has been developed to represent the mass and heat transfers involved in the CSS process. The model takes into consideration the effect of O2 by modeling its reaction with Cd vapors in the space between the source and the substrates. This reaction can decrease the amount of Cd available for condensation near the substrate if the diffusion of Cd from the source to the substrate is not fast enough. One of the factors affecting this reaction rate is the concentration of O2. So a gradient of O2 from the edges to the center of the substrate can result in a laterally non-uniform film. This gradient can be formed if the rate of diffusion of O2 is lower than that of its reaction with Cd. A steady state model will be solved at various temperatures, pressures, and separation distances to determine the optimum conditions for depositing a CdTe film with uniform thickness. Experiments will be carried out at these conditions and results compared with the simulation results. The comparison will help in determining the reaction rate constants, as there is a lot of variation in the values reported in the literature. A transient model will also be developed to better represent the experiments. Finally the model will be modified to represent the vapor transport deposition of CdTe. This abstract is subject to government rights.
3:30 PM - Q2.4
CdTe Thin Film Growth Using High Rate Sputtering for Photovoltaic Applications.
John Walls 1 , Paresh Nasikkar 1 , Hari Upadhyaya 1
1 Electronic and Electrical Engineering, Loughborough University, Loughborough United Kingdom
Show AbstractMagnetron sputtering has a number of important advantages for the deposition of thin films for use in photovoltaic devices. Sputtering provides control over thin film thickness with sub-nanometre precision using time only. This allows the thickness of the CdTe absorber layer to be optimized thereby minimizing materials usage and process manufacturing time. Using the closed field configuration, the thin films are super-smooth (< 1 nm rms roughness). This is especially important in the TCO base layer since roughness of the TCO can break through the CdS layer and cause “shunting” across cells. This paper describes a flexible reactive sputtering process in which adjacent unbalanced magnetrons are constructed of opposite magnetic polarity. The resulting closed magnetic field maintains a high density reactive plasma. In contrast to previous reactive sputtering strategies, the process does not require an auxiliary ion or plasma source and the associated use of high voltage ion acceleration. As a result, the deposition energy is optimized and insufficient to cause damage in the growing thin film. The substrate temperature is typically maintained below 100°C without the need for direct cooling. The thin films exhibit bulk optical properties, they are also dense and super-smooth. The thin films also have typically low compressive stress. The magnetron targets are simple metals or semiconductors for high rate deposition and are converted to compound thin films when required by using the appropriate reactive gas. This paper provides data derived from a medium throughput batch system with a 0.4 m diameter drum substrate carrier and four 0.6m linear magnetrons. However, the process geometry is scalable and adaptable to in-line deposition. The optical and electrical performance of each layer in the CdTe thin film photovoltaic stack will be presented together with preliminary device performance.
3:45 PM - Q2.5
Fabrication and Modeling of Three-Dimensionally Structured CdTe Thin Film Photovoltaic Devices with Self-Aligned Back-Contacts.
Jonathan Guyer 1 , Daniel Josell 1 , Carlos Beauchamp 1 , Suyong Jung 2 , Behrang Hamadani 2 , Lee Richter 3 , John Bonevich 1 , Nikolai Zhitenev 2 , Tom Moffat 1
1 Metallurgy Division, NIST, Gaithersburg, Maryland, United States, 2 Center for Nanoscale Science and Technology, NIST, Gaithersburg, Maryland, United States, 3 Surface and Microanalysis Science Division, NIST, Gaithersburg, Maryland, United States
Show AbstractOur goal is to provide industry with test structures and models ofnext-generation photovoltaics, with an initial focus on CdTe andCuInxGa1-xSe2 (CIS or CIGS) materials. These tools will enableinterpretation of measured external properties affected by geometry, grainstructure, and nanoscale phase separation, which will support improvedprocessing and design of Second Generation (thin film) and Third Generation(nanostructured) photovoltaic devices.
CdTe and CIGS are some of the most stable and efficient photovoltaicmaterials. A wide variety of deposition methods enable novel devicestructures at previously unobtainable dimensions, but optimal structuresand dimensions are unknown. Sensitivity of the microstructure (and,ultimately, the device efficiency) to deposition methods and processingconditions varies, and potential new and cheaper fabrication methods havenot been verified.
We are adapting our experience with the electrochemical "superfill" ofmetal in sub-micrometer-scale trenches and vias [1] to the fabrication ofnovel photovoltaic structures. We have fabricated electrodeposited CdTedevices that use interdigitated back-contact electrodes both forindependent electrode-position of n- and p-type material during fabricationand for collecting the photo-generated current in the fabricated devices[2]. The devices enable quantitative evaluation of bulk and interfaceproperties of 3-d devices through controlled variation of length scales,independent of the absorber thickness, including electrode pitch andcross-section.
To guide and interpret the experimental measurements, we have implemented aphotovoltaic device model that gives us complete control over geometry andmicrostructure. We are using the model, in conjunction with experimentalmeasurements, to extract discrete materials properties from complex structures.We implemented our device model in theFiPy partial differential equation solver package [3]. This freely available package allows us to easilydistribute our photovoltaic codes to researchers in industry and elsewhereas they are validated.
[1] D. Josell, D. Wheeler, W. H. Huber & T. P. Moffat Phys. Rev. Lett. 87 (2001) 016102
[2] D. Josell, C. Beauchamp, S. Jung, B.H. Hamadani, A. Motayed, L. Richter, M. Williams, J.E. Bonevich, A. Shapiro, N. Zhitenev & T.P. Moffat J. Electrochem. Soc., in press
[3] J. E. Guyer, D. Wheeler & J. A. Warren Comput. Sci. Eng. 11 (2009) 6 http://www.ctcms.nist.gov/fipy
4:30 PM - Q2.6
Thin-film ITO/InP Solar Cells on Flexible Plastic Substrates.
Kuen-Ting Shiu 1 2 , Jeramy Zimmerman 2 , Hongyu Wang 2 , Stephen Forrest 2
1 Electrical Engineering, Princeton University, Princeton, New Jersey, United States, 2 Materials Science and Engineering, Electrical Engineering and Computer Science and Physics, University of Michigan, Ann Arbor, Michigan, United States
Show AbstractPhotovoltaic cells made with amorphous silicon, II-VI compounds such as CdTe [1], and copper-indium-gallium-selenide (CIGS) [2] thin-films have been adopted for use in high specific power, light-weight solar cell panel applications. In this work, we demonstrate single-crystal thin-film InP Schottky-type solar cells mounted on flexible plastic substrates by cold-welding a metallic film on the solar cell to one deposited on the plastic sheet. The lightly p-doped InP cell is grown epitaxially on an InP substrate via gas source molecular beam epitaxy. The InP substrate is removed via selective chemical wet-etching after the epitaxial layers are bonded to a metal layer pre-deposited onto the surface of an 25 m thick Kapton sheet, and indium tin oxide (ITO) top contacts are then deposited. The power conversion efficiency under 1 sun is up to 10.2±1.0% and its specific power is 2.0±0.2 kW/kg. Stress tests indicate that the thin film solar cells can tolerate both tensile and compressive stress by bending over a 1 cm radius without damage. Full materials and device characterization will be presented. This work provides an alternative method for adopting III-V semiconductor solar cells for portable and space applications where very high specific power efficiency is required. 1. Romeo, et.al, Solar Energy Materials & Solar Cells, 90, 3407 (2006).2. Otte, et.al, Thin Solid Films, 511-512, 613 (2006)
4:45 PM - Q2.7
Suppression of Edge Recombination in InAs/InGaAs DWELL Solar Cells.
Tingyi Gu 1 , Kai Yang 1 , Mohamed El-Emawy 1 , Andreas Stintz 1 , Luke Lester 1
1 Center for High Technology Materials, University of New Mexico, Albuquerque, New Mexico, United States
Show AbstractRecent interest in using InAs quantum dots (QDs) in the absorbing region of solar cells has focused primarily on the predicted increase in quantum efficiency due to the intermediate band effect or simply larger short circuit current density. However, the three-dimensional carrier confinement inherent to QDs endows them with unique carrier transport capabilities that have not been previously explored in the context of solar cells. In this work, it is observed that InAs/InGaAs dots-in-a-well (DWELL) structures efficiently suppress lateral carrier diffusion. Therefore, not only do the DWELL structures enhance photocurrent by extending the absorption edge, but they should also inhibit the spreading of current to the perimeter of a device where edge recombination can dominate. In this paper, we examine this premise by comparing the dark current behavior of DWELL cells and GaAs control cells of varying area. The results are promising for applications such as concentration and flexible surfaces where shrinking the size of the device while maintaining high charge collection efficiency are of paramount importance.The InAs/InGaAs DWELL solar cell grown by MBE is a standard pin diode structure with six layers of InAs QDs embedded in InGaAs quantum wells placed within a 200-nm intrinsic GaAs region. The GaAs control wafer consists of the same pin configuration but without the DWELL structure. The typical DWELL solar cell exhibits higher short current density while maintaining nearly the same open-circuit voltage for different scales, and the advantage of higher short current density is more obvious in the smaller cells. In contrast, the smaller size cells, which have a higher perimeter to area ratio, make edge recombination current dominant in the GaAs control cells, and thus their open circuit voltage and efficiency severely degrade. The open-circuit voltage and efficiency under AM1.5G of the GaAs control cell decrease from 0.914V and 8.85% to 0.834V and 7.41%, respectively, as the size shrinks from 5*5mmsq to 2*2mmsq, compared to the increase from 0.665V and 7.04% to 0.675V and 8.17%, respectively, in the DWELL solar cells.The lower open-circuit voltage in the smaller GaAs control cells is caused by strong Shockley-Read-Hall (SRH) recombination on the perimeter, which leads to a shoulder in the semi-logarithmic dark IV curve. However, despite the fact that the DWELL and GaAs control cells were processed simultaneously, the shoulders on the dark IV curve disappear in all the DWELL cells over the whole processed wafer. As has been discussed in previous research on transport in QDs, it is believed that the DWELL cells inhibit lateral diffusion current and thus edge recombination by collection first in the InGaAs quantum well and then trapping in the embedded InAs dots. This conclusion is further supported by the almost constant current densities of the different area DWELL devices as a function of voltage.
5:00 PM - Q2.8
Virtual Single Crystalline GaAs Epitaxial Thin Films on Flexible Polycrystalline Metallic Substrates.
Alex Freundlich 1 , Venkat Selvamanickam 1
1 , University of Houston, Houston, Texas, United States
Show AbstractDevelopment of high quality III-V epitaxial layers on inexpensive flexible substrates is a desirable feature to many civilian and military applications. In particular it may be a game-changing enabler toward significantly reducing the cost and increasing the efficiency of thin film solar cells, as it offers the possibility of combining the unsurpassed performance of GaAs based multi-junction technologies (1 sun efficiency >36%) with a conventional roll to roll processing standard of thin film industry as afforded by polycrystalline metallic foil technology.Here we report our recent results on the development of virtual single crystalline GaAs thin film on thin (50 microns) flexible polycrystalline metallic substrates. The flexible poly-crystalline Ni-based substrates were coated with a few hundred nm oxide-ceramic epitaxial buffer, adapted from a previously developed structure for high Tc superconductor wire technology, followed by a very thin (<50nm) Ge epilayer. After introduction in the molecular beam epitaxy chamber Ge native oxide was thermally removed and a subsequent high temperature annealing was implemented resulting in the formation of clear c(2x2) (mixed (2x1)(1x2) ) reconstruction, typical of (001) Ge surface. The GaAs growth was initiated at relatively low temperature (~400C) and a thin nucleation layer of GaAs was deposited , followed by an annealing step under As2, subsequently 1 micron-thick GaAs was deposited in standard growth conditions (growth rate ~ 1ML/sec T~550C). The entire growth sequence was monitored by reflection high energy electron diffraction (RHEED). The self -annihilation of anti-phase boundaries (mixed 2x4, 4x2 RHEED diagram), was observed for thicknesses exceeding 100 nm where a 2x4 RHEED diagram typical of a single domain (001) GaAs was recorded. Epilayers exhibited a specular morphology. High resolution X-ray diffraction analysis confirmed the single crystalline (001) nature of GaAs. Temperature dependent photoluminescence (PL) analysis revealed a strong PL in as-grown samples (Fig. 2). The low temperature photoluminescence was found to be dominated with DA –eA like bands in the 1.4-1.5 eV range and a relatively broad deeper luminescence band at 1.3- 1.35 eV. At low temperature the GaAs excitonic emission was detected at 1.526 eV (FWHM~20 meV) and was found to be slightly red shifted compared to the typical A0X exciton (1.512) in homoepitaxial GaAs. The magnitude of this red-shift (~14 meV) suggested the absence of any significant thermoelastic/lattice mismatch strain in the epilayers. In summary the development of high quality single crystalline (001) GaAs on flexible metal substrates is demonstrated. Samples exhibit high optical and structural quality as stressed by the RHEED, X-ray and luminescence properties of the as-grown epilayers. Development of GaAs based thin film single junction solar cells is underway and preliminary device results will be presented at the meeting.
5:15 PM - Q2.9
Neodymium Luminescent Solar Concentrator.
Phil Reusswig 1 , Carmel Rotschild 1 , Marc Baldo 1
1 EECS, MIT, Cambridge, Massachusetts, United States
Show AbstractLuminescent solar concentrators (LSCs) are promising in photovoltaic applications because they do not need to track the sun to obtain high optical concentration factors. However, loss mechanisms that are associated with optical self-absorption decrease the efficiency of LSCs at increasing values of the geometric gain, G, which is defined as the ratio of the facial area to the edge area. In this work, we demonstrate LSCs based on a lanthanide infrared emitter, neodymium (Nd3+). Neodymium is nearly the optimal infrared LSC material: inexpensive, abundant, efficient, and spectrally well matched to high-performance silicon solar cells. Neodymium is a natural four level system, making it reasonably transparent to its own emission enabling high optical concentrations and geometric gain. Neodymium’s one disadvantage is its absorption. It has relatively poor overlap with the visible spectrum, meaning that it will require sensitization. LG-760 and APG-1 phosphate glass was purchased from Schott with Nd3+ doping concentrations of 1.0%, 2.0% and 3.0% with thicknesses of 1mm and 5mm. For the neodymium glass without sensitizer, peak optical quantum efficiencies (OQE) of 50% have been measured with an estimated power efficiency of 2.5%. The neodymium glass was sensitized by spin casting thin films doped with BASF Lumogen Violet 570, Gelb 083, and Orange 240 in a matrix of poly(methyl methacrylate) (PMMA). Together the dyes exhibit broad absorption below λ = 550 nm. Foerster energy transfer was used to couple the dyes non-radiatively such that all radiative emission was matched to the neodymium absorption lines at approximately λ = 580 nm. Initial results indicate peak OQE of 45% with an increase in estimated power efficiency of 3.5%. We will also discuss the sensitization of neodymium glass using nanocrystals in a polymer matrix.
Q3: Poster Session I
Session Chairs
Tuesday AM, December 01, 2009
Exhibit Hall D (Hynes)
9:00 PM - Q3.1
Electron Reflector Strategy on Micron-Thickness CdTe Solar Cells.
Kuo-Jui Hsiao 1 , James Sites 1
1 Physics, Colorado State University, Fort Collins, Colorado, United States
Show AbstractIncorporation of an electron reflector is a proposed strategy to improve Voc for CdTe thin-film solar cells. An electron reflector is a conduction-band barrier at the back surface, which can reduce the recombination resulting from the electron flow to the back surface. It should be particularly valuable at sub-micron thicknesses. For optimal improvement with an electron reflector, reasonable carrier lifetime (1 ns or above) and full depletion are required. Numerical simulation is used to investigate the electron reflector strategy for thin CdTe cells. Theoretically, a 200 mV increase in voltage and 3% in efficiency should be possible for a micron-thickness CdTe cell with a 0.2-eV electron reflector barrier, assuming 2x1014-cm-3 hole density and 1-ns lifetime, which are currently achieved. For the electron reflector to be beneficial, the CdTe needs to be fully depleted at its typical operating voltage. With the electron reflector, good CdTe cell performance at thicknesses as low as 0.3 μm should be possible.
9:00 PM - Q3.10
Mechano-chemical Synthesis, Deposition and Structural Characterization of CIGS.
Vidhya Bhojan 1 , Velumani Subramaniam 1 , Jesus A.Arenas-Alatorre 2 , Rene Asomoza 1
1 Electrical Engineering, CINVESTAV, Mexico,D.F. Mexico, 2 Institute of Physics, Universidad Nacional Autónoma de México, Mexico ,D.F. Mexico
Show AbstractCuInGaSe2 (CIGS) is a prominent thin-film photovoltaic material. However, commonly used physical vapour deposition and sputtering techniques to fabricate CIGS thin-film photovoltaic (PV) devices are complex and expensive. Therefore non-vacuum deposition techniques such as paste coating, spray pyrolysis and electro deposition are gaining more attention in recent years. Our intention is to choose a low cost non-vacuum technique like mechano-chemical synthesis of CIGS powder, followed by screen printing. Mechano chemical synthesis is a process that induces physical and/or chemical change in the compounds by mechanical energy, such as pulverization, friction or compression. This method has some advantages for the mass production of CIGS solar cells, high productivity and short processing cycle time. In the present work CIGS powders suitable for screen-printing ink has been prepared by ball milling. High purity elemental copper granules (>99.9% pure), selenium and indium powders (>99.9% pure) and fine chips of gallium (>99.9% pure) were used as the starting materials. Ball milling was carried out for an optimized composition of CuIn0.75Ga0.25Se2 using a SPEX-8000 mixer/mill at 1200 rpm for 1.5 hours. X-ray diffraction analysis of the milled powder shows the presence of (112), (220)/ (204), (312)/ (116), (400) and (332) peaks corresponding to CIGS chalcopyrite structure with a preferential orientation along (112) peak. The average grain size calculated by Scherrer’s formula is about 13.8 nm. Crystallographic structure of the prepared CIGS powder was analyzed by Rietveld analysis using X-ray powder diffraction data. Geometry optimization of the structure was performed and the basic structural properties have been evaluated by density functional approximation and compared with experimental results. Density of states has been simulated for the same structure inorder to have a better understanding of the distribution of atomic orbitals. FESEM analysis shows the agglomeration of nano particles. Particle size varied from 11 to 30nm. Final composition of the milled powder studied by Energy dispersive X-ray analysis gives 24.39 at% Cu, 21.42 at% In, 7.22 at% Ga and 46.97 at% Se. HRTEM analysis reveals the presence of nano crystalline particles. The interplanar distance (d-spacing) corresponding to (112), (220)/ (204), (512)/ (417) and (620)/ (604) diffraction peaks has been estimated and compared with standard values and the corresponding diffraction pattern has been simulated with simulaTEM software. CIGS ink is prepared by proper mixing of the powder with a suitable organic binder (ethyl cellulose). Screen printing is carried out on glass substrates, followed by annealing at 5 different temperatures of 300,350,400,450 and 500 degrees, in order to form a porous CIGS film. XRD and SEM analysis were carried out to study the structure and morphology of screen printed CIGS .Cross section of the screen printed CIGS thin film has been analyzed by HRTEM.
9:00 PM - Q3.11
Photo-patternable Polythiophenes Having Methacrylate Pendant Group for Organic Photovoltaics.
Yuna Kim 1 , Jeonghun Kim 1 , Sehwan Kim 1 , Eunkyoung Kim 1
1 Department of Chemical and Biomolecular Engineering, Yonsei University, Seoul Korea (the Republic of)
Show AbstractNew series of poly(3-hexylthiophene) copolymer functionalized with methacrylate group were synthesized for the application to organic photovoltaics. The regioregularity of the polythiophenes was controlled by the ratio of thiophene methacrylate to 3-hexylthiophene. As synthesized copolymer showed good solution processible and photo-patternable properties. Photoconversion of the side chain cross-linking of the polymer was examined by isothermal photo-DSC and FT-IR study under UV irradiation. By the photo cross-linking reaction, the ordering of poly(3-hexylthiophene) copolymer itself and the phase separation in fullerene blended films were significantly changed as observed by UV-Vis absorbance, PL, AFM and XRD. The photo-induced charge transport of polymer films were also changed by the photo cross-linking reaction. The organic photovoltaics were fabricated to show different photo conversion efficiency from the polythiophenes depending on the degree of the side chain cross-linking.
9:00 PM - Q3.13
Thermally and Chemically Stable Transparent Conducting Multi-layered Al-doped ZnO and Its Applications for Dye Sensitized Solar Cells.
Jun Hong Noh 1 , Hyun Soo Han 1 , Sangwook Lee 1 , Hyun Suk Jung 2 , Kung Sun Hong 1
1 Materials Science and Engineering, Seoul National University, Seoul Korea (the Republic of), 2 Advanced Materials Engineering, Kookim University, Seoul Korea (the Republic of)
Show Abstract Thermally and chemically stable NTO/AZO multilayered TCO fabricated by pulsed laser deposition for application in DSSC. Nb doped TiOx (NTO) layer was deposited at room temperature with amorphous and the layer crystallized to anatase and Ti3+ ions in the TiOx layer oxidized Ti4+ ions during air annealing at 450 oC. Owing to the crystallization and oxidation, NTO layer prevented from penetrating oxygen into AZO layer and consequently the AZO layer was in quasi-reducing atmosphere. As if the AZO layer is annealed under reducing atmosphere, oxygen vacancy was produced and substitutional Al atoms were activated in the AZO, thereby the conductivity of NTO/AZO was enhanced. In addition, the NTO layer prevented from forming the aggregate Zn2+-dye molecule on surface of AZO so that high transmittance of the NTO/AZO was maintained. The DSSC using the stable NTO/AZO TCO showed twice efficiency as high as the AZO-DSSC. However, the NTO/AZO-DSSC showed lower fill factor compared to the conventional FTO-DSSC. The cause for low fill factor might be electronic barrier which was formed by the non-degenerated thin AZO:Oi layer.
9:00 PM - Q3.14
Fabrication, Characterization and Optical Studies of Cu(In1-xGax)3Se5 Bulk Compounds.
Dayane Habib 1 , Georges El Haj Moussa 1 , Roy Al Asmar 1 , Michael Ibrahim 1 , Mario El Tahchi 1 , Claude Llinares 2
1 LPA, Lebanese University, Jdeidet Lebanon, 2 Centre Electronique et Micro-optoélectronique de Montpellier (CEM2), University of Montpellier 2, Montpellier France
Show AbstractIn this paper we present the structural and optical properties of Cu(In1-xGax)3Se5 ternary and quaternary compounds crystals fabricated by horizontal Bridgman technique. The Cu(In1-xGax)3Se5 materials were characterized by Energy Dispersive Spectrometry (EDS), hot point probe method, X-ray diffraction, Photoluminescence (PL), and Optical response (Photoconductivity). The Cu(In1-xGax)3Se5 have an Ordered Vacancy Chalcopyrite-type structure with lattice constants varying as a function of the x composition. A good stœchiometry given by the EDS characterization method is well observed in our samples and its magnitude deviation Δy is slight; so, our samples present a nearly perfect stœchiometry (Δy = 0) [1]. X-Ray diffraction patterns show the presence of many preferential orientations according to the planes (112), (220) and (312) of all the samples [2]. Also, it shows a linear shifting of peaks towards the higher magnitudes of 2θ when the x composition increases. These compounds can be of stanite structure [3] or an Ordered Vacancy Chalcopyrite structure (OVC) [4] or Ordered Defect Chalcopyrite Structure (ODC).We observe a large shift of the main PL and optical response emission peak versus x composition. The band gap energy of Cu(In1-xGax)3Se5 compounds is found to vary from 1.23 eV to 1.85 eV as a function of x.[1] Migual A. Contreras, Holm Wiesner, Rick Mtson, John Tuttle, Kanna Ramanathan, Rommel Noufi, Mat. Res. Soc. Symp. Proc. Vol. 426 (1996) 243-254.[2] Ariswan, G. El Haj Moussa, M. Abdelali, F. Guastavino, C. Llinares, Solid State Communications 124 (2002) 391-396. [3] M. Suzuki, T. Uenoyama, T. Wada, T. Hanada, Y. Nakamura, Jpn. J. Appl. Phys. 36 L1139 (1997).[4] Kristjan Laes, Sergei Bereznev, A. Tverjanovich, E.N. Borisov, Tiit Varema, Olga Volobujeva , Andres Öpik. Thin Solid Films 517 (2009) 2286–2290.
9:00 PM - Q3.15
Reactive Sputtering of Magnesium Hydride Thin Films for Photovoltaic Applications.
Charlotte Platzer-Bjorkman 1 , Smagul Karazhanov 1 , Jan-Petter Maehlen 1 , Erik Marstein 1 , Arve Holt 1
1 , Institute for Energy Technology, Kjeller Norway
Show AbstractMetal hydrides have been intensively studied for hydrogen storage [1], battery [2] and smart window [3] applications. For these purposes, fast and repeatable switching of material properties and hydrogen content are crucial. Many metal hydrides are also semiconducting or insulating and have recently been suggested for application in photovoltaic devices either as transparent conducting layers, antireflective coatings or even absorber layers [4]. The advantage of using metal hydrides for these applications are the large abundance of the constituent elements, high hydrogen content possibly improving bulk and surface passivation of silicon-based devices and suitable band gap range of several of the materials. In contrast to the applications based on hydrogenation/dehydrogenation, high stability is required for photovoltaic applications. In this work, we investigate the possibilities for in-situ deposition of MgH2 using reactive sputtering from a metallic target in Ar/H plasma. For low H2/Ar ratio, crystalline MgH2 is formed together with metallic Mg while for increasing ratio films are amorphous, partly transparent and insulating. Previous studies of in-situ deposition of MgHx films by activated reactive evaporation showed difficulties in obtaining single phase MgH2 [5]. In the present study, sputtering process parameters such as RF power, gas ratio, pressure and substrate temperature are varied. Film properties are monitored using x-ray diffraction, resistivity measurements and optical characterization. Depositions on both glass and silicon substrates are reported as well as studies of stability under annealing and light exposure.References1P. Chen, Z. Xiong, and J. Luo, Nature 420, 302 (2002).2L. Schlapbach and A. Zuttel, Nature 414, 535 (2001).3J. Huiberts, R. Griessen, and J. Rector, Nature 380 (1996).4S. Karazhanov, A. Ulyashin, P. Vajeeston, and P. Ravindran. Phil. Mag. 88(16), 2461 (2008).5R. Westerwaal, C. Broedersz, R. Gremaud, et al., Thin Solid Films 516, 4351 (2008).
9:00 PM - Q3.16
Wide Energy Bandgap Materials (LiF and Al2O3) as Hole Blocking Layers in Organic Solar Cells.
Chen Cheng Hong 1 , Wu Jui Chi 1 , Chen Chun Wei 1 , Chen Jen Sue 1
1 Materials Science and Engineering, National Cheng Kung University, Tainan Taiwan
Show AbstractMaterials of different energy bandgaps are investigated as hole blocking layers in organic solar cells with P3HT:PCBM heterojunction active layer and Al cathode. In this study, by inserting LiF (Eg: 14.6 eV) or Al2O3 (Eg: 10.8 eV) between P3HT:PCBM and Al as the hole blocking layer, we demonstrate the power conversion efficiency (PCE) of the solar cell devices reaches 3.05% or 2.57%, respectively, which is much higher than that of the solar cell without the interlayer (only 1.32%). The enhancement of efficiency is associated with the increase of short circuit current and fill factor. The mechanism for the high efficiency is based on the tunneling effect. Owing to the wide energy bandgap of the hole blocking layer, the band offset between HOMO of PCMB and valance band edge of hole blocking layer is significantly large and the barrier height for tunneling of holes will be subsequently elevated. Consequently, the electron-hole recombination at the Al cathode will be appreciably reduced. Therefore, the use of wide bandgap materials as the hole blocking layer in solar cell devices will achieve higher efficiency.
9:00 PM - Q3.17
Ti-doped Gallium Phosphide Layers with Concentrations Above the Mott Limit.
David Pastor 1 , Javier Olea 1 , Maria Toledano-Luque 1 , Ignacio Martil 1 , German Gonzalez-Diaz 1 , Jordi Ibanez 2 , Ramon Cusco 2 , Luis Artus 2
1 Dpto. de Física Aplicada III, Facultad de Ciencias Físicas, Universidad Complutense de Madrid, Madrid Spain, 2 Institut Jaume Almera, Consejo Superior de Investigaciones Científicas (CSIC), Barcelona Spain
Show AbstractIntermediate band solar cell (IBSC) materials exhibit highly attractive properties that could allow exceeding the solar conversion efficiency limit for single junction solar cells. An intermediate band (IB) material is formed by the introduction of a new band inside the bandgap of a semiconductor. In these compounds, photons with energy below the bandgap can be absorbed, pumping electrons from the valence band to the conduction band with the IB as an intermediary step. To form the IB material, deep center impurities with a concentration above the Mott limit (∼5x1019 cm-3) have to be introduced. Theoretical calculations have predicted Ti-doped Gallium Phosphide as one of the most efficient IB materials. For the introduction of such high Ti concentrations, well-above the solid solubility limit of Ti in GaP, here we explore the use of ion beam implantation to produce a GaP-based IB material. For the post-implantation annealing treatment of these heavily implanted samples, non equilibrium thermal processes such as pulsed laser melting (PLM) have to be used instead of thermal equilibrium treatments to avoid impurity out-diffusion. We perform a structural characterization by means of Time of Flight Secondary Ion Mass Spectroscopy (ToF-SIMS), Raman Spectroscopy and Glancing Incidence X-Ray Diffraction (GIXRD) of GaP layers with Ti concentrations above the Mott limit. The samples were produced by implantation of Ti into GaP layers with a dose of 8x1014, 8x1015 and 1.6x1016 cm-2 and subsequent PLM annealing at energy densities of 0.2, 0.7 and 1.2 J/cm2. The electrical properties of the samples were also studied by Hall effect measurements in the Van der Pauw configuration. ToF-SIMS Ti depth profiles after the PLM indicate that a Ti impurity concentration above the Mott limit is achieved. The Raman spectra show forbidden TO modes in all the annealed samples, which indicates that implanted GaP is polycrystalline after the PLM. Both the LO and TO peaks redshift by about 5 cm-1 after PLM at 0.2 J/cm2, which can be attributed to strain induced by the presence of Ti. The observed shifts disappear after PLM at higher energy densities, which may be related to a change of Ti location in the GaP lattice. GIRXD corroborates these results and suggests that TiP of GaTi alloys could be formed after PLM at the highest energy densities. The Hall effect measurements show an increase of the sheet resistivity and a suppression of the mobility.We discuss the possibility to achieve an IB material by Ti implantation and subsequent PLM annealing in GaP. We show that a high concentration of Ti impurities above the Mott limit is still present in the samples after PLM. However, the PLM process alters substantially the crystal quality of GaP, yielding a highly polycrystalline lattice and a deleterious effect on the electrical properties.
9:00 PM - Q3.18
Optical Characterization of Cu2ZnSnSe4 Grown by Thermal Co-evaporation.
Do young Park 1 , Hyeonsik Cheong 1 , Sunghun Jung 2 , Jaeho Yun 2 , Sejin An 2 , Jihye Gwak 2 , Kyeong-hoon Yoon 2
1 Physics, Sogang university, Seoul Korea (the Republic of), 2 Solar Cells Research Group, Korea Institute of Energy Research, Daejun Korea (the Republic of)
Show AbstractCuInGaSe2 (CIGS) is widely studied due to its promise as low-cost material for high-efficiency thin film solar cells. However, since indium and gallium are becoming scarce and expensive, high-efficiency and low cost solar cell absorber materials to replace CIGS such as quaternary Cu2ZnSnSe4 (CZTSe) and Cu2ZnSnS4 (CZTS) have been suggested. In previous years, the direct bandgap energy of CZTSe was suggested as being around 1.5eV, estimated from optical absorption data [1]. However, recent theoretical analysis and photoluminescence measurements suggest it to be around 1.0eV [2, 3]. In this study, we investigated quaternary CZTSe grown by thermal co-evaporation various optical spectroscopic techniques, including Raman scattering, photoluminescence and contactless electroreflectance (CER). Since CER is a modulation technique, the critical points of a semiconductor is pronounced in CER spectra, which helps identify the bandgap energy. Raman spectra of CZTSe show three main peaks at 173 cm-1, 197 cm-1, and 231 cm-1 [3]. The photoluminescence spectra show a peak at around 1.0 eV, which is consistent with previous measurements. There seems to be a correlation between the presence of the 173 cm-1 Raman peak and the intensity of the photoluminescence. Photoluminescence and Raman results are compared with CER measurements. [1] T. Tanaka, T. Nagatomo, D. Kawasaki, M. Nishio, Q. Guo, A. Wakahara, A. Yoshida, and H. Ogawa, J. Phys. Chem. Solids 66, 1978 (2005).[2] S. Chen, X. G. Gong, A. Walsh, and S.-H. Wei, Appl. Phys. Lett. 94, 041903 (2009).[3] M. Grossberg, J. Krustok, K. Timmo and M. Altosaar, Thin Solid Films 517 2489 (2009).
9:00 PM - Q3.19
Influence of Deposition Parameters on Surface Texturing of ZnO:Al Films Prepared by In-line RF Magnetron Sputtering.
Jun-Sik Cho 1 , Young-Jin Kim 1 , Jeong Chul Lee 1 , Sang-Hyun Park 1 , Jinsoo Song 1 , Kyung Hoon Yoon 1
1 Photovoltaic Research Center, Korea Institute of Energy Research, Deajeon Korea (the Republic of)
Show Abstract A systematic study of the effect of sputtering deposition parameters on material properties of Al-doped ZnO (ZnO:Al) films prepared by in-line rf magnetron sputtering and on surface morphologies of the films after wet etching was carried out. After deposition, the as-deposited films were surface-textured to improve optical properties such as haze and angle resolved distribution of scattered light on the film surfaces for application of silicon thin film solar cells.High quality ZnO:Al films with electrical resistivity of 5 × 10-4 ohm cm and optical transmittance of 85% in the visible range are obtained at high substrate temperature and low working pressure. The surface morphologies and optical properties of textured ZnO:Al films are changed significantly depending on the deposition conditions, whereas the electrical properties such as resistivity, carrier density and mobility are seldom affected.The photovoltaic characteristics of silicon thin film solar cells prepared on the textured ZnO:Al films will be presented in comparison of commercial Asahi U glass.
9:00 PM - Q3.2
Incorporation of Sodium into Low-Temperature Deposition of CIGS Flexible Solar Cells.
Hendrik Zachmann 1 , Stefan Puttnins 1 , Thomas Lange 1 , Felix Daume 1 , Andreas Rahm 1 , Karsten Otte 1 , Tobias Eisenbarth 2 , Raquel Caballero 2 , Christian Kaufmann 2 , Hans-Werner Schock 2
1 , Solarion AG, Leipzig Germany, 2 , Helmholtz-Zentrum für Materialien und Energie, Berlin Germany
Show AbstractIt is a well known fact that the quality of Cu(In,Ga)Se2 (CIGS) absorber layers can be enhanced by the addition of sodium. However, the way how sodium influences the electrical and structural parameters of CIGS absorber layers is still under discussion. Different methods for sodium incorporation are known for the use of sodium-free substrates like stainless steel or polyimide foil. In this work we compare sodium free CIGS-layers with CIGS-layers, where sodium was incorporated extrinsically at different growth stages or by a sodium-containing precursor. The CIGS samples were prepared via a roll-to-roll process with ion beam assistance (Solarion) or by a multi-stage low temperature co-evaporation process (HZB) respectively. The elemental depth profiles of the samples, especially for sodium, were studied via secondary ion mass spectroscopy (SIMS). In addition, we investigated a set of solar cell samples by means of quantum efficiency measurements and capacitance spectroscopy. It was observed that the depth profile for sodium is not directly correlated with the gallium gradient when an extrinsic sodium incorporation process is used. For sodium containing precursor layers the sodium gradient follows the gallium depth profile. The electrical parameters of the solar cells are correlated with the amount of sodium supplied either via a precursor, deposited prior to the absorber layer, or by extrinsic incorporation in the deposition process together with the metallic elements. With increasing amounts of sodium an increase of Voc is observed for both techniques. In contrast, Jsc decreases for extrinsic sodium incorporation while it remains unaffected for a precursor layer. Changes in Jsc are correlated with the quantum efficiency in the long wavelength region as well as in the short wavelength region. Higher amounts of sodium lead to a higher acceptor concentration which is usually correlated with a higher built-in voltage. This result is confirmed by capacitance spectroscopy investigations. In summary the beneficial effects of the addition of sodium to the growing CIGS-layer are shown. The efficiency of the CIGS-solar cells on polyimide foil was improved from 10 % (sodium-free) to up to 13.5 % for an optimal sodium content.
9:00 PM - Q3.20
Low Band Gap Poly(Thieno[3,4-b]Furan): A Potential Light-harvesting Material for Organic Photovoltaic Cells.
Tanmoy Dey 1 2 , Jayesh Bokria 1 2 , Michael Invernale 1 2 , Daminda Navarathne 1 2 , Zeki Buyukmumcu 3 , Gregory Sotzing 1 2
1 Department of Chemistry, University of Connecticut, Storrs, Connecticut, United States, 2 Polymer program, University of Connecticut, Storrs, Connecticut, United States, 3 Department of Chemistry, Erciyes University, Kayseri Turkey
Show AbstractThe field of optically transparent conducting polymers is an active research area due to their low density, flexibility and relatively low processing cost. Apart from their high number of conducting applications, these materials are used as hole-injection layers in photovoltaics. Our group has reported the polymerization of thieno[3,4-b]thiophene (T34bT) to yield a low bad gap conjugated polymer with Eg = 0.85 eV. The success of PEDOT-PSS as a commercially available low band gap conjugated polymer motivates us to study the properties of new conjugated polymers with fused ring structures, such as T34bF. Electrochemically prepared Poly(thieno[3,4-b]furan (PT34bF) was anion dominant upon redox cycling; its band gap was measured to be 1.04 eV by spectroelectrochemistry. This indicates the possibility of applications for the polymer as a transparent conductor or an ion-storage layer for electrochromic devices. Density Functional Theory (DFT) calculations using B3PW91 hybrid functional have been done for each of the possible connections between the three open α-positions. Besides optimizing the geometry, properties in relation to band structure such as band (energy) gaps, band widths, and effective masses were calculated for each connection. Based on the calculations, it can be concluded that the 4-6 connectivity is the most probable and dominant structure. The calculated energy gap of 1.01 eV for polymerization via the 4-6 connection corresponds well with the experimentally observed value. PT34bF has a good spectral match to the solar flux and exhibits appropriate band energies for exciton travel. With these two elements in mind, PT34bF could serve well as a potential light harvesting material in the active layer of an organic photovoltaic device.
9:00 PM - Q3.21
Photovoltaic Electrochemical Optical Rectennas of Carbon Nanotubes with Self-assembled Monolayers.
Juan Duque 1 , Matteo Pasquali 1 , Howard Schmidt 1
1 Department of Chemical and Biomolecular Engineering, Rice University, Houston, Texas, United States
Show AbstractCarbon nanotubes (CNT) and CNT-gold hybrid nanostructures are shown to produce rectified photocurrents without pre-bias when illuminated in the presence of aqueous electrolytes containing anionic surfactants. Graphite nanoplatelets, multi-wall carbon nanotube mats, single-wall carbon nanotube (SWNT) arrays and carbon black-loaded polymer all display this novel photovoltaic effect, and share a unique isolated photocurrent peak at ~300 nm. This feature apparently originates with their shared π-plasmon absorption. Rectified photocurrents are obtained only in the presence of anionic surfactants known to form self-assembled monolayers on graphitic surfaces. Carbon nanotube electrodes also produce minor photocurrents in the near infra-red. When coated with 5 nm of gold, single-walled carbon nanotube arrays generate similar photocurrents when treated with thiol-terminated anionic surfactant molecules. The Au-SWNT composite nanostructure displays a broad photocurrent spectrum spanning the entire visible region, along with a pronounced peak at ~400 nm. These features appear to arise from photoabsorption by gold nanorod plasmon hybridized with SWNT π-plasmons. A model explaining these results based on hot carriers traversing a rectifying tunnel barrier is presented and discussed. This mechanism is consonant with the concept of an optical rectifying antenna (rectenna), and could enable development of high efficiency photodetectors, photovoltaics and photochemical converters.
9:00 PM - Q3.22
Effects of Discotic-liquid Crystals as an Additive on the Performance of Polymer Solar Cells.
Seonju Jeong 1 , Cham Kim 1 , Jong Tae Kim 1 , Yun Seon Do 1 , Yoon Soo Han 1
1 , Daegu Gyeongpook Science & Technology, Daegu Korea (the Republic of)
Show AbstractAlthough an polymer solar cell is promising and new renewable resources for generation of electrical energy, its low power conversion efficiency (PCE) is considered as a major obstacle to its commercial utilization. The most promising method to improve power conversion efficiency of polymer solar cell is to control the morphology between the electron donor and the electron acceptor, which are conjugate polymer and fullerene derivatives, respectively, because exciton diffusion length is limited to 10~20 nm within organic polymer semiconductor. Besides, discotic liquid crystals (DLCs) are considered as promising organic semiconductors for application to molecular electronics, optoelectronics, photoconductor, photovoltaic solar cell and organic light emitting diode (OLED) devices. Organic devices with DLCs materials showed their high charge mobilities of up to 1 and 0.5 cm2/Vs in the crystalline solid phase and the columnar liquid crystalline phase, respectively. Especially, p-type discotic liquid crystals, triphenylene derivatives substituted with flexible alkoxy chains are widely studied because they show very high charge carrier mobility in various mesophase. With the concepts, by very simply incorporating a small portion of triphenylene derivatives such as DLC 1 and DLC 2, polymer photovoltaic cells with the configuration of ITO/PEDOT:PSS(30 nm)/P3HT:PCBM:DLC 1 or 2 (120 nm)/LiF (10 Å)/Al (100 nm) were fabricated. It was observed that electric - and optical properties were improved by the addition of a small amount of DLC 1 and DLC 2 into the film composed of P3HT and PCBM, caused by the change of morphology in active layer. Open circuit voltage (Voc) and fill factor (FF) of fabricated photovoltaic cells containing DLC 1 and DLC 2 were improved from 0.59 V and 57.19 (without DLC) to 0.65 V and 63.60 (with DLC 1) and 0.67 V and 64.94 (with DLC 2), respectively. This was considered to be due to the increased charge carrier mobility and the increased contact area between active layer and electrode by the introduction of columnar DLCs, as well as the wider absorption band. As the results, PCEs of the polymer cells were greatly improved from 3.0 % to 3.9 % (DLC 1) and 3.97 % (DLC 2), respectively.
9:00 PM - Q3.23
Effect of Annealing Solvent Solubility on the Performance of Poly(3-hexylthiophene)/Methanofullerene Solar Cells.
Jong Hwan Park 1 , Jong Soo Kim 1 , Ji Hwang Lee 1 , Wi Hyoung Lee 1 , Kilwon Cho 1
1 Chemical Engineering, Pohang University of Science and Technology, Pohang Korea (the Republic of)
Show AbstractThe effect of the solubility of the annealing solvent on the performance of poly(3-hexylthiophene) (P3HT):C61-butyric acid methyl ester (PCBM) solar cells is studied. The short-circuit current (Jsc) and the fill factor (FF) increase remarkably, regardless of the type of annealing solvent, whereas a reduction of the open-circuit voltage (Voc) (of 0.1~0.2V) is observed after solvent annealing. Interestingly, both the value of Jsc and the power conversion efficiency (PCE) are higher for the poor-solvent-annealed devices than for the good-solvent-annealed ones. A good solvent vapor induces better self-organization of P3HT than a poor solvent vapor. However, the exciton loss increases due to excessive phase separation. A study of the space-charge-limited current reveals no significant differences between the carrier mobilities of good- and poor-solvent-annealed devices. Furthermore, the measured photocurrent suggests that the space charges no longer limit the values of Jsc and FF for all the solvent-annealed devices. These results indicate that the higher Jsc and PCE values obtained for the poor-solvent-annealed devices can be attributed to the optimized phase separation of the active layers, which induces balanced carrier mobility and minimum exciton loss.Acknowledgement. This work was supported by a grant (F0004021-2008-31) from the Information Display R&D Center under the 21st Century Frontier R&D Program and Creative Research Initiative-Acceleration Research. (R17-2008-029-01001-0)
9:00 PM - Q3.24
Tuning of Absorbance Spectra in CdS/CdSe Quantum Dot Co-sensitized Solar Cells.
Dongho Lee 1 2 , Jeremy Nevins 1 3 , David Watson 1 3 , Alexander Cartwright 1 2 , Paras Prasad 1 2 3
1 Institute for Lasers, Photonics and Biophotonics, University at Buffalo, State University of New York, Buffalo, New York, United States, 2 Department of Electrical Engineering, University at Buffalo, State University of New York, Buffalo, New York, United States, 3 Department of Chemistry, University at Buffalo, State University of New York, Buffalo, New York, United States
Show AbstractWe investigate the solar power conversion efficiency (PCE) in quantum dot sensitized solar cells (QDSSC) using in-situ precipitation of a) CdS quantum dots, b) CdSe quantum dots, and c) cascade CdS/CdSe nanostructures. The sensitizing CdS and CdSe QDs were deposited directly on the TiO2 surface by chemical bath deposition (CBD) using cadmium, sulfur and selenium sources. Deposition of the CdS nanostructures on the TiO2 resulted in a gradual increase, and red-shift, of the absorbance as the number of CdS CBD cycles increased. Similarly, deposition of CdSe nanostructures on TiO2 resulted in an increase of the absorption for wavelengths less than the bandgap of CdSe (~700nm). Detailed analysis of the deposition of a cascade structure of CdS deposition followed by CdSe deposition resulted in the ability to tune the spectral response of the absorptive material. In general, the cascade structure resulted in better solar power conversion efficiency than the CdS or CdSe single QD structure solar cell due to the spectral response engineering. The shorter wavelength absorption (<550nm) is mainly due to CdS with a contribution from CdSe, while the longer absorption (550~700nm) is due to CdSe. The optimum structure was obtained with a CdS:CdSe cycling strategy of 6:6 cycles. This resulted in a 1.55% PCE, 0.53V open circuit voltage (Voc) and short circuit current density (Jsc) of 8.1mA/cm2 under AM1.5G one sun illumination. In general, increasing the CdSe cycles on the CdS QD layer increases the Voc, Jsc, and fill factor. However, excess cycles of either CdS or CdSe beyond 6:6 decreases the device performance.
9:00 PM - Q3.25
Sputter Deposition of CuInSe2 and CuGaSe2 from Composite Targets on (100) Si.
Okechukwu Akpa 1 , Shoieb Shaik 2 , Trenton Thompson 2 , Tamara Isaacs-Smith 3 , Philip Anderson 4 , Supapan Seraphin 4 , Kalyan Das 2
1 Material Science and Engineering, Tuskegee University, Tuskegee, Alabama, United States, 2 Electrical Engineering, Tuskegee University, Tuskegee, Alabama, United States, 3 Physics, Auburn University, Auburn, Alabama, United States, 4 Material Science and Engineering , University of Arizona, Tucson, Arizona, United States
Show AbstractCopper indium-gallium diselenide (CIGS) is an important material in thin film photovoltaics. Finding better ways to deposit stoichiometric high quality CIGS is a necessary part of furthering the use of this photovoltaic material. The growth and characterization of CuInSe2 and CuGaSe2 p-type material systems on n-type (100) silicon was studied to examine the characterization of chalcopyrite/Si heterojunctions for future use as solar cells. In addition chalcopyrite/Si heterojunctions using a graded CIGS structure were studied and will be analyzed. Film deposition was obtained by RF magnetron sputtering using stoichiometric targets at various temperatures. Visual inspection indicated that sputter deposited CuInSe2 and CuGaSe2 was translucent and highly reflective when deposited in thin layers. Rutherford backscattering spectroscopy (RBS) indicated that CuInSe2 films with a composition of Cu0.85In1.5Se2.5 were deposited on Si. The RBS profile also showed that CuGaSe2 layers were Se-rich and Cu-poor with a ratio of Cu0.5Ga1.5.Se3.0. Transmission electron microscopy (TEM) micrographs of earlier samples showed a high degree of contamination at the Si/chalcopyrite interface. The high levels of contamination were confirmed by energy-dispersive X-ray spectroscopy (EDS) and RBS. Despite the contamination, the materials were polycrystalline in nature and the diffraction patterns confirmed that the crystals were of reasonable quality. Cleaner chalcopyrite/Si interfaces were obtained using Fl terminated Si substrates. Micrographs, from TEM of these samples, showed that the native oxide was completely eliminated and EDS showed that contamination levels were significantly lower. The Hall-mobility experiments and electrical characterization of the heterojunction will be reported. (Work reported here was partially supported by NSF through RISE, CREST and IGERT grants. The authors would like to acknowledge Dr. John Williams of the Auburn University Department of Physics for the use of RBS facilities and Mr. Charles Ellis of the AMSTC Microelectronics Laboratory at Auburn University for providing laboratory facilities)
9:00 PM - Q3.26
Effect of Cu/Ga Molar Ratio on Optical and Electrical Properties of Spin Coated-CuGaSe2 Thin Films for Tandem Solar Cell Applications.
Ik Jin Choi 1 , Bhaskar Chandra Mohanty 1 , Yeon Hwa Jo 1 , Deuk Ho Yeon 1 , Yong Soo Cho 1
1 Materials Science & Engineering, Yonsei University , Seoul Korea (the Republic of)
Show AbstractThe wide-band gap chalcopyrite compound, CuGaSe2 is a promising material to achieve the highest levels of tandem-junction device performance. The CuGaSe2-based photovoltaic devices (PV) were fabricated with transparent back contact for use as a top cell in a tandem junction PV cells. The CuGaSe2 thin films were prepared by spin coating of solutions having different molar ratio of Cu and Ga (0.8 to 1.2) on Al-doped ZnO deposited glass substrate followed by subsequent selenization at various temperatures ranging from 350 to 550 °C. The effects of the Cu/Ga molar ratio and selenization temperature on various properties of the films were investigated. Below the selenization temperature of 450 °C, in addition to the chalcopyrite phase, secondary phases of Cu and Ga selenides were observed in the X-ray diffraction patterns irrespective of the starting compositions. However, phase pure chalcopyrite CuGaSe2 was obtained for films having Cu to Ga molar ratio of only 0.9, 1.0, and 1.1 and selenized at 500°C and above. Grain size of the films decreased with increase in deviation of Cu/Ga ratio from stoichiometric composition, i.e., 1.0. Results of Hall measurements for phase-pure samples indicate that CuGaSe2 films have p-type conductivity. Mobility of these films was found to be higher than those of the samples containing secondary phases. In this work, the highest conversion efficiency of about 4.1% has been achieved under the standard air mass (AM) 1.5 spectrum.
9:00 PM - Q3.27
High Quality Se-Excessive Cu(In, Ga)Se2 Films Prepared by Screen Printing.
Deuk Ho Yeon 1 , Bhaskar Mohanty 1 , Yeon Hwa Jo 1 , Ik Jin Choi 1 , Yong Soo Cho 1
1 Materials Science & Engineering, Yonsei University , Seoul Korea (the Republic of)
Show AbstractCIGS-based photovoltaic devices were prepared by the simple and cost effective screen printing method. The CIGS powder synthesized by the mechanochemical process was mixed with excess Se with Se content varying from 0 to 30 wt %. The excess Se was added with the intention of compensating the Se loss during subsequent process steps. This mixture (i.e., stoichiometric CIGS powder and Se) was added to an organic vehicle to prepare the CIGS paste. The paste was screen-printed on the substrate using 500 mesh screen and the organic residue was removed by sintering the obtained films at 300, 400, 500, and 600 oC in N2. Crystal structure and surface microstructure of the films were studied using X-ray diffraction and scanning electron microscopy. The thickness of the CIGS film was about 5~10μm. It was observed that thickness of the films prepared with increased Se amount was lower than those prepared with lower Se content. This suggests that addition of higher Se helped in densification of the films. Our cell fabricated in Glass/Mo/CIGS/CdS/ZnO structure showed an efficiency of about 1.7 % under AM 1.5 condition.
9:00 PM - Q3.28
Effect of Deposition Temperature on the Structural, Optical and Electrical Properties of ZnO:Al Deposited by Pneumatic Spray Pyrolysis.
Jagadeesh Bellam 1 , Vidhya Bhojan 1 , Arturo Maldonado 1 , Velumani Subramaniam 1
1 Electrical Engineering, CINVESTAV, Mexico .D.F. Mexico
Show AbstractTransparent Conducting Al-doped ZnO(AZO) thin films were prepared on glass substrates by Pneumatic spray pyrolysis method. The effect of substrate temperature on structural, optical and electrical properties of ZnO:Al thin films were investigated. 0.2M of Zinc Acetate Ac(Zn), 0.2M of Aluminum Pentanedionate P(Al) and a mixture of methanol, acetic acid, DI water were used as starting material, dopant source and solvent respectively. The films were grown at different substrate temperatures of 450°C, 460°C, 475°C, 485°C and 500°C. Structural, optical and electrical properties of the deposited films were investgated by X-ray diffraction, SEM, UV-VIS transmittance spectroscopy, Profilometry and Hall measurements. In addition films were also characterized by photoluminanscence(PL) studies and Raman spectroscopy. Simulation studies were carried out by Density Functional Theory (DFT).XRD measurements showed that the films were crystallized in wurtzite phase with preferential (002) orientation. FWHM, grain size, lattice constants and strain in the films were calculated. The structure of ZnO:Al has been simulated with close-packed O layers in h stacking, Zn in tetrahedral voids ,where some of the Zn atoms are replaced by Al atoms. Lattice constants a=3.248 Å and c=5.205Å were employed. Geometry optimization of the simulated structure has been carried out using Local-Density Approximation (LDA). Simulated XRD pattern is found to be in good agreement with the experimental pattern. SEM images revealed the surface morphology of the films. Influence of deposition temperatures on the optical properties of the films has been studied. The average optical transmittances of all films were 85% in the visible range of wavelength 400 to700 nm. The calculated direct band gap value was found to be in the range from 3.320 eV to 3.350 eV.Resistivity of the films varied from 1.42×10-2 Ω-cm to 1.62×10-3 Ω-cm. Minimum resistivity is obtained for the film deposited at 485°C. PL studies showed a strong near band edge UV emission peak at 380nm and a weak deep level emission centered at about 500nm. In the Raman analysis, the active mode of Al doped ZnO films were observed at 125cm-1(low) and 497cm-1(high).
9:00 PM - Q3.29
Effect of Pitch and Height on the Performance of Textured Polymer Solar Cells.
Kanwar Nalwa 2 , Joong-Mok Park 2 , Wai Leung 2 , Kai-Ming Ho 2 , Sumit Chaudhary 2
2 , Iowa State Univ, Ames, Iowa, United States
Show AbstractPolymer based photovoltaics (PPVDs) are attractive for solar-electric conversion due to potentially solution-processible, low-cost, roll-to-roll manufacturing capability. The charge carrier mobility in conjugated polymers remains low, which demands thin (tens of nm) active photovoltaic layers for high efficiencies. However, efficient optical absorption is only ensured by thicker films (100 - 200 nm). Thus, arises a trade-off situation between the optical and electronic phenomena. Introducing micron- and submicron-scale textures in such PPVDs is one of the possible ways to achieve efficient light absorption in very thin active layers. We fabricated PPVDs on such textured (grating-type) surfaces realized by laser holographic lithography. PPVDs on 600 nm pitch and 300 nm height grating showed higher power conversion efficiency than those on 2 micron pitch and 1.5 micron height owing to more uniform spin coated active organic layer and lower series resistance. We also observed that the variation in pitch and height of such grating-type structures has a significant effect on spectral responsivity; indicating that wavelength tunable photodetection can readily be achieved with the same materials deposited on different geometries. For example, 600nm pitch and 300 nm height PPVDs demonstrated better relative spectral response at wavelengths less than 500nm, as compared to the flat PPVDs. Our observations also include the conflicts between the promising optics and challenges of solution processible deposition on textured surfaces.
9:00 PM - Q3.3
Titanium-Silicon Dioxide as a Transparent Conducting Oxide and an Anti-Reflection Contact for Photovoltaic Applications.
John Chivers 1 , Chandler Downs 1 , Thomas Vandervelde 1
1 ECE, Tufts University, Medford, Massachusetts, United States
Show AbstractWe report on the use of Earth-abundant silicon-dioxide and titanium-dioxide as a transparent conducting oxide (TCO) and antireflective (AR) coating. The varied band gap and index of refraction conditions inherent in the SiTiO2 system allow for an easily adaptable set of conditions, conferring the growth of these materials monolithically. TCOs are a critical component in modern photovoltaic devices, used as a front-side contact that won’t block incident light. At present, many TCOs require rare-Earth materials (e.g. Indium), which will likely be problematic for widespread, long-term distribution. The abundant, well characterized materials used here can be integrated into an existing line quickly and cheaply. Some TCOs also act as an AR coating, further increasing the light absorbed. The ideal AR coating would gradually change from the index of refraction of air to that of the underlying semiconductor. Most AR coatings used today make this transition in a small number of steps, which limits their efficacy. In this work, we use a PECVD process that slowly grades the index of refraction while maintaining conductivity and transparency. In the drive towards grid parity, the quality of the TCO and AR coatings could prove crucial. This work would contribute significantly to that effort.
9:00 PM - Q3.30
Optical and Micro-structural Characterization of Printed Cu2ZnSnS4 Thin Films.
Inyoung Kim 1 , Ki Woong Moon 2 , Kwang-Won Jeon 2 , Jongryoul Kim 2 , Jong-Su Yu 1 , Jeongdai Jo 1 , Dong-Soo Kim 1
1 Nano-Mechanical Systems Research Division, Korea Institute of Machinery & Materials, Daejeon Korea (the Republic of), 2 Department of Metallurgy and Materials, Hanyang University, Ansan Korea (the Republic of)
Show AbstractPrinting technologies, inkjet, roll-to-roll, gravure and so on, have been applied to the fabrication of electronics devices such as TFT, OLED, RFID, LCD color filters. Nowadays their application fields have been expanded to the energy field such as solar cell and battery. One of the main reasons, why many researchers have been interested in printing technology as a manufacturing method, is the reduction of manufacturing cost. In particular, the urgent subject of solar cells for triggering the market growth is also the reduction of cost per watt. For the fabrication of printed the solar cells, printing solutions were developed in several ways but they were usually toxic and scarce materials used [1-3]. However, it is inconsistent idea to use these materials for the cheap production of clean energy. In this study, ink for printed Chalcopyrite solar cell was synthesized using modified sol-gel method. For the free of toxic and scarce materials, CuZnSnS (CZTS), was selected. CZTS ink was printed on the Mo coated SLG glasses and sintered at the ranges of 300 ~ 550 degrees Celsius. Parametric studies were conducted in order to obtain printed Cu2ZnSnS4 phase changing precursor synthesis conditions, ink formulation, binder addition and thermal treatment. Optical and microstructural aspects were discussed for the application of photovoltaic devices of printed CZTS films. [1] D. B. Mitzi, M. Yuan, W. Liu, A. J. Kellock, S. J. Chey, L. Cignac, A. G. Schrott, Thin Solid Films 517 (2009) 2158-2162.[2] J. A. Hollingsworth, K. K. Banger, M. H. -C. Jin, J. D. Harris, J.E. Cowen, E. W. Bohannan, J. A. Switzer, W. E. Buhro, A. F. Hepp, Thin Solid Films 431-432 (2003) 63-67.[3] V. K. Kapur, A. Bansal, P. Le, O. I. Asensio, Thin Solid Films 431-432 (2003) 53-57.
9:00 PM - Q3.5
The Effects of Bathocuproine Incorporation on the Properties of Organic Photovoltaic Devices.
Ching-Chun Chang 1 , Chi-Feng Lin 2 , Jian-Ming Chiou 3 , Tzung-Han Ho 4 , Kuei-Hsien Chen 1 5 , Li-Chyong Chen 5
1 Institute of Atomic and Molecular Sciences, Academia Sinica , Taipei Taiwan, 2 Graduate Institute of Photonics and Optoelectronics, National Taiwan University, Taipei Taiwan, 3 Department of Chemical Engineering, National Taiwan University of Science and Technology, Taipei Taiwan, 4 Department of Physics, National Taiwan University, Taipei Taiwan, 5 Center for Condensed Matter Science, National Taiwan University, Taipei Taiwan
Show AbstractIn the present work, we have investigated the effects of cathode buffer layers on the poly(3-hexylthiophene):[6,6]-phenyl C61-butyric acid methyl ester (P3HT:PCBM) blend photovoltaic cells. The results indicate that the performance of the bulk-heterojunction photovoltaic devices is improved considerably by using ultra-thin layer (20 Å) of bathocuproine (BCP). We find that the thin BCP layer increases both the open circuit voltage (Voc) and the fill factor (FF) of the device, yielding an increased power conversion efficiency (PCE). Compared with the bufferless device, obvious enhancements of Voc from 0.35 to 0.62 V and FF from 41 to 65% are obtained, which make the overall device PCE improved by three times (from 1.3% to 4.1%). The significant enhancement may be attributed to the passivation for the contacts by preventing aluminum compounds and dipole moment formation at the interfaces between active materials and Al cathodes. Thus, the introduction of cathode buffer layer such as lithium fluoride (LiF) and tris(8-hydroxyquinolinato)aluminium (Alq3) are also studied for clarification purpose. The interfacial reactions between P3HT:PCBM blending layers and Al cathodes were investigated by using X-ray photoelectron spectroscopy (XPS) depth profiles. For the bufferless device, Al cathode reacts with the underneath active materials to form Al-C compounds, which decreases the built-in potential in the device and thus lower Voc. Nevertheless, for the device with BCP buffer, the outer-diffusion of carbon from P3HT:PCBM blending layer can be prohibited and charge injection will be improved under forward bias. Therefore, the device shows even better performance than the one with commonly used LiF thin film.
9:00 PM - Q3.6
Effects of Non-noble Metal Current-collecting Grids on Internal Resistance for Dye-sensitized Solar Cells.
Seon Hee Seo 1 , Hyun-Ju Kim 1 , Bo-Kun Koo 1 , Dong Yoon Lee 1
1 , Korea Electrotechnology Research Institute, Changwon, Kyungsangnam-do, Korea (the Republic of)
Show AbstractWe studied the effectiveness of various non-noble metals used as current-collecting grids on fluorine-doped SnO2 (FTO) for dye-sensitized solar cells (DSCs). Non-noble metal grids were deposited by dc magnetron sputtering and mechanical adhesion of non-noble metals was improved by a 300-nm-thick Cr interlayer between FTO and grid. The metals were evaluated on the basis of J-V characteristics and impedance spectra. Analysis using an equivalent circuit model showed that such metal grids affect the internal ohmic resistance, which is inversely related with fill factor. A 1-um-thick NiCr grid afford a fill factor of 0.639—as much as that with Ag—but with a slightly lower efficiency of 6.4 %, which supports the feasibility of using stable non-noble metal grids in DSCs, albeit with a 5 % reduction in photovoltaic efficiency.
9:00 PM - Q3.8
Photosensitization of TiO2 by Electron Beam Irradiated Silica and Titania Based Hybrid Polymer for PEC Cell Application.
Seung Hwa Yoo 1 , Sung Oh Cho 1
1 Nuclear and Quantum Engineering, Korea Advanced Institute of Science and Technology, Daejeon Korea (the Republic of)
Show Abstract In this study, mesoporous titania nanoparticle was sensitized by electron beam irradiated silica and titania based polymer. Titania nanoparticle was fully dispersed in ethanol by ultrasonication and was spin coated on the substrate. The prepared titania film was immersed in a dilute solution of the polymer overnight to sufficiently cover the porous film. After drying at ambient condition, the sample was subjected to electron beam irradiation. The beam fluence was controlled to observe the absorbance change of the samples by UV-VIS measurement. The band structure of the irradiated polymer was studied by several methods including UV-VIS spectroscopy, cyclic voltammetry and PES(photoelectron spectroscopy). Yet only the valance band of the irradiated polymer was estimated by PES. This sensitized titania film coated on ITO glass was installed in a PEC(photoelectochemical) cell including a platinum electrode as counter electrode and SCE(saturated calomel electrode) as reference electrode. NaOH aqueous solution (pH 11) was applied as electrolyte. Certain photocurrent was measured under 500W Xe lamp with a 420nm cut-off filter. Estimation of the band structure and chemical structure of the irradiated polymer are remaining questions in this study. Also improvement of the photocurrent generation is required in further study.
9:00 PM - Q3.9
Synthesis of Organic Semiconducting Materials by Electron Irradiation and Their Application to Organic Solar Cell.
Hyeok Moo Lee 1 , Sung Oh Cho 1
1 Nuclear and Quantum Engineering , Korea Advanced Institute of Science and Technology, Daejeon Korea (the Republic of)
Show AbstractElectron irradiation is one of the efficient methods to change the optical properties of organic materials. Chemical structures of polymeric materials are easily changed by electron irradiation. This suggests that physical and chemical properties of the materials, including optical property, can be modified by the irradiation. Based on the fact, we present a novel route to synthesize organic semiconducting materials by electron irradiation. Finally, we proposed a possibility that synthesized semiconducting materials can be used for organic solar cell as an electron donor. Electron irradiation was carried out in a vacuum chamber under a pressure of less than 2×10-5 torr using a thermionic electron gun and energy of an electron beam was 35 keV and the beam current density was 4 μA×cm-2. In order to synthesize proper semiconducting materials which have low bandgap, high hole mobility and high absorption coefficients in visible range various non-conducting polymer, semiconducting polymer and organic hole-transport materials are used as a precursor. The optical properties and electronic band structure of synthesized semiconducting materials were characterized by photoluminescence spectroscopy, UV/Vis absorption spectroscopy, and photoelectron spectroscopy in air. The mock-up organic solar cells are simply fabricated using ITO glass and Al electrode. Power conversion efficiency of solar cell was calculated under AM 1.5G solar illumination at 100mW×cm-2.
Symposium Organizers
Bhushan L. Sopori National Renewable Energy Laboratory
Bernhard Dimmler Würth Solar GmbH & Co. KG
Jeffrey Yang United Solar Ovonic LLC
Thomas Surek Surek PV Consulting
Q4: PV Devices and Module Issues
Session Chairs
Tuesday AM, December 01, 2009
Room 306 (Hynes)
9:30 AM - Q4.1
Investigation on the Optical and Electrical Properties of a Novel Synthesized Copolymer Designed for Efficient Polymer Solar Cells.
Fengling Zhang 1 , Weiwei Li 1 2 , Ruiping Qin 2 3 , Yi Zhou 1 , Fenghong Li 1 , Clemens Veit 4 , Hans-Frieder Schleiermacher 4 , Yi Thomann 5 , Mattias Andersson 1 , Viktor Andersson 1 , Olle Inganas 1 , Uli Wuerfel 4 5 , Zhishan Bo 2
1 , Linkoeping University, Linkoeping Sweden, 2 , Institute of Chemistry, Chinese Academy of Sciences, Beijing China, 3 , Beijing Normal University, Beijing China, 4 , Fraunhofer Institute for Solar Energy Systems ISE, Freiburg Germany, 5 , Freiburg Materials Research Center, Freiburg Germany
Show AbstractWe present a novel synthesized copolymer (HXS-1) with 2,7-carbazole-alt-5,5-(4’,7’-di-2-thienyl-2’,1’,3’-benzothiazole) as the backbone of the electron donor polymer for polymer photovoltaic application. HXS-1 is designed with two octyloxy side chains on the benzothiazole ring and an octyl chain on carbazole ring. The solubility of HXS-1 in 1,2-dichlorobenzene(DCB) is not good, but it can be dissolved by heating DCB. HXS-1 film shows broad absorption from UV to 700 nm. Bulk heterojunction solar cells were fabricated from HXS-1 mixed with PC71BM in DCB mixed with 2.5% 1,8-diiodooctane (DIO). The devices were optimized with varied stoichiometries, thicknesses and DIO concentration. A short-circuit current of 9.5 mA/cm2, open-circuit voltage of 0.79 V, fill factor of 0.72 and a power conversion efficiency of 5.4% under AM1.5 (100 mW/cm2, corrected for spectral mismatch) were achieved with the stoichiometry of 1:2.5 by weight, the thickness of 100±10 nm from DCB:DIO, The FF of 0.72 at Jsc =9.5 mA/cm2 indicates that HXS-1 has decent and balanced charge carrier mobility, as verified by field effect transistor measurements. The high photovoltage, extensive spectral coverage and superior electrical transport is attractive. Combining these merits in one polymer makes HXS-1 a very promising polymer for photovoltaic application. The optical and electrical properties of HXS-1 were intensively investigated with photoluminescence and electroluminescence in neat and blend films. The morphologies of active layers of solar cells were imaged with both AFM and electron tomography.
9:45 AM - Q4.2
Polymer Tandem Solar Cells with 3 Terminal Parallel Connected Structure.
Srinivas Sista 1 , Ziruo Hong 1 , Mi-Hyae Park 1 , Yang Yang 1
1 Materials Science & Engineering, University of California Los Angeles, Los Angeles, California, United States
Show AbstractHere we present a 3 terminal tandem cell where in the two sub cells are connected in parallel as opposed to in series that has been reported in literature so far. In this device architecture the two sub-cells are connected in parallel through a transparent conducting interlayer that acts as a common electrode to the two sub-cells. We fabricated a 3 terminal tandem cell with poly(3-hexyl thiophene) as the front cell and poly[(4,4′-bis(2-ethylhexyl) dithieno[3,2-b:2′,3′-d]silole)-2,6-diyl-alt-(2,1,3 benzothiadiazole)-4,7-diyl] (PSBTBT) as the rear cell. The front P3HT cell has inverted structure while the rear cell fabricated from PSBTBT has regular structure. The interlayer connecting the two cells acts as an efficient hole collector from the two cells and delivers it to the external circuit. The front P3HT:PCBM cell showed power conversion efficiency (PCE) of 3.5% with short circuit current (Jsc) of -10.1mA/cm2, open circuit voltage of (Voc) of 0.58V and fill factor (FF) of 60.0%. While from the rear PSBTBT:PCBM we obtained PCE of 1.3%,, Jsc of -5.1mA/cm2, Voc of 0.6V and FF of 39.5%. The two sub cells were connected in parallel with the interlayer acting as the anode and the cathodes of the bottom and front cells shorted to form the cathode connection. The resulting tandem cell showed a Jsc of -15.1mA/cm2, Voc of 0. 60V, PCE of 4.80% and FF of 52.5%. The main advantage of this structure is that the efficiencies of the sub-cells can be measured independently in the tandem structure and as well as when connected in parallel. Another advantage of this device structure is that the current matching between the two sub-cells is not a necessary criterion for achieving an efficient parallel connected tandem cell. With the help of an efficiently conducting and transparent interlayer, we observe an added up short circuit current from the two sub-cells. This 3 terminal device configuration is particularly important for enhancing the absorption range by utilizing two polymer with complementary absorption range or the absorption itself when the same polymer is used for the two sub cells. Thus this device configuration poses great promise in further enhancing the polymer solar cell efficiencies.
10:00 AM - Q4.3
Analysis of Tunnel Junction Suitable for 4000 Suns in Tandem High Efficiency Solar Cell Structures.
Salah Bedair 1 , John Hauser 1 , Donggeun Jung 2
1 Electrical and Computer Engineering, North Carolina State University, Raleigh, North Carolina, United States, 2 Department of Physics, Sungkyunkwan University, Kyonggi-do Korea (the Republic of)
Show AbstractThe road to achieve ultra high efficiency is through multi junction solar cells operating at high solar concentration, larger than 1000 suns. Critical to the success of this approach is the development of tunnel junction (TJ) that serve as electrically low loss interconnections yet optically transparent using high band gap semiconductor material systems. We have previously reported the fabrication of TJ made of n+-InGaP/ p+-AlGaAs with a band gap about 1.9 eV using Se and C doping respectively. This TJ structure has a peak current density of 80A/cm2 allowing it to be implemented in a three junction cell structure for use at solar concentration as high as 4000 suns (x4000). Almost all reported conversion efficiencies higher than 40% have used this tunnel junction. This unexpected very high peak current density, in these high band gap material systems represents good news for the multi junction solar community. The high current seems to be due the fact that the InGaP/AlGaAs interface has a staggered band gap lineup. We will present the effect of this band line up at the hetero interface and its effect on the width of the depletion region and the peak current density. We will also compare the current result from this hetero structure junction with an artificial homo junction made of n+-AlGaAs/ p+-AlGaAs doped to the same levels as that of the hetero junction. Results from the homo junction showed that peak current density is about one half of that obtained from the hetero junction at the same doping levels. A reasonable match between experimental result and the model was obtained when a value of 150 meV was used for ΔEc, the conduction band discontinuity at the interface. Both experiment and theory predicted that at a current density of 80A/cm2, there is only about a few tens of meV voltage drop across the TJ. This will have minimal effects on the over all efficiency of the tandem solar cell structure when used at high solar concentrations.We will also report on the effect of thermal annealing at 650°C and 750°C for 30 minutes to simulate the post growth conditions after the formation of the TJ, on its electrical properties. We will also report on the effect of inter diffusion of impurities at the junction interface on the width of the deletion region and peak current density. The peak current density was reduced by about 20% due to thermal annealing at 650°C.
10:15 AM - Q4.4
Domain Wall Driven Anomalous Photovoltaic Effect in a Complex Oxide.
Seung-Yeul Yang 1 , Jan Seidel 2 3 , Steve Byrnes 2 3 , Padraic Shafer 1 , Chanho Yang 3 , Marta Rossell 4 , Joel Ager 2 , Lane Martin 2 , Ramamoorthy Ramesh 1 2 3
1 MSE, U.C. Berkeley, Berkeley, California, United States, 2 Materials Sciences Division, Lawrence Berkeley National Laboratory, Berkeley, California, United States, 3 Physics, U.C. Berkeley, Berkeley, California, United States, 4 National Center for Electron Microscopy, Lawrence Berkeley National Laboratory, Berkeley, California, United States
Show Abstract While progress in photovoltaic systems proceeds mainly using standard semiconductor p-n junction or heterojunction-based devices, alternative approaches are beginning to receive a renewed interest. We show evidence for an anomalous photovoltaic effect in the complex oxide material BiFeO3 under visible light illumination leading to photovoltages larger than the band gap of the material (2.67 eV). BiFeO3 films with two different domain patterns, possessing 109° and 71° domain walls, were prepared on single crystalline (110) DyScO3 substrates by metal-organic chemical vapor deposition (MOCVD). In the direction of parallel to the net in-plane polarization, a large photo-induced VOC of 16 V was measured, with in-plane current density Jsc ~ 1.2×10-4 A/cm2 from 200μm electrode spacing. The effect arises from a new mechanism, namely structurally driven steps of the electrostatic potential that occur at ferroelectric domain walls. These potential steps—and, in turn, the photovoltaic effect—can be fully controlled through the domain structure in these films.
10:30 AM - Q4.5
Downconversion for Photovoltaics in Lanthanide Doped Low Phonon Energy Host Lattices.
Timon van Wijngaarden 1 , Karl Kraemer 2 , Hans-Ulrich Guedel 2 , Joren Eilers 1 , Linda Aarts 1 , Bryan van der Ende 1 , Andries Meijerink 1
1 , Debye Institute, Utrecht Netherlands, 2 , Department of Chemistry, Bern Switzerland
Show AbstractState-of-the-art commercial crystalline Si (c-Si) solar cells dominate the market and have energy efficiencies around 15%. The main energy losses (over 70%) are related to the spectral mismatch.[1] IR photons with energies lower than the bandgap are not absorbed while for photons with energies exceeding the bandgap, the excess energy is lost as heat during the fast thermalization. Two methods are capable of reducing spectral mismatch losses: upconversion and downconversion.[2] In case of upconversion, two infrared photons are ‘added up’ to give one higher energy photon that can be absorbed. The opposite process, downconversion, involves ‘cutting’ of one high energy photon into two lower energy photons. This process can reduce energy losses related to thermalization.In the present work we explore the possibilities for downconversion of a single visible or ultraviolet photon into two near infrared (NIR) photons using lanthanide ions doped into low frequency host lattices. The unique and rich energy level structures of lanthanide ions allow for efficient up- and downconversion. The Yb3+ ion has a single excited state some 10,000 cm-1 ground state. The absence of other energy levels allow Yb3+ to ‘pick up’ energy packages of 10,000 cm-1 from other lanthanide ions and emitting ~980 nm photons that can be absorbed by c-Si. Efficient downconversion using Yb3+ as acceptor requires donor ions with an energy level around 20,000 cm-1 and an intermediate level around 10,000 cm-1. Evaluation of the Dieke diagram reveals that potential downconversion couples are (Er3+, Yb3+), (Nd3+, Yb3+) and (Pr3+, Yb3+). Recently we demonstrated efficient downconversion for the (Pr3+, Yb3+).[3] Work on the (Er3+, Yb3+), (Nd3+, Yb3+) couples showed that fast multi-phonon relaxation from the starting level around 20 000 cm-1 prevented efficient downconversion in fluorides. Here we demonstrate the feasibility of efficient downconversion for both(Er3+, Yb3+) and (Nd3+, Yb3+) by doping them into bromides and chlorides. The low phonon energies in these host materials reduce multi-phonon relaxation rates and the level around 20 000 cm-1 is sufficiently long lived to achieve efficient two-step energy transfer. The dynamics and efficiency of the downconversion process are evaluated by steady state and time resolved luminescence measurements. Especially in the bromides (Cs3Y2Br9 and CsCdBr3) efficient downconversion is observed leading to strong infrared emission from Yb3+ around 980 nm upon excitation in the 4F7/2 level of Er3+ or the 4G9/2 level of Nd3+. The low phonon energies also aid in reducing quenching processes for the Yb-emission. This makes these systems promising for efficient downconversion for c-Si solar cells. [1] B. S. Richards, Sol. En. Mat. Sol. Cells 2006, 90, 2329-2337.[2] T. Trupke, M.A. Green, P. Würfel, J. Appl. Phys. 2002, 92, 1668; J. Appl. Phys. 2002, 92, 4117.[3] B. van der Ende, L. Aarts and A. Meijerink, Advanced Mater., in press.
10:45 AM - Q4.6
Photovoltaic Devices Based on Carbon Nanostructured Materials.
Zhongrui Li 1 , Viney Saini 1 , Yang Xu 1 , Enkeleda Dervishi 1 , Meena Mahmood 1 , Alexandru Biris 1
1 , University of Arkansas at Little Rock, Little Rock, Arkansas, United States
Show AbstractSuperior properties of nanostructural carbon materials (carbon nanotubes and graphene) represent attractive materials for photovoltaic devices. Single-wall carbon nanotubes (SWNTs) can be directly configured as energy conversion materials to fabricate thin-film solar cells, with nanotubes serving as both photogeneration sites and charge carriers collecting/transport layers. SWNTs can be modified into either p-type conductor through chemical doping (like thionyl chloride, or just exposure to air) or n-type conductor through polymer (like polyethylene imine) functionalization. The solar cells consist of either a semitransparent thin film of p-type nanotubes deposited on an n-type silicon wafer or a semitransparent thin film of n-type SWNT on p-type substrate to create high-density p-n heterojunctions between nanotubes and silicon substrate to favor charge separation and extract electrons and holes. The high aspect ratios and large surface area of nanotubes could be beneficial to exciton dissociation and charge carrier transport thus improving the power conversion efficiency. Initial tests have shown a power conversion efficiency of >4%, proving that p-n SWNT configuration is a potentially suitable configuration for making solar cells. Another prototype employ the use of conducting polymeric materials combined with graphene layers to achieve higher efficiency of such photovoltaic devices. A higher dissociation/transportation of the charge carriers is expected for such nanocomposite materials while used as p/n layers.
11:30 AM - **Q4.7
Using Accelerated Testing to Reduce PV Module Material Costs without Sacrificing Reliability and Lifetime.
John Wohlgemuth 1 , Zhiyong Xia 1 , Daniel Cunningham 1
1 , BP Solar International Inc., Frederick, Maryland, United States
Show AbstractFor photovoltaics (PV) to compete with other forms of electricity, the cost of PV module materials and processes must be reduced from today’s levels. However, this should be accomplished without negatively impacting module reliability and lifetime. This paper will discuss the types of accelerated testing and engineering analysis that can be done to determine whether proposed changes of materials or processes have reduced the ability of the modules to survive in the field. Outdoor testing is essential for all PV technologies in order to identify failure mechanisms. The long term durability and reliability of today’s crystalline silicon modules means that changes in module performance occur slowly outdoors. Module manufacturers can not wait for 20 to 25 years to see if a new lower cost material (e.g. encapsulant) continues to perform well in the field. Therefore, accelerated stress tests like thermal cycling, humidity freeze in combination with thermal cycling, and damp heat have been developed to assess the modules susceptibility to known failure mechanisms. The critical issue discussed in this paper is how to utilize these accelerated tests to evaluate the potential impact on longevity of using new materials and processes in module construction.
12:00 PM - Q4.8
Self-Cleaning Solar Panels and Solar Concentrators with Integrated Transparent Dust Screens.
Malay Mazumder 2 1 , Mark Horenstein 2 , Rajesh Sharma 3 , Hidetaka Ishihara 1 , Jeremy Stark 1
2 ECE Department, Boston University, Boston, Massachusetts, United States, 1 Applied Science, University of Arkansas - Little Rock, Little Rock, Arkansas, United States, 3 Renewable Energy, Arkansas State University, West Memphis, Arkansas, United States
Show AbstractIf only 1% percent of our land is dedicated to the use of solar panels for harvesting energy with only 10% efficiency, most of need for electricity can be met. The investment will be huge but it will be paid off within a reasonable time period. Long term operation of the photovoltaic (PV) and photothermal (PT) systems with a high efficiency requires that the panel surface and the surface of the optical concentrators should be clean for efficient conversion of solar energy by the solar energy systems. Deposition of dust on the light harvesting surfaces, particularly when these are installed in dusty areas, desert, or along the highways, could severely minimize solar-to-power output efficiency. Under a NASA project, we have developed transparent electrodynamic screens (EDS) for protecting solar panels against obscuration by dust deposition on the surface of the moon and Mars. We will describe the applications of this technology for terrestrial application worldwide. The electrodynamic screens (EDS) are made of transparent polymer sheets, such as PET (for its UV radiation resistance) or glass plates and a set of parallel conducting transparent electrodes made of Indium Tin Oxide (ITO) embedded under a thin film transparent coating. The basic principle of EDS operation is based upon the electrostatic and dielectrophoretic forces applied by the electric field to remove the dust particles deposited on the surface of the panel. Mathematical models and experimental studies show that when the screens are energized, dust removal efficiency higher than 95% can be achieved within a short time period (< 1min.). The method is effective for both charged and uncharged particles over a wide range particle size and particle dielectric constant. The results show that EDS technology is applicable for protecting solar panels and solar panels against dust hazards effectively in both space and terrestrial applications. The power requirements will be approximately 10 watts per square meter of the surface to be cleaned when dust removal is needed. Under normal conditions, dust cleaning of solar panels may not be required for more than a few minutes per day while the average power produced per square meter of the panels during the peak hours is approximately 500 watts. For solar concentrators, such as mirrors and Fresnel lenses, the transparent dust shields can be powered by using solar panels. A brief description of the principles self-cleaning system, experimental results, and their automated operation will be presented.
12:15 PM - Q4.9
Photocharge Generation on the Surface of Mesoscopic PCBM Crystals Grown by Dip-coating as Visualized by Kelvin Probe Microscopy.
Reza Dabirian 1 , Vincenzo Palermo 1 , Luca Ortolani 2 , Vittorio Morandi 2 , Xinliang Feng 3 , Klaus Muellen 3 , Paolo Samori 1 4
1 Istituto per la Sintesi Organica e la Fotoreattività, Consiglio Nazionale delle Ricerche, Bologna Italy, 2 Istituto per la Microelettronica e Microsistemi, Consiglio Nazionale delle Ricerche, Bologna Italy, 3 , Max-Planck Institute for Polymer Research, Mainz Germany, 4 Nanochemistry Laboratory , ISIS/CNRS Université de Strasbourg, Strasbourg France
Show AbstractIn organic electronics, the ability to self-assemble well-defined nanostructures with tunable nanoscale electronic functions, such as charge generation and transport, is fundamental for both basic research and device fabrication. Herein we present the preparation of mesoscopic crystals of a soluble fullerene, PCBM ([6,6]-phenyl C61-butyric acid methyl ester) using a novel dip-coating technique. PCBM is a widely used electron-acceptor in organic electronics.Mesoscopic PCBM crystals, with a diameter ranging from 1 to 100 μm and a thickness from 20 to 500 nm, have been obtained by dip-coating the substrates into a solution containing PCBM. These crystals have been prepared on a wide variety of surfaces such as silanized SiOx, Au, Cu, graphite, ITO and amorphous carbon. The size, shape and distribution of the crystals can be tuned by regulating the temperature and dip coating speed. Their multiscale characterization has been accomplished by atomic force microscopy (AFM), Kelvin probe force microscopy (KPFM), transmission electron microscopy (TEM) and optical microscopy. TEM measurements revealed exclusively large single crystals with a confirmed hexagonal symmetry. AFM studies showed that the crystal surface is made of regular, molecularly flat terraces, up to over 1 μm in size, with a roughness Rrms=0.6±0.2 nm over an area of 1 μm2. The obtained crystals represent an ideal optically-active substrate upon which a second material can easily be deposited, ideally using an orthogonal solvent. By using this multiple layer deposition strategy, the PCBM crystals were partially coated with solution-processed nanofibers of a functionalized electron-donor (hexa-peri-hexabenzocoronene, or HBC), allowing quantitative understanding of the photovoltaic activity in bi-component electron-acceptor and donor systems. The different photovoltage build-up upon illumination on neat PCBM crystals respect to those covered by either a few HBC nanofibers or a tens of nm thick HBC film has been measured by KPFM. The measurements of these nanostructured interfaces in darkness revealed potential differences between the two components that increase by an average of 50 mV upon illumination, providing new insight into the physics of charge generation at the acceptor-donor interface.
12:30 PM - Q4.10
Raman and Rutherford Backscattering Characterization of Ti-implanted Si above the Mott Limit.
David Pastor 1 , Javier Olea 1 , Ignacio Martil 1 , German Gonzalez-Diaz 1 , Jordi Ibanez 2 , Ramon Cusco 2 , Luis Artus 2
1 Dpto. de Física Aplicada III, Facultad de Ciencias Físicas, Universidad Complutense de Madrid, Madrid Spain, 2 Institut Jaume Almera, Consejo Superior de Investigaciones Científicas (CSIC), Barcelona Spain
Show AbstractOne of the most relevant ideas to increase the efficiency in solar cell technology is the intermediate band solar cell (IBSC), which could improve significantly the solar conversion efficiency limit for single junction solar cells. The active material of an IBSC exhibits a new band located between the conventional valence band (VB) and the conduction band (CB) of the semiconductor. This intermediate band (IB) divides the band gap into two low energy sub bandgaps. The electrons generated by photon absorption of energy below the band gap can be pumped from the VB to the CB in two stages using the IB as intermediate step. Given that Si solar cell technology is widely developed, Si-based IB materials are highly attractive. The formation of the IB material requires the introduction of deep centers in the semiconductor with a concentration above the Mott limit (∼5x1019 cm-3). Ti is one of the most promising impurities to form the IB in Si. However, the high concentration required clearly exceeds the solubility limit of Ti in Si (∼4x1014 cm-3). Ion implantation is an appropriate technique to achieve the high impurity concentration needed to form the IB material; however, the lattice damage generated by the ion implantation has to be removed. Non equilibrium thermal processes like pulsed laser melting (PLM) are necessary to recover the crystal lattice and avoid impurity out-diffusion. In the present work, we have used Raman spectroscopy and Rutherford Backscattering (RBS) to asses the degree of lattice recovery achieved after PLM annealings in Ti-implanted silicon with concentrations well above the Mott limit. The analysis of the RBS data provide us information about the location of Ti in the lattice of PLM annealed Si, where IB-related electrical behavior has already been observed in this material. We have analyzed samples implanted with Ti doses of 1015, 5x1015, 1016 and 5x1016 cm-2 and subsequently annealed at energy densities of 0.2 and 0.8 J/cm2. Raman measurements performed with 458 and 514 nm excitation wavelengths show a reduction in the lattice crystallinity with increasing implantation dose. For the sample implanted with the lowest dose, a high degree of lattice recovery is achieved after the PLM annealings at the highest energy density. The RBS data corroborate these results and show that after the PLM annealings most of the Ti impurities are in interstitial lattice sites.In this study, we evaluate the potential of the silicon layers implanted with very high Ti doses and subsequently PLM annealed at different energy densities in order to obtain an IB material. We show that a high quality Si crystal lattice can be achieved for the lowest implanted dose at the maximum energy density investigated. We find a massive interstitial non equilibrium position of Ti impurities after PLM annealing, susceptible of IB formation as predicted by theory.
Q5: Metalization, Microcracks and New Manufacturing Initiatives
Session Chairs
Tuesday PM, December 01, 2009
Room 306 (Hynes)
2:30 PM - **Q5.1
A Review of Advanced Metallization for Si Solar Cells.
Aziz Shaikh 1
1 , Ferro Corporation, Vista, California, United States
Show AbstractContact metallizations continue to play a key role in improvement of conversion efficiency of Crystalline Silicon based Photo Voltaic devices. The standard Si Solar Cell efficiency improvements are achieved by forming low resistance contact to a passivated higher ohmic emitter while maintaining good diode quality and minimizing shadowing associated with metallizations. Achieving good quality contact poses significant materials development challenges in terms of control of microstructure in the contact region. This paper discusses issues associated with making contacts to Si devices with varying surface chemistries.The second set of important factors affecting the electrical efficiency is the strength of Back Surface Field, the reflection from the back surface and the back surface recombination velocity. These properties are governed by alloying of Al with Si wafer during the firing process. This paper discusses the evolution of Aluminum conductors for next generation Si Solar Cells.The solar industry is moving towards a goal of achieving > 24 % Efficiency through low cost fabrication techniques.. To achieve these goals various device schemes are being pursued. These schemes require making contacts to wide range of surfaces such as n-doped, p doped surfaces with varying concentration and, with and without passivation coatings. This paper provides an overview of these schemes.
3:00 PM - Q5.2
Screen-Printed Al Back Contacts on Si Solar Cells: Issues and Some Solutions.
Vishal Mehta 1 , Bhushan Sopori 1 , Robert Reedy 1 , Bobby To 1 , Helio Moutinho 1
1 , National Renewable Energy Laboratory, Golden, Colorado, United States
Show AbstractCommercial Si solar cells are fabricated with screen-printed contacts consisting of gridded, Ag-based front contacts and continuous, Al-based back contacts. Typically, the contacts are co-fired, which yields cells with efficiencies in the range of 14% to 17%. However, many issues remain that prevent achieving optimum performance of the contacts. This paper explores issues related to Al back-contact formation that make it difficult to achieve a high-efficiency cell. Formation of a back contact requires creation of a thick melt of Si-Al alloy that, upon solidification, can produce a stratified structure consisting of a uniform P+ layer as a back-surface field (BSF) and low-resistivity Al layer. Unfortunately, a liquid layer of Al on a Si surface is not stable, and creates bumps, dimples, spheres on the back side of the cell, and shunting. Other problem is that it can cause staining of the back contact. The objective of this work is to study the kinetics of Si-Al alloy formation/solidification, identify mechanisms that cause problems in contact formation, and explore the possibility of controlling them to improve the back-contact performance. Experiments were done on two types of samples: (i) single-crystal Si wafers with evaporated Al layers of different thicknesses, and (ii) screen-printed, multicrystalline-Si (mc-Si) cells. The samples were fired in a static optical furnace under different process conditions. The fired samples were examined to determine structure, composition, and thickness of various layers. Solar cells were characterized by current-voltage measurements and their cell parameters were determined. The following electro-optical characterization techniques were used on planar cross-sectioned cells: SEM, C-AFM, SKPM, TEM, EDX, and SIMS. Results include the following: (a) A rapid diffusion of Si into Al occurs during the temperature ramp-up. Our results show that diffusivity of Si in Al is very high (~10 -8 cm2/s at 500°C), and the solubility is about 1% at 545°C. (b) Si diffusion can be deployed to create a Si-Al eutectic melt as an interface to promote the adhesion and improve stability of the Al melt. (c) Obtaining a thick BSF (~10 microns) requires a controlled cooling of the cell. Our best results (i.e., open-circuit voltages in excess of 620 mV), on mc-Si solar cells, were produced with a cooling rate of 20°C/s from 800°C to 600°C. (d) Presence of glass in paste hinders Al flow toward the interface; thus, the entire amount of Al is not used in forming the Si-Al melt. (e) Series resistance of single-crystal Si wafers with evaporated Al increased by 3%–4% due to rapid Si diffusion. We will present details of our investigation and discuss methods to improve back-contact properties of screen-printed mc-Si solar cells. This abstract is subject to government rights.
3:15 PM - Q5.3
Use of Direct Write Metallization for Photovoltaics.
Maikel van Hest 1 , Alex Miedaner 1 , Calvin Curtis 1 , Robert Pasquarelli 2 1 , John Kreuder 3 1 , Peter Hersh 4 1 , David Ginley 1
1 , National Renewable Energy Laboratory, Golden, Colorado, United States, 2 , Colorado School of Mines, Golden, Colorado, United States, 3 , Rochester Institute of Technology, Rochester, New York, United States, 4 , Heliovolt, Austin, Texas, United States
Show AbstractCurrently photovoltaics are becoming an increasing part of the energy supply mix, however to have a really significant impact they must become cost competitive with more conventional and other alternative energy sources. Direct write methods could help toward this significant cost reduction. We are investigating solution precursors and ink based atmospheric deposition approaches to explore metallization for a variety of solar cell materials. We are studying is inkjet printing of contacts for Si, Cu(InxGa(1-x))Se2 (CIGS) and organic photovoltaics. Metallization materials and requirements for each of these types of photovoltaics are different. We have developed metal organic decomposition inks for metals such as: silver, nickel, copper and aluminum. All of these can be deposited in lines with 30-40 µm width and conductivities close to that of bulk metals. For silicon devices most commonly screen printed silver is used. The screen printing pastes have an additive to facilitate contact formation between the silver and the silicon. We have developed a metal organic ink equivalent of this paste additive, which has resulted in improved electrical properties. For CIGS devices pure silver contacts and bilayer contacts of nickel and silver have been studied. For organic photovoltaic devices pure silver contacts deposited and processed at moderate temperatures (<140°C) were explored. For silicon, CIGS and organic photovoltaic devices with printed contacts, efficiencies similar to those with contacts deposited by alternative means were obtained. Additionally to the metallization results also an overview will be given of the new direct write capabilities for large area deposition (Up to 157 mm x 157 mm) at NREL.
3:30 PM - Q5.4
Investigation of Fired and Non-fired Si-SiNx Interface Properties by Deep-level Transient Spectroscopy measurements.
Chun Gong 1 , Eddy Simoen 1 , Rui Yang 1 , Niels Posthuma 1 , Emmanuel Van Kerschaver 1 , Jef Poortmans 1 , Robert Mertens 1
1 , IMEC, Leuven Belgium
Show AbstractSilicon nitride (SiNx) films deposited by plasma-enhanced chemical vapor deposition (PECVD) are widely used in silicon solar cell fabrication as passivation layers . The use of hydrogenated silicon nitride is one of the most significant technological evolutions that has taken place in solar cells industry, due to its ability to act simultaneously as antireflective coating as well as a source of hydrogen for surface and bulk passivation. Moreover, the firing step after SiNx film deposition is assumed to improve the passivation quality by forming hydrogen terminated dangling bonds. However, until now, most of the published data on the Si-SiNx interface have been based on C-V measurements of MIS structures which mainly give interface states density (Dit) information. The aim is to investigate both fired and non-fired Si-SiNx interface properties by deep-level transient spectroscopy measurements which can provide insight on both Dit and capture cross sections.Monocrystalline FZ 0.7 Ω.cm p- and CZ 2 Ω.cm n-type silicon wafers with a <100> orientation were used. Prior to the plasma deposition, wafers received a standard RCA clean and an HF dip prior to SiNx deposition. The deposition of Si-SiNx films was performed by a direct-plasma PECVD reactor with two parallel plates. The thickness of the deposited films was around 80 nm. After deposition, samples were diced in half, half received an extra firing step. Then circular gate contacts consisting of 1 μm aluminum were deposited onto the SiNx films by e-beam evaporation through a shadow mask. InGa/In was used as the rear ohmic contact. Finally, DLTS measurements were performed and analyzed using a Boonton C-V bridge operating at 1MHz; C-V measurements were also performed at temperatures varying between 77 and 310 K. The temperature scan of non-fired n-Si exhibits a rather deep interface peak at ~Ec-0.7 eV which is close to the Et-Ev = 0.4-0.5 eV position of Schmidt et al., APL, 71, 252 (1997). The non-fired p-type sample show a clear peak at 200K for a saturating voltage pulse which corresponds to B defects published by Schmidt et al., APL, 71, 252 (1997). It should be noted that plasma deposition techniques may vary which cause Si-SiNx interface variation. The frequency scan of fired and non-fired, p- and n-type samples show that the DLTS signal is much lower for fired samples. This indicates that the extra firing step can improve the passivation quality by reducing Dit, due to the formation of hydrogen terminated bonds.In conclusion, using the DLTS technique, B defect states are identified at the Si-SiNx interface fabricated by direct PECVD. Fired samples show much lower DLTS signals, which suggests a lower Dit. Small pulse measurements of the activation energy and capture cross section will be carried out and presented in the final paper.
4:30 PM - **Q5.6
CENTESIL: A Pilot Plant for R&D in Polysilicon.
Antonio Luque 1 3 , Carlos del Canizo 1 3 , Gabriel Ovejero 2 3
1 Instituto de Energia Solar, Universidad Politecnica de Madrid, Madrid, Madrid, Spain, 3 , CENTESIL, Madrid, Madrid, Spain, 2 Departamento de Ingenieria Quimica, Universidad Complutense de Madrid, Madrid, Madrid, Spain
Show AbstractThe growth photovoltaics is explosive nowadays. The main part of it has been based in the silicon solar cells, up to 89,5% in 2008 according to Photon International, and it is to be expected that the introduction in the market of new technologies —thin film of high efficiency concentrators— will probably not deprive this technology of its condition of workhorse of the PV development, at least in the next ten years.This incredible expansion and the relative avidity for silicon of the solar cell technology has resulted in a dramatic change of the polysilicon industry structure. While in the past the polysilicon was manufactured almost exclusively for the semiconductor industry in 2008 47,700 tm of polysilicon were consumed by the solar industry vs.23,300 by the semiconductor industry. The consequence is that while in 2000 virtually only 7 companies supplied all the polysilicon consumed in 2008 there were 11 major suppliers and there are numerous new ventures entering this market.Based on this in 2006 CENTESIL was founded in Spain as a new private-public partnership venture, with vocation of opening to an international partnership, to deal with the polysilicon research. For it, a pilot plant is in advanced state of construction that has been preceded of some laboratory-size implementations. The pilot plant is designed for a production capacity of 100 tm year. The purpose is to allow the photovoltaic companies worldwide to count with an independent research centre to help them to establish their own polysilicon plant. This is a profitable option for companies producing more than 100 MW/year.The purpose of CENTESIL is to produce solar silicon but the very concept of solar silicon is uncertain. In this respect in CENTESIL we have analyzed the permitted concentration of a number of impurities in the feedstock to allow for 15% efficiency solar cells. It has also been determined that improving efficiency in a certain percentage allows to increase polysilicon cost ~6 times this percentage for equal module cost. Therefore CENTESIL is aligned with the production of highly purified polysilicon for high efficiency solar cells.To fulfill these objectives the development contains four areas of activity: (1) synthesis of chlorosilanes, with the silicon tetrachloride, a process subproduct, as the main source of chlorine; (2) chlorosilane purification based on fractional distillation but with additional processes when necessary to remove lifetime-killing impurities; (3) development of a chemical vapor deposition (CVD) reactor for the polysilicon production and (4) recycling (mainly of the CVD outputs) for optimal use and sustainability.These steps and their present status will be discussed in the presentation as well and the cost parameters of the plants to be developed with this technology.
5:00 PM - Q5.7
Strength of Multicrystalline Si Wafers for Various Surface Conditions.
Przemyslaw Rupnowski 1 , Bhushan Sopori 1 , Dan Armentrout 2
1 , National Renewable Energy Laboratory, Golden, Colorado, United States, 2 Mechanical and Materials Engineering, University of Denver, Denver, Colorado, United States
Show AbstractThe photovoltaic (PV) industry is constantly attempting to reduce the thickness of Si wafers for solar cells as a means to lower Si usage and cell cost. However, even at the current thickness of about 200 microns, significant breakage occurs during cell and module fabrication. Solar cell designs that use effective light-trapping can yield efficient cells for much thinner wafers; however, reducing the wafer thickness leads to unacceptably high wafer breakage. Because silicon is a brittle material at room temperature, the reason for wafer breakage is the presence of stress-concentrating structural flaws such as microcracks, voids, and inclusions. These defects can be located in the bulk, at the surface, and at the edges. In terms of wafer breakage, micro-cracks on the surfaces and edges of the wafers are the most fatal. These microcracks remain from incomplete removal of surface damage following ingot sawing.This paper will focus on the effects of edge and surface microcracks on the strength of multicrystalline-Si (mc-Si) wafers. We performed both experimental and theoretical analyses to quantitatively describe the relation between the crystal flaws and mechanical strength. A theoretical model was developed to predict the strength distribution for a given density of microcracks (of different sizes and shapes) on surfaces and edges. We used a multimodal Weibull distribution to describe the fracture response of the wafer as a function of load conditions. To compare with the experimental data, we designed and performed the following experiments: Four sets of specimens were prepared by subjecting them to various surface treatments. The first set was in an as-sawn state; the second and third sets were defect- and texture-etched, respectively; and the last set was polished using chemical-mechanical polishing to produce mirror-like, damage-free surfaces. The wafer fracture load was measured on a hydraulic mechanical tester using a ring-on-ring and 4-line bending fixtures. We found that maximum stress for ring-on-ring and 4-line bending was about 650 MPa, and 350 MPa, respectively—clearly showing that edges have a stronger influence on wafer breakage. A correlation between theory and experiment allowed us to answer critical questions regarding the fracture response of mc-Si wafers. We showed how different surface conditions affect the wafer strength. For the first time, we actually separated the impact of the edge, surface, and bulk defects on the fracture response.This abstract is subject to government rights.
5:15 PM - Q5.8
Evaluating Amorphization Around Micro-Cracks in PV Silicon.
Prashant Kulshreshtha 1 , Khaled Youssef 1 , George Rozgonyi 1
1 Material Science and Engineering, North Carolina State University, Raleigh, North Carolina, United States
Show AbstractMicro-cracks initiate and propagate in silicon for stresses beyond the yield stress. The micro-crack acts as a local stress-raiser and is a main source of wafer breakage. Since, crack initiation and propagation, where impurities, defects, and residual stresses play a highly interactive role [1], it is critical to wafer breakage to evaluate the crack propagation velocity [2], cracking types [3], and associated defect generation [1]. Unfortunately, phase-transformations resulting from stress-release as a crack propagates within silicon have not adequately examined [4]. Minowa et.al. [5] have reported that a surface scratch made on a silicon crystal small induces an amorphous region which is surrounded by dislocated crystalline Si. In this study, we have used nanoindentation, EBSD and Raman spectroscopy technique to document the phase-transformation leading to amorphization in the vicinity of micro-crack. Samples of 2cm x 5mm were cleaved from cast solar grade silicon wafers (≈ 200 µm thick) sliced from the top, middle, and bottom of the ingot with low carbon (<1017 /cm3), metal and doping concentration. A Vickers micro-indenter load of 100 g was applied such that micro-cracks emanate from the indent edges. Hysitron 900TM Triboindenter was then used to perform low load (<10 mN) nanoindentations on deformed region in the vicinity of a micro-crack. Since localized atomic scale perturbations can modify the direction of crack propagation, micro-cracks do not follow a straight line path [6]. Hardness and elastic modulus measurements were carried out to demonstrate the impact of the phase transformations, which lead to amorphization. Hardness measurements made on individual indents within a particular row, for example, in a sample from middle of ingot, showed a gradual drop in hardness from 10.2 GPa to 6.9 GPa for the distance from micro-cracks to the rows of indents decreases. Also, maximum indenter penetration depth jumped from 68 nm to 102 nm for indents near the micro-crack. Interestingly, comparing these hardness values with bulk Si (100) in a micro-crack free region (hardness = 10.6 ± 0.2 GPa) confirms that amorphization is localized in within 10 µm from the micro-crack. Our nano-indentation results suggest amorphization, as verified by AFM 3D-imaging and EBSD measurements with the indistinct diffraction lines. In addition, micro-Raman spectroscopy results using the 521 cm-1 peak height measurements also verify the extent of amorphization in and around a micro-crack. References 1.C. Scandian et.al, Phys. Stat. Sol. (a) 171, 67-82, 1999.2.I. Chasiotis et.al, J. App. Mech., 73, 715, 2006.3.P. Puech et.al, J. Mater. Res., 4 (19), 1273-1280, 2004.4.L. Zhang et.al, Int. J. Mech. Sci., 43, 1985–1996, 2001.5.K. Minowa et.al, Phy. Rev. Let., 2 (69), 320-322, 1992.6.J. R. Kermode et.al, Nature 455, 1224-1227, 2008.
5:30 PM - Q5.9
Comprehensive Modeling of Si Wire Array Solar Cells.
Michael Kelzenberg 1 , Morgan Putnam 1 , Daniel Turner-Evans 1 , Nathan Lewis 1 , Harry Atwater 1
1 , California Institute of Technology, Pasadena, California, United States
Show AbstractPhotovoltaic devices based on arrays of CVD-grown Si nano- or micro-wires are being investigated as potential low-cost alternatives to wafer-based Si solar cells. High-fidelity arrays of Si wires can be grown by a potentially low-cost vapor-liquid-solid (VLS) deposition technique. [1] Furthermore, the radial p-n junction geometry tolerates relatively low-purity Si, and prior modeling of single-wire solar cells has predicted efficiencies exceeding 17%, based on experimentally observed diffusion lengths of ~10 microns. [2] However, this modeling has not addressed the ensemble properties of a wire array -- in particular their optical absorption properties.We have developed a comprehensive, numerical model of Si wire array solar cells by combining optical absorption simulations with device physics simulations. Light absorption was simulated by the full-field, finite-difference time-domain (FDTD) method in 3 dimensions. Bloch boundary conditions were used to simulate the structure of a periodic wire array, and simulations were performed throughout the solar spectrum and across a wide range of incidence angle. Arrays of 3 µm-diameter wires with 14% packing fraction were simulated to absorb over 66% of incident (above-bandgap) solar illumination at normal incidence, and more off-normal incidence. FDTD simulations also provided photogeneration profiles which were used in device physics simulations to calculate a photovoltaic efficiency. Our device physics model was implemented using TCAD Sentaurus (Synopsys, Inc), a comprehensive numerical device physics solver. Throughout our simulations, the physical geometry and material parameters were chosen to match the experimentally-observed structure and properties of our Si wire arrays. Thus, we are able to demonstrate agreement between simulation and experimental results for several experiments; including I-V behavior, light-beam induced current (LBIC) profiles, and optical absorption. Our modeling predicts that Si wire array solar cells with only 14% packing fraction could achieve photovoltaic efficiency of 14.5% based on experimentally observed doping levels, lifetimes, and surface recombination velocities. The low packing fraction of the wire array yields an effective optical concentration, and produces a higher predicted Voc than prior simulations (620 mV). The simulations also provide insight into the design of wire array solar cells. We will discuss effects of wire depletion, wire diameter distribution, shunting, parasitic absorption, and light trapping.[1] Kayes, B. M. et al. Growth of vertically aligned Si wire arrays over large areas (> 1 cm2) with Au and Cu catalysts. Appl. Phys. Lett. 91, 103110-103113 (2007).[2] Kelzenberg, M. D. et al. in Photovoltaic Specialists Conference, 2008. PVSC '08. 33rd IEEE. 1-6.
Symposium Organizers
Bhushan L. Sopori National Renewable Energy Laboratory
Bernhard Dimmler Würth Solar GmbH & Co. KG
Jeffrey Yang United Solar Ovonic LLC
Thomas Surek Surek PV Consulting
Wednesday AM, December 02, 2009
Room 306 (Hynes)
9:30 AM - Q6.1
Characterization of CIGS Structures for Composition, Phase and Impurities using SIMS, Auger, STEM/EDS, XPS and XRD.
Gary Mount 1
1 , Evans Analytical Group, Sunnyvale, California, United States
Show AbstractCu(In,Ga)Se2-based thin film solar cells are showing great promise as they are relatively economical to produce and have shown conversion efficiencies approaching 20%. Keys to high efficiency include compositional grading of Ga that provides a tool to tune the bandgap to capture more of the solar spectrum. Another key is the incorporation of Na into the CIGS layer although the role of Na is not well understood. Phase and grain orientation can affect performance. Impurities can play a role in efficiency reduction. Process development can be aided and sped up by using surface analytical techniques. They provide the ability to accurately and reproducibly measure composition, Ga gradient, phase, orientation, and impurities.We explore the capabilities of various analytical techniques to determine composition, structure and impurities in CIGS materials. Accuracy, depth resolution, and artifacts are evaluated as well as strengths and limitations of each technique.Many research studies reporting analytical results arbitrarily set Se composition to 50 at% and Cu composition to 25at%. We calibrate our composition measurements without such assumptions.SIMS is well known for its capabilities to depth profile trace level elemental concentrations. We use this capability to investigate Na concentration as well as various elemental contaminants including Al, K, Cr, Fe, and Ni. SIMS accuracy is affected by the composition of the matrix material. We investigate and make corrections for Na ion yield as a function of Ga/(Ga+In) ratio.SIMS can provide a quantitative measure of composition under the right conditions,. We evaluate the effectiveness of SIMS to report composition as a function of depth through various CIGS structures.Auger can provide small spot, single grain depth profiling for CIGS elements. Additionally, some Auger instruments have in-situ FIB cross-sectioning capabilities. Auger line-scans done on cross-sections can show composition and a visual correlation with cross-section features such as individual grains and voids.STEM imaging can effectively show CIGS grain structure and combined with EDS, composition can be measured on a very fine scale. This approach can be effective for the evaluation of CdS layer composition as well.XPS can measure chemical state in addition to elemental composition. Chemical state measurements are valid prior to sputter depth profiling.XRD is well known for crystal phase and orientation measurements. We look at the capabilities of XRD to report phase and orientation in thin CIGS layers and near interfaces using Glancing Incidence XRD.
9:45 AM - Q6.2
Dielectric Functions and Growth Dynamics of Cu(In,Ga)Se2 and Cu2-xSe Thin Films for Photovoltaic Applications via Real Time Spectroscopic Ellipsometry.
James Walker 1 , Himal Khatri 1 , Scott Little 1 , Vikash Ranjan 1 , Robert Collins 1 , Sylvain Marsillac 1
1 Center for Photovoltaics Innovation and Commercialization, University of Toledo, Toledo, Ohio, United States
Show AbstractIn situ, real time spectroscopic ellipsometry (RTSE) has been used to study the growth processes and optical properties of the chalcopyrite thin film absorber layer copper indium gallium diselenide, Cu(In,Ga)Se2, and related binary, Cu2-xSe. Bulk layer (db) and surface roughness layer (ds) thicknesses and complex optical dielectric functions (0.75 – 6.5 eV) were extracted during film growth. The ellipsometric determined thicknesses correlate well with ex situ AFM and profilometer measurements while the dielectric functions complement and expand upon the published literature, elucidating critical points near and above 5 eV in photon energy. The high surface roughness of Cu(In,Ga)Se2 and Cu2-xSe thin films can make spectroscopic measurements challenging. We find that dielectric function extraction during the initial stages of film growth (in situ) allows for a unique determination whereas extraction after growth completion, and without mechanical or chemical polishing and complex oxide removal, severely limits the accuracy of the dielectric functions for thicker films (~1 μm). The growth temperature dielectric functions presented in this study are expected to allow for a greater level of control and understanding of the so-called 2- and 3-stage Cu(In,Ga)Se2 fabrication processes in which a Cu2-xSe phase, present at the grain boundary surfaces, acts as a fluxing agent for the growth of photovoltaic quality Cu(In,Ga)Se2. Real time optical feedback via RTSE combined with the growth temperature dielectric functions presented here could play an important role in improving device efficiencies on both the laboratory and industrial scales.
10:00 AM - Q6.3
Optimising Sulfurisation Conditions in the Fabrication of Cu2ZnSnS4 Absorber Layers from Electroplated Precursors.
Jonathan Scragg 1 , Laurence Peter 1 , Phillip Dale 2 , Guillaume Zoppi 3 , Ian Forbes 3
1 Chemistry, University of Bath, Bath United Kingdom, 2 Laboratoire Photovoltaïque, Université du Luxembourg, Luxembourg Luxembourg, 3 Northumbria Photovoltaics Applications Centre, Northumbria University, Newcastle upon Tyne United Kingdom
Show AbstractThe cost reduction potential of the thin-film photovoltaic materials Cu(In,Ga)(S,Se)2 and CdTe compared to silicon is one of the key factors motivating continued research into them. However, in the context of large-scale deployment up to TW levels, it has been suggested that the rarity of Tellurium and Indium will limit their applicability [1].Cu2ZnSnS4 (CZTS), being isoelectronic with CuInSe2, has very similar properties but contains only naturally abundant, non-toxic components. If the fabrication of efficient devices based on CZTS can be achieved, then it promises both significant cost reductions and the potential for sustainable production. To date, the best reported CZTS device, produced by RF co-sputtering with post-annealing, had an efficiency of 6.7 % [2]. We demonstrate the coupling of this inherently low-cost material system with a non-vacuum fabrication process which is well suited to industrial up-scaling. Metallic precursor films are produced by sequential electroplating of Cu, Zn and Sn layers. These precursors are then converted to the quaternary sulfide compound by thermal treatment in sulfur vapour [3]. Industrially, Cu, Zn and Sn electroplating are all turn-key solutions, large areas being produced on a continuous basis using well-understood and controllable processes. The key processing step from a scientific viewpoint is the thermal sulfurisation. The goal is to obtain uniform, adherent single-phase CZTS films, but little is understood about the processes that occur during sulfurisation. For example, while EDS and SIMS show that uniform and complete sulfurisation with total levelling of the elemental profiles occurs within five minutes at 820 K, longer anneal times are usually required to yield good photocurrents, e.g. up to three hours in the ‘standard’ process of Katagiri et al. [2]. A more rapid sulfurisation is desirable, and for this reason, a detailed and systematic study of the various process parameters involved in this step was carried out. The effects of these parameters on phase structure, composition and film adhesion were studied. Since thin-film X-Ray diffraction is not capable of adequately distinguishing the secondary phases Cu2SnS3 and ZnS from CZTS [4], Raman spectroscopy was used to investigate the phase structure. The photoresponse of the films was measured using an electrochemical method which allows analysis of the absorber layer alone without fabricating complete devices, a useful tool in the optimisation of process parameters. Devices were then made from the most effective absorber layers and already demonstrate efficiencies above 3% [5].[1] A. Feltrin et al., Renewable Energy 33 (2008) 180–185;[2] H. Katagiri et al., Appl. Phys. Express 1 (2008) 041201;[3] J. Scragg et al., Phys. Stat. Sol. (b) 245, No. 9 (2008);[4] P.A. Fernandes et al., Thin Solid Films 517 (2009) 2519–2523;[5] P.J. Dale et al., presented at 34th PVSC at the IEEE conference 5th-12th June 2009;
10:15 AM - Q6.4
First-principles Study of Cu2ZnSnS4 (CZTS) as a Potential Photovoltaic Material.
Akihiro Nagoya 1 , Ryoji Asahi 1 , Joachim Paier 2 , Georg Kresse 2
1 , Toyota Central R&D Laboratories, Inc.,, Nagakute Japan, 2 Faculty of Physics, Universitat Wien and Center for Computational Materials Science, Wien Austria
Show AbstractRecently the quaternary Cu2ZnSnS4 (CZTS) has attracted attention as a potential absorber material for realizing low-cost and environmentally amenable photovoltaic devices because all constituents of CZTS are abundant in the crust of the earth and non-toxic. The highest conversion efficiency of CZTS reported so far is 6.7% [1], demanding further improvement; however, fundamental properties of CZTS have not been understood well. We here present first-principles calculations to determine electronic structure, dielectric properties, and defects formations [2].Calculations are performed using the plane-wave projector augmented-wave (PAW) method applying the semi local Perdew-Burke-Ernzerhof (PBE) exchange-correlation functional and the Heyd-Scuseria-Ernzerhof (HSE) hybrid functional as implemented in the Vienna Ab-initio Simulation Package (VASP) code. The defect formation energies are accurately evaluated using an extrapolating technique of the super cell calculations, which also provide a stable range of chemical potentials in CZTS bounded by those in possible precipitations. The total energy calculations for various Cu-Zn interchanged modifications of CZTS showed the kesterite-type structure of I_4 symmetry is the most stable one, possibly along with cation disorder within the Cu-Zn layer. By virtue of the HSE method, the calculated band gap energy of 1.49 eV agrees well with experimental value of 1.4~1.5eV. The valence band maximum (VBM) and the conduction band minimum (CBM) consist of Cu-3d/S-3p bands and Sn-5s/S-3p* bands, respectively, which dominantly contribute to dipole transitions under the solar light absorption. The defect formation energies indicated that the most stable defect is a substitution of Cu atom for a Zn site (Cu@Zn) for whole range of the allowed chemical potentials with a negative formation energy, suggesting the origin of the p-type character of CZTS.
10:30 AM - Q6.5
Defect Formation in CuAlS2 and CuGaS2 Chalcopyrite Semiconductors.
Stanko Tomic 1 , Christine Bailey 1 , Leandro Liborio 2 , Giuseppe Mallia 2 , Nicholas Harrison 1 2
1 Computational Science and Engineering Department, STFC Daresbury Laboratory, Warrington, Cheshire, United Kingdom, 2 Department of Chemistry, Imperial College, London United Kingdom
Show AbstractThin film solar cells based on Cu(In,Ga)(S,Se)2 (“CIGSSe”) chalcopyrite materials exhibit a power conversion efficiency close to 20%. Recent experimental data [1] suggests a depth-dependence of the fundamental energy gap (Eg) within the CIGSSe material. This is an interesting development as it may assist energy harvesting at different wavelengths, currently achieved in “multi junction” solar cells, without the difficult and expensive requirement to grow several pseudomorphically mismatched layers. The depth-dependent variation of the Eg in the CIGSSe is attributed to variation in the concentration of the S anions in the absorber. The S richer regions exhibit larger energy gaps. An alternate method of developing highly efficient thin film solar cells is to introduce an intermediate band (IB) into the Eg of the chalcopyrite [2]. In this concept the larger Eg of the host material is a prerequisit.Although the CuInSe2 and CuGaSe2 members of the I-III-VI2 chalcopyrite family have been studied extensively [3], relatively little is known about CuGaS2 and CuAlS2. These materials have significantly larger band gaps (2.43 eV and 3.49 eV for CuGaS2 and CuAlS2 respectively) and hence have not always been prime candidates for thin-film solar cells. However, they have a huge potential for use in CIGSSe thin film and IB solar cells. In this paper we systematically study the stability and defect chemistry of CuGaS2 and CuAlS2. Our analysis is based on hybrid exchange density functional theory, which incorporates a fraction of Fock exchange in the B3LYP density functional [4]. For CuGaS2 and CuAlS2 this functional results in calculated optical energy gaps of 2.45 eV and 3.50 eV respectively; these are in excellent agreement with the measured values. We calculate the relative stability of CuGaS2 and CuAlS2 with respect to other competing phases (e.g. Cu2S and Ga2S3) as a function of Cu, Ga/Al and S chemical potentials. Defects consisting of Cu vacancies (V:Cu), Ga/Al substitution of Cu ions (Ga/Al:Cu) and S vacancies (V:S) are modelled within 64 atom supercells. Compound defects consisting of 2V:Cu+Ga:Cu/Al:Cu are shown to be energetically favourable and lead to ordered defect compounds (ODC) such as CuGa5S8. It is also shown that Ga/Al:Cu defects can significantly reduce the Eg of these materials by up to 0.5 eV. We also briefly discuss the effect of transition metal doping and demonstrate the importance of an accurate prediction of the energy gap by B3LYP. We show that substituting Ga ions for the transition metals Ti, Cr or Fe leads to intermediate bands in the electronic structure of the materials.[1] M. Bar et al, Appl. Phys. Lett. 93, 244103 (2008)[2] A. Luque and A. Marti, Phys. Rev. Lett. 78, 5014 (1997).[3] C. Persson et al, Phys. Rev. B. 72, 035211 (2005)[4] S. Tomic et al, Physica E 40, 2125 (2008)
10:45 AM - Q6.6
High Efficiency CuIn(Se,S)2 Devices with Tunable Bandgap Deposited Using a Hydrazine-based Solution Approach.
Wei Liu 1 , David Mitzi 1 , Andrew Kellock 2 , S. Jay Chey 2
1 , IBM TJ Watson Research Center, Yorktown Heights, New York, United States, 2 , IBM Almaden Research Center, San Jose, California, United States
Show AbstractPreviously we reported CuIn(Se,S)2 (CIS) devices with a power conversion efficiency of approximately 9% and deposited using a hydrazine-based solution process [1]. Continued efforts have been made to increase the bandgap of this material system by adding more S into the system. As expected, improved open circuit voltage (> 550 mV) has been achieved and overall power conversion efficiency under AM 1.5 radiation has improved to beyond 12%. These high power conversion efficiencies approach values for devices based on more cost-intensive vacuum-grown CIS [2-4], as well as those for previously reported hydrazine-deposited CIGS [5, 6]. Comparing to hydrazine-based CIGS precursors, the CIS precursors are simpler as they do not contain Ga. Yet these precursors maintain the capability to tune the bandgap of deposited films via the variation of the amount of S in the precursor solution. The ability to fabricate high-efficiency CIS solar cells further demonstrates the flexibility of the hydrazine-based approach for depositing metal chalcopyrite-based absorber layers and provides further evidence that this approach may offer a low-cost, high-efficiency route to thin-film PV device fabrication. [1] Liu, W.; Mitzi, D. B.; Yuan, M.; Kellock, A.; Chey, S. J. MRS Fall 08 meeting 2009, in press.[2] AbuShama, J.; Johnston, S.; Moriarty, T.; Teeter, G.; Ramanathan, K.; Noufi, R. Progress in Photovoltaics: Research and Applications 2004, 12, 39.[3] da Cunha, A. F.; Salome, P.; Kurdzesau, F. Materials Science Forum 2006, 514-516, 93.[4] Hollars, D. R.; Dorn, R.; Paulson, P. D.; Miasole, R. Z.; Titus, J. Materials Research Society Symposium Proceedings 2005, 865, 477.[5] Mitzi, D. B.; Yuan, M.; Liu, W.; Kellock, A.; Chey, S. J.; Deline, V.; Schrott, A. G. Adv. Mater. 2008, 20, 3657.[6] Mitzi, D. B. Adv. Mater. 2009, available on-line at http://dx.doi.org/10.1002/adma.200802027
11:30 AM - Q6.7
3-Dimensional Elemental Analysis of Cu(In,Ga)Se2 (CIGS) Solar Cells using Time-of-Flight Secondary Ion Mass Spectrometry.
Roger Michel 1 , Ian Mowat 1 , Bruno Schueler 2 , Larry Wang 1 , Gary Mount 1 , John Moskito 1 , Thomas Fister 1
1 , Evans Analytical Group, Sunnyvale, California, United States, 2 , ReVera, Inc, Sunnyvale, California, United States
Show AbstractThe device performance of CIGS solar cells is believed to be dependent on the amount of Na within the CIGS layer. Contaminants (e.g transition metals) are known have a negative effect on device performance. The localization and quantification of such species plays a crucial part in the understanding and improvement of these devices.Time-of-Flight Secondary Ion Mass Spectrometry (TOF-SIMS) can be used to detect organic and inorganic species from the top few monolayers of a surface, by using pulsed primary ions and detecting secondary ions emitted from the surface. This extremely sensitive technique can be enhanced by using a second, dynamic sputter beam (Cs or O2). By alternating the two sources, depth profiling into the material can be achieved. The use of a highly focused primary analysis beam makes it possible to detect species at sub-micrometer spatial resolution.The sample analyzed was an NREL 3 stage process CIGS layer on stainless steel. These NREL 3 stage process solar cells have shown very promising results with upwards of 19% conversion efficiencies [1]. Layers of CdS, NaF and Mo also make up the sample structure. Standard survey depth-profiling by TOF-SIMS observed 18 elemental species of interest. These species can also be shown as ion image stacks and can be displayed in 3D. In particular, the behavior of Ga, In, Cu, Se, Na, Fe, Cr, F and O is shown and discussed. Na and its diffusion from the Mo/CIGS interface into the CIGS layer is of particular interest, as it is believed to affect conversion efficiencies. Cr was found localized at the Mo/CIGS interface as a potential impurity and was not co-located with other species. Auger electron spectroscopy (for quantification of certain species, sub-micrometer resolution) and dynamic SIMS (for quantification, and lower detection limits) data is also shown in support of a better understanding of the material structure. [1] Diode Characteristics in State-of-the-Art ZnO/CdS/Cu(In1-xGax) Se2 Solar Cells. Contreras, M. A et.al. Progress in Photovoltaics: Research and Applications. Vol. 13(3) 2005 pp209-216
11:45 AM - Q6.8
Inherent ZnS Bufferlayer for the CuInS2 – ZnO Heterocontact Prepared by MOMBE.
Christian Pettenkofer 1 , Carsten Lehmann 1 , Stefan Andres 1
1 E-I4, Helmholtz-Zentrum-Berlin, Berlin Germany
Show AbstractChalcopyrite thin film solar cells have shown efficiencies up to 20%. To optimize the hetero-junction in the device a buffer layer is needed to obtain a suitable band alignment between the chalcopyrite absorber and the ZnO window layer. The ZnO is deposited in two steps beginning with a i-ZnO layer followed by a Al doped ZnO film. For technologically prepared devices etching in a cyanide solution after the vacuum deposition of the absorber to remove Cu-rich phases and a subsequent deposition of CdS as a buffer in a chemical bath is the standard technique. As the deposition of the ZnO TCO layer again is a vacuum process an inline processing of the devices will reduce the process Steps considerably at least for the loading and deloading to the vacuum deposition steps. Here we report on an all vacuum process for CuInS2 devices by applying a MOMBE technique to prepare the i-ZnO layer. In a first inherent step a ZnS bufferlayer forms on the Cu-rich CuInS2 substrate, which shows perfect lattice matching as shown by LEED. After 0.8nm ZnS the i-ZnO starts to grow in registry to the ZnS layer showing a perfect band alignment to the absorber. The deposition technique is tested for a dedicated model system, where the process steps could be followed by highly sensitive electron spectroscopies at the surface and prepared interface. To show the lattice matching (112) oriented epitaxial CuInS2 films were prepared. The buffer and i-ZnO deposition was investigated after each deposition step by LEED, XPS and UPS. The proposed process can be integrated as an in line vacuum process directly after the absorber preparation in an attached vacuum system. The cell is completed by the Al:ZnO sputterdeposition. The wet chemical processes for etching and CdS deposition are omitted.
12:00 PM - Q6.9
High Efficiency Copper Indium Gallium DiSelenide (CIGS) by High Power Impulse Magnetron Sputtering (HIPIMS): A Promising and Scalable Application in Thin-film Photovoltaics.
Ankush Halbe 1 , Paul Johnson 1 , Shen Jackson 1 , Bob Weiss 1 , Upendra Avachat 1 , Alex Welsh 1
1 , DayStar Technologies, Santa Clara, California, United States
Show AbstractA novel method to deposit Copper Indium Gallium Diselenide (CIGS) using High Power Impulse Magnetron Sputtering (HIPIMS) was demonstrated and compared to the existing DC magnetron sputtering process. The metal-ion assisted thin-film growth inherent to a HIPIMS deposition process was used to advantage in depositing CIGS films. The HIPIMS plasma was characterized by measuring ion currents on a Langmuir probe placed into the plasma sufficiently close to the substrate. The high density plasma consisting of both metal and metal ions resulted in CIGS thin-film solar cells of superior conversion efficiencies (~13%) as compared to conventional DC magnetron sputtering (~10%). The efficiency enhancement was attributed to the improvement in the shunt resistance of the solar cell which corresponds to the increase in the density of the CIGS layer. Furthermore, it was also possible to grow large grained CIGS (~1 micron) with high mobility metal-ions from the HIPIMS process. The scalability potential of the HIPIMS CIGS process was also demonstrated by running a 1.5 m long Copper-Indium-Gallium rotatable in a selenium environment using a HIPIMS power supply. The cylindrical magnetron was run at an average power of 7.8 KW and peak powers of as much as 300 KW with controlled arcing. The existence of a HIPIMS plasma was confirmed by the ion currents on the Langmuir probe and the metal signals from a Plasma Emission Monitor (PEM).
Q7: Thin Film and Amorphous Si
Session Chairs
Wednesday PM, December 02, 2009
Room 306 (Hynes)
2:30 PM - *Q7.1
Improving Conversion Efficiency and Decreasing Thickness of HIT Solar Cell.
Hirotada Inoue 1 , Yasufumi Tsunomura 1 , Daisuke Fujishima 1 , Ayumu Yano 1 , Shigeharu Taira 1 , Yasuko Ishikawa 1 , Takeshi Nishiwaki 1 , Toshio Asaumi 1 , Mikio Taguchi 1 , Hitoshi Sakata 1 , Eiji Maruyama 1
1 Advanced Energy Research Center, Sanyo Electric Co., Ltd, Kobe Japan
Show AbstractIn order to reduce power-generating cost of silicon solar cell, it is necessary to use thinner crystalline silicon (c-Si) substrate together with an excellent light trapping and well passivated surface that are the essential for the high efficiency solar cells. The HIT (Heterojunction with Intrinsic Thin-layer) solar cell is an amorphous silicon (a-Si) / c-Si heterojunction solar cell which have these features. We have achieved the world’s highest conversion efficiency of 23.0% with a practical size of 100.4cm2 HIT solar cell, by improving the quality of surface passivation, reducing optical absorption loss and reducing resistance loss. To improve the passivation quality of a-Si layer on c-Si, we have upgraded the PECVD (plasma-enhanced chemical vapor deposition) condition of a-Si layers. With this improvement, the open circuit voltage (Voc) of the solar cell increased from 0.725V to 0.729V, proving the Improved surface passivation quality. To reduce optical absorption loss, the a-Si layers with wider optical band-gap and the TCO layer with less absorbance have been developed. We have controlled the optical gap of a-Si wider to decrease the absorption of short-wavelength region of solar radiation. The deposition condition of the TCO has been optimized to obtain both low absorption coefficient and high conductivity. This new TCO layer enables to make solar cells with lower absorption loss at longer wavelength region with keeping its sheet resistance low enough. As a result of these improvements, the short circuit current (Jsc) is improved from 39.2mA/cm2 to 39.5mA/cm2. To reduce the resistance loss, we have recently developed the lower-resistance electrode material for the grid electrode together with improved printing technology to obtain the higher aspect ratio. With these technologies, the fill factor (FF) was improved from 0.791 to 0.80. The excellent passivation with a-Si on c-Si surface is very important for the thinner c-Si high efficiency solar cells. We have fabricated HIT solar cells using thin c-Si wafers (96~58μm). As the wafer thickness decrease, the Voc increases and reaches extremely high value of 0.745V with 58-μm-thick. The fact indicates that the surface recombination velocity of HIT structure is extremely low due the excellent passivation on the c-Si surface. This technology (increased Voc with decreased thickness) has been reported as a simulated result on ideas solar cells with very low surface recombination. We have experimentally confirmed with the real solar cells which has the same structure as the cells on the market. Further, these cells have no sagging because HIT structure cancels the stress caused by front side and back side materials on c-Si wafer (i.e. a-Si, TCO and grid electrodes).Therefore, the HIT solar cell is suitable for thin wafer.
3:00 PM - Q7.2
Lower Temperature Si Thin Films by Metal Induced Growth for Lower Cost Solar Cells.
Peter Mersich 1 , Wayne Anderson 1 , Rossman Giese 2
1 Electrical Engineering, University at Buffalo, State University of New York, Buffalo, New York, United States, 2 Geology, University at Buffalo, State University of New York, Buffalo, New York, United States
Show AbstractAmid the escalating energy crisis, there has been greatly expanded interest in establishing a viable, low-cost approach to solar cells. Si has been a popular material of choice because it is very abundant, non-toxic, and has a rich history in advanced electronics. While traditional wafer-based crystalline Si solar cells are able to achieve relatively high efficiencies, this advantage is quickly outweighed by the cost of manufacturing. Therefore, many have turned to thin film solar cells. Thin films not only have the advantage of much less expensive fabrication but also utilize far less material. This may be further augmented with the assistance of lower processing temperatures, allowing for the utilization of inexpensive foreign substrates (plastics, etc.). The use of less material is made possible through microcrystalline Si (μc-Si), which has an absorption coefficient about 10 times higher than that of single crystal Si due to different crystal orientations in the grains.
Metal-induced growth (MIG) is a method used to epitaxially grow thin layers of μc-Si. Previous reports have been done using Ni and Co at temperatures of 600 oC or greater. The process is similar to metal-induced crystallization (MIC), which uses a two-step process of depositing amorphous Si (a-Si) and later converting it to crystal form. MIG has the advantage of using a one-step process to form the μc-Si and simultaneous ohmic back contact. Crystallization of Si using Al has been reported to start at temperatures as low as 300 oC. Therefore, MIG done with Al has the potential to significantly lower the processing temperature. When Si is sputtered from a target, a pre-deposited Al film at an elevated temperature acts as a catalyst to crystallize the Si during a slow rate process. This forms the seed layer for crystallization as the sputtering process continues with a higher rate, and the μc-Si film is grown.
In a preliminary study, 700 Å of Al was deposited on oxide-coated Si by e-beam evaporation. Before dc sputtering, the substrate was heated to 525 oC. A power of 50 W was first administered for 30 mins to establish the seed layer. It was then increased to 150 W for 3 hr. Scanning electron microscopy (SEM) and energy dispersive x-ray spectroscopy (EDX) showed a 5-μm thick film with 98.7% Si near the surface. Atomic force microscopy (AFM) demonstrated triangular-shaped structures on the surface reminiscent of <111> oriented Si, while x-ray diffraction (XRD) confirmed the crystallinity of the film with distinct Si peaks at <111> and <220>. Using the Debye-Scherer formula, the approximate crystallite size was 52 nm. Schottky photodiodes using Au/Cr exhibited a p-type Si film with good rectification. While the open-circuit voltage was low, the dark current density showed space-charge limited conduction in the ballistic regime. Further work will be done at temperatures approaching 300 oC with p-n diodes and heterojunction solar cells.
3:15 PM - Q7.3
Orientation-controlled Poly-Si on Glass by Al-induced Layer Exchange Technique.
Masashi Kurosawa 1 , Naoyuki Kawabata 1 , Taizoh Sadoh 1 , Masanobu Miyao 1
1 Dept. of Electronics, Kyushu Univ., Fukuoka Japan
Show AbstractOrientation-controlled large grain poly-Si on glass is necessary to realize high-efficiency solar cells. For this purpose, we have investigated a metal-induced crystallization of a-Si by using Al as catalyst, i.e., Al-induced layer exchange (ALILE), and realized orientation control of large-grain poly-Si.In the experiment, Al films (100nm) were deposited on quartz, where XRD revealed preferential (111)-orientation of Al films. The samples were exposed to air (5min-24h) to form Al oxide layers, and subsequently a-Si films (100nm) were deposited. Finally, they were annealed at 450C.Nomarski microscopy and Raman observations revealed that large grains (>20micron) were obtained, where short annealing (20h) was sufficient for samples with short and medium air exposure (5, 60min), however, long annealing (40h) was necessary for the sample with long exposure (24h). This indicated that thickness of Al oxide layers, acting as diffusion barrier of Si/Al atoms, became thick with increasing air exposure time. To evaluate crystal orientation, EBSD were measured, which indicated preferential (001) or (111) orientation for samples exposed to air for 5min or 24h. Namely, for short air exposure (5min), a high fraction (80%) of grains was oriented to (001). On the other hand, the fraction of (111)-oriented grains increased with increasing air exposure time, and a high fraction (85%) of grains was oriented to (111) for the sample with long air exposure (24h). These indicate that interfacial oxide layers played an important role in alignment of crystal grains.To explain these phenomena, we propose a model considering the phase transition of the Al oxide. The inter-diffusion of Si/Al atoms fast proceeds in samples with short air exposure, because of the very thin diffusion barrier. Thus, the oxide layers should remain amorphous in the stage of Si nucleation. Recently, Schneider reported preferentially (001) oriented nucleation of Si at amorphous-Al2O3/Al interfaces based on free energy calculation [1], which should explain the preferential (001) orientation of the sample with short air exposure. On the other hand, we speculate that the Al oxide change into (111)-oriented gammma-Al2O3 by sold-phase crystallization during long annealing, because the initial Al films were preferentially oriented to (111) in the present study. Since the lattice constant of gamma-Al2O3(111) is close to that of Si(111) (mismatch: 2.4%), nucleation of Si(111) is expected at the gamma-Al2O3/Al(111) interfaces. This well explains the preferentially (111) orientation for long air exposure. This model qualitatively explains the present experimental results of preferential orientation.We demonstrated control of the preferential orientation of large Si grains (>20micron) by interfacial oxide controlled ALILE. This technique should be useful to achieve advanced thin-film solar cells. This work was partially supported by STARC.Ref. [1] J. Schneider, J. Cryst. Growth 287, 423 (2006).
3:30 PM - Q7.4
Cyclohexasilane as a Precursor to Silcon-based Photovoltaics.
Douglas Schulz 1 , Philip Boudjouk 1 , Konstantin Pokhodnya 1 , Justin Hoey 1 , Xuliang Dai 1 , Iskander Akhatov 1
1 Center for Nanoscale Science and Engineering, North Dakota State University, Fargo, North Dakota, United States
Show AbstractA synthetic route to a higher-order silane precursor in the form of cyclohexasilane, Si6H12 or CHS, has been under development at ND State. Recent studies teach the utility of CHS as a precursor for two routes to silicon-based films of potential interest to the PV community. First, the PECVD growth of amorphous silicon thin films from CHS was evaluated and compared to growth using SiH4, Si2H6 and trisilane (Si3H8) under identical conditions. When H2 was used as a carrier gas with the CHS precursor, a growth rate of 50 nm/min was observed – a rate that is two or three times faster than the lower order silanes. Second, CHS-based inks have been used as precursors to printed silicon. These inks may contain polyfunctionalized CHS as a dopant using intermediates such as Si6H11Cl. The CHS inks are amenable to collimated aerosol beam direct write of printed silicon where deposits as thin as 10 microns in width have been realized.
3:45 PM - Q7.5
High Pressure Thermal Chemical Deposition of a-Si:H from Silane within Microstructured Optical Fibers.
Neil Baril 1 3 , Venkatraman Gopalan 1 3 , John Badding 2 , Mahesh Krishnamurthi 3 , Ivan Temnykh 3 , Pier Sazio 4 , Justin Sparks 2 , Rongrui He 2
1 Materials Research Institute, The Pennsylvania State University, University Park, Pennsylvania, United States, 3 Materials Science and Engineering, The Pennsylvania State University, University Park, Pennsylvania, United States, 2 Chemistry, The Pennsylvania State University, University Park, Pennsylvania, United States, 4 Optoelectronics Research Centre, University of Southampton, Southampton United Kingdom
Show AbstractWe report a simple high pressure thermal chemical deposition technique capable of infiltrating high aspect ratio structures such as microstructured optical fibers with a-Si:H tubes and wires. High quality films and structures of a-Si:H are generally produced via PECVD, or hot wire CVD. However, because these techniques rely on remote generation of reactive radicals for deposition they are not suitable for the infiltration of deep high aspect ratio structures. Template organization of materials is a powerful tool for developing new and exciting functionalities and light matter interactions. We show that thermal deposition from silane at temperatures as low as 400°C is possible when using high pressures. Raman and IR spectroscopy give confirmation to the presence of hydrogen in the deposited material. Without optimization of the deposition process we have already achieved optical loss values at 1.55 μm that are greater than a 20 fold improvement as a result of the hydrogenation. The best non-hydrogenated microwires deposited under similar pressure conditions have losses of 17dB/cm, we have already achieved loss values as low as 3dB/cm. Generally high pressure pyrolysis of silane leads to the formation of Si fines that are a result of gas phase homogeneous reactions. We have found that the increased surface area to volume ratio within high aspect ratio templates such as the capillaries of microstructures optical fibers favors heterogeneous surface deposition allowing the formation of highly conformal structures and solid micro wires centimeters in length. We have investigated the deposition within fused silica microstructured optical as well as polyimide and PTFE substrates. Infiltration of a-Si:H into high aspect ratio structures would enable the production of waveguides capable of long, intense light matter interactions. Such interactions could be exploited for the production of new in-fiber photonic, photovoltaic, and optoelectronic devices that would, for example, allow manipulation of light in transmission.
4:30 PM - Q7.6
Electrodeposition of Si in a Non-oxygenated Organic Solvent.
Mikhael Bechelany 1 , Jamil Elias 1 , Laetitia Philippe 1 , Johann Michler 1
1 , EMPA, Thun Switzerland
Show AbstractOne cost-effective potential process route for thin films is electrochemical deposition. In general, the films deposited by this method do not possess the crystalline perfection or low levels of electrically active impurities of single crystal epitaxial films deposited by techniques such as molecular beam epitaxy or chemical vapour deposition. Nonetheless, in crystallised Si thin films used for PV application, a significant amount of defects can be tolerated due to the short diffusion length necessary for the carriers. Furthermore, in applications where larges areas of semiconductors are required, such as photovoltaic power generation or corrosion protection as opposed to integrated circuit fabrication, the low cost and comparatively low material demands make this deposition technology attractive in terms of ultimate commercialization. In addition, for application in solar cells, electrodeposition allows one to easily alter both band gap and lattice constant by composition modulation through control parameters such as applied potential and temperature of the bath. The main problem in electrodeposition of elemental silicon is the large reduction potential of silicon and the high reactivity of most of its compounds to water. Therefore, elemental silicon cannot be obtained from aqueous solutions. This means that the process must be carried out either in high-temperature molten salts or at near-ambient temperature in a nonaqueous electrolyte: ionic liquid or organic solvent under inert atmospheres. However, not only the control of an oxygen free atmosphere is needed during deposition, but also the stabilisation of the layer before exposure to air should be controlled, for limiting possible oxidation of the deposit. In this talk, we will present our efforts in the deposition of oxygen-free silicon, the stabilization of the Si deposit before exposure to air and finally the optimization of the process in order to tune the thickness and the porosity of the layer to have the best photovoltaic absorption.
4:45 PM - Q7.7
Intrinsic and Extrinsic Residual Stresses in Multi-crystalline Silicon Thin Film Solar Cells on Glass by a Novel Combined Diode Laser and Solid Phase Epitaxy Process.
George Sarau 1 , Arne Bochmann 1 , Michael Becker 1 2 , Silke Christiansen 1 2
1 , Institute of Photonic Technology, Jena Germany, 2 , Max-Planck-Institute for Microstructure Physics, Halle Germany
Show AbstractStress engineering in thin film solar cells on foreign substrates represents an important manufacturing issue to prevent cracking or peeling off of the film or even substrate bending. In order to achieve low stress levels in the film, the intrinsic and extrinsic components of the residual stress are addressed separately. The intrinsic stress is minimized by using optimal laser and solid phase crystallization (SPC) parameters, since it depends on the preparation conditions and the resulting microstructure, while the extrinsic (thermal) stress is minimized by choosing a substrate with the thermal expansion coefficient close to that of crystalline silicon. Furthermore, the substrate should become soft during processing to allow stress relaxation and it should have a low thermal conductivity requiring low laser fluence during melting to avoid holes in the film due to overheating. The borosilicate glass (BSG) from Schott, namely borofloat 33, has been found to meet these demands.In the present contribution, we use micro-Raman spectroscopy (μRS) as a non-destructive method to analyze stress states at the micrometer scale in multi-crystalline silicon thin film solar cells on BSG produced by a new laser - SPC process. In this process, an amorphous silicon layer (~ 100 nm thick) is melted and solidified by scanning a beam of a diode laser array (12 mm x 0.1 mm wide line focus) to form a large grained seed layer which is then epitaxial thickened (to >2 μm) by SPC (www.high-ef.eu). Based on literature data on thermal expansion coefficients of crystalline silicon and BSG, the extrinsic stress has been accurately evaluated to be 3 MPa and tensile over the relevant temperature interval (25 - 560 oC). Extrinsic stress is incorporated in the seed layer during cooling from the BSG holder temperature of 690 oC, which is necessary to avoid film and glass cracking in the course of laser crystallization, to room temperature. This stress is transferred to the soft BSG during epitaxial thickening by the SPC process at 600 oC but it is recovered during cooling to room temperature of the complete thin film solar cell (including the pn-junction). The measured stresses by μRS in both the seed layer and the complete solar cell correlate with the resulting crystalline microstructure visualized by electron backscatter diffraction (EBSD) indicating that they have mainly intrinsic origin. Stress mappings show stress concentrations up to 300 MPa at structural defect configurations such as grain boundary triple (multiple) points, a grain boundary separating a large grain from a region with twin lamellas, and in overlapping laser irradiated areas of adjacent scan lines. To conclude, more focused materials optimization is applied by separating the residual stress into its intrinsic and extrinsic components. The novel solar cell material presented here can be produced at competitive costs and it is expected to provide solar cell efficiencies > 10%.
5:00 PM - Q7.8
Intergranular Microstructure and Residual Stresses Investigations by EBSD on Laser-crystallized Polycrystalline Si Thin Film on Glass Substrate for Solar Cells.
Xavier Maeder 1 , Christoph Niederberger 1 , Silke Christiansen 2 , Gudrun Andrae 2 , Annett Gawik 2 , Fritz Falk 2 , Johann Michler 1
1 Laboratory for Mechanics of Materials and Nanostructures, Empa, Thun Switzerland, 2 , Institute for Photonic Technology, Jena Germany
Show AbstractCombined process of diode laser crystallization of an amorphous Si seed layer and solid phase epitaxy (SPE) are used to produce polycrystalline silicon thin film on glass substrate for solar cells. The laser crystallization process offers the advantage to produce large grains of several 100 microns in size while heating the temperature sensitive substrate only for a very short duration. Grain size, orientation distribution, grain boundary population and lattice defects of polycrystalline silicon thin films are investigated by electron backscatter diffraction analyses (EBSD). The low angle boundaries (< 15°) form between 5 and 15% of the grain boundary population. A systematic strong lattice twisting with a rotation axes parallel to the growth direction is observed in the larger grains. This progressive lattice rotation is generally between 10 and 25° from one side of the grain to the other and can be up to 50° in the strongest cases. Intergranular misorientation rate higher than 0.2° per micron, constant for several hundreds of microns have been observed. The intergranular misorientation is associated both with geometrically necessary dislocations and low angle boundaries which can serve as recombination centre for electron hole pairs. Since the lateral grain size is up to two orders of magnitude larger than the film thickness, these dislocations and low angle grain boundaries are becoming an important factor deteriorating the solar cells performance. EBSD cross correlation technique has been used to assess the residual stress in the grains that influence the mechanical integrity of the device. The calculations show residual stress values on the order of several GPa inside the grains with strong intergranular misorientation. The strain analyses show that the progressive lattice bending is composed both of rigid body rotation and lattice distortion.
5:15 PM - Q7.9
Raman Imaging of Si-crystallization Induced by Laser Scribing in Amorphous Thin Film of Photovoltaic Module.
Tomoya Uchiyama 1 , Taisuke Ota 1 , Minoru Kobayashi 1 , Isao Sumita 2
1 , Nanophoton corporation, Suita, Osaka Japan, 2 , V-Technology Co., Ltd., Yokohama, Kanagawa Japan
Show AbstractLaser scribing is widely used for the removal of thin films in photovoltaics. The main purpose of the scribing is the electrical isolation of the individual segments of a monolithically serial connection of solar cells. But it is recently known that the laser scribing sometimes degrades the photovoltaic module and it is inferred that the thermal effect of the ns-laser pulse induce the crystallization and as the results, the electrical isolation becomes imperfect. To make it clear how much crystallization is induced and how the crystallization is distributed, we measured the distribution with the laser Raman microscope. Raman spectroscopy is very powerful tool to evaluate the crystallization but has one weak point. It takes huge time to construct high-definition Raman spectral images. For example, for a Raman image of 256 by 256 pixels, it takes more than 1000 minutes or 18 hours if it takes 1 second for 1 Raman spectrum. To avoid the slow speed problem, we used new-generation, ultra-fast Raman microscope, RAMAN-11 (Nanophoton corp.). The laser Raman microscope is equipped with line-illumination laser to excite the Raman scattering at four hundred points along the line and also with multi-spectra detection system so that the microscopes can measure four hundred spectra at the same time. The microscope scans the line-shaped laser laterally with the laser beam scanner quickly without any vibration. With the combination of the line illumination, multi-spectra detection system and laser beam scanner, the microscope took 80,000 spectra within several ten minutes over the laser scribed thin film and constructed the image of the crystallization distribution. From the Raman image, it was observed that the highly crystallized parts remained sparsely and the transition from the amorphous to the crystal exists along the scribing.
5:30 PM - Q7.10
High Rate Deposition of Cluster-suppressed Amorphous Silicon Films Deposited Using a Multi-hollow Discharge Plasma CVD.
Kazunori Koga 1 , Hiroshi Sato 1 , Yuuki Kawashima 1 , Masaharu Shiratani 1
1 Dept. of Electronics, Kyushu Univ., Fukuoka, Fukuoka, Japan
Show AbstractHydrogenated amorphous silicon (a-Si:H) is most widely employed as a top cell material in thin Si tandem solar cells. Because light exposure reduces their efficiency, the light induced degradation of a-Si:H represents an important issue for its application to Si tandem solar cells [1]. In our recent studies, a-Si:H films with a less volume fraction of clusters in the films have been found to show less light induced degradation. We have obtained stable a-Si:H films at a deposition rate of 0.12nm/s using a multi-hollow discharge plasma CVD method [2]. To increase the deposition rate, we have examined effects of gas residence time and gas pressure using the method. Here, we report the results.Experiments were carried out with a multi-hollow discharge plasma CVD reactor [2]. SiH3 radicals and clusters are generated in the hollows. Clusters are transported to the downstream region by gas flow while SiH3 radicals are transported towards the upstream and downstream region due to their fast diffusion. Therefore, incorporation of clusters into the films deposited in the upstream region can be significantly suppressed [3,4].We have studied dependence of the deposition rate and the volume fraction on the gas residence time τg as a parameter of gas pressure. For 0.1Torr, the deposition rate increases from 0.73nm/s to 0.96nm/s with increasing τg from 0.0019s to 0.0027s, and then it decreases from 0.96nm/s to 0.6973nm/s with further increasing τg from 0.0027s to 0.0037s. For 0.2 and 0.5Torr, the τg dependence of the deposition rate is similar to that for 0.1Torr. For 1.0Torr, the deposition rate monotonously increases with increasing τg for τg <0.062s. The maximum deposition rate realized for each pressure increases exponentially from 0.014nm/s to 0.96nm/s with decreasing the pressure from 1.0Torr to 0.1Torr because of the faster diffusion of SiH3 at the lower pressure. We have also measured dependence of the volume fraction on the deposition rate from the dependence of the deposition rate and the volume fraction on gas residence time. The volume fraction only slightly increases with increasing the deposition rate, suggesting that highly stable a-Si:H films can be deposited at a high deposition rate by using the multi-hollow discharge plasma CVD method.Based on the results, we have deposited a-Si:H films under a low gas pressure of 0.1Torr and measured the defect density of the films using ESR. By using the multi-hollow discharge plasma CVD method, we have succeeded in depositing films with a stabilized defect density down to 4x1015cm-3 at high deposition rate of 1.2nm/s. [1] D. L. Staebler and C. R. Wronski, Appl. Phys. Lett., 31, 292 (1977)[2] K. Koga et al., Jpn. J. Appl. Phys., 44, L1430 (2005).[3] W. M. Nakamura, et al., IEEE Trans. Plasma Sci. 36, 888 (2008).[4] W. M. Nakamura, et al., J. Phys.: Conf. Series 100, 082018 (2008).
Q8: Poster Session II
Session Chairs
Thursday AM, December 03, 2009
Exhibit Hall D (Hynes)
9:00 PM - Q8.1
Plasma Treatment of Indium Compounds to Reduce Their Adverse Health Effects.
Kazunori Koga 1 , Shinya Iwashita 1 , Hiroshi Miyata 1 , Masaharu Shiratani 1 , Miyuki Hirata 2 , Yutaka Kiyohara 2 , Akiyo Tanaka 2
1 Dept. of Electronics, Kyushu Univ., Fukuoka, Fukuoka, Japan, 2 Graduate School of Medical Science, Kyushu Univ., Fukuoka, Fukuoka, Japan
Show AbstractIndium compounds are widely employed as materials for flat panel displays, liquid crystal displays, and solar cells. In the middle of 1990's, they have been pointed out to have harmful effects on animals [1]. Potential risks for the human health increase with increasing amount of them used in industry. The first case of interstitial pneumonia caused by occupational exposure to indium-tin oxide (ITO) was reported in 2003 [2]. Evaluation of toxic effects for the indium compounds is important as well as reducing the effects. In this study, we have studied effects of plasma treatment on the atomic compositions of copper indium gallium diselenaide (CIGS) particles to reduce the toxic effects. We employed N2 dc discharges for the treatment. N2 gas pressure was 26Pa. Discharge current density was 0.51mA/cm2. The electron temperature and density were 2.3eV and 5.7x106cm-3, respectively. The discharge duration was 1 hour. The atomic compositions were measured by EDS. Before the treatment, atomic compositions of CIGS particles are 23% of Cu, 11% of In, 15% of Ga, and 51% of Se whereas, after the treatment, they are 24% of Cu, 11% of In, 18% of Ga, and 47% of Se. We will apply other plasma treatment to reduce the toxicity of CIGS. To evaluate the pulmonary toxic effect of CIGS, we have repeatedly instilled the CIGS particles into the trachea of rats. For the experiments, we have instilled three types of the CIGS into the trachea of the anesthetized rats with ether 3 times a week. 0.5 mg/kg, 5 mg/kg and 50 mg/kg of CIGS particles were suspended in 1ml/kg distilled water. Control rats were given 1ml/kg of only vehicle, distilled water. No rats died during the instillation periods. During the observation period, body weight gain was not suppressed in all CIGS treated groups, compared to the control group. The relative lung weight in three CIGS groups was significantly higher than that of the control group. Moreover, those in the CIGS 5 mg/kg group and the CIGS 50 mg/kg group showed higher compared with CIGS 0.5 mg/kg group. Foci of slight to severe inflammatory lesions were present in all the CIGS- treated groups. These lesions deteriorated among all CIGS-treated rats during examination period. The severity of lung lesions was getting worse with the passage of time among CIGS-treated rats. The results indicate that CIGS particles cause pulmonary damages when instilled intratracheally into rats. Therefore, human exposure to CIGS particles should be severely suppressed.[1] A. Tanaka, "Toxicity of indium arsenide, gallium arsenide, and aluminium gallium arsenide", Toxicol. Appl. Pharmacol. 198,405 (2004).[2] T. Homma, T. Ueno, K. Sekizawa, A. Tanaka, M. Hirata,"Interstitial pneumonia developed in a worker dealing with particles containing indium-tin oxide", J Occup Health 45, 137 (2003).
9:00 PM - Q8.10
Evaluation of GaTlP for Use in Multijunction Photovoltaics.
Chandler Downs 1 , John Chivers 1 , Thomas Vandervelde 1
1 Electrical Engineering, Tufts University, Medford, Massachusetts, United States
Show AbstractTo achieve the highest possible conversion efficiencies in multijunction photovoltaics, the individual layers of the device must both be lattice-matched and have optimal band-gap spacing. Lattice-matched or strain-compensated epitaxy is required for the growth of junctions thick enough to elicit high quantum efficiency. The conversion efficiency is directly linked to how well the band gaps are optimized for absorption of the solar spectrum. Ideally, for spectral matching, one would have an infinite number of junctions that are current-matched; however, fabrication of a large number of junctions is neither easy nor desirable because of problems that arise from series resistance. In the end, it becomes a balancing act where the optimal number of junctions for a high efficiency concentrator cell is 3-6 junctions. Unfortunately, many of the optimal lower band-gaps for these multijunction cells do not occur in the dominant materials system (i.e Ge and mixtures of In, Ga, Al, As, and P). As such, of late there has been a strong push to characterize new materials in hopes of providing more design options for photovoltaic cells. GaTlP is one such material, theorized to be useful as one of the lower junctions of 3+-junction cells while still being lattice-matched to GaAs and Ge. In this research, the change in lattice constant and band gap of GaTlP with varying compositions are investigated first by computational simulation and then with physical devices. New efficiency records should be achievable by incorporating these new optimal junction materials into the design for multijunction cells. This development will help solar concentrator cells achieve grid parity, thereby becoming a viable renewable energy choice.
9:00 PM - Q8.11
Structural and Chemical Analysis of Femtosecond Laser-doped Silicon using Transmission Electron Microscopy.
Matthew Smith 1 , Arthur Reading 1 , Bonna Newman 2 , Joe Sullivan 2 , Mark Winkler 3 , Meng-Ju Sher 3 , Eric Mazur 3 , Tonio Buonassisi 2 , Silvija Gradecak 1
1 Materials Science and Engineering, Massachusetts Institute of Technology, Cambridge, Massachusetts, United States, 2 Mechanical Engineering, Massachusetts Institute of Technology, Cambridge, Massachusetts, United States, 3 Physics, Harvard University, Cambridge, Massachusetts, United States
Show AbstractFemtosecond laser irradiation of silicon in the presence of chalcogens (S,Se, Te) dopes the silicon to non-equilibrium levels. Femtosecond laser-doped silicon exhibits near-unity broadband absorption of visible and near infrared light and therefore is of great interest for photodetector and photovoltaic applications. The fabrication and optimization of such devices necessitates understanding the relationship between processing parameters, optoelectronic properties, and microstructure. It has been shown that post-process annealing (30 minutes, 500-900°C) decreases the IR-absorptance of chalcogen laser-doped silicon, likely due to a diffusive dopant relaxation mechanism. However, the interplay between IR-absorption and dopant relaxation is not well understood in this materials system. Electron microscopy techniques are well-suited for the investigation of the microstructure and dopant distribution. In this work we use transmission electron microscopy (TEM) techniques to study both sulfur and selenium femtosecond laser-doped silicon in order to (1) elucidate the ablation and melting proccesses that occur during the doping process and (2) further our understanding of dopant relaxation from the optically active state. Bright-field TEM and Selected Area Diffraction (SAD) of cross-sectional samples reveal that crystalline, polycrystalline, and amorphous regions develop during laser irradiation. The distribution of these regions varies between selenium and sulfur doping, which we show is a result of the phase of the dopant precursor. From these observations we construct models for the ablation and melting processes that occur during femtosecond laser-doping of silicon with sulfur as well as selenium. Dark-field Scanning Transmission Electron Microscopy (DF-STEM), Aberration-corrected STEM, Energy Dispersive X-ray Spectroscopy (EDX) and Electron Energy Loss Spectroscopy (EELS) are used to study the effects of annealing temperature on dopant distribution. We study the evolution of dopant-rich features with annealing temperature and discuss the possible dopant relaxation mechanisms at play.
9:00 PM - Q8.12
Polysilicon Thin Film Solar Cells on Thermally Stable Aluminum Doped Zinc Oxide Films.
Yoo Jin Lee 1 2 , Kyung Ho Kim 1 , Moon Sik Hwang 1 , Dong Jee Kim 1 , Koeng Su Lim 2
1 R&D Center, TG Solar Corporation, Hwaseong, Gyeonggi, Korea (the Republic of), 2 Department of Electrical Engineering and Computer Science, Korea Advanced Institute of Science and Technology (KAIST), Daejeon Korea (the Republic of)
Show AbstractPolysilicon thin film solar cells on aluminum doped zinc oxide (AZO) films were fabricated and characterized. AZO films with 0.5 um thickness were deposited by sputtering. The polysilicon films were deposited by plasma enhanced chemical vapor deposition (PECVD) method and crystallized by solid phase crystallization (SPC). Bare AZO films were degraded after thermal annealing process. However, AZO films with silicon films as capping layers maintained their electrical and optical properties. The sheet resistance of AZO films with silicon capping layer were 30 and 17 ohm/sq. before and after crystallization, respectively. In this research, we fabricated the polycrystalline silicon thin film solar cells on AZO/glass structured substrate. The photo I-V characteristics were measured at standard illumination condition with AM 1.5, 100 mW/cm2 and Voc (open circuit voltage), Jsc (short cuicuit current density), FF (fill factor), Eff (efficiency) of this cell were 0.3 V, 19 mA/cm2, 0.4 and 2.3 %.
9:00 PM - Q8.13
Deactivation Treatments of Silicon Solar Cells for Efficiency Recovery after Illumination Degradation.
Teng Yu Wang 1 , Terry Wang 1 , Yan Zu Chen 2 , Chwung Shan Kou 2 , Chien Hsun Chen 1 , Wei Lun Chang 1 , Sung Yu Chen 1 , Chen Hsun Du 1 , Wen Ching Sun 1
1 Photovoltaics Technology Center, Industrial Technology Research Institute, Hsinchu Taiwan, 2 Department of Physics, National Tsing Hua University, Hsinchu Taiwan
Show AbstractBoron doped (p-type) single crystalline silicon wafer is wildly used for solar cell. However the energy transfer efficiency can not keep stable under illumination. It has been known that the efficiency degradation under illumination due to the recombination active boron-oxygen complex. Therefore, the reasonable way to avoid illumination degradation is decompose the boron-oxygen complex. In this work, we apply the deactivation treatment to p-type single crystalline silicon solar cells. The method we used includes thermal annealing treatment, illumination treatment, applied voltage treatment, capacitively couple plasma (CCP) treatment, and plasma immersion ion implantation (PIII) treatment. The results show the deactivation treatment is working and the energy transfer efficiency is thereby increased by more than 1% absolute compared to the degraded state.
9:00 PM - Q8.14
Characteristics of Ga-doped ZnO Films Grown by Atomic Layer Deposition using DEZn/TMGa and N2O Precursors.
Yen-Ting Lin 1 , Chien-Hua Chiu 2 , Chi-Ying Hsiao 2 , Sheng-Yu Hsiao 2 , Kuo-Yi Yen 2 , Jyh-Rong Gong 1 2
1 NanoScience, National Chung Hsing University, Taichung Taiwan, 2 Physics, National Chung Hsing University, Taichung Taiwan
Show AbstractGallium-doped ZnO (GZO) films were grown on (11-20) plane sapphire substrates at low temperatures by atomic layer deposition(ALD) using diethylzinc(DEZn)/trimethylgallium (TMG) and nitrous oxide (N2O). The conductive, optical and structural properties of the ALD-grown GZO thin films were investigated by four-point probe method, optical transmission spectroscopy and θ-to 2θ x-ray diffractometry (XRD). It was found that all the ALD-grown GZO thin films exhibited low resistivities in the ranging of low 10-4 Ω-cm along with optical transmittances being as high as 90% in the visible light and near infrared spectra.
9:00 PM - Q8.15
Solar-grade Silicon Refining - Understanding Impurity Behavior in the Arc Furnace.
Sarah Bernardis 1 , Marisa Di Sabatino 2 , Sean Gaal 2 , Mariana Bertoni 1 , Bonna Newman 1 , Sirine Fakra 3 , Jafar Safarian 4 , Merete Tangstad 4 , Tonio Buonassisi 1
1 , Massachusetts Institute of Technology, Cambridge, Massachusetts, United States, 2 Metallurgy, SINTEF Materials and Chemistry, Trondheim Norway, 3 Advanced Light Source, Lawrence Berkeley National Laboratory, Berkeley, California, United States, 4 Materials Science and Engineering, Norwegian University of Science and Technology, Trondheim Norway
Show AbstractAs the demand for solar-grade silicon increases, alternative Si-bearing raw feedstock sources are being explored. In these lower purity alternative raw feedstock materials, impurities detrimental to solar cell performance can be present in abundance, yet little is known about their micrometer scale spatial distribution and chemical state before and during the refining process. In this work, we report on studies of micrometer-scale spatial distributions and chemical states of impurities present in hydrothermal quartz, and their behavior as the quartz reaches melting temperature in the arc furnace. Results from this investigation aim at helping develop novel refining possibilities to extract detrimental impurities in the very first stage of the solar-grade silicon production. We employed synchrotron-based analytical microprobe techniques of micro X-ray fluorescence (µ-XRF) and of micro X-ray absorption spectroscopy (µ-XAS), to probe the microscale distribution and chemical state of embedded impurity-rich particles in hydrothermal quartz. As a reference, bulk impurity concentrations, microscopic distribution, particle sizes, their chemical states, and their clustering behaviors, were first studied at room temperature. Sub-micron sized particles containing Ti, Fe, Ca, K, Mn, and/or P were observed and found to decorate mostly structural defects. In addition, oxidation states of Fe, Ti, and Ca impurity clusters were assessed via µ-XAS. Subsequently, a batch of hydrothermal quartz samples was prepared by heating in a CO(g) controlled environment, as to mimic the process in the arc furnace, and then quenched to room temperature. Each quartz sample was heated to a range of temperatures between 1000oC and 1900oC in order to observe 1) impurity distribution modifications and 2) chemical state transitions, both due to heating and to the SiO2 reactivity with the CO(g) environment. These results were complemented by measurements of dissolved impurity concentrations acquired via bulk analysis technique, such as inductively-coupled plasma mass spectrometry (ICP-MS).
9:00 PM - Q8.16
ToF-SIMS Study of Pulsed Laser Melting Energy Density on Ti Implanted Si for Intermediate Band.
Javier Olea Ariza 1 , David Pastor Pastor 1 , German Gonzalez Diaz 1 , Ignacio Martil de la Plaza 1
1 Dpto. de Fisica Aplicada III, Facultad de Ciencias Físicas, Universidad Complutense de Madrid, Madrid Spain
Show AbstractIn the last few years, an increasing effort has been carried out to improve the efficiency in the solar cell technology. One of the most promising concepts is the Intermediate Band Solar Cell (IBSC), which can overcome significantly the theoretical solar conversion efficiency limit for single junction solar cells. The formation of an Intermediate Band (IB) by means of deep impurity centers inside the bandgap of the semiconductor permits to capture photons with energy lower than bandgap. The absorbed photons can promote electrons from the valence band to conduction band directly or by means of two steps using the IB as intermediary. The transition from the localized deep levels to an IB formed by delocalized states requires an impurity concentration above the Mott limit. Ti doped silicon is a promising candidate to form a IB material, however the high concentration required exceeds the solid solubility limit of Ti in Si (~4x1014 cm-3). Ion implantation is one of the most appropriate techniques to achieve the high impurity concentration needed. Although, this process damages the crystalline lattice of the host semiconductor, thus thermal annealing treatments are required in order to recover the crystallinity. Non equilibrium thermodinamical processes like Pulsed Laser Melting (PLM) are necessary to surpass the Mott limit concentration avoiding the solid solubility limit.In the present work, we have studied the effect of PLM energy density on IB formation in Ti implanted Si. We have analyzed silicon layers with Ti concentrations above the theoretical Mott limit (5x1019 cm-3) implanted in the range 1015-5x1016 cm-2 and subsequent PLM at energy densities from 0.2 to 0.8 J/cm2. Time-of-Flight Secondary Ion Mass Spectroscopy (ToF-SIMS) measurements have been carried out to determine the Ti concentration depth profile after PLM. A push effect is observed in all the Ti profiles. These profiles become steeper with a higher and sharp maximum. The ToF-SIMS measurements allowed us to determine that Ti concentration is above the Mott limit for all the samples annealed with the different energy densities. For high implantation doses a clearly loss of the as-implanted gaussian Ti profile is observed. As implantation dose decreases the push effect is increased. For the sample implanted with the lowest dose, a very strong push effect produces a considerable reduction in the thickness of the layer containing an impurity concentration above the Mott limit. Whereas at low implanted dose the push effect become stronger as PLM energy density increases, at the highest doses the trend is the opposite.In this study, we discuss the effects of PLM energy density over different Ti implanted doses above the Mott limit. We show that all the PLM annealed samples maintain the Ti concentration required to the IB formation. We find that push effect is significantly pronounced for the samples implanted with 1015 cm-2 and annealed at the highest energy densities.
9:00 PM - Q8.17
Reduction of Porous Silica Pellets by Electrodeoxidation in Molten Salts.
Emre Ergul 1 , Ishak Karakaya 1 , Metehan Erdogan 1
1 Metallurgical and Materials Engineering, Middle East Technical University, Ankara Turkey
Show AbstractDirect electrochemical reduction of porous SiO2 pellets in molten CaCl2 salt and CaCl2-NaCl salt mixture were investigated by applying 2.8 V potential. The study focused on the effects of temperature, powder size and cathode contacting material. Starting materials and electrolysis products were characterized by X-ray diffraction analysis and scanning electron microscopy. Due to reactive nature of silicon, different cathode contacting materials were used to test the extent of reactions between silicon produced at the cathode and the contacting material. X-ray diffraction patterns showed that silicon produced at the cathode reacted with nickel and iron in stainless steel to form Ni-Si and Fe-Si compounds respectively. Besides, studies revealed that higher temperature and smaller particle size had positive effects in increasing reduction rate. The results were interpreted from variation of current versus time graphs under different conditions, microstructures and composition of the reduced pellets.
9:00 PM - Q8.18
Electrical Properties of Silicon with Bistable Impurity Complexes.
Smagul Karazhanov 1 , Tine Naerland 1 , J. Mayandi 1 , Rune Sondena 1 , Arve Holt 1
1 Department for Solar Energy, Institute for Energy Technology, Kjeller Norway
Show AbstractMany impurity complexes in silicon such as, e.g., boron-oxygen and iron-boron complexes, are found to be bistable. Commonly for analysis of the carrier lifetime experiments in silicon with the bistable recombinative complexes the Shockley-Read-Hall recombination theory is used, which is valid for stable defects with one configuration and one energy level in the band gap. However, the theory might fail upon considering the recombination centers through bistable defects, which can be in two different configurations separated by a potential barrier. This work presents a study of electrical properties of silicon with bistable impurity complexes. The analysis has been performed for statistics of free electrons and holes, their recombination rate and lifetime. The results have been compared with those obtained from the Shockley-Read-Hall recombination theory.
9:00 PM - Q8.19
Abnormal Dopant Distribution in Textured Crystalline Silicon Solars Cell with n+/p Junctions.
Chel-Jong Choi 1 , Mira Jeong 1 , Yeon-Ho Gil 1 , Myeong-Il Jeong Jeong 1 , Young-Woo Ok 2 , Dong-Hwan Kim 3
1 School of Semiconductor and Chemical Engineering, Chonbuk National University, Jeonju Korea (the Republic of), 2 School of Electrical and Computer Engineering, Georgia Institute of Technology, Atlanta, Georgia, United States, 3 Department of Materials Science & Engineering, Korea University, Seoul Korea (the Republic of)
Show AbstractTransmission electron microscopy (TEM) coupled with selective chemical etching was employed to assess two-dimensional dopant profiles in the n+/p junctions that were formed using phosphorus (P) diffusion from a POCl3 source in textured Si. The selective chemical etching leads to the formation of thickness fringes in the TEM images, which represent iso-concentration contours in n+/p junctions. It was shown that the chemically delineated junction depth in the peak region of texture Si surface is much deeper than that in valley one. Based on the Silvaco simulation using a stress model with the default values, such an abnormal dopant distribution in n+/p junctions could be attributed to the different behavior of P diffusion under the exertion of strains induced by the geometry of textured Si surface.
9:00 PM - Q8.2
Ruthenium Sensitizers for All-Purpose Dye-Sensitized Solar Cells.
Chun-Guey Wu 1
1 , National Central University, Jhong-Li Taiwan
Show AbstractAbstract:Dye-sensitized solar cells (DSCs) have emerged as a promising candidate for photovoltaic technology in virtue of the impressive photon-to-current conversion efficiency and low manufacture cost. An attractive branch of DSC in conjunction with organic hole-transport material, the all-solid-state DSCs (SSDSCs), has been developed since 1998. In SSDSCs, the light-capturing ability of sensitizer self-assembled onto the surface of n-type semiconductor, and the interfacial charge transfer are crucial to the photovoltaic performance of cells. Here we show a well-functionalized ruthenium sensitizer in which a conjugated segment and a hole-transporting moiety connected sequentially to the ancillary bipyridyl ligand for enhancing the light-harvesting capacity of sensitizer and for facilitating the regeneration of photon-oxidized sensitizers. Most importantly, its application in all-solid-state DSCs and the influence of hole-transporting moiety on the physical properties of sensitizer and on cell performance was also explored.
9:00 PM - Q8.21
Optimization of Gettering Processes of Metallurgical-grade Silicon for Solar Cell Applications.
Inna Iskandarova 1 , Igor Zvyagin 1 , Andrey Knizhnik 1 , Andrey Konovalov 1 , Boris Potapkin 1 , Natalya Arutyunyan 1 2 , Alexander Zaitsev 1 2 , Thomas McNulty 3 , Timothy Sommerer 3 , Mohamed Rahmane 3 , Victor Lou 3 , Stanislav Soloviev 3 , Alexei Vert 3 , Svetlana Selezneva 3
1 , Kintech Lab Ltd, Moscow Russian Federation, 2 , I.P. Bardin Central Institute for Ferrous Metallurgy, Moscow Russian Federation, 3 , GE Global Research, Niskayuna, New York, United States
Show AbstractAn approach for optimization of metallurgical-grade silicon purification processes for application in manufacture of solar cells was developed. This approach is intended for modelling of the main gettering processes and for estimation of photovoltaic efficiency.Conversion efficiency of a solar energy in the electric is substantially determined not only by the total impurity concentration, but also by their chemical state in silicon: in particular, conversion of a recombination active impurity from isolated state to the precipitated decreases its activity. Gettering processes, which are included in the technology of solar cell manufacturing, are usually used for such impurity redistribution. As the chemical behaviour of various impurities strongly differs with temperature, optimization of gettering processes is necessary.In order to optimize gettering processes we developed a program tool based on the fundamental physical and chemical laws. System of transport and kinetic equations is used for simulation of impurity diffusion and complexes formation, including precipitates growth and dissolution. The following gettering processes can be modeled: 1) annealing; 2) gettering by an aluminium layer during deposition of electric contact; 3) gettering by a heavy-doped phosphorus layer during formation of p-n junction. The description of physical and chemical behavior of impurities in silicon is based both on known experimental data, and on calculations of necessary parameters by means of modern thermodynamic and quantum-chemical methods. Developed tool helps to choose a gettering regime (a temperature profile, time, getter layer thickness) to optimize gettering efficiency for the given initial chemical composition of the silicon wafer.Besides it, the program tool can analyze the recombination activity of various types of defects in silicon on the basis of carrier lifetime criterion, so it is possible to estimate recombination activity of the silicon wafer after gettering. Using this program tool we investigated gettering efficiency for solar-grade silicon with Cr, Fe, Ni, Co impurities. The results of modeling are in reasonable agreement with available experimental data. We demonstrated that solar cell efficiency can be significantly increased by optimal choice of gettering conditions.
9:00 PM - Q8.22
Photovoltaic Effect in Ar+ Irradiated SrTiO3.
Sang-Yong Kim 1 , Chanho Yang 2 , Jan Seidel 2 3 , Seung-Yeul Yang 1 , Steve Byrnes 2 , Martin Gajek 2 , Ramamoorthy Ramesh 1 2 3
1 Materials Science and Engineering, University of California Berkeley, Berkeley, California, United States, 2 Department of Physics, University of California, Berkeley, Berkeley, California, United States, 3 Materials Sciences Division, Lawrence Berkeley National Laboratory, Berkeley, California, United States
Show AbstractPhotovoltaic (PV) system is one of the most promising environmentally friendly and renewable energy sources at present. In order to make highly efficient and cost effective PV system, many researchers have been focusing on finding new PV materials. Recently, complex oxide materials have been widely studied for various PV applications due to their unique properties such as chemical stability, transparency and tunable electrical characteristics. Here, we report the manifestation of PV effects in commercially available SrTiO3 (STO) single crystals as a result of Ar+ irradiation on the polished surface and sequentially application of electric field along in-plane at elevated temperature. The sample shows a PV effect with an open circuit voltage of about 0.5 V and a short circuit current of 10 nA. Also, it is presented that the short circuit current varies with the Ar+ irradiation time and electric field applying conditions. Possible mechanisms of the observed PV effects will be discussed on the basis of a generation of oxygen vacancies by the Ar+ irradiation and an electric-field-induced redistribution of the oxygen vacancies. This might result in a potential gradient necessary to separate photo-induced electron-hole pairs. Our observations through the oxygen vacancy engineering suggest a viable way to use oxides for PV applications.
9:00 PM - Q8.23
Method for Optical Characterization of Dielectric Thin Films on Pyramid Textured Crystalline Silicon Wafers with Spectroscopic Ellipsometry.
Mario Saenger 1 , Craig Herzinger 2 , James Hilfiker 2 , Jianing Sun 2 , Tino Hofmann 1 , Mathias Schubert 1 , John Woollam 2 1
1 Department of Electrical Engineering, University of Nebraska - Lincoln, Lincoln, Nebraska, United States, 2 , J.A. Woollam Co., Lincoln, Nebraska, United States
Show AbstractThe use of surface textures and antireflecting coatings are two widely used methods to increase the quantum efficiency of solar cells by reducing reflection losses. Deviations of the index of refraction and thicknesses from the design values cause a higher reflection and thus, lower device efficiency. To obtain the maximum performance of a solar cell the knowledge of accurate values of the antireflecting film complex index of refraction, namely, the index of refraction “n” and extinction coefficient “k” is required. However, the high light scattering and depolarization due to the surface texture leads to inaccurate results when standard thin-film optical characterization methods are utilized. Previous attempts to characterize films on pyramid textured silicon wafers used non-traditional measurement geometries and an effective medium approximation for the optical constants [1]. However, the results using the effective medium approximation approach demonstrated discrepancies depending on the measurement geometry. Here we present a new method where no effective medium approximations are used and where the scattering effects of the texture are included in the data analysis. With this new method it is possible to obtain consistent results for the optical constants and thickness values independently of the measurement geometry.
9:00 PM - Q8.24
Performance of Silicon PV Devices Based on a Single Step for Doping, Anti Reflection and Surface Passivation.
Vikram Iyengar 1 , Barada Nayak 1 , Mool Gupta 1
1 Electrical & Computer Engr., University of Virginia, Charolttesville, Virginia, United States
Show AbstractThere is a need for reducing the manufacturing cost of silicon solar cells. This reduction in manufacturing cost can be achieved by simplification of the process. We have examined the feasibility of a single step process where in a thin film of spin on dopant (SOD) serves as dopant source, anti reflective and a surface passivation layer. Results from this simplified process have shown cell efficiency of over 14% and further improvements are achievable. Commercially available spun on dopants are spun on p type single crystal silicon wafers and its thickness was adjusted for quarter wave for it to provide an anti reflection property as well as a dopant source. The good quality of the interface between the spin on dopant layer and silicon substrate after thermal diffusion provides surface passivation effect. The surface dopant concentration in a single diffusion step can be controlled by adjusting the phosphorus concentration in the initial spin on dopant solution. Internal quantum efficiency close to 90% over the spectral range was observed by controlling the phosphorus content in the SOD solution. Detailed PV fabrication and performance results in terms of Voc, Isc, fill factor and efficiency for this single step process will be presented. Results of devices based on this single step process on chemically textured surfaces will also be presented. Minority carrier lifetime, surface recombination velocity and affect of thermal processing will be discussed. Further improvements in efficiency are achievable by process optimization and using wafers with higher minority carrier lifetime.
9:00 PM - Q8.25
Tandem PV Devices Based on Silicon and Germanium Sub-cells.
Mahieddine Emziane 1
1 Materials and Energy, Masdar Institute of Science and Technology, Abu Dhabi United Arab Emirates
Show AbstractWe have considered a tandem PV device consisting of a silicon cell at the top together with a germanium bottom cell. By adding a Ge cell (with a lower bandgap) to a Si single-junction cell, the spectral coverage of the resulting tandem device is extended in the near-infrared region leading to an enhanced overall performance through the conversion of more photons. Furthermore, based on the concept that we developed and published recently [1], we have adopted a three-terminal device configuration which allows an independent operation of the two individual cells without the need for current matching and the corresponding tunnel junctions. This configuration also provides an improved output compared to the same sub-cells used in a two-terminal device as it has been shown for structures based on III-V semiconductors. The advantage of having an additional Ge cell at the bottom is further highlighted by the current-voltage (I-V) characteristics simulated for these Si/Ge tandem PV devices. For the purpose of implementing such Si/Ge tandem devices, a structure optimization was undertaken. The results clearly indicate that an optimal overall performance can be reached for the tandem device within the constraint of keeping it reasonably thin.[1] Emziane M., Nicholas R.J., Journal of Applied Physics, 102 (2007) 074508.
9:00 PM - Q8.26
Prospects for Defect Engineering in Amorphous Silicon Thin Films.
Sebastian Castro-Galnares 1 , Mariana Bertoni 1 , Tonio Buonassisi 1
1 , Massachusetts Institute of Technology, Cambridge, Massachusetts, United States
Show AbstractDespite the promise of amorphous silicon (a-Si) as an active-layer material for photovoltaic devices, thirty years of research have resulted in cells with industrial efficiencies barely above 6% (small-area laboratory record = 9.5% [1]) – well below the Shockley-Queisser efficiency limit. Low efficiencies are usually associated with short minority carrier diffusion lengths, which are influenced by small minority carrier lifetimes and low carrier mobility.In this contribution, we explore the potential for applying defect-engineering techniques developed for crystalline silicon (c-Si) to a-Si. We have produced plasma-enhanced chemical vapor deposited (PECVD) films on conducting and non-conducting substrates, to measure relevant optoelectronic material parameters (absorption coefficient, majority carrier mobility, and minority carrier lifetime). In particular, we explore opportunities during film growth, to manipulate the structural/mechanical properties of the material, towards enhancing carrier transport.[1] Martin Green, Keith Emery, Yoshihiro Hishikawa and Wilhem Warta, Prog. Photovolt: Res. Appl. 2008; 16:435–440
9:00 PM - Q8.27
Efficient Light Trapping by Ultrafast Laser Texturing Process for Higher Efficiency Silicon Solar Cells.
Barada Nayak 1 , Vikram Iyengar 1 , Mool Gupta 1
1 Electrical & Computer Engr., University of Virginia, Charolttesville, Virginia, United States
Show AbstractEfficient light trapping is essential for high efficiency solar cells and it becomes more important for thin film devices. Ultrafast laser texturing process provides an efficient method of texturing for single crystalline, multi-crystalline, as well as thin films. Using optimized laser parameters reflection losses less than 1% have been achieved.Laser texturing process has been applied for n & p-type silicon wafers. Solar cell devices are fabricated using a spin-on-dopant process. Front metal contacts have been made by Ti-Pd-Ag for p-type and Al for n-type wafers. In this presentation the details of fabrication process and photovoltaic performance in terms of open circuit voltage, short circuit current, fill factor and efficiency are presented. Initial results indicate that the laser texturing process is advantageous for n-type wafers in terms of higher open circuit voltage compared to p-type wafers. With high light trapping properties of ultrafast laser textured surfaces, an optimized device processing will lead to much higher efficiency.
9:00 PM - Q8.28
Fabrication of Photo/Beta-Voltaic Diodes from Boron Carbide/Silicon Heterojunction.
Ruqiang Bao 1 , Joseph Sandstrom 2 , Anthony Caruso 2 3 , Douglas Chrisey 1
1 Materials Science and Engineering, Rensselaer Polytechnic Institute, Troy, New York, United States, 2 Center for Nanoscale Science and Engineering, North Dakota State University, Fargo, North Dakota, United States, 3 Department of Physics, University of Missouri-Kansas City, Kansas City, Missouri, United States
Show AbstractThe long-lived and self-supplied beta-voltaic devices are very useful in remote and long term use, such as satellite in the aerospace, as well as conventional batteries in consumer devices, such as cell phones and laptop computers. But beta-voltaic devices made with conventional semiconductors tend to suffer very rapid degradation due to radiation damage. Fortunately, the degradation problem can potentially be alleviated by replacing conventional semiconductors with icosahedral boron carbide semiconductors, which have been proved to survive extensive irradiance due to the self-healing of icosahedral structure. We report that the prototype photo/beta-voltaic diodes based on boron carbide/Si heterojunction have successfully fabricated. The results show that boron carbide thin film by PECVD has amorphous structure, and the remaining H in boron carbide thin film passivates the dangling bond and improves the voltaic performance of device. The quantum efficiency of boron carbide / Si diode is high in the range of 400 nm - 1000 nm, which indicates that boron carbide/Si diode is suitable to be used to make photovoltaic device. In both photo-voltaic and beta-voltaic characterization of boron carbide / Si diode, the repeatable response and sensitivity to the beam intensity were observed.
9:00 PM - Q8.29
Low Temperature Crystallization of Amorphous Silicon Thin Film using Ni Catalyst for Nickel Induced Crystallization Deposited by Metal Organic Chemical Vapor Deposition.
Se Wan Son 1 , Chang Woo Byun 2 , Seung Ki Joo 1
1 Material Sicence and Engineering, Seoul National University, Seoul Korea (the Republic of), 2 Material science and engineering, Seoul National University, Seoul Korea (the Republic of)
Show AbstractMetal Induced Crystallization seems very promising technology for fabrication of thin film solar cells due to its low cost and short processing time. However metal contamination can cause severe effect on the electrical properties of solar cells such as high leakage current and therefore the decrease in Voc and Jsc. It was reported that Ni organic source such as (MeCp)2Ni and acac2Ni can be used as a Ni precursor for metallic chemical vapor deposition of ultra thin Ni layer which can drastically reduce the residual Ni in the crystallized poly silicon thin film. In this study we fabricated the poly silicon thin film using metallic chemical vapor deposition of Ni as a catalyst for crystallization of amorphous silicon and acac2Ni was used as a Ni organic source. The layers of amorphous silicon and Nickel were deposited on glass substrate by means of PECVD and MOCVD(Metal Organic Chemical Vapor Deposition) respectively. The thickness of the amorphous silicon layers were fixed at 2micron and the thickness of the Ni layers were varied from 10 to 100angstron. Post annealing was then carried out in a tube furnace in hydrogen atmosphere at different temperatures(300-600celsius degree) for different times (1-5hours) in order to crystallize the bi-layer films. Band gaps of the samples were measured by means of photoluminescence spectroscopy and crystallized samples have a band gap of 1.1eV which is the band gap of crystalline silicon. It was found that Ni deposited a-si films can be crystallized below 500 celsius degree. Ni concentration in crystallized films will be measured by SIMS(Secondary Ion Mass Spectrometry) to investigate the effect of the thickness of Ni layers on the Ni contamination in the poly crystalline si thin films and The effects of the thickness of the Ni Layer on the grain size of the p-si film will be discussed in detail.
9:00 PM - Q8.3
High Mobility Epitaxial Ge Films Grown by a General Chemical Route.
Guifu Zou 1 , Hongmei Luo 1 , Yingyin Zhang 1 , Darrick Williams 1 , Filip Ronning 1 , Quanxi Jia 1
1 , LANL, Los Alamos, New Mexico, United States
Show AbstractGe is attracting increasing attention because of its wide applications in ultra-large scale integration and nanotechnology. Ge/Si heterostructures have been used in high performance devices because of their higher electron and hole mobilities relative to those of Si. Here, epitaxial Ge films have been grown on Si (001) substrates by a chemical solution deposition technique. X-ray diffraction analysis of the films revealed that the full width at half-maximum (FWHM) of (004) rocking curve was 0.34o. The high-resolution transmission electron microscopy of the epitaxial Ge films further demonstrated the good epitaxial quality of films. The carrier concentration and mobility of epitaxial Ge films are measured to be 3.3*10^13 cm-2 and 1510 cm2V-1s-1 at 300 K, respectively. Its photovoltaic performance stimulates Ge films prepared by this process to be potential application in the industry.
9:00 PM - Q8.30
Stress Imaging of Microdefects in Multicrystalline Silicon Using Infrared Birefringence Imaging.
Sebastian Oener 1 , Sergio Castellanos 1 , Stephan Schoenfelder 2 1 , Vidya Ganapati 1 , Barry Lai 3 , Tonio Buonassisi 1
1 , Massachusetts Institute of Technology, Cambridge, Massachusetts, United States, 2 , Fraunhofer Center for Silicon Photovoltaics, Halle Germany, 3 , Advanced Photon Source, Argonne, Illinois, United States
Show AbstractBulk microdefects can impact the cost of solar cell production in a variety of ways: SiC microdefects can cause disruptions during wire sawing; SiC and Si3N4 microdefects can shunt finished solar cells devices. Dislocations reduce solar cell conversion efficiency and can lead to reduced breakdown voltages (module failure). There is hence a need to detect such defects at a wafer or brick level, in a non-destructive and high-throughput (imaging) manner, and to predict their impact on wafer mechanical properties and electrical performance.In this contribution, we demonstrate the utility, strengths, and weaknesses of infrared (IR) birefringence imaging to detect bulk microdefects (dislocations, grain boundaries, inclusions and precipitates) in multicrystalline silicon (mc-Si) solar cell materials. Unique birefringence signatures can be observed for each of the aforementioned defect types, and even distinguish among sub-classes of defects. Finite element models are presented for each defect type, reproducing with fidelity the magnitude and form of stress distributions observed using IR birefringence. In conclusion, we discuss future prospects for this unique characterization tool to predict the mechanical and electrical behaviors of mc-Si solar cell material in industrial and R&D settings.
9:00 PM - Q8.31
CdTe Solar Cells at CEA / LETI.
Sergio Bernardi 1
1 , CEA / LETI / MINTEC / DOPT / STM, Grenoble France
Show AbstractThe present communication will present the results obtained at CEA/LETI in Grenoble (France) in the frame of an experimental work finalized to directly test the criticalities and chances of this technology row. By means of Closed Space Sublimation (CSS) processes, appropriate annealing treatments and specific CdTe-contacting procedures, we reached the state-of-art of this technology with 15% conversion efficiency under standard illumination conditions in “superstrate” cells configuration for structures deposited on 3 mm thick low-iron containing sodalime glass.
9:00 PM - Q8.5
Ageing of Standard PV Module when Integrated in a V-trough Concentration System.
Filipa Reis 1 2 , Miguel Brito 1 , Victoria Corregidor 2 , Joao Wemans 2 , Gianfranco Sorasio 2
1 DEGGE - SESUL, Faculty of Sciences of University of Lisbon, Lisbon Portugal, 2 , WS Energia, Oeiras Portugal
Show AbstractThe growing interest on concentration photovoltaic (CPV) technologies arises from its potential to significantly reduce the PV electricity cost/kWh. Within concentrators technology, V-trough systems, such as DoubleSun® technology, can be pointed as a shortcut solution for two main reasons: they are less demanding in terms of tracking accuracy, which avoids high costs of the final product; and they make use of standard silicon solar modules, a technology with many years of given proofs and a well standardized industrial manufacture, benefiting from scale production.¶Ideally, the efficiency of conventional PV modules should increase with concentration since the irradiance falling upon the cells is higher. However, such boost will lead to new challenges: higher operating cells temperature, which diminishes the open-circuit voltage (Voc); high series resistance (Rs) losses, due to the higher transverse flow of current from the solar emitter to the front grid (initially designed to work under a maximum irradiance of 1000W/m2); and possibly, accelerated modules’ degradation rate originated by a higher exposure to the sunlight. These issues limit the efficient application of standard modules up to a few suns. This work focuses on the study of the long-term behaviour and reliability of the standard modules on DoubleSun technology applications. The DoubleSun is a V-trough 2x concentration using reflective mirrors and 2-axis tracking.¶We report on the behaviour of standard 1-sun silicon modules when integrated in a reduced-scale prototype similar to DoubleSun technology. As far as ageing is concerned, the concentrator qualification standard IEC62108 procedures were fully and successfully applied to the standard 1-sun silicon modules. In this report particular attention is devoted to ultraviolet conditioning test, outdoor exposure test and off-axis beam damage test. For example, the relative power degradation after exposure to UV accumulation (UVA+UVB) of 50kWh/m2 was found to be 5%.¶The successful IEC62108 tests suggest that, for the analysed range of concentration factors, standard silicon modules may be eligible for concentration.¶
9:00 PM - Q8.6
Novel Determination Method for Solar Cell Series Resistance and Shunt Resistance Considering Voltage Dependent Carrier Collection.
Stefan Puttnins 1 2 , Andreas Rahm 1 , Marius Grundmann 2
1 , Solarion AG, Leipzig Germany, 2 Institut für Experimentelle Physik II, Universität Leipzig, Leipzig Germany
Show AbstractThe reliable determination of series resistance RS and shunt resistance RSh is important to identify loss mechanisms of solar cells. Various methods to determine RSh and RS from IV-curves have been published. These methods rely on a model fit to the IV-curve, determination through comparison of IV-curves measured at different illumination intensities using the superposition principle or use the IV-curve's slope.However, IV-curves of thin film solar cells like Cu(In,Ga)Se2 (CIGSe) or CdTe are often affected by voltage dependent carrier collection, i.e. a non-constant photo-current IPh(V) . This causes the superposition principle for junction current and constant photo-current to fail and changes the slope of the IV-curve.Hence, determination of RS and RSh with common methods is only possible for illuminated cells by fitting to an extended model including voltage dependent carrier collection, for which at least two new parameters have to be included. This makes the fitting routine less robust and reliable.The objective of this work is the introduction of a new determination method for RS and RSh. We first calculate the voltage dependent carrier collection from IV-curves measured at different irradiation intensities without introducing new parameters nor the need of model fitting.The measured IV-curves are then corrected by the influence of the voltage dependent carrier collection and RS and RSh can be iteratively determined by standard methods. The reliability and validity of the calculated RS and RSh are extensively discussed on model IV-curves with given parameters and measured IV-curves of commercial CdTe and flexible CIGSe solar cells on polyimide foil produced by SOLARION AG.
9:00 PM - Q8.7
Investigation on the Electrical, Optical and Plasma Properties of ZnO:Al Films Deposited by Laser Induced High Current Pulsed Arc.
Jin-Bao Wu 1 , Chao-Ying Chen 1 , Jia-Jen Chang 1 , Ming-Sheng Leu 1 , Mei-Yi Li 2
1 , Materials Research Laboratories, Industrial Technology Research Institute, Hsinchu Taiwan, 2 , National Nano Device Laboratories, Hsinchu Taiwan
Show AbstractHighly transparent conductive Al-doped ZnO (AZO) thin film were deposited at 100°C by laser induced high current pulsed arc (LIHCPA) from a Al-Zn alloy target. A pulsed current more than 1 kA was generated on the Al-Zn target in order to produce fully-ionized plasma of high ion energy. The films properties were correlated with the growth conditions, including O2 flow rate, Al doping content of the target and pulsed arc current applied to the target. The properties of the films such as crystallinity, chemical composition, electrical and optical features of the AZO films were investigated. In order to investigate the correlations between the properties of films and plasma characteristics, a quadrupole plasma analyzer was used to identify the positive ion energy distribution (IED). It was confirmed that the pulsed current had a large effect on both the ion energies and ion fluxes generated within the plasma. The experimental XRD results showed that the AZO films had preferred c-axis orientation along the (002) plane. The minimum resistivity was as low as 4.2×10-4 Ω-cm, while the average transmittance of films in the visible range (400-700 nm) was above 84 %. It was found that the O2 flow rate affected not only the electrical properties of the films but also the band gap. The results clearly showed that when the O2 flow rate increased from 30 sccm to 150 sccm, the resistivity increased from 4.2×10-4 to 3.7×10-3 Ω-cm, and the band gap of the AZO films calculated by UV/VIS spectrometer measurement decreased from 3.76 eV to 3.58 eV accordingly. Moreover, by taking deposition of O2 flow rate 30 sccm and 2 kA pulsed current into consideration, the results revealed that the plasma species generated by the Al-Zn alloy target were Zn+, Zn2+, Al+, O+, O2+ and ZnO+ among which that the Zn+ ion energy can achieve more than 50 eV. The relation of ion energy measurements as a function of pulsed arc current is also discussed.
9:00 PM - Q8.8
Feasibility Study of Mesoplasma CVD for Direct Production of High Pure Si Thin Films from Trichlorosilane.
Makoto Kambara 1 , Junichi Fukuda 1 , Takanori Yamamoto 1 , Toyonobu Yoshida 1
1 Materials Engineering, University of Tokyo, Tokyo Japan
Show AbstractRecent global concerns on the energy and environment have spurred development of high quality Si production for solar cells. Siemens method has been employed as a versatile and established technology for years. However, as it attains practically the production yield as small as 30% and also requires substantial energy consumption, several new approaches have been introduced for enhancement of the yield based primarily on an increase in the reaction interface. Still, expecting a radical improvement in the production efficiency, novel processes from innovative standpoints are longed for.One potential approach would be high hydrogen partial pressures in the reduction of SiHCl3 (TCS) for Si production. A large amount of H2 addition to TCS is expected to promote the reduction and thus improve the production yield of Si. However, disproportionation reaction of TCS takes place inevitably in parallel and the overall reduction yield is in principle limited. In fact, according to the equilibrium chemical species in this system, SiCl4 (STC) becomes stable as a primary phase at the temperature of processing, around 1300 – 1500K, and excess addition of H2 would not contribute effectively to the reduction reaction but promote the reproduction of TCS, instead. In contrast to the effect of H2, in a simulated plasma environment that atomic hydrogen exists as a stable phase, thermodynamic analysis turns out that STC is no longer stable at any temperatures and Si vapor becomes a primary stable phase at temperatures higher than 2000K. The second stable silicon-containing phase is ClSi vapor and its relative molar ratio to Si vapor is ~40% at 2500K and 10% at 3000K, respectively. This estimation therefore suggests that the high yield of Si production is possible if Si vapors are to be used effectively in an appropriate film deposition through direct condensation of these vapors.Meanwhile, an intermediate pressure plasma, mesoplasma, is expected to provide unique plasma environment owing to its low electron and gas temperatures and viscous flow pressure regime. Proving its uniqueness of reduced ion and thermal damages and high rate transport of reaction species, we have demonstrated polycrystalline Si film deposition at 1000 nm/sec and epitaxial Si film growth at 70 nm/sec from monosilane (SiH4) as a source gas, attaining reasonably high electric performance. In addition, high flux of atomic hydrogen by several orders of magnitude than the low pressure processing is readily anticipated in the mesoplasma environment. Therefore, upon applying mesoplasma to the TCS reduction process, one can foresee that Si vapor would be formed and high quality Si films are deposited directly from these vapors at high yields. Based on these ideas, in this work, we have attempted a direct deposition of Si thin films from TCS as source gas under mesoplasma condition. Preliminary experimental results will be presented in the session.
9:00 PM - Q8.9
First Principle Calculations on Impurity Effect in Silicon Grain Boundary.
Hiroshi Mizuseki 1 , Ambigapathy Suvitha 1 , Natarajan Venkataramanan 1 , Ryoji Sahara 1 , Yoshiyuki Kawazoe 1
1 , Institute for Materials Research, Tohoku Univ., Sendai, Miyagi, Japan
Show AbstractCurrently, over half of all solar cells produced worldwide are made from polycrystalline silicon (pc-Si). While these devices have comparatively lower efficiencies mainly due to transition metal impurities (1014-1016 cm-3) in most of the pc-Si materials. Metal silicide precipitate is observed to increase with decreasing atomic coincidence within the grain boundary (i.e. increasing Σ values). In the present study we used DFT method to understand relationship between sigma value and impurity precipitation. First we study the dopant position and the nature of interaction between the GB and transition metal. Finally we have studied the electronic changes that occurred up on doping the transition metal impurities in the GB regions using Σ 3, Σ 9 grain boundary of polycrystalline silicon. Σ 3, Σ 9 grain boundary of silicon were constructed using GB studio [1]. The calculations were performed with monkhorst-pack of 4x4x4, using projector augmented wave (PAW) pseudopotentials with a wave cutoff of 240 eV. The Perdew - Wang (PW91) functional is used for the generalized gradient approximation (GGA) as implemented in VASP code. Periodic boundary condition was applied along x and y axis of the super cell containing 96, 288 atoms for Σ 3, Σ 9 GB respectively. To validate our studies we used copper, iron, nickel and chromium as dopants and compare our results with the experimental findings. This work was partially supported by New Energy and Industrial Technology Development Organization (NEDO) of Japan.1). H. Ogawa, Mater. Trans. 47 (2006) 2706.